foundation design and construction in · pdf filefoundation design and construction in...

FOUNDATION DESIGN AND CONSTRUCTION INLIMESTONE FORMATION: A MALAYSIANCONSULTANT’S EXPERIENCE

Tan Yean-Chin and Chow Chee-Meng

G&P Geotechnics Sdn Bhd, Kuala Lumpur

The design and construction of foundations in limestone formation have posed var-ious challenges to geotechnical engineers due to the karstic features of limestonesuch as steeply inclined bedrock, slime zone, cavities, floaters, etc. The designof foundations in such highly irregular ground conditions requires careful planningand execution of the works from preliminary to detailed subsurface investigation,analysis and design, and up to the construction stage where continuous feed-back is essential for the satisfactory performance of the foundations. This paperpresents some aspects of foundation design and construction practice in limestoneformation in Malaysia with particular emphasis on the design of piled foundationssuch as driven piles, jacked-in piles and bored piles. A novel approach for verifica-tion of rock socket for bored piles socketed into limestone bedrock which has beensuccessfully implemented in various projects since 2005 is also presented.

1. INTRODUCTION

The design and construction of foundations in limestone formation have posed variouschallenges to geotechnical engineers due to the karstic features of limestone such as steeplyinclined bedrock, slime zone, cavities, floaters, etc. Karst refers to a characteristic topo-graphic feature or landscape which can be developed by rock undergoing dissolutionby percolating meteoric water (Jakucs, 1977). Karst occurs primarily on limestones (anddolomites), and ground cavities and dissolutional landforms develop best on competent,fractured rocks whose intact strength is generally 30-100 MPa. Weaker limestones, chalkand unlithified carbonate sediments lack the strength to span large cavities, and developlimited suites of karst features that are generally smaller than those on stronger limestones(Jennings, 1968; Waltham and Fookes, 2003). In Peninsular Malaysia, under tropical humidconditions, calcite and dolomite limestones or their metamorphised equivalents developtropical features which show spectacular tall steep-sided hills (Jennings, 1982) and solu-tion features such as pinnacles, sinkholes and cavities. The treacherous and almost unpre-dictable karstic bedrock associated with extremely variable overburden soil properties isa typical feature of limestone (Yeap, 1985), which leads to a variety of geotechnical prob-lems and hazards. Some of these common engineering problems are discussed below (Gue,1999) and Figure 1 shows some typical piling problems in limestone formation.

Advances in Foundation EngineeringEdited by K. K. Phoon, T. S. Chua, H. B. Yang and W. M. ChamCopyright © 2014 Research Publishing Services.ISBN: 978-981-07-4623-0 :: doi:10.3850/978-981-07-4623-0 KN-05 67

68 Advances in Foundation Engineering

Figure 1. Some piling problems in limestone formation (Neoh, 1998).

1.1. Pinnacles

Pinnacles are columns or cones of limestone or marble left by dissolution of the surround-ing rock. Pinnacles with soft or loose overburden immediately above them pose tremen-dous challenges to geotechnical engineers to ensure proper seating of piles on the rock,particularly for driven concrete piles, as well as the difficulty of rock socketing for boredpiles.

1.2. Subsidence

Subsidence is described as several related localized or already widespread phenomenaassociated with the sinking of the ground. The vertical movements ranged from suddencollapse to slow settlement. Generally, subsidence can be caused by the lowering ofgroundwater level, drainage or oxidation of organic soil and surface collapse into natu-ral or excavated subsurface cavities had taken place. The formation of subsidence in theKuala Lumpur and Ipoh areas is often associated with the occurrence of soft mining slimein ex-mining areas upon which housing and roadwork projects were carried out. This isdue to the consolidation of the underlying slime/clay upon loading.

