journal uii yogya

i

Vol.1, No. 1, Mei 2011 ISSN 1979-7303

DAFTAR ISI

Daftar Isi i Comparison of the Behaviour of Pre-Crack and Intact Partially Saturated Stiff Clay Subjected to Plane Strain Testing Miftahul Fauziah ........................................................................................................................... 1 Compressive Strength Assessment of Concrete Structures from Small Core by Point Load Test Achfas Zacoeb and Koji Ishibashi ................................................................................................ 10 Essential Strategies to Promote Earthquake Safer Non-Engineered Housing: An Indonesian Perspective Setyo Winarno .............................................................................................................................. 19 The Implication Of Peatland Built Environment In Urban Drainage System: Case Study Of Sungai Merah Area, Sibu, SarawakFrederik Josep Putuhena, Ting Sie Chun, and Salim Said ........................................................... 29 Perbandingan Kekakuan Struktur Lantai Komposit Kayu Glugu-Beton dengan Berbagai Jenis Penghubung GeserSusastrawan .................................................................................................................................. 38 Masonry Unit Utilizing Aggregate from Construction Demolition Bound with Asphalt I Nyoman Arya Thanaya .............................................................................................................. 47 Pengembangan Peta Hazard Kegempaan untuk Pulau Jawa dan Perbedaannya dengan Hasil-Hasil Studi Terdahulu Lalu Makrup, Sarwidi, dan Susilo ................................................................................................ 58 Shoreline Change Model Using the Epr Method and the Simulation of Coastal Vulnerability in Sambas District-West Kalimantan M. Meddy Danial, Rustamaji, and Eka Priadi .............................................................................. 68 Time-History Response of 2-D Timber Frame Reinforced with Wooden Panel Ali Awaludin ................................................................................................................................ 75 Wavelet-Spectrogram Analysis of Surface Wave Technique: A Novel Procedure for Non-Destructive Measurement on Pavement Surface LayerSri Atmaja P. Rosyidi .................................................................................................................. 83 Analisis Antrian Akibat Daerah Sempit di Jalan Bebas Hambatan Sukarno ......................................................................................................................................... 93

Fauziah: Comparison of The Behaviour of Precrack and Intact Partially Saturated Stiff Clay ...Testing

Comparison of the Behaviour of Pre-Crack and Intact Partially Saturated Stiff Clay Subjected to Plane Strain Testing

Miftahul Fauziah

Civil Engineering Departement, Islamic University of Indonesia, Yogyakarta, Indonesia, email: [email protected]

Abstract: The present modeling of soil is based on principles of continuum mechanics. The existence of cracks bring soils to be non uniform condition and therefore not amendable to analysis by continuum mechanics. On the other hand, there has been considerable interest in the application of fracture mechanics to replicate their behaviour. Partially saturated soil behaviour is different from those of fully saturated soil because of the influence of suction. Results obtained with the strength theory of saturated soil could not be directly applied to solve the partially saturated soil problems. Moreover, the mechanical properties of soil that usually conducted using triaxial apparatus rather than biaxial device will be addressed according to the fact that geotechnical field problems are often trully or close to plane strain situation. This paper will disseminate the result of some experimental testing on the on the properties of pre-crack and intact partially saturated compacted kaolin clay specimen under plane strain condition. Two types of intact specimen and pre-crack specimen had tested under different matric suction and net normal stress. The results demonstrated that a higher failure stress and compressive strength was reached by the intact specimen than the pre-crack specimen. Shear strength of the intact specimens were higher along their axial strain than those of pre-crack specimens. It was also shown that a pronounce failure stress was exhibited by pre-crack specimen rather than intact specimen. Keywords: Pre-crack, partly saturated, matric suction, and plane strain. INTRODUCTION The present modeling of soil is based on principles of continuum mechanics in spite of the fact that discontinuities are known to develop when such geological materials are subject to loading. However, in the case of strong rock, there has been considerable interest to account for such discontinuities using fracture mechanical approach (Ingraffea, 1987). Many studies have been conducted on detailed aspect of such discontinuities, but these are of limited practical application in an actual situation. The existence of cracks and fissures, which are the results of mechanical, thermal and volume-change-induced stresses, such soils are non uniform and therefore not amendable to analysis by continuum mechanics. On the other hand, fracture mechanical theory may be used to advantage to replicate their behaviour. Atkinson and Bransby (1982) proposed the conventional

failure criteria for soils which might be partly appropriate to yield-dominant behaviour, but not this category of brittle fracture. In practice, there is the possibility that soil behaves more like a brittle material. The soil ruptures suddenly under compressive loading like soft rock, starting from the weakest fracture in it. A basic premise of fracture theory is that crack like imperfections are inherent in engineering materials. These defects have the tendency to make stresses higher, which eventually trigger off fractures when a material body is subjected to a critical load or undergoes damage under cyclic loading. This present state of fracture mechanical theory has been summarized by Anderson (2004). Lo et al. (2005) modeled brittle overconsolidated clay accordingly and thereby provided a rational basis for the prediction of such soil behaviour. The behaviour of partially saturated soil is different from those of fully saturated soil because of the influence of suction. It has been

1

Jurnal Rekayasa Sipil (JRS), Vol. 1, No. 1, Mei 2011: 1-9

observed that several stability problems, involving soils used as construction materials are associated with variability of water content that occur periodically in nature and consequently in changes of matric suction. Although soils are generally assumed fully saturated below the groundwater table, they may be semi saturated near the state of full

saturation under certain conditions. The partially saturated soil can be the results of fluctuation of the ground water table being the consequence of natural processes or human activity. Partially saturated soils form the largest category of materials that cannot be classified by classical saturated soil mechanics concepts. The strength theory of saturated soil could not be directly applied to solve the problems of partially saturated soil. Partially saturated soil is generally characterized by three phases, soil solids, water, and air. Fredlund and Morgenstern (1977) was introduced an additional independent phase, a so called the air–water interface or contractile skin. Based on multi phase continuum mechanics, a theoretical stress analysis of an partially saturated soil has been presented (Fredlund and Morgenstern, 1977; Fredlund and Morgenstern, 1976). The analysis concluded that any two of three possible normal stress variables can be used to describe the stress state of an unsaturated soil. This is in contrast to saturated soil, where it is possible to relate the mechanical properties of the soil to the effective stress only. The mechanical properties of soil is routinely interpreted from conventional triaxial testing or axisymmetric conditions; whereas, testing of soil using the plane strain device would be more useful information, as more geotechnical field problems such as landslide problems, failure of soils beneath shallow foundations, and failure of retaining structures are truly or basically occur in these situations. It was reported by Mochizuki et al. (1993) that when soil is tested under plane strain conditions, it, in general, exhibits a higher compressive strength and lower axial strain. The plane strain testing on the behaviour of fined grained sands had been reported (Alshibli and Akbas, 2007, Alshibli et al., 2004; Alshibli and Sture,

2000; Bizzarri, 1995; Han and Vardoulakis, 1991; Hans and Drescher, 1993; Lee, 1970; Marach et al., 1984; and Mochizuki et al, 1993). However, the plane strain testing of clay soils have only been initiated recently (Fauziah and Nikraz, 2008; Fauziah and Nikraz, 2007; Lo et al., 2000; Drescher et.al (1990)) and published data of such tests especially for brittle clay material is very limited. This paper will disseminate the result of experimental study on the behaviour of pre-crack partially saturated clay specimens compare to the intact specimen by the use of plane strain apparatus, although the behaviour of overconsolidated clay (Fauziah and Nikraz, 2007) and fracture characteristics of brittle clay may also be determined by this test apparatus. Discription of the apparatus, specimen preparation, testing method, procedures and data analysis will be presented in the following discussion. Some results of the testing will be compared with the known soil behaviour and previous working. EXPERIMENTAL PROGRAM Material and Specimen Preparation

Plane strain experiments have been performed on remoulded kaolin clay specimens. The material used in this study was commercial kaolin clay, which is a product of UNIMIN PTY LTD, Australia, with a specific gravity Gs =2.6, liquid limit LL of 53.5 %, plasticity index PI of 22.74 % and plastic limit PL of 30.76 %. Firstly, a kaolin clay sample was slurried to a uniform consistency of 1 ½ times its liquid limit using an electrical soil mixer. This slurry was obtained by mixing 8 kg of kaolin powder with 6 kg of water using the electric mixer for about 2 hours. A lubricated steel cylindrical mould with the height of 600 mm and 150 mm in diameter was used to consolidate the slurry using a hydraulic tester in over a period of one to two weeks which the maximum of 300 kPa was applied in three stages. Two circular perspexes were placed at both ends of the slurry in the mould to apply the pressure evenly to the slurry. The slurry was allowed to

2

Fauziah: Com

consolidatepressure wbeing squperspexes pressure wfurther sconsolidatiwas then lubricated specimen plastic filmfaces withwater presthe dehumtested. In the plamembrane specimen The rectanon top of assembly aThe rubbethe porousHAED andrings and specimen, the top aspedestal oapparatus. slipped ovuse of a srigid perspsecured byset up. Thlaser senso Plane Stra

A modifiewas used fof partiallshows the pThe plane cell, with internal diAn initial of 36 mm,72 mm, soon the baslaterally byits out-of

mparison of The

e by its owas appliedueezed out

and the mwas appliesettlement ion process

extruded formers.

and the fom sheeting h liquid wassure, the sp

midifier for a

ane strain was first

with the aingular porou

the specimand the highr membrans plate, thed secured b

rectangulatogether w

ssembly weof the pla

The rubbver the pedeset of O-rinpex plates

y the use of he pressureor set-up we

ain Equipm

ed conventfor the purpoly saturateplane strainstrain arrathe height

iameter andprismatic s

, height of that the asp

se pedestal,y two rigid plane move

Behaviour of P

own weighd to prevet between

mould. A hed when th

change. s had comp

from the Followin

ormer wereafter seali

ax. To obtapecimens wat least 2 d

test set-upt placed oid of a sleus plate wa

men followeh air-entry

ne was nexte top assemby the use oar Perspex

with the porere then plaane strain er membraestal and sengs and cla

were thenclamp set f cell, top

ere then inst

ment

tional triaxose of planeed specimen test arrangangement it of 300 md 30 mm wspecimen w72 mm, andpect ratio is where it wperspex plaement whic

Precrack and In

ht and sment the slu

the circuhigher vertihere was

Once tleted, the s

mould inng this, te wrapped ing both thain even powere placeddays until it

p, the rubbover the teeve stretchas then placed by the tdisc (HAEDt slipped ovmbly and tof a set of clamp. T

rous plate aaced over t

compressiane was neecured by tamp set. Twn placed afor biaxial tassembly atalled.

xial apparae strain testien. Figure gement. s placed in

mm, 200 mwall thicknewith the widd thicknesss 2 was placwas restrainates to restrach make σ,

ntact Partially S

mall urry ular cal no the soil nto the in

heir ore

d in t is

ber test her. ced top D). ver the O-

The and the ion ext the wo and test and

atus ing

1

n a mm ess. dth of ced ned ain , ε.

ThprAcein the1919the2 friwoma

F

Saturated Stiff C

herefore, orincipal stre

30 mm entrally with

Figure 2, te specimen

941; Rowe a964; Bishope specimenor more, soiction and ould be negaintain cons

Figure 1.

igure 2. A sconditio

72 mm

Clay ...Testing

only major esses acting diagonal p

hin the intato simulaten. Previousand Barden

p and Greenn aspect ratio that the ef

the restraigligible andstant strengt

Plane strain

schematic don of the pre

(σ) and on the soil

precrack wact specimene the discons researchen, 1964; Leen, 1965) sugio (height toffect of loaint of loadd the specimth.

n test arrang

diagram of pe-crack spec

36 m

minor (ε)l specimen.

was formedn as shownntinuities inrs (Taylor,e and Seed,ggested thato width) of

ading platending framemen would

gement

plane straincimen

mm

) .

d n n , , t f n e d

3


A high air-entry disc (HAED) was used as the interface between the partly saturated soil and the pore water pressure measuring system to prevent any air from passing through the disc into the measuring system. All surfaces which are in contact with the specimen were greased to avoid the likelihood of scratching and reduced friction. An LVDT (Linear Vertical Displacement Transducer) was used to measure the axial displacements of the test specimen. The global volume change of the water saturated soil specimen was monitored by the use of an automatic volume change unit which is connected to the back-pressure line. A data acquisition system consisting of an MPX 300 data logger and a set of microcomputer were used to record the displacement, loads, pressure and volume change reading of the specimen. WINHOST 2.0 package software was used to convert digital bit data from the ADU (Analogue digital Unit) to engineering units based on the calibration of the relevant measuring unit, which was done before running the plane strain test. A more detail discription of this biaxial compression device can be found in previous working reported by Fauziah and Nikraz (2008) and Fauziah and Nikraz (2007). Testing Method and Procedure

Two types of intact specimen and pre-crack specimen have been tested under plane strain condition. The pre-crack specimens were formed a 30 mm diagonal precrack in the center of the intact specimen to simulate the discontinuities in the specimen. Two specimen of IM (intact specimen) and PCM (Pre-crack specimen) were tested under net normal stress of 0 and maximum matric suction of 500 kPa and two specimen of IN (intact specimen) and PCN (Pre-crack specimen) were tested under matric suction of 0 and maximum net normal stress of 800 kPa. The specimen was first saturated until the B-value of the specimen reached the value of 0.95-0.98, followed by matric suction and net normal stress applied and loading compression processed. An initial net normal stress of 0 and matric suction of 10 were applied to the IM and PCM specimens by set a cell pressure of 600

kPa, back pressure of 590 kPa and pore pressure of 600 kPa. The volume change of the soil skeleton was monitored continuously by the laser sensors. The change in volume of water in the specimen was also monitored continuously by the volume change gauge which was connected to the back-pressure line. Once the changes in the soil and water volumes had ceased, the test specimen was presumed to have fully consolidated under a matric suction of 10 kPa. The matric suction was next increased to 20 kPa by reducing the back pressure to 580 kPa. The corresponding changes in the soil skeleton and water volumes were monitored continuously until they had ceased, at which stage the corresponding void ratio and the water content of the test specimen were computed from the cumulative changes in soil skeleton and water volumes. The entire procedure, that is from increasing the matric suction to the desired value up to flushing out the air bubbles, was repeated for matric suctions of 50, 100, 200, 300, and 500 kPa respectively, which were applied by reducing the back pressure accordingly. Another test apparatus were set up for IN and PCN specimens similarly as before, except that the head was replaced with a porous disc. A cell pressure of 110 kPa, back-pressure of 100 kPa and pore-air pressure of 100 kPa were then applied to the specimen to provide an initial net normal stress of 10 kPa and matric suction of 0. The changes in soil skeleton and water volumes were then monitored continuously and when these changes had ceased, the total changes in soil and water volumes were noted. The net normal stress was first increased to 20 kPa by increasing the cell pressure to 120 kPa. Thereafter, the entire above procedure, starting from applying the net normal stress up to when the changes in soil and water volumes ceased, was repeated for net normal stresses of 50, 100, 200, 300, 500 and 800 kPa. The specimen was then compressed by elevating the base of the confining pressure cell at a constant velocity of 0.08 mm/m with the drainage line closed at at net normal stress of 0 kPa (NNS=0kPa) and matric suction of 500 kPa (MS=500 kPa) for IM and PCM

4


specimens and at matric suction of 0 and net normal stress of 800 kPa for IN and PCN specimens. This loading rate was deduced based on the permeability of adopted kaolin clay suggested by Bishop and Henkel (1962). The data were recorded at 3 minute interval test and it was terminated at the axial strain of about 20 % or sooner. The specimen was then taken out immediately for the purpose of moisture content test. In the analysis of the behaviour of the brittle unsaturated clay, the pore pressure parameter would be required in order to the determined the pore pressure increments and the matric suction. The pore pressure parameters were deduced from the volumetric deformation coefficient, which was obtained by laboratory testing. This procedure was adopted from Fredlund and Rahardjo (1993), although adapted to biaxial conditions. RESULTS AND DISCUSSIONS Figure 3 and Figure 4 present the strain softening response of the intact and pre-crack specimen. The peak stress and strain of the specimen is summarised in Table 1.

Table 1. The peak stress vs axial strain

Specimen name

Peak stress (kPa)

Vertical strain (%)

IN 159.18 3.64PCN 125.34 2.73IM 228.39 3.87

PCM 124.52 3.33 In general, the shear stress curves increase monotonically with the increasing of vertical strain until they reach peak stresses followed by strain softening behaviour. According to Lo et.all (2005), this is the typical phenomenon of specimen of brittle, hard partly saturated soil specimen and demonstrated elastic failure only. As can be seen in Figure 3 and Figure 4, the shear strength of the intact specimen of IM and IN were higher along the vertical strain than that the specimen containing discontinuities or precrack of PCM and PCN specimen. The highest failure stress of

228.395 kPa was reached by the intact specimen of IM, and the lowest failure stress of 124.519 kPa was derived by the pre-crack specimen of PCM. The presence of a fissure or discontinuity makes the soil weaker as the effective area offering resistance to shear is reduced. The shear strength along a surface of discontinuity is thereby less than that of the intact material. It is also shown from the graph and Table that pronounced peak shear strength was occurred to the precrack specimen than the intact specimen. Similar observation of pre-crack overconsolidated clay had been reported elsewhere (Lo et al., 2000).

Figure 3. Stress-strain of specimen under Net Normal Stress (NNS)= 0, and Matric Suction

(MS )= 500 kPa

Figure 4. Stress-strain of specimen under Matric Suction (MS )=0 and Net Normal

Stress (NNS)= 800 Kpa Figure 5 shows the constitutive surface of void ratio versus log net normal stress and log matric suction of the intact specimen, while

0

50

100

150

200

250

300

0 1 2 3 4 5 6 7 8

axial strain (%)

Dev

iato

ric

stre

ss (

kPa)

IM (Intact)PCM (Pre-crack)

0

50

100

150

200

0 1 2 3 4 5 6 7 8

axial strain (%)

Dev

iato

ric

stre

ss (k

Pa)

IN (Intact specimen)PCN (Pre-crack)

5


Figure 6 plots the constitutive surface of void ratio versus log net normal stress and log matric suction of the pre-crack specimen. The slope of the intersection curves are the volume change index Cm and water content index Dm

respectively for the case that net normal stress set to zero, whereas the slope of the intersection curves are the volume change index Ct and water content index Dt when the matric suction set to zero. Ct is the slope of the consolidation curve and is equal to the compressive index of a saturated soil, while Cm is the slope of the shrinkage curve.

Figure 5. Constitutive surface of void ratio of

intact specimen

Figure 6. Constitutive surface of void ratio of

pre-crack specimen The change in void ratio of the IM and PCM specimen in relation with matric suction depicted in Figure 7, while void ratio changes of the IN and PCN specimen in connection

with net normal stress changes presented in Figure 8. It can be shown from the graphs that the curves went down with the increasing of either matric suction or net normal stress for all of the specimens. The value of the volume change index of the IM and PCM specimens were calculated as 0.011and 0.016 respectively. The value of the Ct of the IN and PCN specimens are 0.219 and 0.257 respectively. The higher slope of pre-crack curves of IM and IM than that of intact curve of PCM and PCN indicated that the intact specimen had higher compressive strength than that of pre-crack specimen either tested under net normal stress of zero and matric suction of 500 Kpa or under zero matric suction and net normal stress of 800 kPa. This is in consistent with the observation reported by Lo et al (2000).

Figure 7. Void ratio versus matric suction of specimen under NNS=0 and MS=500 Kpa

Figure 8. Void ratio versus Net normal stress of specimen under MS=0, NNS=800 Kpa

0.30

0.50

0.70

0.90

1.10

1.30

1.50

1.70

1.00 1.50 2.00 2.50 3.00

Log (matric suction, net normal strees)

void

rati

o (%

)

Log M S Log NNS

0.30

0.50

0.70

0.90

1.10

1.30

1.50

1.70

1.00 1.50 2.00 2.50 3.00


void

rati

o (%

)

Log MS Log NNS

1.170

1.175

1.180

1.185

1.190

1.195

1.0 1.5 2.0 2.5 3.0 3.5

Log Matric suction (kPa)

Voi

d ra

tio

(%)

IM (Intact)PCM (Pre-crack)

0.800

0.900

1.000

1.100

1.200

1.300

1.0 1.5 2.0 2.5 3.0 3.5

Log Net Normal stress (kPa)

void

rati

o (%

)


6


The Constitutive surface of water content of intact specimen and pre-crack specimen plotted in Figure 9 and Figure 10, respectively. The constitutive surface of water content is defined by the compressive index Dm and Dt corresponding to the matric suction and net normal stress respectively. The value of compressive index obtained by determining the gradient of the linear portion of the curve of the water content against the log of matric suction for Dm and the log of net normal stress for Dt.. As can be seen from the graphs, the curves went down with the increasing of either matric suction or net normal stress.

Figure 9. Constitutive surface of water content

of intact specimen

Figure 10. Constitutive surface of water

content of pre-crack specimen The curves also indicated that the higher gradient value of 0.071 and 0.068 were reached by precrack specimens of PCM and PCN respectively, than that of the gradient

value of intact specimens of IM and IN which were 0.55 and 0.056 respectively. The changes in water content of the specimen in relation with matric suction depicted in Figure 11, while water content changes versus net normal stress plotted in Figure 9. It were clearly demonstrated from the graphs that under the same net normal stress and matric suction the intact specimen had lower gradient than the pre-crack specimen.

Figure 11. Water content change of specimen

under NNS=0 and MS=500 Kpa

Figure 12. Water content change of specimen

under MS=0, NNS=800 Kpa Similar to the curve of the void ratio changes in Figures 8-9, as well as water content changes in Figures 11-12 and consistent with the stress-strain behaviour of the specimens shown in Figures 4-5, the existance of the crack or discontinuities on the specimen not only weaken its shear strength as well as its compressive strenght but they also quicken the failure of the specimen.

0.10

0.15

0.20

0.25

0.30

0.35

1.00 1.50 2.00 2.50 3.00


wat

er c

onte

nt (%

)

Log MS Log NNS

0.10

0.15

0.20

0.25

0.30

0.35

1.00 1.50 2.00 2.50 3.00


wat

er c

onte

nt (%

)

Log MS Log NNS

0.12

0.15

0.18

0.21

0.24

0.27

1.0 1.5 2.0 2.5 3.0 3.5

Log Matric suction (kPa)

Wat

er c

onte

nt (%

)IM (Intact)PCM (Pre-crack)

0.15

0.17

0.19

0.21

0.23

0.25

0.27

0.29

1.0 1.5 2.0 2.5 3.0 3.5

Log Net Normal stress (kPa)

Wat

er c

onte

nt (%

)


7


CONCLUSSION Based on the data obtained and the analysis carried on the following conclussion might be drawn from this experimental study. 1. The specimens tested demonstrated the

typical of brittle hard clay and exhibit elastic only behaviour .

2. Shear strength of the intact specimens were higher along their axial strain than that the pre-crack specimens.

3. Pronounced and lower value of peak shear strength were occurred to the precrack specimen than that the intact specimen.

4. The compressive strength of the pre-crack specimens were lower than the intacts specimen indicated by their higher void ratio change index as well as their water change index than the intact specimen’s.

ACKNOWLEDGEMENT The author would like to give acknowledgment to all those involved in the project, especially to Professor Hamid Nikraz from Curtin University of Technology, Professor Kwang Wei Lo from National University of Singapore and Dr Min Min Zhao for their valuable advice. REFERENCES Alshibli, K. A. and Akbas, I. S. (2007). “Strain

Localization in Clay: Plane Strain Versus Triaxial Loading Conditions”. Geotechnical Geological Engineering. No. 25, p. 45-55.

Alshibli, K. A., Godbold, D. L., and Hoffman, K. (2004). “The Louisiana Plane Strain Apparatus for Soil Testing”. Geotechnical Testing Journal, ASTM, Vol. 27, no. 4, p. 337-346.

Alshibli, K. A. and Sture, S. (2000). “Shear Bands Formation in Plane Strain Experiments of Sand”. Journal of Geotechnical and Geoenvironmental Engineering, ASCE, No. 21167.

Atkinson, J. H. and Bransby, P. L. (1982). “The Mechanics of Soils: an Introduction

to Critical State Soil Mechanics”. 1st ed., McGraw-Hill, New York.

Anderson, T. L. (2004). “Fracture Mechanics: Fundamental and Aplications”. 3rd ed., CRC Press, London.

Bishop, A. W. and Henkel, D. J. (1962). “The Measurement of Soil Properties in The Triaxial Test”. Arnold, London.

Bizzarri, A., Allersma, H. G. B., and Koehorst, B. A. N. (1995). “Preliminary Tests on Soft Clay with a Biaxial Apparatus”. Proceedings of the 1995 International Symposium on Compression and Consolidation of Clayey Soils, Part 1 (of 2).

Drescher, A., Vardoulakis, I., and Han, C. (1990). “A Biaxial Apparatus for Testing Soils”. Geotechnical Testing Journal, GTJODJ, Vol. 13, p. 226-234.

Fauziah, M. and Nikraz, H. (2007). “Biaxial Testing of Overconsolidated Clay”. Proceeding of The 1st International Conference of European Asian Civil Engineering Forum, Pelita Harapan University, Jakarta.

Fauziah, M. and Nikraz, H. (2007). “Stress-strain Behaviour of Overconsolidated Clay Under Plane Strain Condition”. Proceeding of 10th Australia New Zealand Conference on Geomechanics, AGS, Brisbane, Australia, p. 148-153.

Fauziah, M. and Nikraz, H. (2008). “The Behaviour of Unsaturated Compacted Clay Under Plane Strain Condition”. Proceeding of the 3rd International Conference on Evaluation, Monitoring, Simulation, Management and Remediation of the Geological Environment and Landscape, WIT Press UK, The New Forest UK, p. 77-85.

Fauziah, M. and Nikraz, H. (2008). “Plane Strain Testing on Properties of Unsaturated Compacted Clay”. Proceeding of Geo-Chiang Mai 2008, An International Conference on Geotechnical Engineering, Chiang Mai, Thailand, p. 157-164.

Fredlund, D. G. and Rahardjo, H. (1993). “Soil Mechanics for Unsaturated Soil”. John Willey & Sons, Inc.

8


Fredlund, D. G. and Morgenstern, N. R. (1976). “Constitutive Relation for Volume Change in Unsaturated Soil”. Canadian Geotechnique Journal, Vol. 17, No. 3, p 261-276.

Fredlund, D. G. and Morgenstern, N. R. (1977). ”Stress State Variables for Unsaturated Soils”. Journal of Geotecnical Engineering, ASCE, Vol. 103, No. 5, p 447-466.

Green, G. E. and Reades, D. W. (1975). “Boundary Conditions, Anisotropy and Sample Shape Effects on the Stress-strain Behaviour of Sand in Triaxial Compression and Plain Strain”, Geotechnique, Vol. 25, No. 2, p. 333-356.

Han, C. and Drescher, A. (1993). “Shear Bands in Biaxial Tests on Dry Coarse Sand”. Soils and Foundations, Japanese Society of Soil Mechanics and Foundation Engineering, Vol. 33, No. 1, p. 118-132.

Han, C. and Vardoulakis, I. G. (1991). “Plane Strain Compression Experiments on Water-saturated Fine-grained Sand”. Geotechnique, Vol. 41, No. 1, p. 49-78.

Ingraffea, A. R. (1987). “Theory of Crack Initiation and Propagation in Rock”. In Fracture Mechanics of Rocks, ed by B.K. Atkinson, Academic Press, London, p. 71-110.

Lee, K. L. (1970). “Comparison of Plane Strain and Triaxial Tests of Sand”. Journal of the Soil Mechanics and Foundation Division, ASCE, Vol. 96, No. 3.

Lee, K. L. and Seed, H. B. (1964). “Discussion on Use of Free End in

Triaxial Testing on Clays”. ASCE, Vol. 91, No. 6, p. 173-177.

Lo, K. W., Mita, K. A., and Thangayah, T. (2000). “Plane Strain Testing of Overconsolidated Clay”. Research Report, Department of Civil Engineering, National University of Singapore.

Lo, K. W., Nikraz, R. H., Thangayah, T., and Zhao, M. M. (2005). “An Elastoplastic Shear Fracture Model for Soil and Soft Rock”. Proc of the 11th International Conference on Fracture at www.icf11.com/proceeding/TOPIC/topic.htm

Marach, N. D., Duncan, J. M., Chan, C. K., and Seed, H. B. (1984). “Plane Strain Testing of Sand”. Laboratory Shear Strength of Soil, ASTM STP 740, p. 294-302.

Mochizuki, A., Min, C., and Takahashi, S. A. (1993). “A Method for Plane Strain Testing of Sand”. Journal of Japanese Geotechnical Society, No. 475, p. 99-107.

Rowe, P. W. and Barden, L. (1964). “Importance of Free Ends in Triaxial Testing”. ASCE, Vol. 90, No.1, p. 1-27.

Taylor, D. W. (1941). “7th Progress Report on Shear Strength to US Engineers”. Massachusetts Institute of Technology.

Viggiani, G., Finno, R. J., and Harris, W. W. (1994). “Experimental Observations of Strain Localisation in Plane Strain Compression of a Stiff Clay”. In Localisation and Bifurcation Theory for Soils and Rocks, Chambon et.al., Eds., Balkema, Rotterdam, p. 189-198.

9


Compressive Strength Assessment of Concrete Structures from Small Core by Point Load Test

Achfas Zacoeb1 and Koji Ishibashi2

1Department of Civil Engineering, Brawijaya University, Indoneia, email: [email protected], 2Professor, Department of Civil Engineering, Saga University, Japan, email: [email protected]

Abstract: To assess a compressive strength from existing concrete structures by core drilling are usually gathered with a diameter specimen of 100 mm or three times of maximum coarse aggregate size and examined by uniaxial compressive strength (UCS). It is relatively difficult to gather a large sized core, and a pit place will be limited by main members. To get an alternative solution with smaller specimen, point load test (PLT) has been selected which is a simple test and widely accepted in rock materials research, but relatively new in concrete. The reliability of PLT is examined by extracting a lot of core drilled specimen from ready mixed concrete blocks with maximum coarse aggregate size, Gmax of 20 mm in representative of architectural structures and 40 mm in representative of civil structures on the range of concrete grade from 16 to 50. Compressive strengths were classified into general categories, conversion factors were determined, and scattering characteristics were also investigated. The relationship between point load index (IS and compressive strength of concrete core specimen (f’cc) can be written as linear approximation as f’cc = k.IS – C.

Keywords: Strength assessment, standard specimen, small core, point load index, and linear approximation.

INTRODUCTION One of the reliable tests for assessing insitu strength of concrete is coring. Coring may prove expensive and the holes have to be backfilled, but the resulting data are usually accepted as the best evidence of the condition of the concrete in place. It is established in JIS A1107 (1993) that a core drilled specimen diameter of 100 mm or three times of maximum coarse aggregate size from a concrete structure member should be taken for performing strength evaluation. Small cores are often used as substitutes for large cores to test concrete strength. They have the advantages of being easily drilled and cut, minimum damage to structures, and a lower capacity machine is needed (Ruijie, 1996). The main parameter for characterizing a concrete in engineering practice is compressive strength. Ibragimov (1989), the maximum aggregate size is considerable as played role for affecting the properties of

concrete. The standard laboratory test ussualy requires a standard specimens, so indirect test are needed. The PLT is intended as an index test for the strength classification of rock materials, but it may also be widely used to predict other material strength parameter. It is an attractive alternative method, because it can provide similar data at a lower cost, a simple preparation of specimen, and possibility on field application ISRM (1985), in order to estimate UCS indirectly, index-to-strength conversion factors are developed. Richardson (1989) conducted a point load tests of cast specimens with various diameters. Zacoeb et al. (2007) showed a strong correlation between point load index of core drilled specimen (I) and compressive strength of concrete core (f’). Ishibashi et al. (2008) investigated the influence of h/d ratio and maximum aggregate size (Gmax) on concrete core specimen by using PLT. Many research works had been conducted to acknowledge with regard to PLT and resulted in widely used

10

Zacoeb, Ishibashi: Compressive Strength Assessment of Concrete Structures From Small Core by Point Load Test

and other parameters. However, more experimental works helps to substantiate the existing correlation. FINITE ELEMENT ANALYSIS Broch, et al. (1972) started with a simple formula taking an idealized failure plane of diame-tric core sample as shown in Figure 1. From this figure can be taken into account as conceptual model for derivation on point load index equation as:

where, IS : point load index (MPa) P : load (N) D : diameter of specimen (mm)

By taking the circular area of the core into account, an argument can be made that Equation 1 should be written as:

4

The user of this test soon noticed, that the results of a diametric test were about 30% higher than those for an axial test using the same specimen dimensions. Broch et al. (1972) and ISRM (1985) suggested acknowledge this difference by applying a size correction and introducing the equivalent core diameter of De. Hence, the Equation 2 can be rewritten as:

The variations of IS with specimen size and shape lead to introduce a reference index IS(50)

which corresponds to the IS of a diametrically loaded rock core of 50 mm diameter (Broch et al.1972). Accordingly, initial IS values are re-duced to IS(50) by size correction factors deter-mined from empirical curves as a function of d. It is indicated that the considerably larger shape effect should be avoided by testing spe-cimens with specified geometries. ISRM (1985) proposed a new correction function which ac-counts for both size and shape effects by utiliz-ing the concept of equivalent core diameter (De). This function, known as geometric correc-tion factor F is given by:

where F : the geometric correction factor

50

.