1.3. Sinkhole

This is a common phenomenon in karst areas, especially areas with loose and non-cohesivesands over limestone bedrock. It is commonly known that limestone can be dissolved byacidic solution from rain or polluted groundwater. After a certain period of time, flow orpenetration of groundwater through weak zones in the limestone will develop channelsand voids. These channels will act as passages for water together with the loose sand to

Foundation Design and Construction in Limestone Formation 69

Stage 1

LIMESTONE BEDROCK

Cavity

OVERBURDEN SUBSOIL

Arching

Stage 2

LIMESTONE BEDROCK

Stage 3 Stage 4

LIMESTONE BEDROCK

LIMESTONE BEDROCK

Sinkhole

Figure 2. Typical mechanism of sinkhole formation (Gue, 2013).

Figure 3. Sinkhole in Mata Ayer, Perlis.

flow into voids and cavities in the limestone. The movement of sand into existing cavitiesor voids will then develop empty spaces in the sand layer where soil arching occurred. Thiscontinuous process will reach a critical stage where the roof of the space or the soil archingcan no longer support the weight of the overburden. This will result in the occurrence ofsinkholes or cave-ins as illustrated in Figure 2. Figures 3 and 4 show examples of previousoccurrences of sinkholes in Malaysia. Sometimes, the delicate balance of the arching mech-anism that supports the overburden soil can also be disrupted by lowering of groundwaterlevel due to construction activities (e.g. excavation dewatering) or by dynamic loadingsfrom sources such as earthquakes and railway lines. Sinkholes are hazards to both shallowand deep foundations as their occurrence is sudden and catastrophic and certainly posetremendous challenges to the geotechnical engineering profession.


Figure 4. Sinkhole near Ipoh, Perak.

1.4. Slump Zone

Zones of weakness often occur immediately above the bedrock of limestone. The slumpzone above limestone formation is usually identified by the very low SPT-N values or lowcone resistance where SPT-N values of zero are often detected. The formation is either dueto subsurface erosion as a result of overburden slumping into cavities in the limestone orresidual weathering of ancient karst features (Tan and Ch’ng, 1986).

1.5. Cavern/Cavity

Cavities are voids formed by dissolution of the rock in the limestone which will pose prob-lems if the roof of the cavities is not of sufficient thickness and strength to support thefoundation resting on them, especially for empty or partially filled cavities (Figure 5). Thelargest single cave chamber is over 300 m wide and 700 m long, in a cave in the Mulu

Figure 5. Cavern/cavity exposed after excavation.


karst of Sarawak, Malaysia. All voids in a block of karstic limestone are interconnectedbecause they were formed by through drainage; narrow fissures, wide river passages andlarge chambers are merely elements of a cave system. Though cave morphology may beunderstood in terms of limestone geology and geomorphic history, the distribution of caveopenings in an unexplored limestone mass cannot be predicted. Unkown cave locationsremain a major problem in civil engineering (Culshaw and Waltham, 1987; Waltham andFookes, 2003).

1.6. Steeply Inclined Bedrock Surface

Steeply inclined bedrock surface in limestone posed significant difficulties for piled foun-dations such as inadequate pile capacity for bored piles socketed into limestone (Figure 6)and tilting of piles and pile breakage during installation of driven piles, especially forbedrock with adverse geological features such as vertical joints as shown in Figure 7 whichare common in Malaysia.

BORED PILE

Figure 6. Pile breakage in steeply inclined bedrock surface with adverse geological features.

Figure 7. Vertical joints in exposed limestone in the Kuala Lumpur area.


2. ENGINEERING CLASSIFICATION OF KARST

Waltham and Fookes (2003) proposed a classification system for karst to assist civil/geotechnical engineers identify the degree of variability of the karst features. The classi-fication system is summarized in Table 1 and Figure 8. In tropical regions with high rain-fall such as Malaysia, the karst are usually type kIV or kV with possible cave widths of10 m. As such, the recommended termination criterion for subsurface investigation worksfor design of foundations on limestone is continuous solid rock coring of 10 m in the lime-stone bedrock. This criterion also minimize risk of instability of natural rock roof coverover cavity/cave as informal guideline to the stability of the natural rock roof over a caveis that it is stable if the thickness of rock is equal to or greater than its span; this excludes

Table 1. An engineering classification of karst (Waltham and Fookes, 2003).

An engineering classification of karst. This table provides outline descriptions of selected parame-ters: these are not mutually exclusive and give only broad indications of likely ground conditionsthat can show enormous variation in local detail. It should be viewed in conjunction with Figure 8,which shows some of the typical morphological features. The comments on ground investigationand foundations are only broad guidelines to good practice in the various classes of karst. NSH =rate of formation of new sinkholes per km2 per year.


Figure 8. Typical morphological features of karstic ground conditions within the five classes of theengineering classification of karst (Waltham and Fookes, 2003).

any thickness of soil cover or heavily fissured limestone at rock head (Waltham and Fookes,2003). According to Waltham and Fookes (2003), this guideline is conservative as in typicallimestone karst the rock mass is of fair quality (Class III), with Q = 4–10 on the clas-sification scheme of Barton et al. (1974), and RMR = 40–60 on the rock mass rating ofBieniawski (1973). In such material, a cover thickness of intact rock that is 70% of the cave


Figure 9. The stability of cave roofs in limestone under engineering imposed load, related to thethickness and structural morphology of the roof rock. Data points are derived from destructive testsof laboratory scale models of caves all 4 m wide in limestone with UCS of about 80 MPa, centrallyloaded by foundation pads of 1 m2. Scale factors were calibrated by numerical modelling to a full-scale test, and the required loading capabilities are for Safe Bearing Pressures of 2 MPa and 4 MPamultiplied by a Factor of Safety of 3 (Waltham and Fookes, 2003).

width ensures integrity (Figure 9) under foundation loads that do not exceed 2 MPa –which is half the Safe Bearing Pressure appropriate for sound limestone.

3. FOUNDATION DESIGN

Previously, bored piles and barrettes are generally used for high-rise buildings whiledriven and jacked-in piles are adopted for low-rise buildings. However, with the introduc-tion of higher capacity jacked-in piling systems, jacked-in piles were also adopted for high-rise buildings. Common sizes of bored piles range between 600 mm to 1800 mm diameteralthough 3000 mm diameter has also been installed in Malaysia. Barrettes are sometimesused in thick overburden (e.g. more than 80 m) where bored pile machines have difficultyreaching the required depth. Hand-dug caisson foundations have also been designed andconstructed in Malaysia. Yee and Yap (1998) reported the design and construction of hand-dug caisson foundations of up to 3.0 m diameter for a 33-storey tower in Kuala Lumpurtogether with other measures such as cavity detection, replacement grouting and verifica-tion of cavity grouting to ensure satisfactory foundation performance in the highly unpre-dictable limestone formation.

3.1. Driven and Jacked-In Pile

In this section, some general design considerations associated with driven and jacked-inpiles in limestone formation are discussed. Driven piles have been installed in limestoneformations in Malaysia with varying degree of success. Various problems such as excessivetilting/deflection, rotation, distortion, buckling/bending, cracking, shattering, etc leading


Figure 10. A typical hydraulic jacked-in piling system used in Malaysia.

to high percentage of damaged piles could be expected if inadequate attention is given dur-ing the design and construction of the piles in limestone formation. Recently, high capacityjacked-in piles have also been installed in Malaysia. Pile capacities of up to 3000 kN havebeen installed in Malaysia where the piles are jacked to at least two times of their capacity,i.e. 6000 kN. Pile capacities of up to 3000 kN are normally adopted in formations such asgranite or Kenny Hill while for limestone, lower pile working capacities are usually spec-ified. Figure 10 shows a typical hydraulic jacked-in piling system used in Malaysia wherejacking force of up to 7000 kN can be achieved.