The unique point load index can be obtained by applying a size correction for the specimen as point load index of IS varies with core specimen diameter of De. The size-corrected point load index of IS(50) for each specimen is defined as the value of IS that would have been measured on a standard specimen diameter of De=50 mm. In the case of specimen diameter of DeLABORATORY WORKS other than 50 mm, size correction must be calculated by using of Equation 5. LABORATORY WORKS The concrete block for core specimen extracting were sized of 300 mm x 300 mm x 600 mm made from ready-mixed normal concrete with typical slump range value from 8 to 12 cm for most application as workability control and divided into two groups as shown in Table 1. For curing, all concrete block specimens were covered with plastic sheets and the humidity was set for about a week. Commonly in Japan, for architectural

Figure 1. Specimen diametric of PLT on cores

(2)

(1)

(4)

(5)

(3)

Idealized failure plane

Loading direction

dCentral axial direction

h/2 h/2

11


structures such as building construction is using the maximum coarse aggregate size, Gmax of 20 mm. While for civil structures such as pier, abutment, bridge deck and check dam is using the maximum coarse aggregate size, Gmax of 40 mm.

Table 1. Group of concrete block

Group Gmax (mm)

Grade Cement Type

I 20

16

OPC (Ordinary Port- land Cement)

21 24 36 50

II 40

16

PBSC 21 24 30

Core specimen diameter of 125, 100, 50 and 35 mm were extracted from the above men-tioned concrete block with the electric core pull-ing out machine. The wet type that used by flowing some water during the core drilled process is applied, and the extraction speed was assumed to be about 4 cm/min. The direc-tion of extraction is considered as the direction of concrete placing as vertical direction in assumption of practical work in construction. The situation of core specimen extraction is shown in Figure 2.

The core specimen with h/d ratio of 1.5 and 2.0 were selected as core specimen of PLT in this study. In order to establish a specific h/d

ratio, core specimens were cut both ends with a concrete cutting machine to become a fixed height (h). The total number collection of each specimen is shown in Table 2.

Table 2. Total number of core specimens

Group Grade d (mm) h/d Total

I

16 35

1.5 135 2.0 135

50 1.5 105 2.0 99

21 35

1.5 90 2.0 90

50 1.5 60 2.0 60

24 35

1.5 66 2.0 66

50 1.5 59 2.0 58

36 35

1.5 123 2.0 126

50 1.5 111 2.0 108

II

16 35

1.5 126 2.0 126

50 1.5 85 2.0 87

21 35

1.5 154 2.0 138

50 1.5 82 2.0 79

24 35

1.5 113 2.0 113

50 1.5 87 2.0 86

30 35

1.5 157 2.0 172

50 1.5 113 2.0 108

The specimen in PLT is taken and loaded between two hardened steel cones. The system consists of a small hydraulic pump, a hydraulic jack, a pressure gauge and interchangeable

Concrete placing direction

Figure 2. Core specimen extractions

12


testing frame of very high transverse stiffness. Spherically truncated, conical platens of the standard geometry shown in Figure 3 are to be used with the cylinder area of 14.52 cm2. The platens should be of hard material such as tungsten carbide or hardened steel so that they remain undamaged during testing (ISRM, 1985).

The core specimen in this study is gradually loaded by activating the hand pump until failure and determined this load as P. The point load index of IS was calculated by using Equation (3) and for core specimen diameter of 35 mm was corrected to the standard core diameter as point load index of IS(50) for core

specimen diameter of 50 mm by using Equation (4). The examination is conducted by using PLT machine with oil pressure cylinder type and maximum load capacity of 98 kN as shown in Figure 4. RESULTS AND DISCUSSION Compressive Strength

From a concrete block is extracted a core specimen with diameter of 100 mm as minimum requirement and 125 mm as three times of maximum coarse aggregate size (JIS A1107, 1993), cut both ends of the core with a concrete cutting machine, end face polished, processed it to become specific h/d of 2.0 and examined by UCS test. The mean value of compressive strength of concrete core, f’cc is shown in Table 3, and assumed these values as reference on this study.

Table 3. Compressive strength

Group Grade Age

(days) f’cc

(MPa)

I

16 161 15.6 21 337 35.4 24 73 31.5 36 177 42.9 50 78 51.5

II

16 188 21.6 21 173 22.4 24 532 34.4 30 118 32.2

Point Load Index

The mean values of PLT were computed for both diameter sizes as shown in Table 4. Scattering characteristics were also investigated by mentioning of CoV (coefficient of variation). The CoV is the degree to which a set of data points varies. When assessing precision, the lower of CoV percentage, the better of precision between replicates. For Group I, the level of CoV is almost same or less than that on actuality of ready-mixed concrete product (Saga, 2008) from 10 to 15%. It can be stated that the test results are

Figure 3. Point load cone platen

Figure 4. Setup of PLT

13


satisfy enough. For group II, the CoV is bigger than the requirements (except for concrete grade of 30).

Table 4. Point load index

Group Grade d

(mm)h/d

IS (MPa)

CoV (%)

I

16 35

1.5 2.27 11.4 2.0 2.32 13.3

50 1.5 1.87 10.7 2.0 1.93 10.8

21 35

1.5 3.24 11.4 2.0 3.31 12.4

50 1.5 2.57 9.7 2.0 2.61 11.1

24 35

1.5 3.21 12.4 2.0 3.28 13.1

50 1.5 2.71 8.1 2.0 2.77 9.3

36 35

1.5 3.61 11.9 2.0 3.69 13.5

50 1.5 3.06 10.4 2.0 3.02 10.9

50 35

1.5 3.95 8.1 2.0 4.05 9.6

50 1.5 3.27 8.2 2.0 3.34 8.9

II

16 35

1.5 2.20 26.1 2.0 2.30 31.9

50 1.5 1.85 18.7 2.0 1.95 18.1

21 35

1.5 2.52 23.0 2.0 2.55 24.3

50 1.5 2.12 21.0 2.0 2.00 18.0

24 35

1.5 2.90 26.3 2.0 2.92 27.0

50 1.5 2.31 18.2 2.0 2.43 18.2

30 35

1.5 2.80 24.7 2.0 2.85 19.7

50 1.5 2.43 8.5 2.0 2.47 7.6

For Group I, the level of CoV for h/d of 1.5 is smaller than h/d of 2.0. It can be stated that h/d ratio of 1.5 is better than h/d of 2.0 for making a PLT specimens from core drilled extraction. While for core specimen diameter, d is better using 50 mm than 35 mm, because the level of CoV is also smaller. Beside this reason, it is also fulfilled with the standard core diameter requirements of 50 mm. For all groups, it is possible and acceptable for using a core diameter of 50 mm and h/d ratio of 2.0 as PLT specimen with results in the range of CoV from 8 to 18%. Application of PLT for small diameter of core specimen is not suggested for d/Gmax ratio below 1.25, considering the CoV results for Gmax of 40 mm and d of 35 mm are larger than 20%. Correlation Between Point Load Index and Compressive Strength

Point load index of core specimen diameter of 50 mm, IS(50) is determined as standard value. Hence, the value of different core specimen diameter, IS(35) should be corrected in order to show a relationship with IS(50) by using Equation (4) and (5). By correcting the point load index of IS(35) and assuming as standard core specimen diameter of 50 mm, will add the number of data for analysis of point load index IS(50). The new result for this combination is shown in Table 5 corresponding with the compressive strength of concrete core (f’cc) for each grade.

Table 5. Point load index and compressive strength

Group Gradef’cc

(MPa)

IS of h/d

1.5 2.0

I

16 15.6 1.86 1.93 21 35.4 2.57 2.61 24 31.5 2.71 2.77 36 42.9 3.07 3.03 50 51.5 3.27 3.34

II

16 21.6 1.85 1.95 21 22.4 2.12 2.00 24 34.4 2.31 2.43 30 32.2 2.43 2.47

14


Figure 5. Second order of polynomial regression

Figure 6. Linear regression

Figure 7. Linear approximation for IS(50) to f’cc

15


The correlation between point load index, IS(50) and compressive strength of concrete core, f’cc for both of groups is shown graphically in Figure 5 and 6. It is clearly evident to show the correlation by proposing a second order of polynomial and linear regression, respectively. It is proven by showing the square value of correlation coefficient which judges the effectiveness of a second order of polynomial approximation curve for h/d = 1.5 is thought to be similar for h/d = 2.0. Linier regression also showed the same trend of effectiveness except for group II. JIS A5308 (2003) gives the compressive strength range of ready-mixed concrete in field application from 18 to 45 MPa. Hence, the application of PLT for estimating insitu strength of concrete structure should be confirmed in this range. So, the correlation was limited to this range for core specimen diameter of 35 and 50 mm as shown in Figure 7. When using the linear regression as shown in Figure 6, the approximation line does not intercept in the origin point. However, Figure 7 shown that the fitted curve will pass through the origin which aims to establish the relation of the whole area would be overestimated. It is preferable to using linear approximation than other modes in order to minimize the standard for the assessment of risk. The New Geometric Correction Factor

By considering the Equation (4) and (5) were proposed for rock specimen, so it is not suitable for concrete regarding the issue of homogeneity. The previous section already mentioned that maximum coarse aggregate size in concrete will affect the results of point load index. A new correction factor of F is proposed by following the format of previous Equation as:

50

The value of X can be generated by using data from group I for core specimen diameter of 35mm. The selection of this data was considered more reliable by showing a lower

CV. The solution is simple, because the nature of linear approximation as the origin. The exponent value of X is calculated as 0.53 with coefficient of correlation is 0.982. Finally, the expression geometric correction factor for concrete core specimen is given by:

50

,

Table 6 shows the absolute relative error between experimental and estimation values for point load index of IS(35) to become standard point load index of IS(50) by using Equation (8). The results are satisfied enough by showing a value of absolute relative error less than 5% in the case of d of 35 mm and Gmax of 20 mm. Table 6. Experimental and estimation of IS(50)

f’cc (MPa)

h/dPoint load index (MPa) Error

(%) IS(35) IS(35)a IS(50)

b

15.6 1.5 2.27 1.88 1.87 0.53

2.0 2.32 1.92 1.93 0.52

31.5 1.5 3.24 2.68 2.57 4.28

2.0 3.31 2.74 2.61 4.98

35.4 1.5 3.21 2.66 2.71 1.85

2.0 3.28 2.72 2.77 1.81

42.9 1.5 3.61 2.99 3.06 2.29

2.0 3.69 3.06 3.02 1.32

51.5 1.5 3.95 3.27 3.27 0.00

2.0 4.05 3.35 3.34 0.30

*) a = estimation b = experimental Recalculation procedure is conducted by using a new Equation (8) for correcting point load index of core specimen diameter of 35 mm and performing linear regression analysis to propose a formula of compressive strength esti-mation for equivalent core diameter of 50 mm as f’cc = k.IS - C as shown in Table 7. The coefficient of correlation, R2 also shows an improvement in strong relationship between IS(50) and f’cc.

(6)

(7)

16


In consideration with the standard compressive strength that used as linear approximation for index-to-strength conversion factor k. The k value is calculated by divided the compressive strength (f’cc) with point load index (IS). While C is constant depend on the linear regression equation.

Table 7. Formula of estimation

Group h/d Formula of Estimation R2

I 1.5 f’cc = 24.4IS – 30.3 0.953

2.0 f’cc = 24.9IS – 32.7 0.953

II 1.5 f’cc = 20.8IS – 16.7 0.928

2.0 f’cc = 22.3IS – 22.0 0.979

CONCLUSION Based on this study, it can be concludedthat an approximation curve showed a strong corre-lation between point load index (IS) and com-pressive strength (f’cc) for core diameter of 35 and 50 mm with height and diameter ratio of h/d=1.5 and 2.0. In addition for reference index IS(50), it can really deal with a linear approximation. Considering the issue of homogeneity that concrete is composite material, a new correction factor is proposed for core specimen diameters differ from 50 mm as:

50

,

To estimate a concrete compressive strength can be conducted with proposed equation as f’cc = k.IS - C. There is a prospect that PLT can be applied as indirect method to estimate a compressive strength on concrete structure. Application of PLT for insitu concrete compressive strength estimation should be confirmed in the range of compressive strength of ready-mixed concrete product from 18 to 45 MPa. Considering the maximum coarse aggregate size of Gmax in concrete, new criterion is proposed by determining the minimum value of d/Gmax ratio should not less than 1.25.

ACKNOWLEDGEMENTS The research work for this study was conducted by the first author during Doctor Course in the Laboratory of Structural Engineering and Mechanics, Department of Civil Engineering, Saga University, Japan under supervision of the second author. Thanks to the Government of Japan for financial support trough the Monbukagakusho scholarship. REFERENCES Broch, E. and Franklin, J. A. (1972). “The

Point Load Strength Test”. International Journal of Rock Mechanics and Mineral Sciences, Vol. 9(6), p. 669-676.

ISRM Commission on Testing Methods. (1985). “Suggested Method for Determining Point Load Strength”. International Journal of Rock Mechanics, Mineral Sciences and Geomechanics, Abstract Vol. 22, p. 51-60.

Ibragimov, A. M. (1989). “Effect of the Maximum Size of Coarse Aggregate on the Main Parameters of Concrete”. Journal of Power Technology and Engineering, Vol. 23, p. 141-144.

Ishibashi, K., Zacoeb, A., and Ito, Y. (2008). “Influence of Coarse Aggregate Size on The Estimation of Compressive Strength of Concrete by Point Load Testing”. Journal of Structures and Materials in Civil Engineering, Japan, Vol. 24, p. 108-115.

Japanese Industrial Standard. (1993). “JIS A1107: Method of Sampling and Testing for Compressive Strength of Drilled Cores of Concrete”. Referenced JIS Standard, Japan.

Japanese Industrial Standard. (2003). “JIS A5308: Ready-mixed Concrete”. Referenced JIS Standard, Japan.

Richardson, D., N. (1989). “Point Load Test for Estimating Concrete Compressive Strength”. ACI Materials Journal, Vol. 86(4), p. 409-416.

Ruijie, K. L. (1996). “The Diameter-compression Test for Small Diameter

(8)

17


Cores”. Journal of Materials and Structures, Vol. 29(1), p. 56-59.

Saga Fresh Concrete Industrial Union. (2008). “Ready-mixed Concrete Products in Saga Prefectures, Fiscal Year of 2007”. Quality Audit Report, Japan.

Zacoeb, A., Ishibashi, K., and Ito, Y. (2007). “Estimating the Compressive Strength of Drilled Concrete Cores by Point Load Testing”. Proceeding of the 29th JCI Annual Meeting, Sendai, Japan, July 11-13, 2007, p. 525-530.

18

Essential Strategies To Promote Earthquake Safer Non-Engineered Housing: An Indonesian Perspective

Essential Strategies to Promote Earthquake Safer Non-Engineered Housing: An Indonesian Perspective

Setyo Winarno

Senior lecturer at Islamic University of Indonesia, Yogyakarta, Indonesia, email: [email protected]

Abstract: Current evidence suggests that the most effective measure to protect people’s lives against major earthquakes is to build safer houses, particularly on non-engineered residential houses. Promoting earthquake safer housing requires a thorough understanding and synergetic efforts on both controlling by seismic codes and regulation and addressing the local practices, opportunities, limitations, and problems. This paper explores some appropriate strategies to promote earthquake safer housing by improving non-engineered construction practices that ensures safety against disasters. Information and data were gathered in Indonesia through a direct observation to the stricken areas following the 2006‘s Yogyakarta earthquake, a series of in-depth interviews with multidisciplinary stakeholders, and also two workshop events. This study found that promoting safer non-engineered housing requires a comprehensive strategy that includes technical and non-technical measures. In the top priority, the role of government is extremely crucial in tackling the compelling problem by institutionalisation of seismic risk reduction activities at national and local levels. For grass root communities, wide ranging reform in non-engineered construction practice should be delivered by a locally adapted technique, culturally accepted and compatible, local resource-based, not burdensome, and less bureaucratic. This fits into the existing community structures without any friction, value everyone’s unique contribution, and also break powerful psychological barriers as well as emphasize the importance of sustainability. If people perceive that the implementation of the codes is practically possible and achievable and they can control it, this strategy will have a tremendous effect on how well they can cope with change. In all, the most challenging part of successful promotion of earthquake safer housings is not finding the tools, but realizing and accepting that a seismic event is simply a real fact for all people who live in a seismic prone area. Therefore it is better to grasp it rather than denying or ignoring it. Keywords: Earthquake, essential strategies, and non-engineered housings.

INTRODUCTIONS Recently, strong earthquakes have occurred throughout the world. Protracted study of the true nature of the seismic risk has shown that most of the loss of life has occurred due to the collapse of non-seismic resistant buildings made of heavy materials, classified as 'non-engineered' - in simple terms: unsystematically designed and poorly built structures, in both developing and developed countries (Mansouri et al., 2002; Corpuz, 1990; Lee et al., 2003; Sarwidi, 2001). An example can be seen when a severe ground shaking hit Yogyakarta and Central Java, Indonesia in May 2006, and left over 150,000 private houses, mostly low-rise

non-engineered houses, totally destroyed and claimed to around 6.000 fatalities (BAPPENAS, 2006). The strong correlation between the large number of deaths and the collapse of non-engineered buildings suggests it is imperative to improve seismic resistance for both existing and new non-engineered housing. While seismic-related building codes for simple housing have been available for use in Indonesia for many years (Boen, 1978), as in other countries, current evidence has shown that earthquakes still continue to cause tragic events with high death tolls. Obviously, there is a broad gap between the existence of appropriate building codes and their effective

19


widespread application (Comartin et al., 2004). It is clear that unless something is done quickly to significantly implement seismic codes in Indonesia, earthquakes will continue to tragically inflict greater human and economic losses. Based on the fact that non-engineered buildings as the majority of building stocks (Winarno and Sarwidi, 2010), the existence of seismic codes, the reluctance of grass root people to implement seismic codes, and the growing concern for the seismic safety, this paper reports on research that has explored fundamental strategies to promote the implementation of seismic codes on non-engineered construction in Indonesia. This paper elaborates a subject that has not yet been sufficiently researched in Indonesia. By better understanding the varied facets of the particular construction type, the problems in the non-engineered construction practice for the implementation of seismic codes can be better resolved. Hence, non-engineered buildings without seismic codes will eventually be replaced by more reliable constructions built with seismic resistant attribute. The findings presented herein should not be seen as conclusive, but rather for the development of practical initiatives to encourage ways in which the various disciplines can become more involved. In this paper, the criterias of earthquake safer non-engineered houses focus on self-construction of non-engineered houses under support,

supervision, and training from engineer or expert and also the use of local labour and material. The overarching aim of the research study was to examine the essential strategies to promote earthquake safer non-engineered housing within the context of current practice in Indonesia. In sequence with the aim of the research, the objectives of the research were (1) to examine best practice of a range of stakeholders regarding sustainable and effective dissemination of implementation of seismic codes at grass root level, and (2) to study the potential role from various stakeholders as agents of change. Through understanding these important strategies, all stakeholders can take a part to prevent a widespread persistence of grass root communities not to implement seismic codes. Indeed, the implication of a general assessment of such grass root communities is that the manual of seismic features is understood in different ways and not through standard curricula. Therefore, all perspectives need to be involved deliberately to resolve this problem. METHODOLOGY Based on wider literature, involvement of multidisciplinary stakeholders in the seismic risk reduction of non-engineered buildings is imperative to ensure sustainability. Those

Figure 1. NVivo screen display of the nodes created from the interview and workshop events

20


stakeholders includes nine types of multidisciplinary stakeholders, i.e. researchers/scientists, small and medium contractors, foremen, government officials, businessmen, educators, Non Government Organizations (NGOs), community leaders, and reporters (IUDMP, 2001; CEEDEDS, 2004; SCEC, 2002; GREAT, 2001; Dixit, 2003). Thus, the above stakeholders were selected as respondents in the research. The appropriateness of each respondent was determined by their role, responsibilities and normal activities within their own organization and the level of experience in the specific subject. Information and data were collected through a direct observation to the stricken areas following the 2006‘s Yogyakarta earthquake, a series of in-depth interviews with multidisciplinary stakeholders, and also two workshop events. Firstly, field investigations in regions damaged by the Yogyakarta earthquake were undertaken during the reconstruction stage to obtain a more enhanced understanding of the problems involved. Through this investigation, nine respondents, representing a mix of nine types of stakeholders, were chosen as interviewees. The interview method was appropriately chosen in this study to collect factual in-depth information, opinions, and the story behind the respondents’ experiences. Then, the interview findings were augmented via two workshop events conducted in Yogyakarta City (with 13 participants) and Bengkulu City (with 12 participants), which were attended by key people and at the leading edge of decision making within each city, in order to ensure that individual's assessment was factual and logically sound. The workshop events took place under the name ‘Workshop on Seismic Risk Management of Non Engineered Buildings’ in collaboration with Master Program in Civil Engineering, Universitas Islam Indonesia, Yogyakarta and Civil Engineering Department, Universitas Bengkulu. The overall data collected was analyzed with the assistance of NVivo software programme. The software helped to code the data and identify themes and/or patterns generated.

FINDINGS During the interview and workshop events, three guiding principles for a sustainable and successful dissemination process emerged, which may have been greatly neglected in the past. The above three principles confirm that: (1) the government role is the backbone of the dissemination initiative – mentioned by 8 interviewees and 22 workshop participants, (2) the dissemination channel through the existing social bond is imperative – mentioned by 8 interviewees and 16 workshop participants, and (3) the message should convince people that they have control over the implementation of seismic codes - mentioned by 6 interviewees and 17 workshop participants. If the three principles are applied rigorously, the dissemination process will obtain great achievement. This is truly beyond technical capacity. However, sometimes the key government officials and the disseminators, who are usually from a wealthy background, do not persevere in conducting dissemination to the low-medium income population, as they are not on the ‘poor’ people’s side and also do not gain any clear financial and political benefit. Figure 1 depicts the research findings with the assistance of NVivo software. Overall, the findings from the research data address each of the objectives stated previously, as follows: The Current State of Non-Engineered Construction Practice and Guiding Principles of Sustainable and Effective Dissemination in the Grass Root Initiatives

Field investigation and interviews with local practitioners undertaken in the stricken area after the Yogyakarta earthquake revealed some inadequate construction materials. The quality of materials, in fact, rested largely on the knowledge of the local builders, foremen, masons, and carpenters. They continued to practice the traditional approach and were often reluctant to accept new techniques. They felt that what they did was in line with their belief and there was no strict regulation which confronted their practices. The resulting

21


mortar and concrete were generally of poor quality. All too often, there was incorrect beam-column connection detail. Interviewees from government staff, researchers/scientists, contractors, and foremen revealed that information on building using seismic codes with seismic codes has been available for public. In reality, the seismic codes were not implemented widely, particularly within non-engineered buildings. From the point of view of contractors and foremen, it was true that there were regulations with limited enforcement and no accountability. The government has not been able to implement even the existing seismic codes because of a lack of suitable implementation mechanism and limited resources for building inspection and control. Also, the respondents who were represented by the community leaders, NGOs, teachers, and businessmen said that the building code was not disseminated down through to the grass root communities. To overcome the limited resources on government, obviously, this requires a comprehensive strategy that includes various multidiscipline interventions. Combining data from interviews and workshop event, three guiding principles of sustainable and effective dissemination in the grass root initiatives emerged to break a widespread reluctance of grass root communities to implement seismic codes, as follows. Government should enact as a proactive backbone of the dissemination initiative

To disseminate the benefits of seismic codes in real construction successfully and continuously, almost all interviewees and workshop participants mentioned the importance of the government’s role, and even mentioned it as the most important factor. Although the government is not necessarily solely resposible for disseminating information regarding seismic codes, and there is a new opportunity for multidisciplinary involvement in this matter, nevertheless, the government itself should act as a proactive backbone of the dissemination initiative. The ability of government to listen, respect, encourage, and

motivate grass root programs initiated by community groups is indispensable. At the same time, the government should generate campaigns or dissemination programs down to the community members to motivate people to gradually prepare for the next disaster, highlighting a shared risk and also enforcing seismic codes. The government’s role is to be the focal point for law enforcement (regulation), communication, coordination, programme monitoring, delivery of knowledge, and accountability. Also, government should institutionalise seismic risk reduction activities at national and local level. It is believed that, if the government makes a huge effort to rigorously enforce seismic codes today in many model houses so people see what they look like and feel very proud, many more model houses would soon be replicated everywhere. When a strong earthquake strikes, the government will make only a little effort in response because the blossom of the implementation of seismic codes thus the collapse of buildings/houses would be rare, even avoided. Furthermore, it would ensure faster recovery and reconstruction. The dissemination method should utilize existing social bonds and/or indigenous methods

Three out of nine interviews and ten workshop participants stated the importance of social bonds and indigenous methods when dealing with communal grass root dissemination. This finding argues that when government staff, researchers, scientists, and/or other disaster management experts gather to elucidate the lay people, this process will be effective if it uses or merges into the formal or informal existing tradition of community meetings. Innovative initiatives, new synergies, and networks are easily absorbed over those already established. This will fit into the existing community structures without any friction and also value everybody’s unique contribution. People will be comfortable, happier, and less worried about being involved in the dissemination process if they are among people they have worked with in the past and with whom they have developed a long-term

22


relationship and have similar beliefs. This approach also emphasizes the importance of sustainability. This process breaks powerful psychological barriers and continues to build up trust amongst them and also acts as encouragement for others who have not been involved in any initiatives in the past. It is a better and less expensive method for including communal actions. In addition, the existing social bond has shaped the community’s capacity in their daily lives. Through this bond, community members can better adjust their own practices to improve performance. If this dissemination mechanism is well done and streamlined over time before disaster strikes, obviously this channel will be sustainable and will also work well in the critical stages of disaster response. Perhaps, the current dissemination method initiated by government officials uses government standards which are often not flexible enough to adjust to various people’s needs. The government tends to be just interested in promoting their own status, and the use of existing social bonds is often greatly neglected. The above discussion is about channeling dissemination by giving a talk or elucidation to grass root communities by ‘visiting’ the community group. On the other hand, the dissemination technique by ‘inviting’ individuals to a meeting, discussion, or seminar is very suitable for people from medium-top management in organizations who have experienced better education because they have an appropriate level of expertise and knowledge. This is why the dissemination to the grass root community by ’inviting’ them to a formal forum, such as a discussion in government offices, always fails to motivate them into concrete action. This is likely to be because the prescriptive mechanisms are not compatible with their beliefs. For grass root communities, wide ranging reform in non-engineered construction practice should be by a locally-adapted technique, culturally accepted and compatible, local resource-based, not burdensome, and less bureaucratic. This principle is closely related to the idea of a sense of place, meaning that

feeling is attached to a place, and feeling secure in all the things that make life truly meaningful is important for people’s identity as mentioned by Covenry and Dutson (2006). The message should be achievable under control of the people

After the correct dissemination channel is diligently decided, here, two interviews and three workshop participants elaborated that the message about the implementation of seismic codes should be achievable under the control of the people. This finding underpins what Lustig (1997) has found: that for a disaster-management system to be sustainable, it should be designed not only to convey the message to the members of the disaster-prone community that they are in control, but also that the system is actually under their control. Although the seismic event itself, when it occurs and how big the magnitude will be, is uncontrollable, most of the components of seismic risk people face are absolutely controllable. Community members who live in a seismic hazard prone area can control the suffering from seismic tremor, for example, through implementing such codes in their houses/buildings and making any other appropriate preparedness. If people rationalize that the implementation of seismic codes in their own houses is the most effective strategy for minimizing losses, and at the same time, they perceive it is practical, possible and achievable and they can control it, this strategy will have a tremendous effect on how well they can cope with it. It is an unwise solution to force people to leave their beloved homeland because of the high level of uncertain but inevitable threats of an earthquake event, as an inconvenient truth. Thus, the dissemination message should convince people to devote themselves to living harmoniously with seismic risk through matters they perceive they can control. The ideal solution is to give people a better understanding about seismic hazard and risk in a reasonable and rational manner and then convinces them they can cope and control it with the proper implementation of seismic codes. As a result, people become happier and

23


don’t worry about living in seismic prone areas because they have some control over events. Better understanding of seismic hazard and risk will encourage people to implement seismic codes voluntarily without coercion. Although disseminating new ideas to people is not welcomed at first, even fought against, efforts must be made to present persuasive arguments of the soundness of the protective and cost-effective measures. It is important to educate people that the implementation of seismic codes is simple,

economically feasible, achievable, and culturally acceptable to obtain their sense of control over their destinies. If people have a sense of control and are clever enough to implement seismic codes properly, then they can also achieve a sense of ownership. This sense of ownership is really needed, not only to encourage and maintain actions voluntarily in order to generate a culture of prevention, but also to make community members feel part of the effort. As a result, people will actively make contributions to reduce vulnerability.

Figure 2. Putting multidisciplinary stakeholders together as agents of change to share the seismic risk of non-engineered buildings

24


Through a greater sense of ownership, more people tend to recognize that reducing seismic risk needs shared responsibility and shared effort. In conclusion, the above finding suggests three guiding principles for a sustainable and successful dissemination process, which may have been greatly neglected in the past. If the three principles are applied rigorously, the dissemination process will obtain great achievement. This is truly beyond technical capacity. A Map of Multidisciplinary Stakeholders as Agents of Change

This section maps out multidisciplinary stakeholders as agents of change, highlighting a shared risk and responsibility and focusing on grass root initiatives with local wisdom, as later elaborated in Figure 2. Based on the data collected and the general facts, many forms of people and organizations can act as agents of change and have unique roles, which are highly varied due to the their nature of duty of care. Therefore, it is important to depict the influence level of many stakeholders as agents of change in seismic risk reduction of non-engineered buildings. The principal aim of this depiction is to give a better understanding that many people or organizations have an important stake in steering the seismic risk management process. It is hoped that a full spectrum of tasks and activities and broad banded partnerships in relation to the implementation of seismic codes to non-engineered buildings are well done and assimilated continuously, integratively, and harmoniously by a new configuration of various people and organizations to ensure a new generation of sustainable seismic risk reduction. Most arguments in the interview and workshop events state directly that the government’s role greatly influences the successful seismic risk reduction of non-engineered buildings under their principal aim of public safety in general. In particular, the top management in local government, together with the legislative council, has the authority and the power to

give approval or disapproval for all development related issues. The top local authority can communicate with different department heads and can make them act on issues of seismic risk which need to be accomplished. Essentially, local government should rigorously disseminate and communicate to their local seismic risks and the benefit of seismic codes by educating people to implement them voluntarily; at the same time, they should enforce the codes through effective regulation. In addition, this includes rules for the control of development, land use regulations, and suitable compliance mechanisms for building construction. The quality of their leadership is an indispensable component of success. They serve as a source of hope. Here, the role of the government appears to be as major contributors to the successful seismic risk reduction of non-engineered buildings. Finally, the role of the government (together with the role of the legislative council) and their concrete actions in disaster reduction remain the biggest and top challenge for effective seismic risk reduction today. On the other side, sometimes, difficulties appear when local builders have to change their traditional approach to fall in line with a new technique of seismic feature implementation. Therefore, changing the practice of builders is very necessary to make more seismic resistant buildings/houses and, as a result, they constitute the second agent of change. Researchers and scientists can be both the first and second agents of change. For example, the researcher/scientist role could be as the most important of the availability of earthquake data and development of seismic codes, and they are the first agent of change in this sense. The researcher who conducts the cost-effectiveness of seismic feature implementation can perform at the second level after the government role to support law enforcement of seismic codes. Researchers and scientists can inspire many aspects of good seismic risk management of non-engineered buildings, such as the development of seismic codes suited to the specific area, introducing earthquake facts for

25


land use planning and a seismic awareness program. Moreover, social and psychology researchers can innovate on how to deliver information, leading to a change in attitude and a change in behaviour. They also can serve as a source of new knowledge and inspiration for those implementing the changes. When discussing the effectiveness of social bonds and how to break psychological barriers of community perception to achieve a feeling of control (as mentioned in the previous findings), it is clear that the role of community leader is very important to explain the whole thinking process. In certain cases, the construction process of residential houses/buildings is directly under the guidance of the owner, who often does not have sufficient skill in the concept of seismic codes or neglects the workmanship of the builders due to the time completion pressure and the overwhelming need for steel bar reinforcement. Again, in this situation, the community leader can contribute to the education of those people and also act as an intermediary between the government and the house owner when government staffs enforce seismic codes during the construction process of the residential house. In certain cases, the house owner often acts as a self-builder due to limited funding available. Here the community leaders (who also represent the house owner) are the third agent of change. The last agent of change belongs to the groups of educators, NGOs, businessmen, reporters, and others since many respondents’ comment appeared to illustrate aspects of non technical intervention, where the involvement of many parties related to non-technical intervention (excluding government officials) was reasonably less important. Obviously, the picture of the agents of change in Figure 1 is merely a general assessment as an indicative explanation drawn from the research data. Probably, certain cases will illustrate different patterns of the degree of the influence within larger parties as condition changes over time. Subsequent investigation is truly needed to explore further for better understanding and to validate against any

misunderstanding or bias due to strong personal opinions. CONCLUSIONS In order to reduce seismic risk of non-engineered buildings, the proper implementation of seismic codes is imperative, because flawed construction practice often lies at the heart of implementation of seismic codes. Thus, the implementation should be an integral component of the total construction process in the first place. Realistically, a combination of skilled non-engineered construction participants and the proactive involvement of government and non-government organizations as well as many other parties (technical and non-technical intervention) are indispensable in promoting safer non-engineered housing. It needs to be controlled by seismic codes and regulation. In the top priority, the role of government is extremely crucial in tackling the compelling problem by institutionalisation of seismic risk reduction activities at national and local level. There are three principles for a sustainable and effective way towards the dissemination of seismic codes for non-engineered construction practice for the grass roots level: (1) the government should act as a proactive backbone of the dissemination initiative – mentioned by 8 interviewees and 22 workshop participants, (2) dissemination mechanisms should use the existing social bond and/or indigenious method mentioned by 8 interviewees and 16 workshop participants, and (3) dissemination message should convince the communities that the implementation of seismic codes is easily achievable under their control - mentioned by 6 interviewees and 17 workshop participants. The successful promotion of safer non-engineered structures needs to sensitise policy makers (government) toward seismic risk, followed by the involvement of many technical and non-technical actors, such as researchers, scientists, contractors, foremen, masons, carpenters, businessmen, educators, NGOs, community leaders, reporters, and others. Wider recognition is needed that

26


building a culture of disaster prevention should become everybody’s duty of care on a daily basis to ensure sustainability. Underlying the research finding, the most challenging part of successful seismic risk management of non-engineered buildings is not finding the tools, but realizing and accepting that a seismic event is simply a real fact for all people who live in a seismic prone area. We cannot predict when an earthquake will occur or how big the magnitude will be. Therefore it is better to grasp it rather than denying or ignoring it. A balance should be struck between communicating local seismic risk and the importance of seismic codes to the people, and enforcing them to implement the codes; whilst some intense regulation is imperative, it should be non-prescriptive. ACKNOWLEDGEMENT This research would not have been possible without the support of the Universitas Islam Indonesia. We thank to Professor Dr. Sarwidi, and Professor Dr. Widodo for their tireless contributions to this study. REFERENCES BAPPENAS/National Development Planning

Agency. (2006). “Preliminary Damage and Loss Assessment: Yogyakarta and Central Java Natural Disaster”. The Consultative Group, Jakarta, Indonesia.

Boen, T. (1978). “Detailer’s Manual for Small Buildings in Earthquake Areas (Manual Bangunan Tahan Gempa (Rumah Tinggal))”. Yayasan Lembaga Penyelidikan Masalah Bangunan, Bandung, Indonesia.

CEEDEDS/The Center for Earthquake Engineering, Dynamic Effect, and Disaster Studies. (2004). “The Manual of Earthquake Resistant Building”. Project Report between CEEDEDS and Japan Government, Yogyakarta.

Comartin, C., Brzev, S., Naeim, F., Greene, M., Blondet, M., Cherry, S., D’Ayala, D., Farsi, M., Jain, S. K., Pantelic, J., Samant, L., and Sassu, M. (2004). “A Challenge to

Earthquake Engineering Professionals”. Earthquake Spectra, Earthquake Engineering Research Institute, November 2004, Vol. 20, No. 4.

Corpuz, A. (1990). “Some Implications of the July 16, 1990 Earthquake on Urban and Regional Planning in the Philippines”. School of Urban and Regional Planning, University of the Philippines, Philippines at http://www.phivolcs.dost.gov.ph/Earth-quake/1990LuzonEQ_Monograph/pp287/pp290_291.html; accessed on 14 October 2004.

Covenry, I. and Dutson, T. (2006). “Sense of Place in Northern England”. International Journal of Biodiversity Science and Management, Vol. 4, No. 2, p. 213-217.

Dixit, A. M. (2003). “The Community Based Program of NSET for Earthquake Disaster Mitigation”. The International Conference on Total Disaster Risk Management, Nepal.

GREAT/Gujarat Relief Engineering Advice Team. (2001). “Repair and Strengthening Guide for Earthquake Damaged Low-Rise Domestic Buildings in Gujarat, India”. GREAT Publication, India.

IUDMP/Indonesian Urban Disaster Mitigation Project. (2001). “Increasing the Safety of Indonesian Cities from Earthquake Disaster Threat”. Asian Disaster Preparedness Center.

Lee, G. and Loh, C. H. (2003). “Human and Institutional Perspective of the 921 Earthquake in Taiwan: Lessons Learned”. at http://mceer.buffalo.edu/publications/ reports/docs/00-0003; accessed on 29 October 2004.

Lustig, T. (1997). “Sustainable Management of Natural Disaster in Developing Countries (in Fundamental Risk Analysis and Risk Management edited by Vlosta Molak)”. Lewis Publisher, USA.

Mansouri, B., Aghda, F., and Safari. (2002). “Preliminary Earthquake Reconnaissance Report on the June 22, 2002 Changureh (Avaj), Iran Earthquake”. at http://www.iiees.ac.ir/English/Publication/eng_publication_journal_12.html; accessed on 8 November 2004.

27


Sarwidi. (2001). “Reconnaissance Team CEEDEDS”. The Center for Earthquake Engineering, Dynamic Effect, and Disaster Studies, Yogyakarta.

SCEC/Southern California Earthquake Center. (2002). “Earthquake as Extreme Events”. Extreme Events Workshop, California.

Winarno, S. and Sarwidi. (2009). “Earthquake Vulnerability of Redential Houses in Yogyakarta City”. Proceedings of 1st

International Conference on Rehabilitation and Maintenance in Civil Engineering (ICRMCE), Solo, Indonesia.

Zacoeb, A., Ishibashi, K., and Ito, Y. (2007). “Estimating the Compressive Strength of Drilled Concrete Cores by Point Load Testing”. Proceeding of the 29th JCI Annual Meeting, Sendai, Japan, July 11-13, 2007, p. 525-530.

28

Putuhena, et al.: The Implication of Peatland Built Environment in Urban Drainage System: Case Study... Sarawak.

The Implication of Peatland Built Environment in Urban Drainage System: Case Study of Sungai Merah Area, Sibu, Sarawak

Frederik Josep Putuhena1, Ting Sie Chun2, and Salim Said3

1Universiti Malaysia Sarawak, Malaysia, email: [email protected], 2Universiti Malaysia Sarawak, Malaysia, email: [email protected], 3Universiti Malaysia Sarawak, Malaysia, email: [email protected]

Abstract: The main purpose of this project is to conduct an analysis to the flooding of Sungai Merah residential area, Sibu, Sarawak. The total catchment area is approximately 13.74 ha with relatively flat and typically of a lowlying floodplain which covered with peat. Poor maintenance, severe conditions and undersized culvert on the current drainage system in Sungai Merah residential area causes flooding to the area in addition to the waterlogged soil. Postflood forensic analysis of Sungai Merah residential area presented an approach using InfoWorks Collection System (CS), coupled with its embedded Geographic Information System (GIS) applications, to identify hydrology of the drainage pattern. The digital map featuring Sungai Merah area was used to create GIS map using ArcMap from Arcgis 9. Hydrology and hydraulic data collected is used for model calibration and verification for 2.4 km-long of drainage network in Sungai Merah residential area during December 26, 2008 and February 26, 2009 flood events. InfoWorks CS is satisfactorily capable of providing a clear picture of flood event through model simulation. Besides, excess rainfall do not influence the surface-runoff of peatlands which been verified by InfoWorks CS. From the study, the root of flood in Sungai Merah area is triggered by the backwater from Seduan River and the current existing drainage is fail to manoeuvre the current situation. Furthermore, InfoWorks CS demonstrates its capability for flood inundation modelling on Sungai Merah area. Thus, it is recommended that further improvement on the drainage network and floodplain in Sungai Merah residential area to create a better understanding of the drainage’s flowpath. Keywords: Peatland, urban drainage, flood, and InfoWorks CS. INTRODUCTION Peat soils are identified as “soft soils” with more than 75% organic matter content (Tang 2009). Compared to mineral soils, peat has a much higher infiltration capacity, drainable pore space and hydraulic conductivity (up to 30 m per day) of the surface (approximately first 1 metre) of peat layer but have lower capillary rise, bulk density and plant-available water (Wӧsten et al., 2003). In Malaysia, 1.7 million hectares or 63% of peat swamp are in delta and coastal plains of Sarawak. In Sarawak, approximately 1.66 million hectares or 13% of the state’s total land area are covered with peatlands (Singh and Bujang, 2003). The aerial extent of peat swamp forest in Sarawak is shown in Figure 1. About 90% of these peat

areas are classified as deep peat with peat layer of more than 1 m in depth (Tie and Kueh, 1979).

Figure 1. Peat swamps in Sarawak (Staub et al., 2000)

29


The lowland peat swamps in Sarawak are purely rain-fed and waterlogged most times of the year. Drainage is the leading principle of the water management practice in order to curb flooding by evacuation of excess rainfall within a period of time which is mainly removed by surface runoff. Furthermore, variation in tidal level may have a direct and consequential bearing on the water level in drainage. Therefore, forensic investigation is the key to identify the causes of the floods in peatland.

BACKGROUND OF THE PROJECT AREA Sibu Town is the capital of Sibu District In

Sungai Merah, Sibu, Sarawak which is about 5 km from Sibu Town. Sungai Merah is located on Seduan catchement with an area of 116.6km2. Water level measuring gauges are also located at Sungai Merah, which record water level continuously in every interval of 30 minutes. The project area is fairly flat and relatively

Sibu Town

Seduan

Igan River

Outlet of study

Sungai Merah Residential AreaRajang River

Inlet of study area

Figure 2. Location of the study area.

30


lowying with ground levels ranging from below 2.0 to 3.0 MSL. Its locality is located in the central part of Seduan catchement. The peat’s properties covers Sungai Merah residential area with a shear stress of 4.87 kPa, compressibility index of 1.045 and bulk density of 0.08g/m (Ting 2009). The shocking floods in Sibu division, Sarawak with its confluence of three major rivers namely the Rajang River, the Seduan River and the Igan River. The research site (Figure 2) is located at Sungai Merah December 26, 2008 and February 26, 2009 had struck the middle and lower part of Seduan catchement due to extreme rainfall with addition to high tidal events. The urban catchment of Sungai Merah covers an area of 13.74 ha of residential area. Sungai Merah residential area is facing frequent flood, whereby water rises and subsides in matter of hour. Flood might be caused by poor maintenance of existing drainage or from earth drain which has waterlogged soil. Figure 3 highlights a few conditions of existing river and drainage. The objectives of conducting this study are the reconstruction of December 2008 flood to analyze and comprehend floods behaviors in Sungai Merah area, studying the hydrology of the drainage pattern in Sungai Merah area and lastly engaging a comprehensive analysis of the performances on the exisiting peatlands drainage system for future development purposes.

RESEARCH METHODOLOGY The research involved data collection, model development, calibration, verification and analysis (Figure 4). Nowadays, technology such as computer software namely Info Works, MOUSE, MIKE and the SWMM Models are used to develop computer models in order to study on the drainage or sewer system and river flow. InfoWorks Collection System (CS) is used to develop 1 Dimensional Urban Drainage System (Figure 5) with its embedded GIS applications modeling Sungai Merah catchment. InfoWorks CS supports the input

of all network data from models such as HydroWorks, DHI/MOUSE and SWMM (InfoWorks CS Technical Review, 2009). Data required in the model development comprises of hydrology data such as the rainfall data and tide data from the Department of Irrigation and Drainage (DID), Sarawak.

Poor Maintenance

Severe Inundation

Water Logging

Figure 3. Current condition of existing drainage and contributing drainage system

31


Apart from that, catchment characteristic, soil type and hydraulic data (water level, discharge and channel manning roughness coefficient) is taken into consideration. The United States Soil Conservation Service (SCS) Method is preferred for rainfall runoff analysis in the model simulation. This analysis is being adopted because of its simplicity and applicability to those watersheds of ungauged catchment with minimum hydrologic information of soil type, land use and treatment, surface condition, and antecedent moisture condition (AMC) (Leow et al., 2008). Hydrodynamic of Sungai Merah residential drainage network is simulated using numerical modeling. The hydrodynamic modeling used to predict the flood depth and performance of the existing drainage system. MODEL BUILD-UP Sungai Merah catchment and its floodplan 7.68 ha were modeled with 1.3 km-long

drainage network using InfoWorks CS. A 1:50 000 scaled digital topographical map of 5m contour intervals featuring Sungai Merah area which is requested from the Land and Survey Department, Sarawak. In addition, 1:10 000 scaled Seduan River sounding is bought from Sarawak Marine Department and digitized using ArcGIS 9. The Sungai Merah Catchment modeling first took place with ground modeling of Sungai Merah residential area using Geographic Information System (GIS) technology. GIS later dis played terrain features of Triangulated Irregular Network (TIN) dataset. The digital map featur ing Sungai Merah area was used to create GIS map using ArcMap from ArcGIS 9. The TIN model was exported to InfoWorks CS as a Digital Terrain Model (DTM) where the ground surface information was used for network build ing. The resulting TIN consisted of approximately 125,000 triangles based on contour and mass points. InfoWorks CS version 9 was used in this study. The model is a hydrology and hydraulic mode, which has its capability in urban drainage components (Liew et al., 2008). Sustainable Drainage System (SUDS) components were also included in the model necessary in small scale residential area. Firstly, DTM is exported into InfoWorks CS through its implanted GIS tool-Geographical Plan and later used to generate and display ground level contours, drainage, roads, buildings and ultility of Sungai Merah catchment. The sub-catchment of each compartment of housing areas was delineated based on available contour lines and inventory data. Nodes and conduits are then digitally placed along the existing drainage to form drainage network. RESULTS AND DISCUSSION Calibration and Verification

The model was calibrated against flood level of a selected historical flood event on February 26, 2009 whereas model verification event on December 26, 2008 is tabulated in Table 1 and illustrated in Figure 6 and Figure 7. Additionally, runoff from the

Figure 4. Flowchart of research methodology

32


sub-catchment is calibrated and verified for both events as listed in Table 1 and shown in Figure 8 and Figure 9. The calibration method was made by adjusting certain network component to obtain the effect of the model. This scenario was due to lack of information

on the areas and flood records. Relevant information such as flood depth was obtained from local authorities. Among the adjusted parameters are the SCS Curve Number (CN=87), Manning coefficients, η (η=0.03) and catchement slope. The selected parameters

Figure 5. Schematic diagram of model building

Model Development, Calibration, and Verification Reconstruction of December 2008

Digital Model of scale 1:50 000 TIN model featuring Sungai Merah catchment ground level

TIN surface model of Sungai Merah’s houses, drainage, road and

DTM of Sungai Merah residential Area

ArcGIS 9 Application

33


of certain network component can be adjusted by percentage to view the effect of changes on the model. Peatlands Implication on Drainage Network

Two different types of flood scenarios were simulated in Sungai Merah drainage system namely internal flood and external flood. Internal flood is the scenario, where heavy rainfall was occurred and there was no tide (normal water level). The flow and the velocity of the existing drainage network were observed whether the drain can cater for the amount water runoff during high intensity

storms. External flood is the scenario, where heavy rainfall is occurred in coincidence with the high tide. The peatlands are abundant with extensive waterlogged soils (Gore, 1983). Research done by Ong and Yogeswaran, 1991, stated that the high infiltration, water holding and transmitting capacity of peat influences the relatively small fluctuations in the water table. Besides, as external flood, it is influenced by King Tide levels along Seduan River with extreme rainfall intensity.

Table 1. Results of model calibration and

verification

Event Peak Water Level (meter)

Observed Simulated Differences

Feb. 26, 2009

0.55 0.48 0.07

Dec. 26, 2008

0.60 0.53 0.07

Event Peak Runoff (m3/s)

Calculated Simulated Differences

Feb. 26, 2009

5.13x10-4 5.37x10-5

2.3x10-5

Dec. 26, 2008 9.48x10-3

9.50x10-3 2.0x10-5

From the simulation that has been carried out, results demonstrate that the most critical

Figure 6. Calibration result for Sungai Merah area during February 26, 2009

Figure 7. Verification result for Sungai Merah area during December 26, 2008 (Long section)

(a) Calibration and verification point (During normal flow)

(b) Observed and simulated water level during February 26, 2009 (Long section)

Simulated Level: 0.48m Observed Level: 0.55m

34


scenario is when the heavy rainfall and fluctuation of water levels in Seduan River coincided. This will affect the water level in the drainage and causes flooding to the Sungai Merah residential area. In the nut shell, flooding in Sungai Merah lowlying areas depends much on the tide influences. The simulation results (Table 2) offers a way to identify the problems, sources of flooding and potential problems to Sungai Merah resident. The results were taken from the culvert in Figure 6(a) whereby the inlet and outlet is connected to the earth drain. The simulation results (Table 2) offers a way to identify the problems, sources of flooding and potential problems to Sungai Merah resident. The results were taken from the

culvert in Figure 6(a) whereby the inlet and outlet is connected to the earth drain. During internal flood, the 0.21 m3/s of flow in drain with 0.156m/s in velocity is higher than external flood of which the flow is 0.14 m3/s with a velocity of 0.11 m/s as illustrated in Figure 10 and Figure 11. This study strongly suggests that the influences of backwater effect from the Seduan River significantly affect the flood in Sungai Merah area. Thus, present studies by Ritzema and Wosten 2002, confirmed that in the lower-lying areas, drainage may be possible only during low-tides.

Furthermore, Ritzema and Wosten (2002) and Holden et al. ( 2006) added that excess rainfall will not be removed as surface runoff

Figure 8. Model calibration (Runoff event on February 26, 2009)

Figure 9. Model verification (Runoff event on December 26, 2009)

Figure 10. Comparison of flow for internal and external flood in December 2008

Figure 11. Comparison of velocity for internal 2008

35


but mainly as groundwater runoff for peatlands area. The surface-runoff (Figure 12) of Sungai Merah area is tabulated in Table 3. The result is verified using InfoWorks CS as shown in Figure 13. In addition, the model demonstrates that the Sungai Merah area will still be flooded due to internal flood resulting from excessive rainfall which is not influence by high tide. The impact from high tide will make the flood in Sungai Merah area more dreadful. Hence, fundamental aspect of flood in Sungai Merah area is triggered by the Seduan River’s backwater as demonstrated from the comparison results of the two events as stated in Table 2.

Table 3. Comparison of surface-runoff of internal and external flood

Events Internal

flood External flood

Surface-Runoff

(m3/s)

0.011

0.011

Therefore, the results simulated using Info- Works CS is believed to be sufficiently strong to provide information of flood in Sungai Merah area. In this study, InfoWorks CS has successfully identified the implication of peatlands on flood. Hence, further improvement on the drainage network system in the Sungai Merah area can be studied for a better implement. CONCLUSION The simulation demonstrates its capability for flood study on urban storm drainage in peat lowlands area. The current results of simulation give a clearer picture of the drainage flows, velocities, surface-runoffs and flood depths of an area along with tide level from the river channel. Further improvement on current existing drainage system in Sungai Merah residential area needs to be reconsidered in order to endure extreme rainfall and King Tide from Seduan River. The created model also allows better understanding of the drainage’s flowpath for future

EventsInternal flood (influence by heavy

rainfall)External flood(influence by heavy

rainfall and high tide)Peak Flow (m3/s) 0.21 0.14

Peak Velocity (m/s) 0.156 0.111

Average Flood Depth (m) 0.31 0.44

Table 2. Comparison of flow, velocity, surface-runoff and average flood depth for internal and external flood

Figure 12. The location of the surface- runoff in Sungai Merah area

Figure 13. Comparison of surface-runoff for internal and external flood in December 2008

Catchment area Surface- runoff

36


development and research work. However, simulation on larger study area, where more observed hydrological flood data are available is recommended for better understanding on the drainage behavior on the peatlands environment. REFERENCES Gore, A. J. B. (1983). “Introduction In: Gore,

A.J.B. (Ed.), Ecosystems of the World. Mires: Swamp, Bog, Fen, and Moor”. General Studies, Elsevier, Amsterdam, The Netherlands, Vol. 4A, p. 1–34.

Holden, J., Evans, E. G., Burt, T. P., and Horton, M. (2006). “Impact of Land Drainage on Peatland Hydrology”. Journal of Environmental Quality, p. 1764-1777.

Infoworks CS Technical Review. (2009). Available online from: http://www.wallingfordsoftware.com/products/infoworks_cs/infoworks_cs_technical_review.pdf; accessed on 31 October 2009.

Leow, C. S., Rozi, A., Zakaria, N. A., Ghani, A. A., and Chang, C. K. (2008). “Modelling Urban River Catchment: a Case Study in Malaysia”. Proceeding of the Institution of Civil Engineers, Water Management (162) 2009, p. 27-36.

Liew, Y. S., Selamat, Z., and Ghani, A. A. (2009). “Urban Stormwater Drainge System Study Using Numerical Modelling”. International Conference on Water Resources (ICWR 2009) 26-27 May 2009, Bayview Hotel, Langkawi, Kedah, Malaysia.

Majewski, W. (2008). “Urban Flash Flood In Gdansk-2001; Solutions and Measures for City Flood”. International Journal of River Basin Management, Vol. 6, No. 4, p. 357-367.

Ong, B. Y. and Yogeswaran, M. (1991). “Peatland as a Resource for Water Supply in Sarawak”. Proceedings of the International Symposium on Tropical Peatland, Kuching, Sarawak.

Ritzema, H. and Wösten, H. (2002). “Hydrology of Borneo’s Peat Swamps”. STRAPEAT-Status Report Hydrology.

Singh, H. and Bujang, B. K. H. (2003). “Basic Engineering Geology for Tropical Terrain”. Universiti Malaysia Sarawak.

Staub, J. R., Among, H. L., and Gastaldo, R. A. (2000). “Seasonal Sediment Transport And Deposition in the Rajang River Delta, Sarawak, East Malaysia”. Sedimentary Geology, V. 133, p. 249-264.

Tang, V. C. K. (2009). “Sustainable Construction on Soft Soils in Sibu: A Practical Perspective”. Engineering Seminar on Peat: Soft Soils:Challenges and Sustainable Solutions, October 15-16, 2009, Sibu, Sarawak.

Tie, Y. L. and Kueh, H. S. (1979). “A Review of Lowland Organic Soils of Sarawak”. Technical Paper, Soils Branch, Department of Agriculture Sarawak, No. 4, pp. 34.

Ting, S. C. (2009). “Relationship between Road Construction Methods And Post-Construction Rate of Settlement: Case Study on Kota Samarahan and Sibu Roads”. Degree Thesis, Universiti Malaysia Sarawak.

Wösten, J. H. M., Ritzema, H. P., Chong, T. K. F., and Liong, T. Y. (2003). “Potentials for Peatland Development”. In: Chew, Dr. D. and Sim Ah Hua (Ed.). Integrated Peatland Management for sustainable development. A Compilation of Seminar Papers, Sarawak Development Institute, Sarawak, Malaysia, p. 233-242.

37


Perbandingan Kekakuan Struktur Lantai Komposit Kayu Glugu-Beton dengan Berbagai Jenis Penghubung Geser

Susastrawan

Jurusan Teknik Sipil, Universitas Islam Indonesia, Yogyakarta, email: [email protected]

Abstract: There are several types of known composite structures, one of them is wood-concrete composite structure. It is relatively good building structure and cheap since wood materials are largely available in most markets and also cheaper if compared with steel materials. Coconut palm wood as part of wood-concrete composite structure has shown a good functioning. In order to make wood and concrete work together as a composite structure, needs shear connector as fastening devices are required such as nails, screws, or wood nails. This research investigates the effect shear connector as fastening devices at a composite structure. The result of laboratory experiment has indicated that wood-concrete structure using nails shear connector has the highest stiffness compared with similar structure employing other shear connector as fastening devices. Keywords: Shear connector and composite. PENDAHULUAN Salah satu persoalan pemukiman di kota-kota besar adalah pengadaan rumah tinggal untuk masyarakat, khususnya masyarakat berpenghasilan menengah kebawah. Ada kecenderungan bahwa orang akan berusaha untuk tinggal/ bermukim di tempat yang dekat dengan dimana dia bekerja. Sebagai akibatnya muncul pemukiman yang kumuh. Mengingat lahan didaerah perkotaan sangat terbatas dan relatif mahal maka dapat ditempuh cara dengan membangun rumah susun yang sangat sederhana. Salah satu cara untuk dapat membangun rumah susun yang relative murah dan aman adalah dengan membangun rumah susun dengan lantai Komposit Kayu-Beton. Dengan metode ini harga bangunan dapat dihemat sampai pada batas yang optimal dan struktur masih cukup aman. Penelitian sejenis yang pernah dilakukan Nor Intang Setyo H. dan Gathot Heri Sudibyo, Juli 2005 menggunakan Bambu-Keruing pada lantai beton dan Suwandoyo Siddiq dan Siti Zubaidah Kurdi menggunakan Kayu Kamper dan Kruing pada lantai beton. Pada penelitian ini akan diteliti Komposit Kayu-Beton, dengan bahan kayu Glugu dan sebagai penghubung geser mengunakan:

pasak, baut dan paku. Dari ketiga macam penghubung ini akan diteliti mana yang terbaik. Lantai komposit kayu-beton ini diharapkan dapat memberikan keuntungan, diantaranya : lebih ringan, kedap air dan suara, bersifat dekoratif, mudah dipasang, beaya lebih ringan, sedang kerugiannya struktur ini hanya efektif untuk menahan momen positif saja dan perlu perawatan yang baik agar kayunya tidak mudah lapuk. LANDASAN TEORI Aksi Komposit

Badan LITBANG PU Dep. PU menyatakan, struktur komposit merupakan struktur yang terdiri dari dua jenis bahan konstruksi bahan yang berbeda yang disatukan dengan bagian

Gambar 1. Struktur balok non komposit

Terjadi slipTerjadi slip

38

Susastrawan: Perbandingan Kekakuan Struktur Lantai Komposit Kayu Glugu-Beton

penyambung yang lebih dikenal dengan penghubung geser (shear connector). Penghubung geser dipasang untuk meng-hubungkan dua bahan tersebut sehingga secara bersama-sama dapat memikul beban yang bekerja pada struktur. Untuk memahami aksi komposit, pertama-tama ditinjau aksi komposit tidak sempurna (non-komposit) seperti yang disajikan pada Gambar 1. Gesekan diantara pelat beton dan balok kayu diabaikan seperti terlihat pada Gambar 1. Dua element struktur, pelat beton dan balok kayu masing-masing memikul sebagian beban secara terpisah. Permukaan bawah beton mengalami perpanjangan akibat deformasi lentur, sedangkan permukaan atas kayu akan mengalami perpendekkan akibat deformasi tekan. Diagram regangan yang bekerja pada struktur tidak komposit disajikan dalam Gambar 2. Dengan meninjau Gambar 2, terlihat bahwa kasus ini terdapat dua garis netral. Garis netral pertama berada pada titik berat pelat beton dan garis netral yang kedua terletak pada titik berat balok kayu. Selanjutnya ditinjau struktur bekerja komposit sempurna, maka slip antara pelat beton dengan balok kayu tidak akan terjadi. Konsep analisis penampang komposit penuh didasarkan pada dua kondisi, yaitu kondisi elastis dan kondisi plastis. Kondisi elastis adalah kondisi dimana baik pelat beton maupun balok kayumasih berada dalam batas-batas elastis. Penelitian ini dibatasi pada kondisi elastis. Beberapa batasan dalam analisis struktur komposit ini diantaranya adalah:

1. Defleksi vertikal mempunyai nilai yang sama untuk kedua elemen, hal ini berarti tidak ada gap antara pelat beton dengan balok kayu.

2. Penampang tetap rata baik sebelum maupun sesudah dibebani, deformasi geser antara dua elemen diabaikan;

3. Jarak antara penghubung geser adalah sama;

Sistem komposit sempurna tidak akan terjadi gelincir relatif diantara pelat beton dan balok kayu seperti yang disajikan pada gambar 3.

Gaya-gaya horisontal (geser) terjadi dan bekerja pada permukaan bawah slab tersebut, sementara gaya-gaya tersebut juga bekerja pada permukaan atas balok kayu. Pada kasus komposit sempurna, diantara pelat beton dan balok kayu tidak akan terjadi gelincir dan diagram tegangan yang disajikan pada Gambar 3. Terjadilah sumbu netral tunggal yang terletak di bawah sumbu netral pelat beton dan diatas garis netral balok kayu.

Penghubung Geser

Dalam tinjauan balok lentur hendaknya dipertimbangkan pula, bahwa pada saat yang sama balok mengalami gaya horizontal yang timbul akibat leturan. Gaya horizontal yang timbul antara pelat beton dan balok kayu selama pembebanan harus ditahan agar aksi komposit bekerja secara monolit (lihat Gambar 4 dan 5). Dalam keadaan terbebani, antara elemen kayu dan beton diberi “shear connector” sebagai alat penyambung geser, agar kedua element tersebut mengalami deformasi yang sama. Adapun jenis-jenis alat penyambung geser yang akan diteliti adalah paku, baut, dan pasak. Tegangan geser yang timbul dapat dilihat keseimbangan elemen pp1 nn1 pada Gambar 5 yang dipotong dari sebuah balok antara dua

Gambar 3. Struktur balok komposit sempurna

Gambar 2. Diagram tegangan penampang balok non komposit

P

T

T

C

C

σ

beton

kayu

39


penampang berdekatan mn dan m1n1 yang terpisah oleh jarak dx. Permukaan atas pada bagian yang diarsir bekerja tegangan-tegangan geser horizontal (τ) pada pemukaan-permukaan ujung elemen bekerja tegangan-tegangan lentur normal (σx) yang dihasilkan momen-momen lentur (Singer dan Pytel,1985).

Ditinjau sebuah elemen balok seluas dA pada jarak y dari sumbu netral pada gambar 5.