Some design aspects related to driven and jacked-in piles in limestone formation are asfollows:

a) Steeply inclined bedrock surfaceb) Floater/boulderc) Cavity

3.1.1. Steeply Inclined Bedrock Surface

In Malaysia, the design of driven/jack-in pile foundations to cater for highly erraticbedrock profiles and inclined bedrock typically associated with limestone formationinvolves the following:

i) Provision of compensation piles within the pile group (if necessary) to ensure that theinduced rotation within the group is within tolerable limits, i.e. within the bendingmoment capacity of the pile and pilecap-column connection and no piles within thegroup are overstressed. This also applies in situations where significant differencesin pile length are observed within the same pile group due to the highly irregularbedrock profile of limestone formation. Such large differences in pile length will induce


bending moments and also uneven distribution of loads within the pile group due todifferent magnitudes of elastic shortening of the piles. Therefore, provisions for higherpercentages of compensation piles are required for driven/jacked-in piles foundationin limestone formation due to the complex geological settings of limestone formation.In addition, the risk of pile damage during driving is also higher in limestone forma-tion. This, however, can be minimised with competent site supervision andexperienced contractors. In order to reduce occurrences of damaged piles duringinstallation and minimise compensation piles due to overstressing of piles, a generalrule-of-thumb which is commonly adopted for driven/jacked-in pile foundations is todowngrade the pile working load within the range of 50–70% of the allowable struc-tural capacity depending on the bedrock profile, expected pile length, soil properties,etc.

ii) Provision of Oslo-point rock shoes (Bjerrum, 1957) in areas where the overburden soilsare very soft or loose (e.g. slime zone above the bedrock) to prevent pile deflection dur-ing installation and to ensure the pile toe is properly secured to the rock as illustratedin Figure 11. The hardness of the hardened steel used for Oslo-point rock shoes shouldbe larger than 300 (Brinell hardness) and the yield strength of the rock shoe shouldnot be less than 760 MPa. The rock shoe should be designed to take the full requiredload at the contact and extra care should be taken during construction to prevent alter-ing its properties, in particular, by welding. Typical details of Oslo-point pile shoesare shown in Figure 12. Gue (1999) reported successful use of Oslo-point rock shoesto reduce pile damage where the percentage of damaged piles has been controlled to1.5% from the initial 15%. The case involves some 5000 pile points of reinforced con-crete piles of 250 mm × 250 mm and 300 mm × 300 mm with length ranging from 17 mto 76 m and average length of 30 m.

iii) If Oslo-point rock shoes are not provided, the use of conical shoes should be avoided asit encourages slippage during construction and it is better to provide flat-ended shoes(Omar and Hon, 1985).

Figure 11. Mechanism of Oslo-point.


Figure 12. Typical Oslo-point rock shoe details.

3.1.2. Floater/Boulder

The effect of founding the pile foundations on floater/boulder as idealised in Figure 13 isthat the axial load in the piles founded on the intended founding layer increases, whilethe axial load in the piles on the floater decreases. This will induce uneven settlement inthe pile group and hence, rotation is developed which will induce bending moments inthe piles at the pile heads and also causes potential overstressing of piles founded on theintended founding layer.

The following are normally carried out for piles on floater/boulder:

a) Preboring through the floater prior to installation of driven/jacked-in piles. This is toensure that the piles reach the intended founding layer.

b) In the event that the driven/jacked-in piles terminate prematurely on the floater, provi-sion of compensation piles should take into consideration the reduced capacity of thepiles founded on the floater and the pilecap/tiebeam should be sufficiently stiff andadequately designed to span across the piles founded on the floater.


Figure 13. Pile foundation on floater/boulder.