Gayanormal σ dA (1)

Tegangan lentur,

σ (2)

Gaya-gaya elemental melalui luas permukaan pn,

F ∗ dA (3)

Dengan cara yang sama p1n1,

F ∗ dA (4)

Gaya horizontal F3 yang bekerja pada permukaan pp1 adalah,

F τ ∗ b ∗ dx (5)

Gaya-gaya F1,F2,F3 harus berada dalam keseimbangan statis dalam arah sumbu x maka,

F3 = F2 – F1 (6)

τ ∗ b ∗ d dA dA

τ *

τ ∗

∗ (7)

Aliran geser,

τ ∗ (8)

Hubungan antara momen lentur dan tegangan lentur, dengan asumsi:

1. Bidang penampang balok, bidang semula, tetap bidang.

2. Balok homogen dan mengikuti hukum Hooke.

3. Modulus elastisitas tarik dan tekan sama. 4. Balok lurus dan penampang tetap. Hubungan Momen–Kelengkungan Persamaan differensial defleksi balok

d’ d

g y

( b )

Gambar 6. Sistem Balok Terlentur

Gambar 4. Diagram Regangan dan Tegangan penampang komposit sempurna

Gambar 5. Diagram tegangan elemen balok

beton

C

σ kayu

T

F

h

dA

F

dx

mm

nn

p1

p y y1

c

( a )

dx P

R2R1

d b

a

k h

f e

c’

b

c a

ρ

40

Susastrawan: Perbandingan Kekakuan Struktur Lantai Komposit Kayu Glugu-Beton

Gambar 6 (a) memperlihatkan dua penampang yang berdekatan, ab dan cd, dipisahkan oleh jarak dx. Penampang ab dan cd berotasi relative satu sama lain sebesar dθ, seperti nampak pada gambar 6 (b), tetati tetap lurus. Serat ac akan memendek dan serat bd akan memanjang. Serat ef tetap tidak berdeformasi karena terletak di garis netral. Kemudian dari titik f ditarik garis sejajar ab, sehingga diperoleh garis c’d’ dan bd’=gh=ac’=dx. Kemudian ditinjau serat gh sejauh y dari garis netral. Sebelum deformasi gh=ef=dx, setelah deformasi serat gh bertambah sebesar hk. ∆ Oef dan ∆fhk sebangun, sehingga dapat diperoleh perbandingan sebagai berikut :

(9)

Oe = ρ ,ef = gh dan hf = y, maka:

(10)

Sehingga diperoleh ;

(11)

, maka:

(12)

Dari Hukum Hooke :

σ E E y (13)

Persamaan diatas menunjukkan bahwa tegangan pada setiap serat bervariasi langsung dengan kedudukan y dari garis netral dan karena modulus elastisitas E dianggap sama untuk tarik dan tekan serta nilai ρ tidak tergantung dari kedudukan serat y. Hal ini berlaku untuk kondisi tegangan masih di bawah batas proporsional, karena hukum Hooke hanya berlaku pada tegangan di bawah batas proporsional. Untuk menurunkan rumus lentur, digunakan kondisi keseimbangan yang terjadi pada suatu tampang. (lihat Gambar 7):

∑Fx = 0, maka Mluar = Mdalam (14)

dM = (σxdA)y (15)

Keseimbangan momen terhadap sumbu Z harus sama dengan nol ( ∑Mz = 0 ). Hal ini berarti momen lentur diimbangi oleh tahanan momen, yaitu M = Mr. Tahanan momen terhadap sumbu netral adalah = y(σxdA), sehingga total tahanan momen adalah :

dM = (σxdA)y (16)

(17)

Karena σx = σ = (E/ρ)y, maka:

(18)

Karena didefinisikan sebagai I, maka:

(19)

Atau:

(20)

Tan θ = dy/dx, karena sudut θ sangat kecil sekali maka tan θ = θ, sehingga:

θ (21)

Gambar 7. Tegangan pada tampang balok

Permukaan netral

Y

τxzdA

τxydA

σxdA X

Z

z y

41


Dan,

(22)

Dari gambar 1b diperoleh : ef = ρ dθ dan ef = gh = bd’ = dx, maka: dx = ρ dθ atau :

(23)

Sehingga:

(24)

Maka:

(25)

Penyelesaian persamaan deferensial diatas: (Chajes, 1974)

Dari gambar 8 diatas dapat diperoleh :

∆ i (26)

∆ f ∆ ∆f∆ / /

∆ f (27)

Dari persamaan batang lentur diatas akan dapat diperoleh:

(28)

(29)

TEST LABORATORIUM DAN PEMBAHASAN Pengujian Geser terhadap Alat Sambung Geser

Pada penelitian ini dipakai baut diameter 10 mm, pasak dg kayu bambu diameter 10 mm dan paku diameter 4,57 mm dengan panjang masing-masing 12 cm. Masing-masing benda uji terdiri dari 12 buah penghubung geser(lihat Gambar 9). Dari hasil test laboratorium diperoleh hasil bahwa penghubung geser dengan paku memberikan hasil yang paling baik (paling kaku) dibandingkan baut dan pasak (lihat Gambar 10). Gambar 10 menunjukkan bahwa masing-masing jenis penghubung geser dapat menahan beban sebagai berikut, penghubung geser paku Pmax = 3000 kg, penghubung geser pasak Pmax = 2000 kg dan penghubung geser baut Pmax = 2400 kg. Dengan demikian maka kekuatan penghubung geser adalah:

Baut: Pmax = 2400/12 = 200kg/buah Pasak: Pmax = 2000/12 = 166,67 kg/buah Paku: Pmax = 3000/12 = 250 kg/buah

Pemindahan gaya pada penghubung geser paku dapat berlangsung lebih baik dari pada baut dan pasak. Hal ini disebabkan karena paku-paku tersebut dipukulkan begitu saja kedalam lubang yang lebih sempit yang dibuat sebelumnya, sehingga paku dapat menyatu dengan kokoh pada kayu (lihat Wiryomartono, 1976). Selanjutnya dibuat benda uji balok komposit yang terdiri dari balok kayu Glugu ukuran 6x12 cm2 dan pelat beton ukuran lebar = 40 cm dan tinggi 6 cm dengan penghubung geser masing-masing Baut, Pasak dan Paku. Bentang masing-masing benda uji adalah sebesar 3,4 m, seperti nampak pada Gambar 11.

Gambar 8. Model Finite Difference

x= i

f(i-h)

f(i)

f(i+h) h) f(x)

f(x)

X

hh

42

Susastrawan:

½ P

kayu

Perbandingan

( a ) Baut

P

beton

Kekakuan Stru

Gambar 10

G

G

A

½ P

kayu

ktur Lantai Kom

0. Grafik da

Gambar 9. U

rafik Be

Alat Sam

½ P

kayu

mposit Kayu Gl

aya dukung

Uji Geser A

eban-Def

mbung G

( b ) Pasak

P

P

beton

lugu-Beton

alat penghu

lat Sambun

fleksi

Geser

½ P

kayu

ubung geser

ng

½ P

kayu

r

( c ) Pak

P

P

beton

ku

½ P

kayu

43

Gamb(

Gam(

Pengguna

bar 12. Grafalat sambun

mbar 13. Gra(alat sambu

aan Alat Sa

fik Beban-Dng geser bau

afik Beban-ung geser pa

Gambar

( b )

Penghubugeser bau

ambung Ge

Defleksi ut)

Defleksi asak)

11. Struktur

dial

Pege

ung ut

J

eser pada S

340 cm

r Balok Kom

(a)

P

( c )

enghubung eser pasak

Jurnal Rekayas

Struktur Ko

mposit Kay

Gambar(ala

Gamb

(

Penggese

sa Sipil (JRS), V

omposit Ka

yu-Beton

r 14. Grafikat sambung

bar 15. Dimkompo

tw

btr

( d )

hubung er paku

Vol. 1, No. 1, M

ayu-Beton

k Beban-Defgeser paku

ensi tampanosit

kayu

beton

Mei 2011: 38-46

fleksi u)

ng

t

hw

6 44

Susastrawan:

Dari uji baf’c = 20 MMOR = 52 Dengan dimb = 40 cmbw = 6 cm hw = 12 cmt = 6 cm h = 18 cmmaka dapteoritis adaФu = 0,493

Tabe

Penghubuy

geser

Baut

Pasak

Paku

Dari hasil pada Gambtampak b

Perbandingan

ahan beton dMPa, maka F2,395 MPa,

mensi tampm dan btrans =

m

m pat dihitungalah: 34 rad/m

l 1. Kekaku

ung BebanPy

(kN)

1 40

2 22

1 30

2 22

1 38

2 38

test laborabar 12, 13, bahwa ba

Gamb

Kekakuan Stru

dan kayu diFc = 17 MPa

maka Fw =

pang (lihat G= 171,2 cm

g kelengku

uan Balok K

n

)

DefleksΔy

(mm)

55.28

48,61

42,8

30,25

36,6

31,74

atorium yan14, dan 16

alok komp

bar 16. Graf

ktur Lantai Kom

iperoleh haa.

= 41,92 MPa

Gambar 15)

ungan ultim

Komposit

si KekakuaPy/Δy(kN/m)

723,59

452.58

700,93

727,27

1038,25

1197,22

ng ditunjukkserta Tabel

posit deng

fik hubunga

mposit Kayu Gl

asil:

a.

:

mit

an )

9

8

3

7

5

2

kan l 1, gan

peya

B

P

P

ApyadenilnilultBapadibgekadeleblaisajka

an Beban-de

lugu-Beton

enghubung gang paling b

Tabel 2. K

PenghubungGeser

Baut 1

2

Pasak 1

2

Paku 1

2

pabila dilihaang terjadi engan penglai kelengkulai kelengkutimet teoritialok kompaku mempunbanding bal

eser baut darena peminengan penghbih baik dain, karena pja, sehingg

ayu (Wiryom

efleksi komp

geser paku besar.

Kelengkunga

g Mom

Maksim(kNm

21,6

12,9

18,3

12,9

28,0

28,0

at dari perh(lihat Tabe

ghubung geungan yangungan terseis = 0,4934.osit denganyai kekakulok komposdan pasak. ndahan gayahubung geari pada pepaku-paku tga terikat/mmartono, 19

posit Kayu-

mempunya

an Balok K

men mum m)

Kel

60

96

36

96

08

08

hitungan kelel 2), balokeser paku mg paling keebut << kel. an penghubuan yang psit dengan p Hal itu a pada baloser paku b

enghubung ersebut dipumelekat er

976).

-Beton

ai kekakuan

omposit

lengkungan (1/m)

0,0222

0,0321

0,0633

0,0777

0,0030

0,0041

lengkungank kompositmempunyaicil, dimanalengkungan

bung geserpaling besarpenghubungdisebabkank komposit

berlangsunggeser yangukul begiturat dengan

n

n t i a n

r r g n t g g u n

45


SIMPULAN 1. Komposit Kayu Glugu-Beton dengan

penghubung Geser Paku menghasilkan struktur yang paling kaku dibandingkan dengan struktur komposit Kayu-Beton dengan penghubung baut dan pasak.

2. Pelaksanaan pembuatan komposit kayu-beton mudah. Karena mudah tentunya juga menjadi lebih murah.

SARAN 1. Alat penghubung geser mempunyai variasi

yang cukup banyak, sehingga diperlukan penelitian lebih lanjut untuk penghubung geser yang lain.

2. Perlu diteliti lebih lanjut struktur komposit kayu-beton dengan menggunakan kayu yang mempunyai kelas awet lebih tinggi,

misal dengan kayu jati atau kayu Kalimantan.

DAFTAR PUSTAKA Chayes and Alexander. (1974). “Principles of

Structural Stability Theory”, Prentice-Hall, Inc, Englewood Cliff, N.J.

Singer, F. L., Pytel, A., and Sebayang, D. (1985). “Teori Kokoh - Strength of Materials”, Erlangga.

Setyo, N. I. H. and Sudibyo, G. H. (2005). “Balok Komposit (Glulam) Bambu Keruing pada Lantai Beton”. Media Teknik Sipil.

Wiryomartono, S. (1976). “Konstruksi Kayu”. Fakultas Teknik UGM, Jilid 1.

Siddiq, S. and Kurdi, S. Z. (1995). “Struktur Lantai Komposit Kayu-Beton”. Majalah Ilmiah Teknik, ISSN: 0854-2279, Jilid 1.

46

Thanaya: Masonry Unit Utilizing Aggregate From Construction Demolition Bound with Asphalt

Masonry Unit Utilizing Aggregate from Construction Demolition Bound with Asphalt

I Nyoman Arya Thanaya

Udayana University, Denpasar, Bali, Indonesia, email: [email protected]

Abstract: In line with increasing pressures to reduce the exploration of natural aggregates, utilization of waste aggregates materials for construction industry had been encouraged. One alternative material that can be used as masonry unit is aggregate from construction demolition with asphalt as the binder (an alternative to cement). Construction demolition may available due to various reasons such as demolition for reconstruction, war, natural disasters etc. The investigation was carried out in the United Kingdom (UK), which is applicable else where in the world especially in oil producing countries where large amount of asphalt is available. The as phalt used as the binder was 100pen grade asphalt. The objective of the investigation was to produce masonry unit with performance equal to concrete block commonly used in the United Kingdom with compressive strength between 2.810 MPa. This requirement is slightly higher than the minimum compressive strength of 2.5 MPa in line with Indonesian standard. The specific creep strain targeted was less than 100 microstrain. The masonry unit requires suitable particles size proportion, in order to obtain the expected results: stable during handling with low compaction effort with satisfactory compressive strength and to meet demand in using minimum bitumen content. The materials were proportioned, hot mixed, compacted than heat cured. The masonry unit requires sufficient heat curing. It was found that construction demolition materials (CDM) were suitable to be used for making masonry unit with asphalt as the binder. Compaction level of 2 MPa and curing regime of 200°C for 24 hours were sufficient and gave an overall satisfactory result. It was also found that the masonry unit volume stability was affected by relative humidity (moisture). Its volumetric movement due to moisture was not fully reversible, but highly reversible on thermal exposure. The compressive strengths of the unit produced using locally available construction demolition material, well meet the Indonesian standard, i.e. minimum of 25 kg/cm2. Keywords: Masonry, construction, demolition, aggregate, and asphalt. INTRODUCTION From masonry Each house on average requires Demands on aggregates fot construction approximately 200 m of building block industries worldwide continue to increase. For example, aggregate demand of the United Kingdom (UK) raise from a level of 270 million tonnes in 1989, to a predicted demand of 420-490 million tonnes by 2011 (Whitbread et al., 1991). Meanwhile there has been an increasing work resulting in approximately 350 million blocks being manufactured each year. Waste or by product materials such as steel slag, crushed glass, and coal fly had been

incorporated into masonry building blocks (Forth et al., 2006). Another alternative material that can be used is construction demolition waste (CDW) from pressure to reduce the exploration of natural building or road pavement (Craighill et al. aggregates). This situation encourages the utilization of waste and secondary aggregate materials for construction industry. Currently, 160,000 new homes are built achear in the UK of which 90% are constructed (Forth et al., 2006). Around 17% of the total UK waste arises from the construction and demolition industries (DoE, 1994). Although a large proportion of

47

the waste most of itsuch as aconly 4% is(HumphreyThere arematerials, perceived strength. confidencematerials, identifyingappropriatedisseminatpublicationof the aggreplaced bCDW agaffecting thal., 2004)

The objecwithin thiswith perfoblocks coKingdom between 3of 2.8 Mcreep strai1985). The applicwhen largedemolitioninevitable

FigureWas

created is t is used ccessroads s used to repys et al., 19

e many pohowever thrisks inv

There is e in the

which cang, undertae demonsting thens and semiregate for c

by up to aggregates he compres

tive of thes paper is

ormances atommonly (UK) wit

.57 MPa (SMPa (BS60in less than

cation of te amount C

n of old bnatural d

e 1. The Conste (CDW) a

recycled ifor low grwithin landplace prima94). otential usehey are deolved due

a need use of CDn only be aking andstration pe resultinars. The cconcrete miapproximate

without ssive streng

e investigatto produc

t least equaused in

h compresSear 2005) 0731981),

n 100 micro

this researcCDW is avbuildings, adisasters, su

nstruction Daggregates m

in some wrade purpodfill sites aary aggrega

es for CDeterred by to its l

to increaDW recyc

achievedd monitoriprojects, ats throucoarse fractiixtures can ely 20% w

significangth (Soutsos

tion describce CDWbloal to concr

the Unissive streng

or minimuand speci

ostrain (Tap

ch is possiailable due

and/or due uch as ea

Demolition material

way, ses and ates

DW the ow ase led by

ing and ugh ion be

with ntly s et

bed ock rete ted gth um ific sir,

ble e to

to arth

quattesplotcaavareproselbit(TThtesan M M

Thcowomabuthithaclarecproas sizmaThgrabit

Jurnal Rekaya

uake, flood,tractive inpecially in t of asphaltn be used

vailable. Theas can be oviding blves.tumen.tumen ha

Thanaya et ahe propertiested in line

nd are given

METHODS

Materials

he aggregaonstruction ould be ofay come fro

uildings or cis investigaat consisteday brick maclaim asphaoportion wcan be see

ze distributaximum parhe bitumenade bitumetumen.

Figure 2. Waste (

100 90 80 70 60 50 40 30 20 10

0 0.01

asa Sipil (JRS),

cyclone, etareas clo

oil producit (oil distilas an altern

he people intrained to m

building m Utilizatio

ad given al., 2006).

es of the Ce wit the Bn in Table 1.

ate used demolition non homog

om demoliticonstructionation was td of a mixtuasonry, sanalt pavemen

with minor cen in Figuretion is showrticle size oused as th

en. This is

The Constr(CDW) agg

original grad

0.1

Vol. 1, No. 1, M

tc. This woose to oiing countriellation residnative binden unfortunamake CDWmaterials on of so

satisfactor

CDW aggreBS 812 (BS.

is aggregwaste (CD

geneous maion of varions. The CDtaken from ure of brokend cement mnt (RAP) inclay brick

e 1. The CDwn in Figuf 10 mm.

he binder wrelatively

ruction Demgregates mat

1 10

Mei 2011: 47-57

rks is moreil refinery,es, where adue) whicher is widelyate disaster

Wblocks forfor themfter gradery results

egates wereS812 1995),

gate fromDW), whichaterial as itous types ofW used fora supplier

en concrete,mortar, andn a randomcomponent

DW particleure 2, with

was 100 pensoft grade

molition terial

0 100

7

e , a h y r r

m e s

e ,

m h t f r r ,

d m t e h

n e

48


Cl

tiP

i

Referring to Table 1, the properties the CDW aggregates were found very comparable with commonly aggregates available for building industries. However, as the CDW is processed from waste materials, therefore the homogeneity of the aggregate component may not be consistent.

Table 1. Properties of the CDW aggregates

Initial Trials

Building block units bound with asphalt can be produced with either using continuous or gap aggregate grading. When using continuous aggregate grading, higher compaction effort commonly needed (Forth et al., 2006). Within this investigation gap aggregate gradation was selected in order to enable the use of lower compaction effort. As there is no standard specifically available for building block aggregate grading bound with asphalt, Hot Rolled Asphalt (HRA) aggregate gradation was referred to as an initial reference, then modified based on trials results. During the initial trials, the CDW material particles sizes were graded (sieved) into coarse fraction of 105mm and 52.36mm, and fine fraction of all passing 2.36 mm. Initially the aggregate grading used for the CDWblock was the same as the Bitublocks (similar block incorporating various waste aggregate materials) previously produced with a gap graded aggregate grading (Forth et al. 2008), but with maximum particle size of 10mm. The aggregate composition was: 40% coarse fraction, 50 % fine fraction, and 10 % coal fly ash filler as shown in Figure 3, which is completed with a hot rolled asphalt (HRA) grading of the BS 594 (BS5941 2005), for a

general comparison, and for a better appreciation on the aggregate composition.

The bitumen used as the binder is a commercially valuable by product material from crude oil refinery industry, therefore the CDWblocks were produced with lowest bitumen content possible, that can give sufficient or adequate bitumen coating with satisfactory shape stability during handling and satisfactory performances when compacted at low compaction effort. Minimum bitumen content of 5 % by weight of total mixture was initially tried. The static compaction effort was 1 MPa for 1 minute. The degree of bitumen coating was satisfactory; however the shape of the blocks was not stable. The corner sides or edges were easily taken off. The surface texture was rather smooth, and some aggregate particles on the edge sides of the samples were taken off. This had caused some part of the samples had an open texture, therefore water absorption after 24 hour immersion was found high (6.9%), as shown in Tables 2b (Mix A). Gradation Modification and the Properties of the CDWblock

The specific gravity of the CDW materials were not the same with the Bitublock previously made (Forth et al., 2008), so it affects the volumetric composition of the materials. This was found to give effect of the

Properties Unit Coarse CDW (> 2.36mm)

Fine CDW (< 2.36mm)

Density (bulk) g/cm3 2.415 2.341

Density (ssd) g/cm3 2.478 2.433

Density (app) g/cm3 2.569 2.578

Water abs. % 2.5 3.9

BS594 upper limit

BS594 lower limit

Bitublock grading

CDWblock grading

100 90 80 70 60 50 40 30 20 10

0 0,01 0.1 1 10 100

Figure 3. The CDW block aggregate grading compared with the Bitublock

grading and the BS594

49


Mix Comp. effort

(MPa) Water Abs*

(%)

Comp. Strength (MPa)

uncured cured

Mix A 1 6.9 2.0 8.1

Mix B

1 3.3 2.8 10.2

2 2.7 3.8 17.5

4 1.6 5.4 25.8

Mix Comp. effort (MPa)

Density (g/cm3)

Porosity (%)

IRS (kg/m2.min)

Mix A 1 1.925 15.1 0.105

Mix B

1 1.872 17.5 0.028

2 1.950 14.0 0.021

4 1.992 12.2 0.018

sample’s shape stability (compactness). In order to improved the CDWblock shape stability, the aggregate grading was slightly modified. The max particle size used was 10mm, instead of 14 mm as used for the Bitublock. The filler content was reduced from 10 % to 5%, but increasing the fine fraction from 50 % to 55 %, where the coarse fraction remains at 40 % (Figure 3). This gradation modification was found to give more compact samples with surface texture neither too smooth nor too rough. Overall, the aggregate gradation of the mix became coarser, hence theoretically it has lower total surface area. Even at the minimum bitumen content of 5 % as initially tried the asphalt film thickness would increase. In order to improve impermeability, asphalt film need to be made thicker. For this reason the bitumen content was increase from 5 % to 5.5 %, in addition to the reduction of aggregate surface area as mentioned above. The shape stability during handling was found satisfactory and the surface texture of the newly produced sample was found neither too smooth nor too rough. By visual observation, the asphalt coating was also satisfactory, and theoretically with thicker asphalt film. This was considered necessary in order obtain satisfactory overall performances. The increase of the bitumen content was also done for anticipating the variation in quality of the CDW materials which are very likely of various absorption properties, as they are indeed a waste material. CDWblock with modified gradation (Mix B in Tables 2a and 2b), were then produced with compaction effort of 1, 2, and 4 MPa, with bitumen content of 5.5 % by weight of total mixture.

Table 2a. The density, porosity, and IRS of the CDWblocks

Parameter that can provide an indication of the effect of the unit on the sand cement mortar.

Units with high IRS require very plastic mortar (high water/cement ratio), while units with lower IRS need stiffer mortar (Vekey, 2001). In addition to the data presented in Table 2, the compressive strength of Mix B are also plotted in a graph as shown in Figure 4. Table 2b. The density, porosity, and IRS of the

CDWblocks

Heat Curing

Asphalt is a viscoelastic material. It has viscous and elastic component. The viscous component of the asphalt would cause creep when loaded. It had previously been found that curing regime played a very significant role for hardening the asphalt due to the evaporation of the volatile component and increasing the asphaltene component of the asphalt (Forth et al., 2006, Whiteoak, 1991), hence can reduce creep deformation due to static load (Forth et al. 2006). It had been previously investigated that when using a 50 pen bitumen and cured in oven at 160 °C, the curing duration required to

Figure 4. Compaction level vs compressive strength 24 hours immersion in water

Cured

Un-Cured

0 1 2 3 4 5

30

25

20

15

10

5

0

50

Thanaya: Mas

satisfy cr(Thanaya curing dusamples whours, andsatisfactory Initial Rat

IRS test wsample insecond. Thsample waarea in coIRS is a pperformancgrading compared content, abitumen coB, at 1 MPlower poroabsorptionthinner. Thwas better coarser anlogical, thimproved regime hadimprove tsamples. The water a reasonab5.5% bitummore impfound relavalues of ckg/m2/minCDWblocklower wate Expansion

This test expansion different equipment with its supoints. Tweach of tvertical (V

sonry Unit Util

eep perforet al., 200

uration, in were curedd the sampy creep resi

te of Suctio

was carriedn 3mm dehe weight oas then calcuontact withpareferring tce of Mix which cwith Mix

and also ontent. AsPa compac

osity than M was higherhe compresthan Mix A

nd of thickhat the perf

at higher cd been expthe compr

absorption ble level, men conten

permeable. atively lowclay brick fon). This ks require ser cement ra

n/Shrinkag

was carriedand shrinkenvironm

used was upporting f

wo Demec the four si

V) direction,

lizing Aggregat

rmance wa06). In ord

this inved at 200°Cple was foistant.

on (IRS)

d out by imepth of wof water abulated and d

h water (Bto Tables 2

A was afcontained x B with affected b

s Mix A fition level M

Mix B. Howr as the bitussive strengA, as Mix Bker bitumeformance ocompaction perienced toressive stre

of Mix B wwhich indint, the samThe IRS

er than theound in the

indicatessand cemenatio (Vekey

ge Test

d out by mkage of th

ment cona 50 mm

facilities supoints we

ides of theas shown in

te From Constr

as 72 hoder to reduestigation C just foround to ga

mmersingwater forbsorbed by divided byS3921 198

2a and 2b,ffected by

10% fil5 % fil

by its lowiner than MMix A wasever the wa

umen film wgth of MixB was slighen film. Itof Mix B w

level. Curio significanength of

was overallicated thatmples beca

values we typical IUK (0.25–2

s that nt mortar w, 2001).

measuringhe samplesdition. TDemec gau

uch as demere pastede samplesn Figure 5.

ruction Demol

urs uce the 24

ave

the 60 the the

85). the its

ller ller wer Mix of

ater was x B htly

is was ing

ntly the

l in at

ame were

RS 2.0 the

with

the at

The uge mec

on in

F

Wfor3.5CDthethekgbeaspsuexus Cr

Threstome6)crecre

lition Bound wi

Figure 5. Thm

Where commr concrete b57 MPa (SeDWblock we range of Ie United

g/m2/min).ecause the aphalt whichggest that

xperiment we stiffer mo

reep Test

his test wasistance of

static loadeans of a s, with 1 Meep stress)eep test (Ta

Figure 6. A

ith Asphalt

he CDWblomeasuring e

mon comprblocks founear, 2005). were found IRS values f

KingdomLow IRS

aggregates wh has hydro

the CDWwould requirortar.

as done in the sample

d. The samsimple arm MPa stress.is common

apsir, 1985)

Arm load matest loa

ock sample aequipment

ressive strend the UK The IRS vasomewhat

for clay bricm (between

values werwere evenlyphobic char

Wblock testre or more

order to ees to defor

mples wereload mach

. This strenly applied .

achine used ading

and strain

ength valueis between

alues of thelower thanck found inn 0.252.0re obtainedy coated byracter. Thisted in this

suitable to

evaluate thermation duee loaded byhine (Figuress (specificin masonry

for creep

e n e n n 0 d y s s o

e e y e c y

51


Expansion Due to Moisture Absorption

Before doing creep test, the cured samples were tested for their expansion at room environment condition (21±0.5 °C and 46 % relative humidityRH). It was found that the samples expanded then stabilized after about 7 days. The samples with higher compaction level gave lower expansion as shown in Figure 7.

RESULTS AND DISCUSSION Compressive Strength and IRS

Referring to Figure 4, it is revealed that the cured compressive strength of the samples were satisfactory, i.e well exceeded 7 MPa, In order to evaluate the performance of the samples at different relative humidity, a further volume stability test was done. Two samples were initially conditioned at room environment (at a different room from previous experiment). The temperature was relatively constant at 21.0 ± 0.5 C°, but the humidity fluctuated at 62±2% RH. The samples were also conditioned at 12%RH and 85%RH which were carried out by using desiccators filled with lithium chloride and potassium chloride hygrostatic solution respectively. The expansion reading was taken at certain time interval until the expansion stabilized. Then the conditioning was changed. The results are shown in Figure 8 where the starting of conditioning changes are coded from A to E.

The samples were initially left overnight at room environment with 62±2% RH before the first strain reading was noted. After conditioning at room environment, the samples were then conditioned in a desiccator with 12% RH, using lithium chloride hygrostatic solution (start of conditioning A). The samples gradually shrunk then stabilized at +100 microstrain. Starting on day 11th, the samples were taken out from the desiccator and left at room environment (62±2% RH), (start of conditioning B). The samples slowly expanded then stabilizedat 20 microstrain. Starting from day 24th the samples were put back into a different desiccator with 85%RH, using potassium chloride hygrostatic solution (start of conditioning C). The samples expanded (towards negative strain values) then stable at -340 microstrain. Starting for day 44th

and then the following days, the samples were consecutively condition until stabilized at room environment (start of conditioning D), then at again at 12% RH (start of conditioning E). cules onto the surface of the particles reduces the surface energy on the capillary system, hence reducing the balancing internal compressive stress leading to volume increase or swelling (Domone, 1994). This is also described by Neville (Neville 1991), that during water adsorption, the water molecules act against cohesive forces and tend to force the cement gel particles further apart. The ingress of water also decreases surface tension, and results in swelling. It was also observed that the samples did not crack which indicates that the expansion was not excessive. The expansion of the unit would be neutralized by the shrinkage of the sand cement mortar joints in wall construction. The expansion can also give a prestressed condition to the wall structure which can improve the ability of the wall to receive horizontal load. Thermal Expansion

Thermal expansion test was carried out by conditioning the samples in oven at 70 °C for 3 hours. This time was sufficient to generate the targeted heat on the core of the samples

0 1 2 4 6 8 10 12 14 16

Figure 7. The vertical expansion test results

1 MPa 2 MPa 3 MPa

-25 -50 -75

-100 -125 -150 -175

0

52


(tested using a thermocouple inserted into one of the samples).

Figure 8. Average movement of the samples (sp) in vertical (v) direction

Figure 9. The strain profile of the samples (sp) during heating at 70 °C for 3 hours in vertical

(v) direction This further test confirmed that the volume stability of the samples was affected by changes in relative humidity. Conditioning to lower relative humidity caused the samples to shrink and vice versa. However, the magnitude of expansion and/or shrinkage was found not proportional to the changes in RH. The results indicated that the samples movement were partly reversible and irreversible. Quiet a large portion of the movement was irreversible. This situation is similar to clay brick (Vekey, 2001). The results suggest that the expansion of the samples were of similar mechanism with cement paste or concrete, i.e. due to moisture adsorption. Due to adsorption of water mole After heating, the samples were left at

room environment (21.0 ± 0.5 °C, with 50±2% RH) until stabilized. The results are presented in Figure 9. Referring to Figure 9 (strain in vertical direction), within the first heating cycle the samples expanded to 750 microstrain (10-6). Then the samples were taken out from the oven and conditioned at room environment for 2 days. Within the fist day at room environment the samples shrunk almost to its original position and then slightly expanded on the next day due to the moisture absorption from the environment (as had been experienced). Similar procedures were carried out on the next two cycles. Thermal expansion of the blocks was found highly reversible, similar to concrete masonry (CST, 2007). The coefficient of thermal expansion on the Building Blocks was around 600-700 microstrain or (10-6) per 70°C, or about 8.6-10x10-6 /°C. This coefficient should have affected by the size of the samples (100x100x65mm) and the level of curing regime applied. The coefficient is comparable to the coefficients of expansion of concrete masonry units, i.e.7.2 to 9.0x10-6/°C (Drysdale et al., 1994). Creep Performance

After the expansion test with results as in Figure 7 (after the volume of the samples stable), the samples were then tested for creep at the same environment. The stress applied was 1 MPa. This stress is commonly applied in masonry. The asphalt used was of 60/70 pen, and the content was varied and gave optimum at 5.5%. At this stage the curing regime applied was 200 °C for 24 hours as shown in Figure 11. Then at optimum asphalt content, the heat curing regime applied was 200 ° were tested C for 4, 8, 16, and 24 hours. The samples at dry condition and after soaking in water for 24 hours. The results are given in Figure 12. Some additional properties of the samples are: 1 creep strain = total strain–elastic strain

shrinkage or expansion. * the samples were tested for creep after the

expansion stabled (at zero expansion). The creep test results indicated that the all of

0 1 2 3 4 5 6 7

200

0

-200

-400

-600

-800

sp1 sp2 avg 300 200 100

0 -100 -200 -300 -400

0 10 20 30 40 50 60 70

sp1 sp2 avg

53


0 4 8 12 16 20 24

5

4

3

2

1

Comp Level (MPa)

Total Strain (µε)

Elastic Strain (µε)

Creep Strain1

(µε)

Exp. at creep test*

(µε)

1 173.25 54.45 118.8 *

2 143.55 49.5 94.05 *

4 113.85 34.65 79.20 *

the samples gave creep strain at least equal of concrete block commonly used in the UK (Sear, 2005, BS6073 1981). In order to ensure a better deformation resistant, creep strain of less than 100 microstrain is recommended (Tapsir, 1985), therefore compaction level of at least 2MPa is suggested.

Figure 10. The creep strain of the samples compacted at different compaction level

hammer RECENT EXPERIMENT Table 3. Creep performance of the CDWblock

samples

The author had carried out recent experiment on masonry block unit using locally available construction demolition material in Bali. After carrying out trials, the aggregate particles composition was determined to consists of 40% coarse particles (14-2.36) mm, 50% fine particles (2.36-0.075) mm and 10 % filler (passing 0,075mm). Waste concrete was used for the coarse particles. The fine particle

portion consisted of 25% waste concrete, 50 % concrete blocks, and 25 % clay brick. Rice husk ash was used as the filler. Compaction was carried out by applying 50 blows Marshall experiments in order to evaluate specific creep, i.e. creep strain per MPa unit stress. The creep test results are shown in Figures 10 and summarized in Table 3. Given in Table 3. The average porosity, water absorption, and initial rate of suction (IRS) of the samples curedat 200°C for 4-24 hours is summarized in Table 4.

Figure 11. Compressive strength of the

samples at varied asphalt (60/70 pen) content, cured at 200 °C for 24

Figure 12. Compressive strength of the sample at 5,5% optimum asphalt content, cured at 200

°C with varied curing duration In line with the Indonesian National Standard the (SNI) for masonry concrete block/unit, the minimum compressive strength is 25 kg/cm2 or 2.5 MPa (SNI 03-0348, 1989). Referring to the results in Figures 11 and 12,

0 4 8 12 16 20 24

160

120

80

40

0

1 4 4,5 5 5,5 6 6,57

5

4

3

2

sp1 sp2 avg

sp 1 sp 2

54


the compressive strength of the block unit well meets the standard. There is no requirement on porosity, water absorption and IRS. The IRS values in Table 3 suggest the use of mortar with higher water content, as the unit gave relatively high IRS and water absorption. CONCLUSIONS Some conclusion can be withdrawn from the investigation, i.e.:

1. The CDW materials were found very suitable for producing asphalt bound construction demolition waste masonry block (CDWblock).

2. Compaction level minimum of 2 MPa and the curing regime applied (200 °C for 24 hours) were found to give satisfactory performances (satisfy compressive strength and creep performance). The performances of the CDWblocks were found at least equal to the concrete blocks commonly used in the United Kingdom (UK).

3. The volume stability of the CDWblock is affected by relative humidity (RH). Higher RH environment tends to cause higher excuring duration (hour) pansion, and vice versa.

4. The volumetric movement of the CDWblocks due to environment moisture is not fully reversible, but highly reversible due to thermal exposure.

5. The compressive strength of the masonry unit recently tried using locally available construction demolition material,well meet the minimum Indonesian standard of 25 kg/cm2.

ACKNOWLEDGMENT The author would like to express his gratitude and appreciation to The School of Civil Engineering Leeds University-United Kingdom, the Engineering and Physical Sciences Research Council (EPSRC)-United Kingdom for providing funds, and to his research supervisors Dr. S. E. Zoorob and Dr. J. P. Forth who had given guidance

during the author’s post doctoral research in the year of 2004-2007. REFERENCES

British Standard Institution BS6073, Part 1. (1981). “Precast Concrete Masonry Unit”. Specification for Clay Bricks, British Standard (BS) 3921.

British Standard BS 812. (1995). “Testing Aggregates, Part 2: Methods of Determination of Density”.

British Standard (BS) 594-1. (2005). “Hot Rolled Asphalt for Roads and Other Paved Areas-Part 1: Specification for Constituent Materials and Asphalt Mixtures”.

CEB-FIP. (1993). “CEB-FIP Model Code 1990”. Comite Euro-International du Beton, Netherlands.

Chen, W. F. (1982). “ Plasticity of Reinforced Concrete”. McGraw-Hill International Company, Washington, D.C.

Craighill, A. and Powell, J. C. (2006). “A Lifecycle Assessment and Evaluation of Construction and Demolition Waste”. Research Project Report funded by the Engineering and Physical Sciences Research Council (EPSRC)-UK, Centre for Social and Economic Research on the Global Environment University of East Anglia and University College London, ISSN 0967-8875.

CST/Cast Stone Institute. (2007). “Design Tips – Technical Bulletin #52 Allowing for Movement of Masonry Materials”. At http://www.caststone.org/bulletins/tb_52.pdf ; accessed on 24 October 2007.

De Witte, F. C. and Kikstra, W. P. (2008). "DIANA Finite Element Analysis, User's Manual: Release 9.3”. TNO Building and Construction Research, Delft, The Netherlands.

DoE/Department of the Environment. (1994). “Guidelines for Aggregates Provision in England”. Minerals Planning Guidance Note 6, HMSO, London.

Domone, P. L. (1994). “Part Three: Concrete, in Construction Materials Their Nature

55


and Behaviour”. Edited by J.M. Illston, Second Edition, E&F Spon, p. 87-195.

Drysdale, R. G., Hamid, A. A., and Baker, L. R. (1994). “Masonry Structures: Behavior and Design”. Prentice Hall.

Forth, J. P., Zoorob, S. E., and Thanaya, I. N. A. (2006). “Development of Bitumenbound Waste Aggregate Building Blocks”. Proceedings of the Institution of Civil Engineers, Construction Materials 159, Issue CMI, Thomas Telford-London, February 2006, p. 23-32.

Forth, J. P., Zoorob, S. E., and Thanaya, I. N. A. (2008). “The Time-Dependent Performance of Bitublock Single Leaf Masonry”. Proceedings of 14th International Brick and Block Masonry Conference, Manly Pacific Hotel, Sydney, Australia.

Goto, Y. and Joh, O. (2004). "Shear Resistance of RC Interior Eccentric Beam-Column Joints”. 13th World Conference on Earthquake Engineering (13 WCEE), Vancouver, Canada, August 1-6, 2004, p. 13.

Indonesian National Standard. (2002). “Earthquake Resistant Design Standard for Building Structures”. SNI 03-1726-2002, National Standards InstituteHumphreys, H., Partners Consulting Engineer. (1994). “Managing Demolition and Construction Wastes”. Report of the study on the Recycling of Demolition and Construction Wastes in the UK, Report for the Department of the Environment, HMSO, London.

Naito, C. J., Mosalam, K. M., and Moehle, J. P. (2000). “Evaluation of Reinforced Concrete Bridge-Joints”. 12th World Conference on Earthquake Engineering, Auckland, New Zealand, p. 1-8.

Neville, A. M. (1991). “Properties of Concrete”. 3rd Edition, Longman Scientific and Technical, p. 373375.

New Zealand Standard. (1995). “Concrete Structures Standard and Commentary NZS 3101, Part I and Part II”. SANZ (Standard Association of New Zealand), Wellington.

Park, R. (2002). “A Summary of Results of Simulated Seismic Load Tests on Reinforced Concrete Beam-Column Joints, Beams and Columns with Substandards Reinforcing Details”. Journal of Earthquake Engineering, Vol. 6, No. 2, 147174.

Paulay, T. (1989). “Equilibrium Criteria for Reinforced Concrete Beam-Column Joints”. ACI Structural Journal, Vol. 86 No. 6 (November-December), p. 635-643.

Sear, L. (2005). “Blocks Made in UK – 100 % Utilization of Bottom Ash”. ECOBA Conference Paper, United Kingdom Quality Ash Association (UKQAA) at http://www.ukqaa.org.uk/Papers/ECOBA%20Conference%20FBA%20in%20block%20-%20Sear.pdf; accessed on 08 February 2006.

Soutsos, M. N., Jones, N., and Millard, S. G. (2004). “Precast Concrete Building Blocks Made with Aggregates Derived From Construction and Demolition Waste”. Proceedings on Use of Recycled Materials in Buildings and Structures, Barcelona.

Standar Nasional Indonesia (SNI) 030348. (1989). “Mutu dan Cara Uji Bata Beton Pejal”.

Tajima, K., Mishima, T., and Shirai, N. (2004). “3-D Finite Element Cyclic Analysis of RC Beam/Column Joint using Special Bond Model”. 13th World Conference on Earthquake Engineering (13 WCEE), Vancouver, Canada, August 1-6, 2004, p. 15.

Tapsir, S. H. (1985). “Time-Dependent Lost of Post Tensioned Diaphragm and Fin Masonry Wall”. Ph.D Thesis, Civil Engineering Department, University of Leeds, UK.

Teguh, M. (2007). “Seismic Performance of Prestressed Concrete Pile-to-Pile Cap Connections”. Wednesday 17th May 2006, ISBN CD-Rom 0-Ph.D Thesis, Department of Civil and Environ-9553490-0-1 , 978-0-9553490-0-3, Edited by Dr. Lindon Sear, Paper Ref. A16.

Teguh, M., Duffield, C. F., Mendis, P. A., and Hutchinson, G. (2006). “Seismic

56


Performance of Pile-to-pile Cap Connections: An Investigation of Design Issues”. Electronic Journal of tructural Engineering (eJSE), Vol. 6, p. 8-18.

Thanaya, I. N. A., Forth, J. P., and Zoorob, S. E. (2006). “Incorporation of Fly Ash and Furnace Bottom Ash in Bitublock”. Proceedings of AshTech 2006, International Coal Ash Technology Conference, The Birmingham Hippodrom Theatre, Birmingham, West Midlands, UK, Sunday 15th the Environment, Arup Economics and Planning, HMSO, London.

Vekey de, R. C. (2001). “Brickwork and Blockwork, Construction Materials, Their nature and Beahviour”. Third Edition, Editted by J. M. Illston and P. L. J. Domone, Spon Press, London and New York, p. 288.

Whitbread, M., Marsay, A., and Tunnell, C. (1991). “Occurrence and Utilisation of Mineral and Construction Wastes”. Report for the Department of Metal Engineering, University of Melbourne, Australia.

Whiteoak, D. (1991). “The Shell Bitumen Handbook”.

57


Pengembangan Peta Hazard Kegempaan untuk Pulau Jawa dan Perbedaannya dengan Hasil-Hasil Studi Terdahulu

Lalu Makrup, Sarwidi, dan Susilo

Staf Pengajar Program Studi Teknik Sipil, UII Yogyakarta dan Alumni Program Magister Teknik Sipil UII Yogyakarta (Pegawai Pemkab Kulon Progo), email: [email protected], [email protected]

Abstract: The island of Java with high population density, tectonically it is an area with a fairly high seismicity in Indonesia. Seismic hazard analysis for Java was done using the theory of total probability. Results of calculation and development of PSHA map show that seismic hazard with risk exceeded 10% during 50-year building useful life or nearly 500-year return period of earthquake show that maximum soil acceleration at bedrock is between 0.1 g to more than 0.5 g. Range of this acceleration is similar to a study conducted by Irsyam et al., (2008) and ISO 2010 and is different from SNI 2002. However, different patterns of acceleration map exist between calculated and others. Java earthquake map is also developed with return period of 500 years (10% probability of exceedance in 50 years) and 2,500 years (2% probability of exceedance in 50 years) with spectra period 0.0 seconds, 0.2 seconds and 1.0 second. Keywords: Hazard analysis, probabilistic seismic hazard, peak ground acceleration, and Java

earthquake map. PENDAHULUAN Pulau Jawa dengan jumlah penduduk dan tingkat kepadatannya yang tinggi secara tektonik merupakan kawasan dengan tingkat aktivitas kegempaan yang cukup tinggi di Indonesia. Kondisi ini disebabkan karena pulau ini berdekatan dengan zona subduksi selatan Jawa. Disamping rawan gempabumi akibat aktivitas tumbukan lempeng tektonik pada zona subduksi, Pulau Jawa juga rawan gempabumi akibat aktivitas sesar-sesar lokal di daratan. Kondisi tektonik semacam ini menjadikan Pulau Jawa sebagai kawasan seismik aktif yang kompleks. Salah satu upaya untuk mengurangi dampak akibat bencana alam gempabumi adalah melalui pengembang an peta bahaya akibat gempa bumi. Peta bahaya (hazard) kegempaan atau disingkat peta gempa dikembangkan berdasarkan analisis probabilitas bahaya kegempaan (probabilistic seismic hazard analysis). Peta ini merupakan peta percepatan tanah maksimum di batuan dasar (peak ground acceleration) yang dapat digunakan untuk perencanaan struktur bangunan tahan gempa.

Peta hazard kegempaan dibuat berdasar kan data catatan kejadian gempa yang pernah terjadi sebelumnya dan diolah dengan prinsip-prinsip statistik dalam analisis seismic hazard probabilistik (PSHA). Sejak diterbitkannya peta wilayah gempa pada tahun 2002, sampai dengan saat ini di Indonesia telah banyak terjadi gempa dengan kekuatan signifikan yang menimbulkan kerusakan dan korban jiwa. Sebagai contoh adalah gempa Yogya Mei 2006 dan Tasikmalaya September 2009. Gempa Yogya M6.3 memakan ribuan korban jiwa dan harta benda di Yogya dan Jawa Tengah, sedang gempa Tasikmalaya M7.3 dengan puluhan korban jiwa dan harta benda. Dalam studi ini dilakukan analisis hazard gempa untuk Pulau Jawa dengan bantuan Program PSHA Seismic Risk Model (Makrup, 2009). Model sumber gempa yang digunakan adalah model tiga dimensi dikombinasi dengan model sumber seismisitas background. Sebelum ini sekitar tahun 2002 kebawah di Indonesia, perhitungan hazard gempa dilakukan dengan menggunakan program komputer yang mengaplikasikan model sumber gempa area (dua dimensi). Sebagai

58

Makrup, dkk: Pengembangan Peta Hazard Kegempaan untuk Pulau Jawa dan ... Terdahulu

contoh adalah peta gempa dalam SNI 2002. Peta gempa ini dikembangkan oleh 4 tim yang berbeda. Yang pertama adalah peta gempa yang dikembangkan oleh Shah dan Boen, (1996) dari kosultan swasta, menggunakan program GSHAP (The Global Seismic Hazard Assessment Program), yang kedua oleh Engkon Kertapati, (1999) dari pusat geologi, Bandung menggunakan program EQRisk (McGuire, 1976), yang ketiga oleh Nayoan, (1999) dari Direktorat Jendral Pengairan Depatemen Pekerjaan Umum menggunakan program GSHAP dan yang keempat Firmansyah dan Irsyam (1999) dari ITB Bandung menggunakan program EQRisk (McGuire, 1976). Karena tidak ada kesepakatan tentang peta gempa yang mana yang akan digunakan dalam SNI 2002, maka nilai numerik dari keempat peta tersebut lalu dirata-rata, dan peta gempa dengan nilai rata-rata inilah yang digunakan sebagai peta gempa untuk SNI 2002 (Gambar 1).

Gambar 1. Peta gempa PGA Jawa dengan level hazard 10 % probabilitas terlampaui

dalam 50 tahun atau mendekati perioda ulang gempa 500 tahun (SNI, 2002)

Peta gempa Jawa dengan pemodelan sumber gempa tiga dimensi yang digabungkan dengan model sumber gempa sismisitas background dua dimensi, telah dikembangkan oleh Irsyam et al., (2008). Program yang digunakan dalam analisis adalah program EZ Frisk (McGuire, 2005) dengan tidak mengakomodasi fungsi atenuasi terbaru NGA (The Next Generation Attenuation) karena pada saat pengembangan peta ini EZ Frisk belum mengakomodasi NGA (Gambar 2). Peta gempa Jawa dalam rencana peta gempa SNI 2010 dikembangkan dengan pemodelan

sumber gempa tiga dimensi dan digabung dengan model sumber gempa sismisitas background Frankel, (1995).


dalam 50 tahun atau mendekati perioda ulang gempa 500 tahun (Irsyam et al., 2008)

Peta ini dikembangkan menggunakan program USGS (Asrurrifak, 2010) dengan mengakomodasi fungsi atenuasi terbaru NGA (Gambar 3).


dalam 50 tahun atau mendekati perioda ulang gempa 500 tahun (SNI, 2010)

Model sumber gempa tiga dimensi yang digabung dengan model sumber gempa sismisitas background dua dimensi, dan fungsi atenuasi generasi baru (NGA) diakomodasi dalam program kegempaan SR-Model (The Seismic Risk Model, Makrup, 2009). Program ini digunakan dalam PSHA untuk studi mengembangkan peta gempa Jawa dalam paper ini. Hasil analisis akan diberikan pada sebelum bagian akhir dari paper ini.

59


ANALISIS SEISMIK HAZARD PROBABILISTIK Konsep probabilistik telah memberi ruang bagi ketidakpastian dari site, lokasi, dan laju keberulangan gempa serta variasi karakteristik gerakan tanah secara eksplisit dapat dipertimbangkan dalam melakukan evaluasi terhadap hazard kegempaan. Analisis hazard kegempaan probabilistik (PSHA) menyediakan kerangka yang memungkinkan ketidakpastian dapat diidentifikasi, kuantifikasi, dan dikombinasikan dalam cara-cara yang rasional untuk menyediakan gambaran yang lebih lengkap tentang hazard kegempaan (Kramer, 1996). Tahapan analisis dalam hazard kegempaan dengan metode Probabilistic Seismic Hazard Analysis (PSHA) secara tipikal diuraikan menjadi empat tahap proses (Reiter, 1990) yang dapat dijelaskan sebagai berikut ini.

1. Tahap pertama adalah identifikasi dan karakterisasi sumber gempa, termasuk didalamnya adalah karakterisasi distribusi probabilitas dari lokasi rupture yang berpontensi menyebabkan gempa pada sumber. Dalam kebanyakan kasus, diterapkan distribusi probabili tas yang sama untuk masing-masing zona sumber. Hal ini secara tidak langsung menyatakan bahwa gempa mungkin sama-sama akan terjadi pada setiap titik dalam zona sumber gempa. Distribusi ini, dikombinasikan dengan bentuk geometri sumber untuk mendapatkan distribusi probabilitas yang sesuai dengan jarak sumber ke site.

2. Langkah berikutnya adalah karakterisasi seismisitas atau distribusi perulangan kejadian gempa. Hubungan empiris perulangan kejadian gempa (recurrence relationship), yang meng ekspresikan laju rata-rata (average rate) suatu gempa dengan magnitude yang berbeda akan terlampaui. Hubungan ini digunakan untuk meng karakterisasi seismisitas masing-masing zona sumber gempa. Hubungan empiris ini dapat mengakomodasi magnitude maksimum gempa.

3. Gerakan tanah yang terjadi disuatu lokasi

akibat adanya gempa dengan besar gempa berapapun dan lokasi kejadian dimanapun dalam masing-masing zona sumber gempa, dapat ditentukan dengan menggunakan predictive relationships.

4. Langkah terakhir adalah mengkom bina-sikan ketidakpastian dari lokasi gempa, magnitude gempa dan prediksi parameter gerakan tanah untuk mendapatkan probabilitas dimana parameter gerakan tanah akan terlampaui selama perioda waktu tertentu.

Untuk satu kejadian gempa tertentu, probabilitas satu parameter gerakan tanah Y akan melampaui satu nilai y* tertentu dapat dihitung menggunakan teorema probabilitas total (Kramer, 1996). Bentuk fungsi dari teorema probabilitas total (McGuire, 1976) adalah:

drdm)r(f)m(fr,m|*yYP*yYP RM (3)

dengan r,m|*yYP diperoleh dari hubungan prediktif probabilitas parameter Y yang akan melampaui nilai y* pada lokasi yang ditinjau untuk kejadian gempa dengan magnitude M dan jarak hiposenter R, sedangkan fM(m) dan fR(r) adalah fungsi kerapatan probabilitas (PDF) masing-masing untuk magnituda dan jarak. Untuk memperoleh nilai tahunan rata-rata ter lampaui ( = annual rate of exceedance) persamaan (3) dapat ditulis menjadi :

drdmrfmfrmaAPa RMA )()(,|)( (4)

dengan adalah rate dari kejadian gempa. Penyelesaian persamaan (4) dalam domain sumber gempa sangat rumit, sehingga secara umum penyelesaian tersebut di tuangkan dalam bentuk software. METODE Tempat dan Data

Data gempa yang digunakan dalam penelitian ini adalah seluruh kejadian gempa tercatat yang pernah terjadi di Pulau Jawa dan sekitarnya dengan magnitude minimal Mw = 5 dan kedalaman hipocenter gempa maksimal

60


250 km. Data gempa yang dipakai dalam penelitian ini diperoleh dengan men-download dari Advanced National Seismic System (ANSS) dan National Earthquake Information Center–United States Geological Survey (NEIC-USGS) dengan rentang waktu sampai dengan kejadian gempa bulan Desember 2010, meliputi data episenter gempa yang terjadi di Pulau Jawa dan sekitarnya yaitu yang berada pada rentang koordinat 100° – 120° bujur timur dan 2° – 13° lintang selatan. Konversi Skala Magnitude

Supaya didapatkan hasil analisis seismic hazard yang akurat maka dalam analisis seismic hazard diperlukan data kejadian gempa yang memiliki skala magnitude yang sama, dalam hal ini adalah skala magnitude moment Mw. Konversi skala magnitude ke dalam skala momen magnitude untuk gempa-gempa di wilayah Indonesia dilakukan dengan menggunakan persamaan yang diperkenalkan oleh Hendriyawan (2001), sebagai berikut:

Mw = 0,99MB + 0,253 (5)

Mw = 0,08MS2 + 0,04MS + 3,01 (6)

Dengan menggunakan persamaan (5) dan (6), data gempa yang mempunyai skala magnitude selain Mw kemudian dikonversi ke dalam skala magnitude Mw sehingga diperoleh skala magnitude yang seragam. Pemilihan Gempa Utama

Pemisahan gempa utama dari gempa-gempa ikutan (foreshock dan aftershock) dilakukan agar diperoleh data gempa dengan kejadian yang independen. Kejadian gempa dependen atau gempa ikutan, seperti foreshock dan aftershock yang terjadi dalam suatu rangkaian kejadian gempa harus diidentifikasi dan di lepaskan dari katalog sebelum katalog tersebut digunakan dalam analisis. Dalam studi ini, untuk mengeliminasi foreshock dan aftershock dari katalog gempa digunakan kriteria Gardner dan Knopoff (1974). Hasil dari proses analisis pemilihan gempa utama (dependency analysis) adalah data gempa utama yang kemudian digunakan pada proses analisis selanjutnya.

Analisis Kelengkapan Data Gempa

Analisis kelengkapan data gempa dilakukan untuk mengetahui kelengkapan data gempa dalam rentang waktu tertentu. Kelengkapan data diperlukan mengingat hasil proses analisis probabilistik yang akan dilakukan sangat tergantung dari data tersebut. Analisis seismik hazard memerlukan data gempa dalam kurun waktu tertentu, dimana kejadian gempa independen dalam rentang magnitude tertentu dapat dikatakan lengkap dalam suatu katalog gempa. Data pencatatan kejadian gempa historis untuk kejadian-kejadian gempa besar lebih lengkap dibanding kejadian-kejadian gempa kecil. Hal ini disebabkan pada masa lalu jumlah alat pencatat gempa belum banyak dan teknologinya belum maju seperti sekarang, sehingga alat-alat tersebut hanya mencatat kejadian-kejadian gempa besar. Jika data yang tidak lengkap digunakan dalam analisis resiko gempa, maka hasil yang didapat akan terlalu kecil (underestimated) untuk gempa-gempa kecil dan terlalu besar (overestimated) untuk kejadian gempa besar (Aldiamar, 2007). Identifikasi dan Pemodelan Sumber Gempa

Sumber gempa yang digunakan dalam pembuatan peta hazard kegempaan Pulau Jawa meliputi zona sumber gempa subduksi dan sumber gempa sesar di daratan. Secara umum digunakan tiga zona sumber gempa dalam studi ini yaitu (1) zona megatrust yang merupakan sumber gempa dangkal pada daerah subduksi dengan kedalaman maksimum 50 km, (2) zona benioff yang merupakan sumber gempa pada daerah subduksi dengan kedalaman antara 50 km hingga 250 km, dan (3) zona shallow crustal pada daerah transform fault. Bentuk zona ini adalah patahan yang biasa disebut sesar (fault) dengan kedalaman hingga 20 km. Sumber gempa subduksi dibagi menjadi dua yaitu zona subduksi megatrust dan zona subdaksi benioff. Kejadian gempa akibat thrust fault, normal fault, reverse slip dan strike slip yang terjadi sepanjang pertemuan lempeng dapat diklasifikasikan sebagai zona subduksi. Zona subduksi megatrust terletak pada pertemuan antar lempeng tektonik yaitu pertemuan antara

61


lempeng samudera yang menunjam masuk ke bawah lempeng benua dan memiliki sudut penunjaman yang kecil (landai) pada kedalaman sampai dengan 50 km, sedang kan zona subduksi benioff adalah kelanjut an dari zona subduksi megatrust dengan kemiringan sudut penunjaman yang relatif lebih besar. Tabel 1. Zona sumber gempa Pulau Jawa dan

sekitarnya

No Zona Sumber Gempa A Megatrust

1. SouthernSumatra Megatrust 2. Jawa-1 Megatrust 3. Jawa-2 Megatrust 4. NusaTenggara Megatrust

B Benioff 1. SouthernSumatra Benioff 2. Jawa-1 Benioff 3. Jawa-2 Benioff 4. NusaTenggara Benioff

C Shallow Crustall 1. Sesar Sumatra 2. Sesar Cimandiri 3. Sesar Baribis 4. Sesar Bumiayu 5. Sesar Semarang 6. Sesar Sampanahan 7. Sesar Opak 8. Flores Back Arc

Gambar 4. Zona sumber gempa untuk Pulau Jawa

Gempa pada fault umumnya terjadi pada zona dangkal dengan mekanisme strike-slip,

reverse, oblique, atau normal. Oleh karena itu zona fault hanya sampai kedalaman 20 km. Zona sumber gempa subduksi dan fault yang secara umum sudah dikenal sebagai sumber gempa aktif di Jawa dapat dilihat pada Tabel-1 dan Gambar 4. Atenuasi

Salah satu model fungsi yang harus tersedia dalam PSHA adalah fungsi atenuasi. Fungsi atenuasi merupakan for mula yang menunjukkan hubungan antara parameter gerakan tanah misalnya percepatan (a), magnitude (M), jarak (R) dan kondisi tanah site (fs) dari kejadian gempa pada suatu wilayah. Fungsi atenuasi kebanyakan dikembangkan di Amerika Serikat dan sangat sedikit dari tempat lain. Oleh karena itu hingga saat ini belum ada fungsi atenuasi yang dihasilkan secara spesifik bagi wilayah Indonesia karena kurangnya data-data yang dibutuhkan untuk menghasilkan fungsi atenuasi tersebut. Fungsi atenuasi yang digunakan di Indonesia diambil dari negara lain. Dalam penggunaan fungsi atenuasi ini diasumsi kan bahwa kondisi tektonik dan site di Indonesia mirip dengan kondisi tektonik dan site dimana fungsi atenuasi itu dikembangkan. Dalam studi ini fungsi atenuasi yang digunakan untuk zona sumber gempa subduksi adalah fungsi atenuasi Youngs (1997) dan fungsi atenuasi Atkinson – Boore (2003). Sedangkan zona fault digunakan fungsi atenuasi Boore, Joyner dan Fumal (1997) dan fungsi atenuasi Boore – Atkinson (2006) NGA. HASIL ANALISIS Perhitungan dengan PSHA dilakukan atas dasar level hazard 10% dan 2% probabili tas terlampaui dalam jangka waktu masa guna bangunan 50 tahun atau masing-masing mendekati periode ulang gempa 500 tahun 2500 tahun. Perioda spektra yang ditetapkan dalam perhitungan adalah 0 detik (PGA), 0.2 detik dan 1.0 detik. Hasil perhitungan dengan periode ulang gempa mendekati 500 tahun menunjukkan bahwa percepatan tanah maksimum pada

62


batuan dasar untuk Pulau Jawa adalah antara 0.1g hingga lebih dari 0.5g. Sedangkan untuk periode ulang gempa mendekati 2500 tahun menunjukkan bahwa percepatan tanah maksimum pada batuan dasar berada diantara 0.15g hingga lebih dari 0.8g.

Gambar 5. Peta gempa PGA Jawa dengan perioda ulang gempa mendekati 500 tahun

Gambar 6. Peta gempa Jawa dengan perioda spektra 0.2 detik dan perioda ulang gempa

mendekati 500 tahun


mendekati 500 tahun

Gambar 8. Peta gempa PGA Jawa dengan perioda ulang gempa mendekati 2500 tahun

Hasil pengembangan peta gempa untuk pulau Jawa adalah berupa peta kontur percepatan gerakan tanah dibatuan dasar dengan periode ulang gempabumi 500 tahun dan 2500 tahun dengan periode spektra 0 detik (PGA), 0.2 detik dan 1.0 detik. Peta hasil analisis dapat dilihat dalam Gambar-5, 6 dan 7 untuk perioda ulang gempa 500 tahun serta Gambar-8, 9 dan 10 untuk perioda ulang gempa 2500 tahun. PEMBAHASAN Yang akan dibahas disini adalah perbedaan peta gempa PGA Jawa dengan perioda ulang 500 tahun antara peta hasil studi dan peta hasil studi terdahulu. Secara visual perbedaan ini ditampilkan dalam bentuk gambar.


mendekati 2500 tahun


mendekati 2500 tahun Gambar 11, 12 dan 13 adalah tampilan perbedaan antara peta gempa pulau Jawa hasil studi dengan peta gempa pulau Jawa dari SNI 2002, Irsyam et al., (2008) dan SNI 2010. Peta gempa PGA Jawa hasil studi dan SNI 2002 (Gambar 11) mempelihatkan perbe daan yang relatif kontras terutama pada wilayah Jawa yang ditempat tersebut dinyatakan ada

63


fault. Pada gambar peta gempa hasil studi terlihat adanya konsentrasi percepatan pada dari wilayah Jawa Tengah sampai wilayah Jawa barat. Sebagai contoh pada lokasi fault Cimandiri (Gambar 11A) terlihat bahwa konsentrasi percepatan gerakan tanah ≥ 0.3g.

Bila dibandingkan dengan peta gempa PGA Jawa SNI 2002 maka terlihat dalam peta tersebut seolah-olah tidak ada konsentrasi percepatan pada wilayah tertentu yang dianggap sebagai pusat sumber gempa. Inilah perbedaan nyata yang dapat diperlihatkan

Gambar 11. Perbedaan peta gempa PGA Jawa hasil studi dan SNI 2002

Gambar 12. Perbedaan peta gempa PGA Jawa hasil studi dan Irsyam et al., (2008)

64


secara visual dari peta hasil studi (Gambar 11A) dan peta SNI 2002 (Gambar 11B). Perbedaan lain yang juga dapat disampaikan disini adalah bahwa peta Gambar 11A memperlihatkan percepatan gerakan tanah berkisar antara 0.1g sampai dengan 0.5g, sedang peta SNI2002 berkisar antara 0.03 sampai dengan 0.3g. Perbedaan ini terjadi barangkali karena perbedaan model sumber gempa yang digunakan yaitu pada peta hasil studi menggunakan model sumber gempa tiga dimensi, sedang pada SNI 2002 menguna kan model sumber gempa dua dimensi. Begitu juga mengenai penggunaan fungsi atenuasi. Dalam pengembangan peta gempa hasil studi (Gambar 11A) sudah digunakan fungsi atenuasi mutakhir NGA 2006/2007 sedang pada peta SNI 2002 (Gambar 11A) tentunya menggunakan fungsi atenuasi di bawah tahun 2000. Walaupun di tempat-tempat tertentu terlihat ada kesamaan per cepatan tanah gempa antara Gambar 11A dan 11B namun luasan yang dicakup berkaitan dengan kesamaan ini tidak terlalu signifikan. Perbedaan antara peta gempa PGA Jawa hasil studi dengan hasil Irsyam et al. (2008) dapat dilihat pada Gambar 12. Sebagaimana perbedaan antara peta hasil studi dan SNI 2002

(Gambar 11), maka pada Gambar 12 masih tetap memperlihatkan bahwa perbeda an yang relatif kontras terjadi pada posisi keberadaan fault yang digunakan dalam analisis. Sebagai contoh pada posisi fault Semarang dan Yogya yang digunakan untuk mengembangkan peta Gambar 12C terlihat adanya konsentrasi percepatan gerakan tanah, sementara dalam Gambar 12D ditempat tersebut tidak terjadi konsentrasi percepatan karena pengembangan peta Gambar 12D tidak menggunakan fault Semarang dan Yogya sebagai dasar perhitungan. Di Jawa bagian barat dalam Gambar 12C terlihat adanya konsentrasi percepatan pada posisi fault Cimandiri dan Baribis. Sementara dalam Gambar 12D tidak terlihat hal yang serupa. Memang ada diperlihatkan gambar konsentrasi percepatan menggelembung pada Jawa Barat Bagian Timur, yang itu mungkin karena penggunaan fault Baribis, namun tetap saja penggelembungan ini tidak memperlihatkan pola yang sama dengan konsentrasi percepatan yang ada pada Gambar 12C. Adanya perbedaan peta gempa PGA Jawa antara peta hasil studi (Gambar 12C) dan hasil dari Irsyam et al. (2008) (Gambar 12D) barang kali disebabkan terutama oleh adanya penggunaan

Gambar 13. Perbedaan peta gempa PGA Jawa hasil studi dan SNI 2010

65


fault yang berbeda dan panjang fault yang berbeda walaupun juga ada fault yang sama. Disamping itu juga oleh penggunaan fungsi atenuasi yang berbeda. Peta Gambar 12C melibatkan fungsi atenuasi NGA sebagai dasar pengembangan sedangkan peta Gambar 12D tidak. Secara keseluruhan peta Gambar 12C dan Gambar 12D memberikan rentang per cepatan gerakan tanah yang sama yaitu berkisar antara 0.1g sampai dengan 0.5g, namun percepatan gerakan tanah yang sama hanya terjadi pada tempat-tempat tertentu saja, sehingga perbedaan antara kedua peta dalam Gambar 12 tetap masih cukup signifikan. Perbedaan antara peta gempa PGA Jawa hasil studi dengan hasil SNI 2010 dapat dilihat pada Gambar 13. Sebagaimana perbedaan antara peta hasil studi dan SNI 2002 (Gambar 11) serta antara peta hasil studi dan Irsyam et al., (2008) maka perbedaan antara peta hasil studi dan peta SNI 2010 memperlihatkan pola perbedaan yang tidak terlalu jauh berbeda dengan apa yang diperlihatkan sebelumnya. Pada peta SNI 2010 Gambar 13F terlihat konsentrasi percepatan pada fault Cimandiri memiliki pola konsentrasi yang berbeda dengan fault Cimandiri dalam peta hasil studi Gambar 13E. Begitu juga dengan konsentrasi percepatan pada fault Lembang sekitar Bandung (Gambar 11F) berbeda dengan pola yang terjadi pada Gambar 13E, karena dalam Gambar 13E tidak digunakan fault Lembang. Disekitar Semarang gabungan kontribusi fault Semarang, Pati dan Lasem memperlihat konsentrasi percepatan (Gambar 13F) yang berbeda dengan yang dikontribusi oleh fault Semarang saja pada Gambar 13E. Juga perbedaan semacam ini terjadi juga ditempat-tempat yang lain dalam kedua Gambar 13E dan 13F. Kesamaan percepatan gerakan tanah dapat dijumpai hanya pada tempat-tempat tertentu dan dalam keseluruhannya memliki rentang percepatan yang sama antara Gambar-13E dan 13F yaitu dengan nilai percepatan dari 0.1g sampai dengan 0.5g. KESIMPULAN Dari peta hazard kegempaan Pulau Jawa dengan resiko terlampaui sebesar 10% selama

masa guna bangunan 50 tahun atau setara dengan periode ulang gempa 500 tahun memperlihatkan bahwa percepatan tanah maksimum (PGA) pada batuan dasar untuk Pulau Jawa adalah antara 0.1g hingga lebih dari 0.5g. Hal sama terjadi pada peta gempa PGA Jawa hasil dari Irsyam et al. (2008) dan SNI 2010. Namun hal ini berbeda dengan yang ada pada peta gempa PGA yang ada dalam SNI 2002 dengan percepatan berkisar antara 0.03g sampai dengan 0.3g. Sehingga dapat dinyatakan bahwa peta hasil studi, peta Irsyam et al. (2008) dan SNI 2010 memberikan nilai percepatan tanah PGA dibatuan dasar yang lebih tinggi bila dibanding dengan peta gempa pada SNI-03-1726-2002. UCAPAN TERIMA KASIH Terima kasih kami ucapkan kepada Biro Perencanaan Kerjasama Luar Negeri (BPKLN) Kemendiknas, atas beasiswa yang telah diberikan pada mahasiswa bimbingan kami saudara Susilo. Tesis saudara Susilo telah diterbitkan dalam Jurnal ini yang ditulis secara bersama-sama dengan dosen pembimbing. DAFTAR PUSTAKA Aldiamar, F. (2007). “Analisis Risiko Gempa

dan Pembuatan Respon Spektra untuk Jembatan Suramadu dengan Pemodelan Sumber Gempa Tiga Dimensi”. Thesis, Institut Teknologi Bandung, 2007.