3.1.3. Cavity

If the extent of the cavity is limited, treatment of cavity using grout can be carried out. Theessential steps for successful treatment of cavity and slumpzone involve:

a) Cavity and slumpzone probingb) Injection of grout/mortarc) Verification of cavity grouting

Cavity and slumpzone probing should be carried out using a suitable drilling machineto a minimum depth of 10 m into solid limestone if no cavity is encountered or 10 m belowthe last cavity encountered. Based on the results of cavity and slumpzone probing, therequired treatment should be carried out using grout/mortar according to the followingsequence:

a) If there is more than one drill hole for treatment, generally mortar injection should com-mence around the perimeter of the treatment zone and then proceeding toward thecentre. Each hole should be drilled and grouted before moving to the next hole.

b) In the case of multiple cavities or multiple limestone layers in any drill hole, treatmentshould proceed from the lowest cavity and completed for that cavity before proceedingto the next higher cavity.

c) If required, packer(s) could be adopted to prevent flow out of the grout/mortar beforeachieving the required criteria of acceptance or pressure specified. Each drill hole forgrout treatment may be accompanied by at least one vent hole or pressure release holeof similar depth and size.

Acceptance criteria for cavity treatment using grout/mortar are commonly based on thefollowing:

a) For soils within the treatment zone, the individual SPT-N values at any point are notless than 20 and the average SPT-N value is not less than 25

b) No void is encountered


Figure 14. Typical drilling machine and pump for cavity treatment using grout/mortar.

c) Unconfined compressive strengths of the cores (if required) are in excess of 2 N/mm2or other strength requirements as per design

Figure 14 shows typical drilling machine and pump for cavity treatment using grout/mortar.

Good construction practice is also very important to ensure successful installation ofdriven/jacked-in pile foundations in limestone formation especially for sites where steeplyinclined bedrock and floaters are expected. Some good construction practices aresummarized below:

a) Based on available boreholes and cavity probing points, interpretation of the bedrockprofile is required. The interpreted bedrock profile will serve as reference during piledriving where the hammer height is reduced when approaching the interpreted bedrockprofile to prevent pile slippage. However, the Engineer should be aware that the inter-preted bedrock profile is only a rough guide as the limestone is usually highly irregularin depth and therefore, good engineering judgment must be exercised. When the pilepoint has come into contact with the rock surface which normally can be recognizedby a sudden change in the response of the hammer, pile driving is then continued withvery small heights of drop of the ram (typically about 100 mm to 200 mm). After the pilehas been subjected to a series of blows until the penetration of the pile is negligible, thefall is increased to double the height. The steps are repeated until the required termina-tion criterion is achieved. This procedure is intended to socket the pile into competentbedrock and to prevent sliding of the pile point at the contact with the rock surface. Thesame approach is also applicable for jack-in piles where small jack-in pressure is usedinitially until contact with the rock surface is made. Thereafter, the jack-in pressure isincreased gradually until the required termination criterion.


b) Diesel hammers should be avoided since they cannot be stopped at will and the dropheight of hammer tends to be excessive at high driving resistance (Omar and Hon,1985).

c) High strain dynamic pile test should be used to calibrate the permissible drop heightto prevent damage to piles during installation of driven piles. It also serves as a usefultool for quality control during pile installation and detection of damaged piles. Forpreliminary estimation of pile driving criteria, methods based on wave equations, e.g.using software such as GRLWEAP, should be used to determine the permissible dropheight and set criteria. The use of dynamic formulas (e.g. Hiley) is strongly discouragedas the method is inherently inaccurate for assessing whether a required pile resistancehas been achieved by means of measuring the ‘set’. This is because of the uncertaintiesassociated with the input parameters for the equation such as the hammer efficiency,various components of elastic movement of pile, etc. which, over a period of a fewhours during driving, the characteristics of both the hammer and the cushion may altersignificantly. Thus, piles driven to a given set at the start and end of the period mayhave widely varying capacities.

d) In the event of premature pile termination due to the existence of intermediate hardlenses (high SPT-N value) or small boulders, the problem can be overcome by applyinga higher jack-in force or increasing the driving energy (a heavier hammer is preferredto higher drop heights to reduce potential pile damage). Again, the use of high straindynamic pile tests is recommended to monitor pile stresses during installation of drivenpile in such situations to ensure the compressive and tensile stresses induced in the pileare within tolerable limits. Similarly, software such as GRLWEAP is also useful to studythe pile driving characteristics in such situations prior to implementation at site.