Asrurrifak, M. (2010). “Peta Respon Spektra Indonesia untuk Perencanaan Struktur Bangunan Tahan Gempa dengan Model Sumber Gempa Tiga Dimensi dalam Analisis Probabilistik”. Disertasi, Institut Technologi Bandung, Indonesia.

Atkinson, G. M. dan Boore, D., M. (2003). “Empirical Ground Motion Relations for Subduction-Zone Earthquakes and Their Aplication to Cascadia and Other Region”. Bulletin of the Seimological Society of America, Vol. 93, No. 4, p. 1703-1729.

Boore, D. M., Atkinson, G., M., dan NGA/Next Generation Attenuation. (2007). “Ground Motion Relations for

66


Geometric Mean Horizontal Component of Peak and Spectral Ground Motion Parameters”. PEER Report 2007, Pacific Earthquake Engineering Research Center, College of Engineering University of California, Berkeley, California, USA.

Boore, D. M., Joyner, W. B., and Fumal, T. E. (1997). “Equation for Estimating Horizontal Response Spectra and Peak Acceleration from Western North America Earthquakes: A Summary of Recent Work”. Seismological Research Letters, Vol. 68, No. 1.

Firmansjah, J. dan Irsyam, M. (1999). “Development of Seismic Hazard Map for Indonesia”. Prosiding Konferensi Nasional Rekayasa Kegempaan, Bandung.

Frankle, A. (1995). “Mapping Seismic Hazard in the Central and Eastern United States”. Seismological Research Letters, Vol. 6, No. 4.

Gardner, J. K. L. And Knopoff. (1974). “Is the Sequence of Earthquakes in Southern California, with Aftershocks Removed, Poissonian?” Bull. Seismol. Soc. Am. 64, p. 1363–1367.

GSHAP. (1999). “The Global Seismic hazard Assessment Program (GSHAP) 1992–1999”. Ann. Geofis., p. 957–1230, (summary volume edited by D. Giardini).

Irsyam M., Dangkua D. T., Hendriyawan, Hoedajanto, D., Hutapea B. M., Kertapati E. K., Boen T., dan Petersen, M. D. (2008). “The Proposed Seismic Hazard Maps of Sumatra and Java Islands and Microzonation Study of Jakarta City, Indonesia”. J. Earth Syst. Sci. 117, S2, p. 865–878.

Kertapati, E. K. (1999). “Probabilistic Estimates of Seismic Ground-Motion

Hazard in Indonesia”. Prosiding Konferensi Nasional Rekayasa Kegempaan, Bandung.

Kramer, S. L. (1996). “Geotechnical Earthquake Engineering”., Prentice-Hall, Inc., Upper Suddle River, New Jersey.

Makrup, L. L. (2009). “Pengembangan Peta Deagregasi Hazard Untuk Indonesia Melalui Pembuatan Software Dengan Pemodelan Sumber Gempa Tiga Dimensi”. Disertasi Doktor, Institut Teknologi Bandung, Bandung.

McGuire, R. K. (1976). “FORTRAN Computer Program for Seismic Risk Calculations”. US. Geol. Surv. Open-File Rep. 76-67, p. 90.

Najoan, T. F., Suharjoyo, A., Buditomo, A., Wibowo, S., dan Rizaldi, Nasution, R. B. (1999). “Peta Zona Gempa Indonesia untuk Penentuan Percepatan Gempa Maksimum di Permukaan”. Prosiding Konferensi Nasional Rekayasa Kegempaan, Bandung.

Reiter, L. (1990). “Earthquake Hazard Analysis-Issues and Insights”. Columbiai University Press, New York, p. 254.

Shah, H. C. dan Boen, T. (1996). “Seismic Hazard Model for Indonesia”, Unpublished.

SNI/Standar Nasional Indonesia. (2002). “Tata Cara Perencanaan Ketahanan Gempa untuk Bangunan Gedung (SNI 03-1726-2002)”. Badan Standardisasi Nasional.

Youngs, R. R., Chiou, S. J., Silva, W. J., dan Humphrey, J. R. (1997). “Strong Ground Motion Attenuation Relationships for Subduction Zone Earthquakes”. Seismological Research Letters, 68, p. 58-73.

67


Shoreline Change Model Using the Epr Method and the Simulation of Coastal Vulnerability in Sambas District-West Kalimantan

M. Meddy Danial, Rustamaji, and Eka Priadi

Engineering Faculty of Tanjungpura University, Indonesia, email: [email protected], email: [email protected], email: [email protected]

Abstract: Potential threats to the coastline increase from year to year because of global climate change that increases the temperature of the earth and sea level rise. West Kalimantan has a long coastline of 982 km. Potential length of this beach should be maintained in order not to decrease. District of Jawai and Jawai Selatan is one of the coastal areas, which suffered severe erosion. This study aims to model the pattern of coastline changes and the index of coastal vulnerability to current conditions and future conditions using physical parameters such as waves, tides, bathymetry, land cover and sea level rise (SLR). This research is conducted using satellite image map data is processed with ArcGIS 9.3, Erdas and Autocad. Shoreline changes were analyzed by End Point Rate method. From the results of the shoreline change analysis obtained information that the length of the beach erosion from Jawai Selatan to SB Nilam is about 4 km and the eroded area is 9557.546 m2. Sambas coastal areas in the district Jawai Selatan until the district Jawai are susceptible to the existing conditions (first scenario). Vulnerability score on shoreline occupied criteria "High", where erosion rate on average per year is about 6 m. In the second scenario with tidal elevation is 3 m, wave height is 3 m, and SLR is 0.5 m, obtained results of a vulnerability index is generally dominated by the conditions of very high vulnerability. In the third scenario with tidal elevation is 4 m, wave height is 4 m, and SLR is 1 m, obtained results of a vulnerability index is generally dominated by the conditions of very high vulnerability. Keywords: Shoreline changes, coastal vulnerability index, recent condition, and future model

analysis. INTRODUCTION The coast in Jawai Selatan until Jawai is low-lying area and the topography is quite flat. Prior to 1989, Coastal area of Jawai Selatan was vegetated by mangrove tree and since 1998 had damaged. In 2009, there are several indicators of coastal damage in district Jawai and Jawai Selatan, such as collapse of many coconut tree and shoreline retreat. Erosion problem in Jawai and Jawai Selatan is due to the fishpond conversion, sand mining and position of breakwater which is too close from the shoreline that can cause serious con- sequences of rapidly increasing erosion behind the breakwater. Therefore, it is important to investigate the shoreline changes, erosion rate per year and

how these condition affect the vulnerability of coastal area in district Jawai and Jawai Selatan. OBJECTIVES This study aims to identify and model the pattern of coastline changes between 1992 until 2006. The result can be used to predict the shoreline changes for the future. The second objective is to identify the index of coastal vulnerability againts current conditions and future conditions using physical parameters such as waves, tides, bathymetry, topography, land cover and sea level rise (SLR). This study is expected to provide valuable input to local government in overcoming the problem of coastal vulnerability and coastal hazards.

68

Danial, e.t al.:

STUDY A The study tan coast toPontianak.Selatan antopographylow-lying aCharacteriscomprised can be fouJawai coasIn Jawai damage duand occurlocation of METHOD The methosensing tosatellite imchange canon the Endvulnerabili

: Shoreline Cha

AREA

area is loco SB Nilam The coas

nd SB Nilay in this ararea. stic of of sandy b

und in Ramst is domina

Selatan, tue to mismrs erosion f study can b

DOLOGY

od of this ro extract s

mage using sn be analyzd Point Ratity index (

Figure 1.

ange Model Usin

cated on thm coast of Sst length bam is aboutrea is flat a

Jawai Sebeach and mmayadi beacated by sandthe fishponmanagementbehind brebe seen in F

research is shoreline psoftware Erzed using Ae (EPR) me(CVI) can

. Study Area

ng the Epr Met

he Jawai SeSambas distretween Jawt 31 km. Tand constit

elatan comangrove trech, whereasdy beach. nd is alreat of fishpoeakwater. TFigure 1.

using remposition frordas. ShorelAutocad basethod. Coas

be analyz

a

thod and the Sim

ela- rict wai The tute

oast ees

s in

ady ond The

mote om ine sed stal zed

uswa Re

Deusreqshare19Thto Thus(A EP

ThtioEPshpoim

mulation of Coa

sing ArcGisas roposed b

emote Sens

etection ofsing remote quire sate

horeline poe two satel

992 and 200he softwareobtain the

he flow chsing ErdasAlesheikh, e

PR Method

he next steon from thePR methodhoreline chosition that hmages.

Figure 2. St

astal Vulnerabili

s 9.3 basedby Gornitz

sing

shoreline csensing me

ellite imagosition. In llite image 06. used for saposition of

hart for pros can be t al., 2007).

d

ep is analyse result of re

is used tohange basehave been e

teps of imag

ity in ... Kaliman

d on the eqet al. (1991

hange can bethod. Rem

ge data tthis resea

data, name

atellite dataf the coastlinocessing sa

seen in .

se the shoremote sensio calculate ed on twoextracted fr

ge satellite p

ntan

quation that).

be analysedmote sensingto identifyarch, thereely between

a processingne is Erdasatellite data

Figure 2

reline posi-ing processthe rate of

o shorelinerom satellite

processing

t

d g y e n

g . a 2

- . f e e

69


EPR method is based on the concept of empirical equation shows that Shoreline changes in the future can be determined from the linear relationship between the Shoreline positions and times in the past or in previous years. EPR method can be written in the following equation (Li, R., et al., 1999).

Y = MX + B (1)

where:

Y = position of the coastline changes in year n (the desired end of the year),

m = rate of erosion/year = (Yn-Y1)/(Xn-X1), X = year used as a bench mark (the early

years of the start of the calculation), B = point coordinates are used as a bench

mark in the early years of the start of the calculation.

The Eq. 1 can be modified as follows:

m EPR = (Yn − Y1) - (X n − X1) (2)

where the EPR intercept is:

BEPR = Y1 – mEPR . X1 = Yn – mEPR . Xn (3)

because the line end point can be added at the opposite point of recent point (Xn, Yn), hence the equation can be rewritten into:

YEPR = mEPR . (X – Xn) + Yn (4) Coastal Vulnerability Index

Vulnerability is defined as a specific condition that can increase the likelihood of danger or disaster resulting damage and loss. Coastal vulnerability index is intended to predict

vulnerability due to coastal hazard (Szlafsztein, 2005). The CVI is calculated by using some physi-cal parameters of topography, coastal erosion rate per year, the slope of the shore base, rising sea level, wave height and tidal elevation (Abuodha and Woodroffe, 2006). The parameter of CVI can be modified according to the situation and condition (Cambers, 2001). The CVI can be obtained using the following equation:

. . . . .

There are three scenarios of data input for coastal vulnerability index on the beach Sam-bas that can be shown in Table 2.

Table 2. Three scenarios of data input for

coastal vulnerability index

RESULTS AND DISCUSSION

Coastline Change

Figure 3 show the pattern of shoreline change between 1992 and 2006, and the prediction of shoreline change in the future can be known. From the shoreline change analysis using remote sensing can be seen that between the

Scenario

1 Scenario

2 Scenario

3

Tide level 2 m 3m 4 m

Wave height

2 m 3 m 4 m

Sea level rise

0.1 m 0.5 m 1 m

Very low Low Moderate High Very high

Variable/ Score 1 2 3 4 5Topography (m) ≥ 0.1 20.1 – 30.0 10.1-20.0 5.1-10.0 0 – 5.0

SLR change (m/tahun) ≤ -1.0 -1.0 - 0.9 1.0 – 2.0 2.1 – 4.0 ≥ 4.1

Shoreline change (m) ≥ 2.1 1.0 – 2.0 -1.0 - +1.0 -1.1 - -2.0 ≤ -2.1

Tidal range (m) ≤ 0.99 1.0 – 1.9 2.0 – 4.0 4.1 – 6.0 ≥ 6.1

Annual max wave height (m) 0.0 – 2.9 3.0 – 4.9 5.0 – 5.9 6.0 – 6.9 ≥ 7.0

Table 1. Matrix for determination of coastal vulnerability index (CVI) from Gornitz (1991)

(5)

70

Danial, e.t al.: Shoreline Change Model Using the Epr Method and the Simulation of Coastal Vulnerability in ... Kalimantan

Ramayadi beach (sub distric Jawai Selatan) to Sentebang beach (sub district Jawai) had experienced a severe erosion of the coastline changes in average 6.642 m/year. Coastal areas in the district Jawai had became a place of sand mining, but it was banned and dismissed by the local government since 1997. From the results of the shoreline change analysis are obtained information that the length of the beach erosion from Jawai to SB Nilam is about 4.6 km in segment 1 with an area undergoing erosion about 9557.546 m2. In segment 2, the length of erosion is about 4.3 km with 11310 m2 of land area. The process of sedimentation in segment 1 has the length of sedimentation about 20.6 m and land area has experienced erosion about 133 m21, while in the segment 2 the length of sedimentation is about 5630 m with erosion of land area about 2798 m2. Location of sedimentation process starting from

coordinates 282005, 148207 (in UTM format) to coordinates 281865.2173, 155155.0894. Sediment process occur in coastal area along the coast in the sub district Jawai, from the village Nyirih to the beach in the villages SB Nilam is because there are mangrove trees that grow there. Erosion and sedimentation rates along the coast between Jawai Selatan and Jawai are varying depending on the interaction of waves, currents, defense building (such as breakwater, jetty, etc), mangrove plants, and land use (such as fishpond, sand mining, etc). Rate of erosion in segment 1 and segment 2 can be seen in Table 3.

Modeling of Coastal Vulnerability Index

Coastal vulnerability index (CVI) is the most common worldwide method to determine coastal vulnerability. However, the CVI should not be regarded as a constant value because it can change with time and require a

Figure 3. shoreline change from 1992 to 2006 in Segment 1

Figure 4. Shoreline change from 1992 to 2006 in Segment 2

71


simulation of some data to get a more realistic vulnerability. Scenario 1 has data H = 2 m, tide level = 2 m and sea level rise = 0.1 m. The scenario 1 is simulated as as daily condition or existing condition.

In Figure 5, can be shown the coastal vulnerable index is dominated by moderate and high value. The land area of coastal is dominated by moderate vulnerability value where this area has flat topography and low-lying area, generally. CVI has high vulnerability value on the shoreline due to along the shoreline has experienced a very severe erosion. Along the coastal regions have a very low vulnerability index which are found in high topography, such as rocky coastal. Figure 6 shows the scenario 2 with wave height = 3 m, tidal elevation = 3 m, and SLR = 0.5 m, the CVI, is generally dominated by very high for both shoreline and coastal land. Sea

level rise of 0.5 m or 50 cm is predicted for the year 2050 (Lewsey, 2002).

In Figure 7 is for scenario 3 with wave height = 4 m, tidal elevation = 4 m, and SLR = 1 m, where the CVI score generally dominated by very high both shoreline and coastal land. There are only three CVI value in scenario 3, namely moderate, high and very high. This scenario simulated for the year 2100 where the sea level rise increased to 1 m or 100 cm (Lewsey, 2002). Scenarios 2 and 3 is very useful and important to analyze the impact of storm surge at- tacks that have effects such as sea level rise on coastal areas. Storm surge is the event of sea level rise due to high winds and constant storms accompanied by heavy rains and could damage coastal areas. From the simulation results of both the scenario 2 and scenario 3, the topographical condition in the coastal region along the coast from Jawai Selatan until Jawai is very flat and

Figure 5. Scenario 1


72

Danial, e.t al.: Shoreline Change Model Using the Epr Method and the Simulation of Coastal Vulnerability in ... Kalimantan

low-lying land, so very vulnerable to the threat of large waves and tidal fluctuations.

Verification

In Figure 8, can be seen that this area is the former location of fishpond that have been damaged.

Figure 9 shows that condition of shoreline between beach in Jawai Selatan and Jawai had experienced erosion. Many coconut trees that had collapse and indication of coastal erosion can be seen clearly.

There are two main factors cause erosion in Jawai Selatan. First, land conversion of coastal area to fishponds. Fishpond can cause a large void that susceptible from wave attack. Second, position of breakwater is too close to fishponds area that can cause erosion quickly due to the wave diffraction. Breakwater position should be located more far from shoreline (offshore breakwater) to reduce diffraction effect. Offshore breakwater is more adequate and more safely to protect fishpond from erosion because it can create salient and tombolo behind the structure. From the results of studies in the field, the beach will be safe from erosion if there is a mangrove plant that can reduce wave energy and increase the supply of beach sediment transport. Mangrove trees can be found starting from the village along the coast Nyirih (Jawai district) to the beach in the village of SB Nilam.


Figure 8. Field verification

Figure 9. Field verification

73


CONCLUSIONS From the results of this research, there are several conclusions that can be given as follows:

1. From the results of the shoreline change analysis obtained information that the length of the beach erosion from Jawai Selatan to SB Nilam is 3985.04 m (4 km) and the eroded area is 9557.546 m2 in segment 1. In segment 2, the length of erosion is about 4.3 km with 11310 m2 of land area.

2. Shoreline erosion is very strong with 5 to 6 m/year of average erosion rate for segment 1 and 10 m/year for segment 2. The rate of sedimentation is 0.34 m/year and 11 m/year for segment 1 and segment 2, respectively.

3. Sambas coastal areas in the sub district Jawai Selatan until the sub district Jawai susceptible to the (scenario 1). Sambas coastal region was quite vulnerable to the existing conditions, although CVI value is dominated by the moderate vulnerability, while the shoreline was in danger condition with CVI value is high vulnerability.

4. In the second scenario with tidal data = 3 m, wave = 3 m, and sea level rise = 0.5 m, obtained results of a vulnerability index is generally dominated by the conditions of very high vulnerability. On along the coast, the criteria of vulnerability index changed to "very high vulnerability ". Sea level rise 0.5 m is predicted for the year 2050.

5. In the third scenario with tidal data = 4 m, wave = 4 m, and an increase in sea level = 1 m, obtained results of a vulnerability index is generally dominated by the conditions of very high vulnerability. Sea level rise of 1 m was predicted for the year 2100. This scenario aims to detect coastal vulnerability in the future.

6. The shoreline retreat in Jawai Selatan due to breakwater too close to the shoreline, can be overcome by removing the breakwater into the sea with a certain distance from the shoreline to grow salient and tombolo. Furthermore, the combination

of Breakwater and mangrove plant, are expected to increase the supply of sediment to the eroded beach areas.

ACKNOWLEDGEMENTS This research was carried out with the financial support of DP2M DIKTI. REFERENCES Abuodha, P. A., Woodroffe, C. D. (2006).

“International Assessments of the Vulnerability of The Coastal Zone to Climate Change, Including an Australian Perspective”. Research Report, Department of the Environment and Heritage, Australian Greenhouse Office.

Alesheikh, A. A., et al. (2007). ”Coastline Change Detection Using Remote Sensing”. Int. J. Environ. Sci. Tech., 4 (1), p. 61-66.

Cambers, G. (2001). “Coastal Hazards and Vulnerability”. Sea Grant College Programme, University of Puerto Rice, Mayaguez, Puerto Rico.

Gornitz, V., White, T. W. and Cushman, R. M. (1991). “Vulnerability of the U.S. to Future Sea-Level Rise”. Coastal Zone '91, Proceedings of Seventh Symposium on Coastal and Ocean Management ASCE, p. 2354-2368.

Lewsey, C. (2002). “Climate Change Impacts on Land Use Planning and Coastal Infrastructure”. National Oceanic and Atmospheric Administration’s.

Li, R., Liu, J. K., and Felus, Y. (1999). ”Spatial Modeling and Analysis for Shoreline Change Detection And Coastal Erosion Monitoring”. The Proceedings of Geoinformatics '99, Conference, Ann Arbor, 19-21 June, 1999.

Szlafsztein, C. F. (2005). ”Climate change, Sea-level rise and Coastal Natural Hazards: A GIS-Based Vulnerability Assessment, State of Pará, Brazil”. An International Workshop, Department of Geology, Center of Geosciences, University of Pará, Brazil.

74

Awaludin: Time-History Response of 2-D Timber Frame Reinforced with Wooden Panel

Time-History Response of 2-D Timber Frame Reinforced with Wooden Panel

Ali Awaludin

Civil and Environmental Engineering Department, University of Gadjah Mada, Bulaksumur – Yogyakarta, Indonesia, email: [email protected]

Abstract: Most engineered timber (frame) buildings under great earthquake would undergo large deformation with their beam and column members are still elastic. This fact is potentially due to large flexing ability and ductile behavior of their mechanical connections. Since inelastic deformation occurs locally at their connections, non-linear analysis of the timber frame is often referred to a localized-nonlinearity problem. This paper demonstrated a non-linear time-history analysis of 2-D timber frame model using SAP2000 computer program. In this numerical analysis, a collection of non-linear link elements was used to produce similar load-deformation relation of semi-rigid joints of the frame model. The response of the frame model also was investigated when a lateral force-resisting element such as wooden panel was installed. Keywords: Non-linear analysis, seismic response, timber frame, and wooden panel. INTRODUCTION Indonesia sits on three active tectonic plates: Pacific, Eurasia, and Indo-Australia that often leads to a very catastrophic earthquake disaster such as the December 2004 great Sumatra earthquake (Saatcioglu, 2006). After this earthquake, USGS recorded that about four to

five strong earthquakes (Richter scale of 6.0 or more) shook Indonesia annually. As a result, the provision of an adequate structural system that enables a building to absorb and distribute the earthquake energy without collapse or catatastrophic failure is essential. Since it is very costly to construct strong buildings without damages during earthquakes, modern

Figure 1. Test set-up of timber joint under reversed-cyclic load

75


timber buildings would be an important solution. The modern timber buildings are characterized with the provision of ductile connections, hold-down devices that firmly connect the upper structure to foundation, and lateral force-resisting elements such as wooden panels or diagonal members. Through a combination of these structural elements, the timber buildings are enforced to perform monolithically under any loadings including the earthquake loads. Wood assemblies offer high strength-to-weight ratio over those built with steel and concrete. This results in low inertial or lateral force during earthquake. Wood systems are inherently more flexible than other building materials such as masonry and concrete that have to be carefully designed and detailed to ensure good seismic performance. Due to anisotropic properties of wood and hundreds of possible building geometry, modeling of timber buildings even of a simple frame construction always requires experimental works. Many timber engineers and researchers are currently developing various analytical techniques to predict both elastic and inelastic response of timber structures with minimum laboratory works. And the required numerical data could be derived from relatively simple tests. In this presented study, a seismic response of 2-D timber frame model under specified ground excitation was numerically evaluated using SAP2000 (1997). Moreover, the frame model can also accommodate wooden panel installment as one of important lateral force-resisting elements. The static-cyclic properties of timber joints and wooden panel were obtained from full-scale tests. TIMBER JOINT TEST In general adhesive joints should not be aplied in timber buildings to ensure ductile behavior under earthquake loads. Mechanical timber connections such as nails, lag screws, dowels, and bolts are often used in modern timber building because of their fast joint fabrication. A timber joint model shown in Figure 1 was selected for this study. SPF (spruce-pine-fir)

species was used with 4-mm-thickness steel plates and 12-mm-diameter bolts. The joint was tested cyclically until lateral displacement measured at the hydraulic cylinder (see Figure 1) reached 70-mm. And after that, the joint was subjected to monotonic loading until faiure. A typical hysteresis loop obtained from the static-cyclic test as shown in Figure 2 indicated that joint resistance was decreasing when the applied moment was repeated at the same rotation level. Less resistance at rotation close to zero points describes pinching mechanism due to wood fiber crushing beneath the bolts. At this stage, the joint resistance is solely given by the bending strength of the bolts. Hysteretic damping or area enclosed by the loop of the joints decreased as the number of cycles increased and its reduction occurred with a greater rate when the joints were cycled at higher rotation levels. As the plastic embedment of wood under the bolts progresses, the pinching region or range of joint rotation with relatively low stiffness increases. Improvement of initial joint stiffness and hysteretic damping of bolted timber joints through initial bolt pretension had been intensively reported Awaludin et al. (2008a, 2008b). For bolted joints with wood plates as both main and side members, the improvement can be carried out by gluing or embedding reinforcement materials at the joint shear planes. Among many available reinforcements, natural fiber–based materials such as plywood, plybamboo, and densified veneer wood (DVW) are probably the most widely used Leijten (2007), Awaludin et al. (2007). WOODEN PANELS Besides diagonal members or bracing, wooden panels are often found in modern timber buildings around the world. The panels usually consist of plywood or oriented strand board (OSB). Wooden panels or diagonal members are expected to increase the shear rigidity of the timber buildings during earthquakes. In some countries with extreme weather, the walls are often constructed in two layers with

76


gap in between. Polystyrene-based insulators fill the gap, insulate acoustic and improve thermal comfort of the building. Test results of wooden panel under static-cyclic load described here is summarized from the work of Honma et al. (2007). The panel consisted of 805x2595 mm2 oriented strand board with 9.5-mm in thickness as can be seen in Figure 3(a). At each edge, the board was connected to SPF lumber of 105x105mm2 through nails CN-50 (2.8-mm-diameter and 50-mm-length). The nail spacing was about 100-mm. Before a monotonic loading until failure, the panel was subjected to a reversed-cyclic load to investigate its racking behavior. The lateral load-rotation curve derived from the test is presented in Figure 3(b). Panels with an opening was also fabricated and tested to examine strength degradation related to panel opening. The opening area was about 46 percent. The simplified test setup and experimental load-rotation curve of the panel with opening are shown in Figure 4. A combination of bending and shear forces was introduced to the wooden panel by the applied-load. The test showed that the failure of the panel without opening was due to the failure of the beam-column connection at the top left-corner. The curve shown in Figure 3

clearly shows the wooden panel has small ductility ratio as a typical characteristic of mortise-tenon connection system. After reaching the maximum resistance, the curve is decreasing immediately and with minimum apparent-yielding zone. When there are some openings in the OSB element, the shear rigidity of the OSB element is decreasing. Test results of the panel with opening indicated that the failure mode was a partial slide-off of the OSB element from the column or beam member. Significant bending deformation of the nails and bearing damage of OSB or wood fibers around the nail fasteners were observed. Following this failure mode, in some tests, the nail fasteners were totally moved-out from the OSB element. The withdrawal resistance of nail connections was exceeded. This failure mode produces high ductility ratio as can be clearly seen in Figure 4. The initial stiffness and lateral resistance of wooden panel without opening were significantly greater than those of the panel with opening. Under reversed-cyclic loads, both panels experienced strength degradation due to bending deformation of nails and irreversible embedment of wood fibers around the nail fasteners.

-16

-12

-8

-4

0

4

8

12

16

-120 -80 -40 0 40 80 120 160

Average rotation (0.001 rad)

Mom

ent (

kNm

)

2

Figure 2. Test set-up of timber joint under reversed-cyclic load

77


However, the hysteretic damping of the panel without opening was much greater than that of the panel with opening. The opening of the OSB element definitely leads to a decrease of shear resistance of the panel. NUMERICAL MODEL OF TIMBER FRAME To elaborate the test results of timber joint and wooden panel described previously, a 2-D timber frame model shown in Figure 5 was developed. Hinge support was adopted since,

in many practice, this type of support can perfectly model a combination of shallow foundation behavior and common timber jointing method. At the top of each column, a concentrated weight of 10 kN was assumed as the frame load gene-rated by the gravity loads. In this numerical analysis, N-S component of Elcentro earthquake 1940 shown in Figure 6 was used as the ground excitation input. The equilibrium equation of motion of the system with nonlinear change in stiffness under general time-history load p(t) is expressed as shown in the following equation:

Figure 4. (a) Simplified test setup and (b) Hysteresis loop of panel with opening (Honma et al. 2008)

Figure 3. (a) Simplified test setup and (b) Hysteresis loop of panel without opening (Honma et al. 2008)

78


)()()()()( 0

...tvktptvktvctvm d

…………………………...…………………… (1)

)()( tYtv ......………………………………………………………………………….. (2)

)()()()()(...

tFtPtYKtYCtYM sd

...……………………………………….. (3)

m

pppd

Tn

Tnnnnnnn tYktptYtYtY

1

2...

)()()()(2)( …………….……………... (4)

m

p

kppd

Tn

Tn

knn

knnn

kn YkpYYY

1

)1()(2)(...

)( 2 ……………….……………….. (5)

In the above equation, [m], [c], and [k0] is mass, damping, and stiffness matrices of the system, respectively. {v} is the displacement vector of the system. In the equation, a single overdot denotes the first derivative with respect to time, which is velocity, and obviously double overdots denote the second derivative with respect to time, that is, acceleration. [kd] is being determined from the force-displacement relationship of a connection. Uncouple Equation 1 by means of the natural modes of the system defined by the left-hand side of the equation gives and {Y(t)} is modal matrix and generalized modal coordinates, respectively. For a proportionally damped system, Equation 3 is uncoupled and the integration is actually carried out on the equation. The last term in the Equation 4 is modal pseudo-forces that are functions of all

modal coordinates of the system. and are respectively, damping ratio and natural frequency correspond to mode shape n, . The equilibrium in one step for mode shape n in the k cycle of iteration is shown in Eq. 5. In each cycle of iteration, the equations are integrated and the displacement and velocity at the end of a time step are determined. The analysis was carried out using SAP2000 educational version. Only the first mode shape with damping ratio of 5% was considered for this study. The connection between column and beams was modeled by a set of four NL-link elements and one imaginary node as shown in Figure 7(a). Detail of semi-rigid joint modeling using SAP2000 can be found in Mesic, 2003. The properties of the four NL-link elements were derived by some trials so that it can produce a hysteresis loop similar to the one given in

Figure 5. Timber frame model of this study Figure 6. N-S component of Elcentro earthquake 1940

nn

n

79


Figure 2. An example of moment-rotation curve under one cyclic load is shown in Figure 7(b), where the properties of the four NL-link elements are summarized in Table 1. A different shape of hysteretic loop can be derived by supplying different numerical values of the four NL-link elements. This technique accommodates variation of hysteretic loops as they are essentially

influenced by joint geometry including the number of mechanical fasteners. Since the objective of this paper is to demonstrate the time history non-linear analysis of frame without and with wooden panel, therefore the properties of the four NL-link elements given in Table 1 and Table 2 are selected after some trials.