Based on the Authors’ experience, jack-in piles have some advantages over driven pilesin limestone formation due to the following factors:

a) As the application of jack-in force is more gradual compared to the sudden impact of ahammer for driven piles, the tendency for pile damage due to sliding of the pile pointat the contact with the rock surface is reduced for jack-in piles.

b) As the total numbers of load-unload cycles during pile installation is lesser comparedto driven piles, shaft friction in soil for jack-in piles are found to be higher comparedto driven piles (White and Lehane, 2004 and Chow and Tan, 2010). Due to the highercontribution of shaft friction, the pressure onto the limestone bedrock will be reducedand as such, reduces the risk of pile slippage and collapse of cavity roof.

c) As each jack-in pile will be subjected to a minimum jack-in pressure of two times itsworking load, each pile is akin to being subjected to a static load test to verify itsgeotechnical capacity. Even though the loading duration is short, the maximum jack-inpressure which the pile can support during installation is found to give good indica-tions of its geotechnical capacity (Deeks et al., 2005 and Chow and Tan, 2009).

3.2. Bored Pile

In Malaysia, bored pile design in limestone is heavily dependent on semi-empirical meth-ods. Generally, the design rock socket friction is a function of the surface roughness of rock


sockets, the unconfined compressive strength of intact rock, the confining stiffness aroundthe rock socket in relation to fractures of rock mass and socket diameter, and the geome-try ratio of socket length-to-diameter. Roughness is an important factor in rock socket piledesign as it has significant effects on the normal contact stress at the socket interface duringshearing. The normal contact stress increases due to dilation, resulting in increased socketfriction. The degree of dilation is mostly governed by the socket roughness. The secondfactor on the intact rock strength governs the ability of the irregular asperity of the socketinterface transferring the shear force, otherwise shearing through the irregular asperity willoccur due to highly concentrated shear forces from the socket. The third factor will governthe overall performance of strength and stiffness of the rock socket in jointed or fracturedrock mass and the last factor is controlled by the profile of socket friction distribution.It is very complicated to quantify all of these aspects in rock socket pile design. There-fore, based on local experience, some conservative semi-empirical methods have evolvedto facilitate quick and practical rock socket design taking into considerations the factorsdiscussed above. In most cases, roughness of socket is only qualitatively assessed dueto lack of systematic and reliable methods of assessment. The other three factors can bequantified through strength tests on the rock cores and point load tests on the recoveredfragments, RQD values of the core samples and some analytical method of assessing thesocket distribution. It is also customary and important to perform preliminary and work-ing load tests to verify the rock socket design using such semi-empirical methods. A fullyinstrumented preliminary test pile is recommended as it enables proper assessment of themobilised rock socket friction and base resistance for optimisation of pile design. A safetyfactor of two is a common requirement for rock socket pile design. Table 2 summarisestypical design/working socket friction values for limestone formations in Malaysia.

Another more systematic approach developed by Rosenberg and Journeaux (1976),Horvath (1978) and Williams and Pells (1981) is also referred to in Malaysia. The followingsimple expression is used to compute the rock socket friction, fs with consideration of thestrength of intact rock and the rock mass effect due to discontinuities:

fs = α ∗ β ∗ quc (1)

wherequc = unconfined compressive strength of intact rock

α = reduction factor with respect to quc (Figure 15)β = reduction factor with respect to rock mass effect (Figure 16)

During borehole exploration, statistics of quc can be compiled for different weather-ing grades of bedrock and the rock fracture can be assessed through the Rock Quality

Table 2. Summary of rock socket friction design values for limestone formation in Malaysia (Neoh,1998).