Table 1. Properties of the NL-link elements for semi-rigid joint used in SAP2000

Element Type Opening (rad)

ko (kNm/rad)

Yield Moment (kNm) a exp

NL1 Plastic 1 NA 700 2.0 0.01 1.0 NL2 Plastic 1 NA 400 4.0 0.05 1.0 NL3 Gap 0.01 25 NA NA NA NL4 Hook 0.01 25 NA NA NA

k0 : elastic (rotational) stiffness a : ratio of the post-yield stiffness to k0

exp : indicates the sharpness ratio of yielding curve NA : not available

Table 2 Properties of the NL-link elements for wooden panel used in SAP2000

Element Type Shear stiffness, ko (kN/m)

Shear resistance (kN) a exp

Panel 1 Plastic 1 230 12 0.25 10 Panel 2 Plastic 1 80 7 0.25 10

k0 : elastic (shear) stiffness a : ratio of the post-yield stiffness to k0 exp : indicates the sharpness ratio of yielding curve

Figure 7. (a) Semi-rigid joint modeling and (b) hysteresis loop of the joint due to one cyclic load

80


Time-history lateral displacement of the frame without and with wooden panel subjected to the Elcentro earthquake 1940 is presented in Figure 8 and Figure 9, respectively. The installment of wooden panel was modeled with one NL-link element at the middle of the beam element. The properties of this NL-link element (panel 1 or panel 2), which are shown in Table 2, were derived based on the hysteresis loops shown in Figure 3 and Figure 4, respectively, for the panel without and with 46% opening. Figure 9 indicates that the presence of wooden panel significantly

decreases the lateral displacement of the frame; the lateral displacement had been decreased by 32 percent for the panel without opening and by 20 percent for the panel with 46% opening. Decrease of lateral displacement due to wooden panel installment is a clear sign of overall improvement of the seismic performance of the frame model. This developed analysis can also be used to further evaluate the effect of different joint or wooden panel hysteresis loops on seismic performance of the timber frame.

Figure 8. Model of timber frame without wooden panel and its time-history displacement response due to N-S component of Elcentro earthquake 1940

Figure 9. Time-history lateral displacement of the frame model with wooden panel

81


CONCLUSIONS The seismic performance of timber buildings is substantially governed by their connections behavior. Therefore, performing a full-scale test of timber frame might not be so essential when the cyclic properties of their connections are available. In this study, seismic response of 2-D timber frame model was analyzed numerically using SAP2000 computer program assuming that nonlinearity problems occurred only at the connections. Four non-linear link elements and one imaginary node were necessary to appropriately simulate load-deformation relation of semi-rigid joints of the frame model. The numerical results showed that a significant decrease of lateral displacement of the frame can be attained by installing wooden (shear) walls. REFERENCES Awaludin, A., Hayashikawa, T., Hirai, T., and

Sasaki, Y. (2007). "Splitting Strength of Plybamboo-Reinforced Timber Joints Under Loading Perpendicular-to-Grain". Proceedings of the International Conference on Sustainable Infrastructure and Built Environment, Bandung, West Java, November 2-3. (CD ROM)

Awaludin, A., Hirai, T., Hayashikawa, T., Sasaki, Y., and Oikawa, A. (2008a).

"Effects of Pretension in Bolts on Hysteretic Response of Moment Carrying Timber Joints". Journal of Wood Science, Vol. 54, No. 2, p. 114-120.

Awaludin, A., Hirai, T., Hayashikawa, T., and Sasaki, Y. (2008b). "Load-Carrying Capacity of Steel-to-Timber Joints with a Pretensioned Bolt". Journal of Wood Science, Vol. 54, No. 5, p. 362-368.

Honma, C., Teranishi, M., Sasaki, Y., and Hirai, T. (2007). "Earthquake-proof Reinforcement of Deteriorated Framing of Existing Wooden Houses with Shear-Resistant Components". Mokuzai Gakkaishi, Japan Wood Research Society, Vol. 54, No. 3, p. 123-131 (in Japanese).

Leijten, A. J. M. (2007). "Densified Veneer Wood Reinforced Timber Joints with Expanded Tube Fasteners". Delft University Press, The Netherlands.

Mesic, E. (2003). "Analysis of Timber Frames with Localized Nonlinearities". Facta Universitatis, Series: Architecture and Civil Engineering, Vol. 2, No. 5, p. 307-320.

Saatcioglu, M., Ghobarah, A., and Nistor, I. (2006). "Performance of Structures In Indonesia during the December 2004 Great Sumatra Earthquake and Indian Ocean Tsunami". Journal of Earthquake Spectra, Vol. 22, No. S3, S295-S319.

SAP2000 (1997). "Manual, Analysis Reference", Computers and Structures, Inc., Berkeley, California, USA, Vol. 1.

82

Rosyidi: Wavelet-Spectrogram Analysis Of Surface Wave Technique: A Novel Procedure ... Layer

Wavelet-Spectrogram Analysis of Surface Wave Technique: A Novel Procedure for Non-Destructive Measurement on Pavement Surface Layer

Sri Atmaja P. Rosyidi

Department of Civil Engineering, University of Muhammadiyah Yogyakarta, KampusTerpadu, Jl. Ring Road Selatan, Bantul 55183, Yogyakarta, email: [email protected]

Abstract: Reliable assessment of in situ pavements stiffness is an important aspect in effectively managing a pavement system. The aim of this paper is to propose the new procedure, namely the wavelet-spectrogram of surface wave (WSSW) technique for non-destructively measurement of elastic modulus on surface layer of a pavement system. Using two receivers, surface wave waveform on pavement surface was recorded and transformed into in frequency domain by wavelet analysis. For this analysis, a derivative Gaussian wavelet was selected as an appropriate mother wavelet for seismic waveform propagating along pavement surface. Thus, an interactive 2-D plot of time-frequency spectrogram consisting of wave-energy spectrum was simultaneously generated. CWT-filtration method was implemented in order to reduce the effect of noisy signal recorded during measurement. From selected wave spectrogram, the unwrapped phase different-spectrum was generated to obtain phase velocity which was performed by least-square linear regression. Finally, the elastic modulus of pavement surface layer was calculated from a modified relationship between phase velocity, Poisson ratio and density of pavement surface layer. The results show that the proposed technique is able to measure in situ elastic stiffness of the surface layer. In addition, the change of the surface layer stiffness is also able to be monitored. However, the stiffness produced by the WSSW technique is classified as a modulus at very low strain level. Keywords: Elastic modulus, surface layer, pavement system, and wavelet-spectrum of surface

wave technique. INTRODUCTION Seismic surface wave methods are well-known of non-destructive techniques in soil dynamics and geotechnical investigation. Over past decade, seismic surface wave methods have been improved and subsequently, it has been utilized in different applications, i.e., spectral analysis of surface wave, SASW [site characterization (Stokoe et al. 1994), soil density (Kim et al. 2001), ballast density (Zagyapan & Fairfield, 2002), sloping surface (Kim et al. 2007), pavement characterization (Rosyidi et al. 2007)], multichannel analysis of surface wave, MASW [residual soil characterization (Ariestyanti, 2009), shear wave velocity profiling (Park et al. 1998), near-surface anomaly detection (Miller & Xia, 1999)], continuous surface waves, CSW [soil profiling (Rosyidi, 2006)].

The surface wave methods use the dispersion property of surface waves (dominantly by R-wave) to determine material stiffness of each layer at soil profile. In order to generate the elastic/shear modulus, an elaborated data process should be performed. The process is divided into three steps: collecting seismic data, constructing an experimental dispersion curve, and conducting an inversion process of the dispersion curve for generating a proper profile. An impact source on a ground surface is used to generate seismic waves at a point. These waves are detected at two or more receiver locations and the signals recorded using a spectrum analyzer for post processing. Several sets of test with different receiver spacings are required to sample different depths. The experimental dispersion curve can be developed from phase information of the transfer function at frequency ranges satisfying the coherence criterion.

83


In inversion process, the advanced algorithm is required to produce the stiffness profile froma dispersion curve. Reliable inversion procedure using stress-wave propagation theories, i.e., the transfer matrix method (Thomson 1950, Hanskell 1953); the dynamic stiffness matrix method (Kausel & Röesset, 1980); and the finite difference method (Hossain & Drnevich, 1989) should be employed. All the methods require an initial profile model consists of a set of homogeneous layers extending to infinity in the horizontal direction. The last layer is usually considered as a homogeneous half-space. To each layer, a thickness, a shear wave velocity, a Poisson's ratio (or compression wave velocity), and a mass density are assigned. Based on the initial profile, a theoretical dispersion curve is calculated using one of these wave propagation theories. The theoretical dispersion curve is then compared with the experimental dispersion curve. If the two dispersion curves do not match, the initial profile (number of layers, layer thickness, shear wave velocity, or any combination) is adjusted, and another theoretical dispersion curve is calculated. The trial-and-error procedure is repeated until the two curves match, and then the associated assumed profile is considered the real profile. In addition, when the surface wave method is performed on irregular stiffness profile, i.e. pavement structure, the trial-and-error process becomes a difficult analysis. Due to complexity in surface wave analysis, application of these methods on pavement evaluation is still relatively limited. On the other hand, in pavement evaluation system, the need of accurate, cost-effective and non-destructive evaluation is becoming ever important because the rehabilitation and management of roads is becoming increasingly difficult due to the increasing number of aging roads and limited budgets. As well, for practical purpose, pavement engineer usually needs a quick and relatively effortless analysis for determining the structural condition of pavement surface layers. In this paper, a new procedure in a surface wave methodusing continuous wavelet transform and simple algorithm, namely the

wavelet-spectrogram of surface wave (WSSW) technique, for obtaining the phase velocity is introduced. Modified algorithmon phase different calculation in this analysis aims to avoid the use of a complex inversion algorithm to obtain the elastic modulus of the pavement surface layer. In proposed analysis, the time-frequency decomposition of CWT on seismic signals is employed to characterise the phase information of transfer function spectrum. In previous surface wave method, the data analysis in frequency domain has been carried out by fast Fourier transforms (FFT). However, due to Fourier transform works by expressing any arbitrary periodic function of time with period as sum a set of sinusoidal, this analysis is unable to preserve the time dependence and describe the evolutionary spectral characteristics of non-stationary processes (Rosyidi et al. 2009). Wavelet analysis is becoming a common tool for analyzing localized variations of power within a time series. By decomposing a time series into time-frequency spectrum (TFW), one is able to determine both the dominant modes of variability and how those modes vary in time. A typical result from a case study is presented herein for the structural assessment of an existing asphalt concrete (AC) pavement at Cikampek-Purwakarta road, West Java, Indonesia. THEORY OF CONTINUOUS WAVELET TRANSFORM (CWT) The continuous wavelet transform has been used in numerous studies in geophysics, including the El Niño–Southern Oscillation (Wang and Wang, 1996), the dispersion of ocean waves (Meyers et al., 1993), and wave growth and breaking (Liu 1994). A complete description of geophysical applications can be found in Foufoula-Georgiou and Kumar (1995) and Rosyidi et al. (2008, 2009), while a theoretical treatment of wavelet analysis is given in Daubechies (1992). The continuous wavelet transform (CWT) technique is an alternative tool for localizing the interested frequency of seismic signal processing particularly in non-stationary

84

Sukarno: Teori Kendaraan-Pengikut yang Disederhanakan (Studi Kasus Persimpangan Bersinyal, ... Yogyakarta)

problems. A presentation of basic wavelet theory may be found in several literatures such as Daubechies (1992), Kaiser (1994), and Farge (1992). Wavelets dilate in such a way that the time support changes for different frequency. When the time supports increases or decreases, the frequency support of the wavelet is shifted toward high or low frequencies, respectively. Therefore, as the frequency resolution increases, the time resolution decreases and vice versa (Mallat, 1989). This optimal time-frequency resolution property makes the CWT technique useful for non-stationary seismic analysis. A wavelet is defined as a function of (t) L2() with a zero mean, which is localized in both time and frequency. By dilating and translating the wavelet (t), it can be used to produce a family of wavelets as:

t1, (1)

Where is the dilation parameter or scale and is the translation parameter (, , and 0). Unlike Fourier transform, the wavelet has various wavelet shape used for signal analysis which is called the mother of wavelet. The choice of appropriate wavelet shape used in signal analysis depends on the seismic waveforms. The family of wavelets can be divided into two groups (Soman and Ramachandran 2005) i.e. continuous wavelet transform (CWT) and discrete wavelets transform (DWT). Some mother wavelets which are commonly used in the CWT are Gaussian, Morlet, Paul and Mexican Hat. The CWT is defined as the inner product of the family wavelets Ψ, (t) with the signal of f(t) which is given as:

dtt

tfttfFW

1,, , (2)

where is the complex conjugate of ,

FW(,) is the time-scale map. The convolution integral from Equation (2) can be computed in the Fourier domain. In order to reconstruct the function f(t) from the wavelet

transform, Calderon’s identify (Daubechies 1992) can be used and is obtained as:

ddtF

Ctf W 2

,1

(3)

dC

2ˆ2 (4)

where ̂ is the Fourier transform of (t). The integrand in Equation (4) has an integrable discontinuity at = 0 and implies

that 0 dtt .

In this study, the mother wavelet of the Gaussian Derivative (GoD) was used. The real component of the GoD wavelet in the time and frequency domains is defined as follows:

22

2

1

1 1

0

t

ed

d

m

tm

mm

(5)

22

2

1ˆ0

s

es

m

is m

m (6)

Where m is the wave number and is the Gamma function. The complex wavelet is generated by the addition of a Heaviside function in the frequency domain. This wavelet decays with the square root of the gamma function. The Gaussian Derivative has wavelet's derivative order that can be varied in order to get the best resolution of the waveform. A NOVEL PROCEDURE FOR CALCULATING ELASTIC MODULUS OF PAVEMENT SURFACE LAYER A proposed procedure of surface wave analysis using continuous wavelet transform and elastic linear model for elastic modulus measurement on pavement surface layer is described as follows:

1. The surface wave data are collected from field measurement using a configuration of the mid point receiver spacings. In this configuration, some instruments, i.e., an impact source of small ball bearing, a data acquisition unit and two acceleration

85


receivers (accelerometers) are used for recording seismic wave propagation. Discussion of field set up is presented in the following section.

2. The time-frequency spectrum analysis based on continuous wavelet transform of Gaussian Derivative (DoG) mother wavelet is carried out for two signal waveforms recorded from field measurement.By decomposing a time series of seismic waveform into time-frequency (TF) spectrogram, the dominant modes of variability and how those modes vary in time can be determined very well. In this study, the TF spectrogram is generated by wavelet transform (Equation 2) with a mother wavelet of GoD (Equation 6).

3. Phase information of the transfer function (phase spectrum) are then determined from both TF spectrograms. The data informs the time difference of wave propragating from first to second receiver. A mathematical expression for calculating phase spectrum from TF spectrogram is defined by (Rosyidi, 2009):

dtt

tX

dtt

tY

W

W

fX

fYfH

su

Xf

su

Yf

*1

*1

),(

),(

................................................................. (7)

where,

X(f) = spectrum input of signal, X(t), from first receiver,

Y(f) = spectrum output of signal, Y(t), from second receiver,

uti

su

Yf e

s

utg

stYW

1),(

(8)

uti

su

Xf e

s

utg

stXW

1),(

(9)

From Eq.7, the phase spectrogram in time-fequency domain can be obtained by:

),(),(

),(

,

,,

*

,,

suWsuW

esuW

suW

suWsuH

Xf

Xf

babaiXYf

XXf

XYf

XY

............................................................... (10)

Thus, phase different is obtained from the ratio of the imaginary to real part of the phase spectrogram which is expressed as:

),(

),(tan 1

suH

suH (11)

4. The coherence function is used to visually inspect the quality of signals being recorded in the field and have a real value between zero and one in the range of frequencies being measured. The value of one indicates a high signal-to-noise ratio (i.e. perfect correlation between the two signals) while values of zero represents no correlation between the two signals. The coherence function is a ratio of the output power caused by the measured input to the total measured output which is defined as (Rosyidi 2004):

fGfG

fGfGf

yyxx

yxyx

*

2 (12)

5. A linear relationship between the phase different and frequency from the transfer function spectrum is then derived. The phase velocity is examined as a function of distancefrom the slope value (m). This relationship can be written as:

= phV

D360f = mf (13)

By Equation 13, one can easily determine Vph by performing a least-square linear regression over the frequency range in the transfer function spectrum and the slope of the best-fit line (m) can be obtained.The phase velocity of surface wave propagation is independent of the wavelength for up to a wavelength approximately equal to thickness of the uppermost layer (Nazarian et al., 1996). The range of wavelength to be used can be estimated from the phase velocities (Vph) of the material anticipated at the site:

f

Vph

(14)

where f is the frequency.

86


6. If one simply generates high frequency waves and assumes that properties of the uppermost layer are uniform, the dynamic elastic moduli of the pavement materials can easily be determined as follows:

E = 2 μ1γ 2 SVg

= 2

phKVg

γ (15)

K = (1.13 – 0.16μ)

)(

21

12 (16)

where E is the dynamic elastic modulus, respectively, VS is the shear wave velocity, Vph is the phase wave velocity, g is the gravitational acceleration, γ is the total unit weight of the material and μ is the Poisson’s ratio.

EXPERIMENTAL SET UP Impact source of 5 to 15 g in weight of ball bearings on a pavement surface was used to generate surface waves. These generated waves were detected using two high frequency accelerometers where the signals were recorded and transferred using an analog digital acquisition to a notebook computer for post processing (Figure 1). Several configurations of the receiver and the source spacings were required in order to sample different depths. The configuration used in this study was the mid point receiver spacings (Heisey et al. 1982) as described in Figure 2. However, the receiver spacing (d2) is configurated as a distance which is less than and/or equal to the thickness (H) of pavement surface layer. The distance between a source and first receiver (d1) is set as equal to receivers spacing (d2). Due to the interest pavement layer in this study is asphaltic surface layer, the short receiver spacings of 4 and 8 cm with a high frequency source (ball bearing) were only used. For each receiver spacing measurement, the testing should be repeated at lease 3 to 5 times in order to minimise the effect of internal phase shift between receivers and a good average of received signals can be generated. The surface wave data from SASW testing was carried out at selected sites on Cikampek-

Purwakarta road in West Java, Indonesia. The pavement profile obtained from core drilling consists of an asphalt concrete (AC) layer (183 mm), crushed stone of base course (71 mm), sub-base course (190 mm) over a subgrade layer.

Figure 1. Field equipment for surface wave recording: (a) spectrum analyzer with an acquisition unit; (b) accelerometers for

receiving the seismic wave displacement (Widodo & Rosyidi, 2009)

Figure 2. WSSW experimental configuration

87


RESULTS AND DISCUSSIONS Elastic Modulus from the WSSW Technique

The WSSW tests were performed on existing asphalt pavements. A typical received signals from field measurement by spectrum analyzer instruments (Figure 1) at selected location in Cikampek-Purwakarta roads was shown in Figure 3. The waves were generated from impact source arriving in the first accelerometer (receiver 1) and its propagations were then received in the second one (receiver 2). The body waves (Primary and Secondary wave) and surface wave (Rayleigh wave) are clearly shown in both signal recordings. Therefore, the Gaussian Derivative (GoD) continuous wavelet transform (CWT) as mentioned in Eq. 7 was employed on both signals for generating a time-frequency (TF) plot. These CWT plots may overcome on identification problem of the spectral characteristics of non-stationary signals measured in two receivers. The typical CWT spectrogram for received signals with an improved time-frequency resolution is shown in Figure 4. The TF of GoD-CWT provides good resolutions at high frequencies of signals. It is also effective in the detection of frequency bandwidth of wave groups using various derivation order of this mother wavelet (GoD).

Figure 3. Signal recording from measurement on a pavement structure

From Figure 4, several energy events at different frequency bands are clearly detected which may result in interference at fundamental and higher mode of seismic signals at the interface between surface layer and base course; and modes from reflected body waves. Based on both TF spectrograms (Figure 4), the phase different between two signals for every wave frequency were then calculated (Eq.10 to 11). As a result, a transfer function spectrum on wrapped phase different curve was obtained as given in Figure 5. The phase data oscillates between - and radian (-180 and 180 degrees). This is the standard illustration of spectrum presenting phase data because the detail variation in data can be observed in a small space.

(a) CWT Spectrogram from receiver 1

(b) CWT Spectrogram from receiver 2

Figure 4. Time-frequency plot of received signals from measurement (Fig.3)

Figure 5 also shows that the phase different curve reveals a smooth trend of variation in phase with frequency up to a frequency of 25 kHz. It indicates the high-frequency surface

88


waves were detected representing the high stiffness of the asphalt concrete surface layer on a pavement structure. The quality of phase data is also controlled by the coherence function. As shown in Figure 6, the phase data up to frequency of 20 kHz have the value of coherence magnitude above 0.98.

Figure 5. Comparison between raw data and best-fit curve ofwrapped transfer function

spectrum from measurement

Figure 6. Coherent function spectrum of

received signals from measurement In order to obtain the modulus value of the surface layer, the smoothed fitting process of weighing function was used. The fitted curve between the raw and the smoothed phase spectrum is shown in Figure 5. Thus, the phase data is unwrapped by adding the number cycles to each phase. The unwrapped of the raw phase spectrum is also shown in Figure 7. The unwrapped phase spectrum is smoothed by the linear regression as the best fit curve to the raw data. The slope of the line is more or less constant with frequency. A line is fitted to

the curve in the range of the frequency corresponding to wavelengths shorter than the thickness of the surface layer. Slope value of the line can be used to determine the elastic modulus of the surface layer of the pavement profile using Equation 13, 15 to 16. From Figure 7, the slope (m) of best-fit curve is found to be 0.0140. Consequently, the phase velocity can becalculated using Eq.13 which is found to be 1028.57 m/s. Based on the phase velocity, field configuration data and material parameters, such as of receiver spacing (d2) of 4 cm, Poisson’s ratio of asphaltic layer of 0.25 and unit weight for pavement material (AC) of 2,200 kg/m3, the elastic modulus of asphaltic (AC) surface layer is obtained as 845,662,040.80 kg/m2 or 8456.62 MPa.

Figure 7. Comparison between raw data and best-fit curve ofunwrapped transfer function

spectrumfrom measurement and slope analysis to obtain m value

The result shows that the elastic modulus of AC layer can be easily determined using the WSSW technique. However, the value of elastic modulus presented is relatively high. It is due to the seismic technique measures the dynamic stiffness at very low strain level (less than 103%). In this level, the material modulus behaviour can be assumed as a constant and have only a maximum value (Nazarian & Stokoe, 1986). In order to illustrate the usefulness and sensitivity of this approach in testing the surface layer, changes in the stiffness of the existing surface and the overlay layer of the pavement were measured in situ. Figure 8

-180

-120

-60

0

60

120

180

0 5000 10000 15000 20000 25000

Frequency, Hz

Ph

ase

Dif

fere

nt

(wra

pp

ed)

Raw data

Best-fit curve

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 5000 10000 15000 20000

Frequency, Hz

Coh

eren

ce

� = mf = 0.014fR² = 0.960

0

50

100

150

200

250

300

350

400

0 5000 10000 15000 20000 25000

Frequency, Hz

Ph

ase

Dif

fere

nt

(un

wra

pp

ed)

Raw data of unwrapped phasespectrum

Best-fit curve (linear regression)

89


shows that the different stiffness of an existing and overlay surface layer from the measured profile can be investigated well. Based on these results, it can be summarized that the stiffness changes in the surface layer were easily, non-destructively, and fast measured by the WSSW technique.

Figure 8. Comparison between elastic modulus of an overlay and existing surface layer at Cikampek-Purwakarta Roads, Indonesia

Validation with the SASW Method

In order to validate the results from the WSSW test, the spectral-analysis-of-surface-wave (SASW) analysis was conducted at same locations of a road pavement. In this method, a set of transient impact sources was used to generate surface wave energy that propagates horizontally near the surface layer of the pavement. The phase differences of signal data was obtained from the cross-power spectrum. Thus, the phase information was then unwrapped to produce the dispersion curve of the pahse velocity versus wavelength. An inversion process was then iteratively employed to confirm the experimental dispersion curve from the theoretical model established. A 3-D stiffness matrix model (Rosyidi, 2007) was employed in the SASW inversion analysis. Final profil was obtained after 16 times of iteration with the root-mean-square error (rms) of 35.47 m/s or average deviation of about 5.92%.

Figure 9. The shear wave velocity profile from inversion of the experimental dispersion curve

Figure 10. The dynamic elastic modulus of the pavement profile from the SASW method and

its comparison with the WSSW method. The equivalent shear wave profile from the result of the inversion is shown in Figure 9 and using the dynamic material equation, its equivalent dynamic elastic modulus profile is given in Figure 10. A modulus profile as shown in Figure 10 is only given to 10 cm depth of pavement structure due to the asphaltic layer was indicated at this level. A good agreement of elastic modulus resulted from the WSSW and the SASW method is also presented in Figure 10. The result shows that

-1,39E-1

0,01

0,02

0,03

0,04

0,05

4000 6000 8000 10000

Dep

th, m

Elastic Modulus, MPa

New pavement

Existing pavement

Pave

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

0 200 400 600 800 1000 1200 1400

Shear Wave Velocity, m/s

Dep

th, m

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.1

0 2000 4000 6000 8000 10000

Elastic Modulus, MPa

Dep

th, m

the SASW method the WSSW method

From 4 cm receiver spacing

From 8 cm receiver spacing

90


the difference between both methods is calculated at 0.01 % and 3.14 % for first and second layer of pavement surface layer, respectively. CONCLUSIONS This paper summarizes a new procedure in the wavelet-spectrogram analysis of surface wave (WSSW) technique in order to in situ evaluation the elastic modulus of asphalt concrete surface layer on an existing pavement profile. By the simple calculation on phase spectrum from the surface wave data, the elastic modulus of the surface layer can be obtained without the complex calculation of the inversion process. The calculation is easy and can be simply implemented. This technique is also a very sensitive non destructive testing to monitor the change of the modulus of the existing surface and overlay layers. ACKNOWLEDGMENT This work is part of research project funded by Directorate of Higher Education, Ministry of National Education, Indonesia and the University of Muhammadiyah Yogyakarta (UMY) through hibah bersaing grant 2009. Their supports are gratefully acknowledged. We would also like to thank Dr. Siegfried (PUSLITBANGJATAN, Bandung) and research assistants at University of Muhammadiyah Yogyakarta for their assistance during field works. REFERENCES Ariestianty, S. K., Rosyidi, S. A., Chik, Z.,

Taha, M. R., Nayan, K. A. M., and Syamsudin, A. R. (2008). “Investigasi Profil Kekakuan Tanah 2-D Menggunakan Metode MASW”. Prosiding Pertemuan Tahunan ke 12 (PIT-XII) Himpunan Ahli Teknik Tanah Indonesia, Bandung, p. (VI) 1-5.

Daubechies, I. (1992). “Ten Lecturers on Wavelets”. Society of Industrial and Applied Mathematics.

Farge, M. (1992). “Wavelet Transform and Their Application to Turbulence”. Annu. Rev. Fluid Mech., 24, p. 395-457.

Foufoula-Georgiou, E. and Kumar, P. (1995). “Wavelets in Geophysics”. Academic Press.

Haskell, N. A. (1953). “The Dispersion of Surface Waves in Multilayered Media”. Bull. Seismol. Soc. Am., Vol. 43, No. 1, p. 17-34.

Heisey, J. S., Stokoe II, K. H., and Meyer, A. H. (1982). “Moduli of Pavement System from Spectral Analysis of Surface Waves”. Transportation Research Record 852, p. 22-31.

Hossain, M. M., Drnevich, V. P., Bush III, A. J., and Baladi, G. Y. (1989). “Numerical and Optimisation Techniques Applied to Surface Waves for Back-Calculation of Layer Moduli in Nondestructive Testing of Pavements And Back-calculation of Moduli”. Eds, Special Technical Publication 1026, American Society for Testing dan Materials, p. 649-669.

Kaiser, G. (1994). “A Friendly Guide to Wavelets”. Birkhäuser.

Kausel, E. and Röesset, J. M. (1981). “Stiffness Matrices for Layered Soils”, Bull. Seismol. Soc. Am., 72, p. 1743-1761.

Kim, D. S., Shin, M. K., and Park H. C. (2001). “Evaluation of Density in Layer Compaction Using SASW Method”. Soil Dynamic and Earthquake Engineering, Vol. 21, p. 39-46.

Kim, S. W., Kim, D. S., and Bang, E. S. (2007). “Evaluation of the Applicability of SASW Method to the Sloping Surface-Numerical Study and Model Test”. Proceeding of the twentieth KKCNN Symposium on Civil Engineering, p. 378-381.

Liu, P. C. (1994). “Wavelet Spectrum Analysis And Ocean Wind Waves”. Wavelets in Geophysics, Foufoula-Georgiou, E. and Kumar P., Eds., Academic Press, p. 151-166.

Meyers, S. D., Kelly, B. G., and O’Brien, J. J. (1993). “An Introduction to Wavelet Analysis in Oceanography and

91


Meteorology: with Application to the Dispersion of Yanai Waves”. Mon. Wea. Rev. 121, p. 2858-2866.

Miller, R. D. and Xia, J. (1999). “Feasibility of Seismic Techniques to Delineate Dissolution Feature in the Upper 600 ft at Alabama Electric Cooperative’s Proposed Damascus Site: Kansas Geological Survey”. Open-file Report, Kansas Geological Survey, No. 99-3,

Nazarian, S. and Stokoe II, K. H. (1986). “In Situ Determination of Elastic Moduli of Pavement Systems by Spectral-Analysis-Of-Surface-Wave Method (Theoretical Aspects)”. Research Repot 437-2, Bureau of Engineering Research. Center of Transportation Research, The University of Texas at Austin.

Nazarian, S., Yuan, D., and Baker, M. R. (1996). “Quality Control of Portland Cement Concrete Slabs with Wave Propagation Techniques”. Transportation Research Record 1544, p. 91-98.

Park, C. B., Miller, R. D. and Xia, J. (1998). “Imaging Dispersion Curve of Surface Waves on Multi-Channel Record: Technical Program with Biographies”. Proceeding of SEG 68th Annual Meeting, New Orleans, Lousiana, p. 1380-1388.

Rosyidi, S. A., Taha, M. R., Chik, Z., and Ismail, A. (2008). “Seismic Response Identification of Surface Waves on Residual Soils Using Wavelets Analysis”. Proceeding of International Conference on Geotechnical and Highway Engineering, Geotropika, Kuala Lumpur.

Rosyidi, S. A., Taha, M. R., Ismail, A., and Chik, Z. (2009). “Signal Reconstruction of Surface Waves on SASW Measurement Using Gaussian Derivative Wavelet Transform”. Acta Geophysica, Vol. 57, No. 3, p. 616-635, DOI: 10.2478/s11600-009-0015-8.

Rosyidi, S. A. (2004). “Evaluation of The Dynamic Stiffness of Pavement Layers Using SASW Method”. Master Thesis, Universiti Kebangsaan Malaysia.

Rosyidi, S. A. (2006). “Penggunaan Metode Continuous Source for Surface Wave Analysis (CSSWA) dalam Kajian Struktur

Tanah”. Proseding Konferensi Regional Teknik Jalan ke-9, Makassar.

Rosyidi, S. A. (2007). “Comparison between 2-D And 3-D Stiffness Matrix Model Simulation of Sasw Inversion for Pavement Structure”. Civil Engineering Dimension, Journal of Civil Engineering Science and Application, Vol. 9, No. 1, p. 42-48.

Rosyidi, S. A. (2009). “Wavelet Analysis of Surface Wave for Evaluation of Soil Dynamic Properties”. Ph.D. Thesis, Universiti Kebangsaan Malaysia.

Rosyidi, S. A., Taha, M. R., Nayan, K. A. M, Chik, Z., Ismail, A., and Siegfried. (2007). “Development of VS-CBR-DCP Empirical Model for Determining Dynamic Stiffness of Pavement Base Layer Using SASW”. Proceedings of International Conference on Advanced Characterisation of Pavement and Soil Engineering Materials, p. 895-902.

Soman, K. P. and Ramachandran, K. I. (2005). “Insight into Wavelets from Theory to Practice”. Prentice-Hall of India.

Stokoe II, K. H., Wright, S. G., Bay, J. A., and Roesset, J. M. (1994). “Characterization of Geotechnical Sites by SASW Method. Geotechnical Characterization of Sites”. R.D. Wood, Ed., Oxford and IBH Pub. Co., New Delhi, India, p. 15-26.

Thomson, W. T. (1950). “Transmission of Elastic Waves through a Stratified Solid Medium”. Jour. of Appl. Phys., Vol. 21, p. 89-93.

Wang, B. and Wang, Y. (1996). “Temporal Structure of the Southern Oscillation as Revealed by Waveform and Wavelet Analysis”. J. Climate 9, p. 1586-1598.

Widodo, W. and Rosyidi, S. A. (2009). “Experimental Investigation of Seismic Parameters and Bearing Capacity of Pavement Subgrade Using Surface Wave Method”. Sciencetific Journal of Semesta Teknika, Vol.1, No. 12, p. 67-79.

Zagyapan, M. and Fairfield, C. A. (2002). “Continuous Surface Wave and Impact Methods of Measuring the Stiffness and Density of Railway Ballast”. NDT & E International, Vol. 35, p.75-81.

92

Sukarno: Analisis Antrian Akibat Daerah Sempit di Jalan Bebas Hambatan

Analisis Antrian Akibat Daerah Sempit di Jalan Bebas Hambatan

Sukarno

Staf Pengajar Jurusan Teknik Sipil, Universitas Islam Indonesia, Yogyakarta, email: [email protected]

Abstract: Traffic congestion which is a common phenomenon in big cities sometime is found on a freeway as well. Congestion on a freeway happens when arrival flow is greater than departure flow, for example when there is a road construction or an accident at downstream or greater capacity at upstream compared than downstream. Two models, queuing and shockwave, are accepted as a tool to analyze narrow area caused by over capacity. However, two models give different results. The objectives of this study are to investigate those different results and to give some suggestions if any on the next use. Parameters of queuing, which were indicators of their characteristics, were determined using predetermined data, then the differences of their results were discussed through concepts of queuing, delay, and flow stage changing. The modified shock wave model was tested using the same data and the results show that modified model is slightly different with queuing model. Keywords: Queuing, narrow area, freeway, queuing model, and shockwave model. PENDAHULUAN Kemacetan terjadi hampir di setiap jalan perkotaan tidak terkecuali di jalan bebas hambatan. Kemacetan yang parah akan menyebar dengan cepat dan berlangsung sangat lama. Tujuan diadakannya jalan bebas hambatan adalah untuk perjalanan dengan kecepatan tinggi maka mengatasi masalah kemacetan di jalan bebas hambatan menjadi sangat penting. Menurut Papacostas dan Prevedouros (1993) salah satu diantara banyak masalah yang memerlukan perhatian dalam

manajemen kemacetan di jalan bebas hambatan adalah perlunya prosedur analitis untuk mempelajari efek kemacetan akibat penyempitan (Gambar 1). Menurut May (1990) untuk menganalisis kemacetan akibat penyempitan didaerah bebas hambatan biasa digunakan dua metode yaitu metode antrian dan metode gelombang kejut. Metode antrian memperlakukan lalu lintas yang akan masuk daerah sempit sebagai persoalan antrian tertentu sederhana,sedangkan metode gelombang kejut menganalogikan arus lalu lintas sebagai arus fluida sehingga ketika

Gambar 1. Daerah sempit dengan arus berbeda di jalan bebas hambatan

93


fluida memasuki daerah sempit akan timbul gelombang fluida yang datangnya mendadak (karena itu disebut gelombang kejut). Gelombang fluida inilah yang dipakai untuk memodelkan arus lalu lintas yang berubah kecepatan, volume, dan kepadatannya ketika memasuki daerah sempit (Garber and Hoel, 2009). Morales (1986) mengatakan bahwa meskipun perambatan gelombang utamanya berbentuk non-linier, para ahli biasa menyederhanakan menjadi berbentuk linier. McShane dan Roess (1990) mengatakan terdapat perbedaan pada hasil yang diperoleh pada kedua model. Dengan menggunakan contoh numeris, mereka menunjukkan bahwa jumlah antrian dan tundaan yang dihasilkan model antrian 60% lebih rendah dibandingkan dengan yang dihasilkan oleh model gelombang kejutsedangkan lamanya kemacetan oleh model antrian diprediksi 15% lebih rendah dibandingkan hasil prediksi model gelombang kejut. Shandutan (2011), yag meneliti penentuan lokasi penyempitan dengan metode gelombang kejut berhasil secara grafis menjelaskan perambatan gelombang di hulu daerah sempit, namun belum membandingkan hasilnya dengan metode antrian. Walaupun terdapat perbedaan pada kedua metode namun dalam penerapan keduanya selalu digunakan untuk memecahkan persoalan penyempitan di jalan bebas hambatan. Para ahli lalu lintas mengakui kesahihan kedua model walaupun terdapat perbedaan hasil yang cukup besar. Oleh karena itu, kedua model perlu dievaluasi lebih dalam agar hal-hal yang menjadi perbedaan dapat diketahui. Tujuan dari karya ilmiah ini adalah untuk mengevaluasi model antrian dan model gelombang kejut dalam menganalisis kemacetan akibat penyempitan di jalan bebas hambatan kemudian mengusulkan apabila memungkinkan modifikasi model agar kedua metode sebanding. TINJAUAN MODEL Model Antrian

Penyempitan di bagian jalan bebas hambatan terjadi ketika arus yang melewatinya melebihi kapasitas bagian jalan bebas hambatan

tersebut. Terjadinya penyempitan ini dapat disebabkan oleh dua hal. Pertama adalah meningkatnya arus di hulu bagian jalan yang ditinjau hingga melebihi kapasitas normal (kapasitas di bagian jalan yang ditinjau). Penyempitan ini biasanya bersifat tetap dan efeknya terlihat hanya bila ada lalu lintas tinggi di hulu. Kedua, penyempitan yang terjadi karena pengurangan kapasitas yang sifatnya sementara di bagian jalan bebas hambatan yang ditinjau, misalnya karena ada kecelakaan atau pengerjaan jalan. Andaikan bagian jalan yang ditinjau adalah X dengan kapasitas Qc. Bila arus lalu lintas sebesar Q1 (>Qc) datang, jalan di bagian hulu X akan mengalami kemacetan. Andaikan Q1 berlangsung selama t1 (disebut arus tahap-1) dan kemudian berubah menjadi Q2 (< Qc) selama t2–t1 (disebut arus tahap-2) maka berdasarkan teori antrian tertentu (selanjutnya disebut metode antrian), yang mengasumsikan kedatangan kendaraan (A) dan keberangkatan kendaraan (D) adalah linier, bentuk kurva kedatangan dan keberangkatan dapat dilihat pada Gambar 2. Dari Gambar 2 diperoleh:

a. Untuk kedatangan

Arus tahap-1 (berlangsung selama t1) : A1(t) = Q1 t dan, Arus tahap-2 (berlangsung selama (t2-t1): A2(t) = Q2 t

b. Untuk keberangkatan

Arus tahap-1 (berlangsung selama t1) : D1(t) = Qc t dan, Arus tahap-2 (berlangsung selama (t2-t1): D2(t) = Qc t

Jadi keberangkatan adalah konstan baik pada arus tahap-1 maupun tahap-2. Kendaraan yang tidak dapat memasuki daerah sempit selama arus tahap-1 akan tertahan di awal daerah sempit dan membentuk antrian. Laju tertahannya kendaraan dalam antrian adalah Q1-Qc. Pada arus tahap-2 tidak ada lagi kendaraan yang tertahan karena Q2<Qc dan kendaraan yang antri berkurang dengan laju Q2-Qc. Saat t = t1 jumlah kendaraan yang tertahan dalam antrian mencapai maksimum (nm) dan besarnya,

nm = A1(t1) – D(t1) [kendaraan] (1)

94


Bila arus pada tahap-2 berlangsung cukup lama, antrian akan hilang pada saat t = t2. Waktu t2 ini juga merupakan lamanya kemacetan akibat adanya daerah sempit di jalan bebas hambatan dan dapat dicari dari lamanya arus tahap-1 (t1) ditambah lamanya arus tahap-2 (t2-t1).

)Q(Q

nt

)Q(Q

a)-(bt)tt(tt

2p

m1

2p11212

[jam] )Q(Q

ntt

makamacet saat t tt tbila

2p

m1j

jj2

Total kendaraan yang berangkat selama waktu macet (=tj = t2) adalah D(t2)

Bila D(t2) = Nj maka,

Nj = Qc tj [kendaraan] (3)

Tundaan tiap kendaraan digambarkan oleh jarak horizontal antara kurva kedatangan (kurva permintaan) dan kurva keberangkatan (kurva kapasitas) dengan asumsi tidak ada penyalipan. Dengan demikian tundaan maksimum (dm) terjadi pada kendaraan terakhir pada arus tahap-1 yaitu,

1

11

p

11m Q

)t(A

Q

)t(Ad

atau,

jadi,

p

mm Q

nd [jam] (4)

Tundaan total (D), yaitu tundaan semua kendaraan yang dipengaruhi oleh kemacetan,dapat diperoleh dengan menghitung luas segitiga (bagian yang dihitamkan) pada Gambar 2 yaitu,

D = ½ nm t1 + ½ nm (t2-t1)

atau,

D = ½ nm{(t1+(tj-t1)} [kendaraan jam] (5)

Tundaan rata-rata (d) dari semua kendaraan yang terpengaruh kemacetan adalah tundaan total (D) dibagi semua kendaraan yang terpengaruh kemacetan (Nj),

jN

Dd [jam] (6)

Model Gelombang Kejut

Dimisalkan hubungan arus (Q), kepadatan (K), dan kecepatan (U) dibagian hulu daerah sempit mengikuti metode Greenshield yaitu (lihat Gambar 3a, 3b, dan 3c) yaitu,

a. Untuk hubungan U-K: KK

UUU

j

ff

Dari hubungan Q = U K diperoleh,

b. Untuk hubungan Q-U: UKUU

KQ j

2

f

j

c. Untuk hubungan Q-K: KUKK

UQ f

2

j

f

keterangan: Uf = kecepatan bebas dan Kj = kepadatan maksimum

Bila suatu saat arus menjadi Q1 (>Qc) dengan kepadatan K1, di awal daerah sempit akan timbul gelombang kejut dan akan merambat kedaerah hulu dengan kecepatan s1{besarnya s1 sesuai kemiringan garis (K1,Q1) dan (Kc,Qc) pada Gambar 3c} yaitu,

Gambar 2. Kurva kedatangan (permintaan) dan keberangkatan (kapasitas) pada model antrian

(2)

95


p1

p11 K-K

-QQs [km/jam] (7)

Panjang antrian akan meningkat kearah hulu hingga waktu t1 seperti ditunjukkan Gambar 4. Setelah arus tahap-1 (lamanya t1) selesai, arus tahap-2 mulai. Pada tahap ini gelombang kejut lain terbentuk dan akan merambat kehilir dengan laju s2 {besarnya s2 sesuai kemiringan garis (K2,Q2) dan (Kc,Qc) pada Gambar 3.c} yaitu,

p2

p22 K-K

-QQs [km/jam] (8)

dalam keadaan ini antrian memendek dengan laju s2 dan akhirnya antrian hilang saat waktu mencapai t2 (Gambar 4). Antrian terpanjang (maksimum) terjadi ketika t = t1 dengan panjang,

y1 = s1 t1 [km]

Bila ym = y1 maka,

ym = s1 t1 [km] (9)

kepadatan kendaraan dalam antrian adalah Kc, maka jumlah kendaraan maksimum dalam antrian (nm) adalah,

nm = Kc ym [kendaraan] (10)

kemacetan di hulu antrian berlangsung hingga t2 (=tj) dan besarnya adalah,

tj = t1 + (t2-t1)

tj = t1 + 2

m

s

y [jam] (11)

Total kendaraan yang berangkat (lepas dari antrian) selama waktu macet (=tj = t2) adalah Nj maka,

Nj = Qc tj [kendaraan] (12)

Tundaan Total (D) dapat diperoleh dengan menghitung luas segitiga,

D = ½ t2 y1

karena t2 = tjdan y1 = ym maka,

D = ½ tj ym [km jam] (13)

kalau diinginkan satuan tundaan total dalam kendaraan jam maka,

D = ½ t2 y1Kc [kendaraan jam] (14)

Tundaan rata-rata (d) dari semua kendaraan yang terpengaruh kemacetan adalah tundaan total (D) dibagi semua kendaraan yang terpengaruh kemacetan (Nj),

Gambar 3. Hubungan Q-U-K untuk analisis gelombang kejut

96


jN

Dd [jam] (15)

Tundaan maksimum (dm) adalah tundaan yang dialami kendaraan pada antrian terpanjang dan dapat dicari melalui pembagian jumlah kendaraan maksimum dalam antrian (nm) dengan Qc,

p

mm Q

nd [jam] (16)

Penerapan Model

Kasus 1

Pandang jalan 1 lajur yang menyempit karena perbaikan jalan. Misalkan kapasitas daerah sempit Qc = 1200 kend/jam (kendaraan per jam), arus tahap-1 Q1 = 1500 kend/jam, dan arus tahap-2 Q2= 900 kend/jam. Misalkan pula arus tahap-1 berlangsung 1 jam (t1=1), kecepatan arus bebas (Uf) = 100 km/jam, dan kepadatan maksimum (Kj) = 100 kend/km.

a. Model antrian

A1(t)=1500 t, A2(t)=900 t, dan D(t)=1200 t

Panjang antrian maksimum = nm = A1(t1) – D(t1) = 1500(1) – 1200(1) = 300 kend

Lamanya kemacetan =

jam 2 )900-(1200

3001

)Q(Q

ntt

2p

m1j

Kendaraan total yang berangkat ketika kemacetan berlangsung (Nj),

Nj = Qctj =1200 (2) = 2400 kend

Tundaan maksimum

jam 0,251200

300

Q

nd

p

mm

Tundaan total (D)

D = ½ nm{(t1+(tj-t1)}

= ½ 300{1+(2-1)}= 300 kend jam

Tundaan rata-rata (d),

jam 125,02400

300

N

Dd

j

b. Model gelombang kejut

Uf = 100 km/jam dan Kj= 100 kend/km

Hubungan U-K:

Hubungan Q-K:

Hubungan Q-U:

Qc=1200 kend/jam 1200 = - K2+100K Kc=86 kend/km

Qc=1200 kend/jam 1200 = - U2+100U Uc=14 km/jam

atau U = 100 – K = 100 – 86 Uc = 14 km/jam

Q2=900 kend/jam 900 = - K2+100K K2=10 kend/km

Q2=900 kend/jam 900 = - U2+100U U2=90 km/jam

atau U = 100 – K = 100 – 10 Uc = 90 km/jam

Q1=1500 kend/jam 1500 = - K2+100K K1=18 kend/km

Q1=1500 kend/jam 1500 = - U2+100U U1=82 km/jam

Gambar 4. Laju panjang antrian akibat jalan menyempit di jalan bebas hambatan

97


Tabel 1. Karakteristik antrian akibat penyempitan jalan bebas hambatan

Parameter Model

Perbedaan Antrian Gelombang kejut

Panjang antrian maks (nm) 300 kendaraan 380 kendaraan 27%

Lamanya antrian (tj) 2 jam 2,12 jam 23%

Jumlah kend. Berangkat (Nj) 2400 kendaraan 2554 kendaraan 6%

Tundaan total (D) 300 kendaraan jam 402 kendaraan jam 34%

Tundaan rata-rata (d) 0,125 jam 0,375 jam 28%

Tundaan maksimum (dm) 0,25 jam 0,32 jam 28% Sumber: Hasil perhitungan (2011)

atau U = 100 – K = 100 – 18 U1 = 82 km/jam

Kecepatan gelombang yang menuju ke hulu adalah s1

Kecepatan gelombang yang menuju ke hilir adalah s2

Panjang antrian maksimum (dalam km)

ym = s1 t1 = 4,41 (1) = 4,41 km

atau (dalam kendaraan)

nm = Kc ym = 86(4,41) = 380 kendaraan

Lamanya antrian (=tj),

tj = t1 + 2

m

s

y=1 +

3,95

4,41 = 2,12jam

Total kendaraan yang berangkat (lepas dari antrian) selama waktu macet (Nj) adalah

Nj = Qc tj = 1200 (2,12) = 2544kendaraan

Tundaan Total (D) adalah

D = ½ tj ym = ½ (2,12) (4,41) = 4,47 km jam

atau dalam kendaraan jam,

D = ½ tj ym Kc = ½ (2,12) (4,41) (86)

= 402 kend jam

Tundaan rata-rata (d) adalah,

jam375,02520

945

N

Dd

j

Tundaan maksimum (dm) adalah,

jam 0,321200

380

Q

nd

p

mm

Ringkasan perhitungan karateristik antrian dari kedua model dapat dilihat pada Tabel 1. Dari Tabel 1 terlihat bahwa panjang antrian maksimum hasil model gelombang kejut 27% lebih tinggi daripada hasil model antrian, sedangkan tundaan hasil model gelombang kejut rata-rata 30% lebih tinggi daripada hasil metode antrian. Jumlah kendaraan berangkat hasil gelombang kejut 6% lebih tinggi daripada hasil metode antrian, sedangkan lamanya antrian hasil gelombang kejut 23% lebih tinggi daripada hasil metode antrian.

Kasus 2

Bila dicoba (untuk jalan 3 lajur) Qc = 4000 kend/jam, arus tahap-1 Q1 = 5000 kend/jam, dan arus tahap-2 Q2 = 3000 kend/jam. Misalkan pula arus tahap-1 berlangsung 1 jam (t1=1), kecepatan arus bebas (Uf) = 61 km/jam, dan kepadatan maksimum (Kj) = 393 kend/km.

Semua perhitungan dilakukan per lajur

a. Model antrian

A1(t)=⅓ (5000) t, A2(t)=⅓(3000) t, dan D(t)=⅓(4000) t

98


Panjang antrian maksimum (nm)

nm = ⅓[A1(t1) – D(t1)] = ⅓ [5000(1) – 4000(1)] = ⅓(1000) kend = 334 kend

Lamanya kemacetan (tj)

Kendaraan total yang berangkat ketika kemacetan berlangsung (Nj),

Nj = ⅓ Qctj =⅓(4000) (2) = ⅓(8000) kend

= 2667 kend

Tundaan maksimum (dm)

jam 0,254000

1000

31

31

Q

nd

p

mm

Tundaan total (D)

D = ½ ⅓ nm{(t1+(tj-t1)}

= ½ ⅓ 1000{1+(2-1)}=⅓(1000) kend jam

= 334 kend

Tundaan rata-rata (d),

jam 125,08000

1000

31

31

N

Dd

j

a. Model gelombang kejut

Uf = 61 km/jam dan Kj = ⅓ (393) = 131 kend/km

Hubungan U-K:

Hubungan Q-K:

Qc = ⅓ 4000 kend/jam ⅓ 4000 = - 0,5K2+61 K Kc = 102,5 kend/km

U = 61 – 0,47 K = 61 – 0,47(102,5) Uc = 12,82 km/jam

Q2=⅓3000 kend/jam ⅓ 3000 = - 0,47K2+61 K K2=19 kend/km

U = 61 –0,47K = 61 –0,47(19) U2 =52,07 km/jam

Q1=⅓5000 kend/jam ⅓ 5000 = - 0,47K2+61 K K1=38,5 kend/km

U = 61 –0,47 K = 61 – 0,47(38,5) U1 = 42,90 km/jam

Kecepatan gelombang yang menuju ke hulu adalah s1

Kecepatan gelombang yang menuju ke hilir adalah s2

Panjang antrian maksimum (dalam km)

ym = s1 t1 = 5,2 (1) = 5,2 km

Tabel 2. Karakteristik antrian akibat penyempitan jalan bebas hambatan

Parameter Model

Perbedaan Antrian Gelombang kejut

Panjang antrian maks (nm) 334 kendaraan 533 kendaraan 60%

Lamanya antrian (tj) 2 jam 2,3 jam 15%

Jumlah kend. Berangkat (Nj) 2667 kendaraan 3067 kendaraan 15%

Tundaan total (D) 334 kendaraan jam 613 kendaraan jam 84%

Tundaan rata-rata (d) 0,125 jam 0,2 jam 60%

Tundaan maksimum (dm) 0,25 jam 0,4 jam 60%

Sumber: Hasil perhitungan (2011)

99


atau (dalam kendaraan)

nm = ⅓ Kc ym = ⅓ 307,5(5,2) = 533 kendaraan

Lamanya antrian (=tj),

tj = t1 + 2

m

s

y=1 +

4

5,2 = 2,3 jam

Total kendaraan yang berangkat (lepas dari antrian) selama waktu macet (Nj) adalah

Nj = ⅓ Qc tj = ⅓ 4000 (2,3) = 3067 kendaraan

Tundaan Total (D) adalah

D = ½ tj ym = ½ (2,3 ) (5,2) = 5,98 km jam

atau dalam kendaraan jam,

D = ½ tj ym Kc = ½ ⅓ (2,3) (5,2) (307,5) = 613 kend jam

Tundaan rata-rata (d) adalah,

jam2,03067

613

N

Dd

j

Tundaan maksimum (dm) adalah,

jam 0,4(4000)31

533

Q

nd

p

mm

Ringkasan perhitungan karateristik antrian dari kedua model dapat dilihat pada Tabel 2.

PEMBAHASAN Dari Tabel 1 dan 2 terlihat bahwa terdapat perbedaan nilai untuk parameter yang sama hasil prediksi dari kedua model baik untuk kasus yang sama maupun berbeda.

Panjang antrian: Model gelombang kejut 27% lebih tinggi daripada hasil model antrian (Tabel 1), sedangkan pada Tabel 2 hasil gelombang kejut 60% lebih tinggi daripada hasil model antrian.

Tundaan: Hasil model gelombang kejut rata-rata 30% lebih tinggi dari pada hasil model antrian sedangkan (Tabel 1), sedangkan pada Tabel 2 hasil gelombang kejut rata-rata 68% lebih tinggi daripada hasil model antrian.

Lamanya antrian: Hasil model gelombang kejut rata-rata 30% lebih tinggi dari pada hasil model antrian sedangkan (Tabel 1), sedangkan pada Tabel 2 hasil gelombang kejut rata-rata 68% lebih tinggi daripada hasil model antrian.

Perbedaan hasil ini dapat dijelaskan karena adanya perbedaan konsep antrian, tundaan, dan tahap arus yang dipunyai model antrian dan model gelombang kejut. Konsep Antrian

Model antrian menggambarkan kendaraan yang akan memasuki daerah sempit sebagai antrian berhenti yang hanya bergerak ketika dilayani oleh daerah sempit dengan laju sesuai kapasitas daerah sempit. Jadi model ini tidak melihat bagaimana kendaraan bergerak dalam antrian, mereka hanya menunggu dalam antrian dan bergerak ketika dilayani. Panjang antrian maksimum (Nm) diperoleh dengan dengan mengurangkan arus berangkat dari arus datang, apa yang terjadi dalam antrian tidak dilihat hingga hasilnya pasti berbeda dengan model yang mempertimbangkan bahwa antrian sebenarnya bergerak. Model gelombang kejut menggambarkan antrian yang mendekati daerah sempit sebagai antrian bergerak dan pergerakan arus antrian ini memenuhi hubungan arus (Q), kecepatan (U), dan kepadatan (K) di daerah bebas hambatan yaitu Q = U K. Antrian bergerak muncul ketika arus datang melampaui kapasitas. Saat itu timbul gelombang dengan laju tertentu dan berarah ke hulu. Panjang antrian maksimum diperoleh ketika arus datang yang lebih kecil daripada kapasitas daerah sempit tiba. Ini ditandai dengan bergeraknya gelombang yang tadinya kehulu sekarang berbalik kehilir. Kedua model akan menghasilkan panjang antrian yang sama ketika kapasitas daerah sempit (Qc) = 0 atau terjadi penutupan sempurna didaerah sempit. Konsep Tundaan

Karena berbeda dalam konsep antrian, tundaan yang dihasilkan dari kedua model tentu saja juga berbeda. Model antrian mengasumsikan kendaraan dalam keadaan diam saat mengantri sehingga,tundaan dapat dihitung dengan benar

100


karena mewakili waktu hilang ketika kendaraan berhenti. Tidak demikian dengan model gelombang kejut yang menggambarkan antrian dalam keadaan bergerak, tundaan yang diperoleh sebenarnya adalah tambahan waktu perjalanan karena melewati antrian. Oleh karena itu, untuk mencari tundaan waktu perjalanan tanpa hambatan harus diketahui terlebih dahulu. Konsep Tahap Arus

Pada model antrian maupun model gelombang kejut, arus tahap-1 berlangsung selama t1. Asumsi ini cocok untuk model antrian karena model ini tidak mempermasalahkan bergeraknya antrian atau dengan kata lain hanya mengacu pada dimensi waktu. Tidak demikian halnya untuk model gelombang kejut, model kedua ini mengasumsikan kendaraan bergerak dan mempunyai hubungan yang ditetapkan sebelumnya yaitu Q=UK, artinya selain mengacu pada waktu model kedua ini juga mengacu pada ruang. Oleh karena itu, sangat penting untuk menyatakan dimana letak perubahan arus tersebut di dalam ruang. Misalkan kendaraan pertama pada arus Q1 tiba di awal daerah sempit atau disebut titik 0 (titik ketika waktu antrian dimulai) setelah menempuh waktu - ∆1 dari suatu tempat sejauh z1 di hulu daerah sempit. Kecepatan arus Q1 ini adalah U1 (=z1/∆1). Kendaraan terakhir pada arus Q1 ini seharusnya tiba di titik 0 setelah t1 namun kenyataannya tidak, hal ini dikarenakan ada rambatan gelombang berarah kehulu yang berawal dari titik 0. Kendaraan terakhir pada arus Q1 misalnya bertemu gelombang yang merambat ke hulu di titik yp pada waktu tp (tentu saja yp<y1 dan tp<t1).Waktu tp dapat dicari dari perjalanan kendaraan terakhir maupun dari perjalanan gelombang.

Dari kendaraan terakhir:

Dari perjalanan gelombang:

Diperoleh:

yp adalah panjang antrian maksimum. Bila pm

p yy maka,

Tundaan maksimum ( pmd ) diperoleh dengan

mengurangkan waktu perjalanan kendaraan terakhir arus tahap-1 (kecepatan U1)(bila tak ada penyempitan) dari waktu perjalanan kendaraan terakhir ketika ada penyempitan (kecepatan Uc) yaitu,

Tundaan total (Dp) ditentukan dengan mengurangkan waktu perjalanan semua kendaraan yang tidak terpengaruh kemacetan (kecepatan U1) dari waktu perjalanan semua kendaraan yang terpengaruh kemacetan (kepadatan Kp) yaitu,

.................................................................. (22)

Tundaan rata-rata (dp)

keterangan:

Penerapan Model Gelombang Kejut Modifikasi

a. Dari data Kasus 1

Q1 = 1500 kend/jam, K1=18 kend/km, U1=82 km/jam, Qc = 1200 kend/jam, Kc=86

101


kend/km, Uc=14 km/jam, Q2=900 kend/jam, K2=10 kend/km, U2=90 km/jam, t1=1, Uf=100 km/jam, Kj=100 kend/km, s1=-4,41 km/jam, dan s2 = 3,95 km/jam.

U1=82 km/jam dan t1 = 1 jam, maka z1 = 82 km

Panjang antrian maksimum

km 18,441,482

(4,41) 82

sU

szy

11

11pm

atau dalam satuan kendaran

kend 360)18,4(86yKn pmc

pm

Lama antrian

Total kendaraan yang berangkat (lepas dari antrian) selama waktu macet (Nj

p) adalah,

kend 2408)006,2(1200tQN pjc

pj

Tundaan maksimum

jam 25,082

4,18

14

4,18

U

y

U

yd

1

pm

c

pmp

m

Tundaan total = Dp


jam 126,02408

302,61

N

Dd

pj

pp

b. Dari data Kasus 2

Data ditampilkan per lajur: Q1=⅓(5000) kend/jam, K1=38,5 kend/km, U1=41,75 km/jam, Qc=⅓(4000) kend/jam, Kc=102,5

kend/km, Uc=9,75 km/jam, Q2=⅓(3000) kend/jam, K2=19 kend/km, U2=51,5 km/jam, t1=1 jam, Uf=61 km/jam, Kj=⅓(393) kend/km, s1=-5,2 km/jam, dan s2=4 km/jam

U1=42,90 km/jam dan t1 = 1 jam, maka z1 = 42,90 km

Panjang antrian maksimum

km 64,42,542,90

(5,2) 42,90

sU

szy

11

11pm

atau dalam satuan kendaran

kend 476)64,4(5,102yKn pmc

pm

Lama antrian

Total kendaraan yang berangkat (lepas dari antrian) selama waktu macet (Nj

p) adalah,

kend 2736)052,2(400031tQN p

jcpj

Tundaan maksimum

jam 25,042,90

4,64

12,82

4,64

U

y

U

yd

1

pm

c

pmp

m

Tundaan total = Dp


jam 130,0

2736

356

N

Dd

pj

pp

Ringkasan hasil gelombang kejut modifikasi dengan hasil model antrian baik untuk kasus 1 maupun kasus 2 dapat dilihat pada Tabel 3.

102


Tabel 3. Perbandingan karakteristik antrian dari dua metode (antrian dengan gelombang kejut)

Parameter

Model Perbedaan

Antrian Gel. Kejut Modifikasi

Kasus-1 Kasus-2 Kasus-1 Kasus-2 Kasus-1 Kasus-2

Panjang antrian maks (nm) 300 334 360 476 20% 42%

Lamanya antrian (tj) 2 2 2 2,05 0% 2,5%

Jumlah kend. berangkat (Nj) 2400 2667 2408 2736 0,3% 2,6%

Tundaan total (D) 300 334 303 356 1% 6,6%

Tundaan rata-rata (d) 0,125 0,125 0,126 0,13 0,8% 4%

Tundaan maksimum (dm) 0,25 0,25 0,25 0,25 0% 0%

Sumber: Hasil perhitungan (2011) Dari hasil perhitungan ulang dengan metode gelombang kejut modifikasi terlihat (Tabel 4.) bahwa perbedaan dengan metode antrian turun namun belum hilang sama sekali (terutama parameter panjang antrian yang masih berbeda cukup banyak yaitu rata-rata 30%). Perbedaan ini disebabkan pengabaian interaksi antara akhir arus tahap 1 dan awal arus tahap 2. Bila interaksi diperhitungkan maka bentuk rambatan gelombang kehilir tidak lagi linier namun menjadi non linier yang membawa akibat panjang antrian memendek. Pengabaian dilakukan karena bentuk rambatan gelombang, baik ke hulu maupun ke hilir, sudah diasumsikan linier. SIMPULAN DAN SARAN 1. Model antrian, yang modelnya berbasiskan

waktu, memodelkan kemacetan akibat penyempitan di jalan bebas hambatan lebih sederhana dibandingkan dengan model gelombang kejut, yang modelnya berbasiskan ruang dan waktu. Walaupun begitu model gelombang kejut lebih realistis karena modelnya dapat mendekati keadaan sebenarnya di lapangan.

2. Hasil prediksi terhadap parameter kemacetan dari kedua model berbeda. Walaupun begitu perbedaan dapat diperkecil dan dapat dikatakan sebanding apabila dalam perhitungannya dipergunakan metode gelombang kejut modifikasi.

3. Untuk studi selanjutnya disarankan untuk

mempelajari bentuk interaksi antara akhir arus tahap 1 dan awal arus tahap 2.

4. Karena hasil prediksi kedua model berbeda (walaupun tidak signifikan), penggunaan praktis dilapangan (model mana yang dipakai) harus diserahkan pada ahlinya. Konfirmasi dengan keadaan sebenarnya perlu dilakukan agar revisi model (apabila diperlukan) dapat dilakukan hingga model yang betul-betul mewakili dapat ditemukan.

DAFTAR PUSTAKA Garber, N. J. and Hoel, L. A. (2009). “Traffic

and Highway Engineering”. 4th Ed., Toronto, Canada.

May, A. D. (1990). “Traffic Flow Fundamentals”. Englewood Cliffs: Prentice-Hall Inc., New Jersey.

McShane, W. and Roess, R. P. (1990). “Traffic Engineering”. Englewood Cliffs: Prentice-Hall Inc., New Jersey.

Morales, J. M. (1986). “Analytic Procedures for Estimating Freeway Traffic Congestion”. Public Roads, Vol. 50, No.2, Washington, D.C.

Papacostas, C. S. and Prevedouros, P. D. (1993). “Transportation Engineering and Planning”. Englewood: Prentice-Hall Inc., New Jersey.

Shandutan, D. A. S. (2009). “A Queuing Analysis of Bottleneck Formation and Shockwave Propagation”. Master Thesis, (Unpublished), University of Minnesota, Minnesota.

103

journal uii yogya

Documents

pengikut yang

net normal

sungai merah

lepas dari

selama waktu

grass root

net normal

net normal