Working Rock Socket Friction Remarks

300 kPa for RQD < 25% The working/design values given are subject to 0.05600kPa for RQD = 25–70% × minimum of (quc, fcu), whichever is smaller.1000 kPa for RQD > 70% quc = unconfined compressive strength of intact rock

fcu = compressive strength of concrete/grout for piles


Figure 15. Rock socket reduction factor, α (Tomlinson, 1995).

Figure 16. Rock socket reduction factor, β (Tomlinson, 1995).

Designation, RQD on the rock core recovered or by interpretation of pressuremeter mod-ulus in the rock mass against the elastic modulus of intact rock, which is equivalent to themass factor, j which is the ratio of elastic modulus of rock mass to that of intact rock.

For bored piles construction in Malaysia, the Authors have introduced a practical newapproach for verifying rock quality during construction, using point load test at site toassess and verify rock strength on recovered rock fragments during bored piling afterproper calibration with borehole results. This relatively new approach is further discussedin the following section.

In general, the contribution of base resistance in bored piles should be ignored due todifficulty in proper cleaning of base especially for wet hole construction (with drillingfluid). The contribution of base resistance can only be used if it is constructed in dry holes,if proper inspection of the base can be carried out, or if base grouting is implemented.Tan et al. (1998) reported low values of mobilised base resistance for bored piles in tropicalresidual soils where Kbu values of between 7 and 10 were obtained. Kbu is the ultimate base


resistance factor for the semi-empirical correlations of base resistance with N-values fromStandard Penetration Tests (SPT-N values) given in the equation below:

Ultimate base resistance, fbu = Kbu× SPT-N (in kPa)

The relatively low Kbu values are most probably due to the soft toe effect which is verymuch dependent on workmanship and pile geometry. This is even more pronounced inlong piles. In view of the difficulty of proper base cleaning, the Authors suggest ignoringthe base contribution in bored pile design unless proper base cleaning can be assured andverified.

The construction method is also an important consideration in the design of bored pilesin limestone formation. In Malaysia, the two most common methods of forming rock sock-ets are rock coring with rock cutting bits and chiselling by mechanical impact. Both meth-ods have their own merits and need skillful operators to form a proper rock socket. Ingeneral, the rock coring method will form a smoother, but intact socket surface while chis-elling will form relatively rougher sockets, but these rock sockets could be more fractureddue to dynamic disturbance to existing discontinuities in the bedrock. Therefore, chisellingis usually not recommended in highly fractured limestone formations to prevent the riskof further fracturing the rock mass.

In addition, construction of bored piles in limestone formation often requires goodcollaboration between the design engineer and the contractor. This is due to the highlyvariable ground conditions which require significant input from site personnel and in addi-tion to good geotechnical design, it is recommended that the “observational approach”be adopted for bored pile construction in limestone formation. Such an arrangement willenable any unexpected geological formations and uncertainties to be detected and changesto the design can be made immediately to ensure safe and cost effective design. In order forthe successful implementation of the observational approach, the designer should antici-pate and identify potential difficulties and measures that need to be carried out. For exam-ple, “unexpected” geological formation which causes large differences in pile length dueto irregular bedrock profiles should be anticipated (even though SI during design may notdetect such features) and criteria for compensation piles already in place. Therefore, foun-dation construction in limestone formation is expected to involve significantly more inputfrom the designer during the construction stage as compared to other less complicatedgeological formations.

Construction method for bored piles in limestone formation often needs modification toensure proper formation of the piles. Figure 17 shows a modified rock coring tool used forbored pile construction in limestone formation. Such a tool enables the casing to penetrateor reamed into the required rock socket length and thus prevents problems such as thecollapse of loose soil or slime surrounding the bored hole normally associated with theconstruction of rock socketed piles as illustrated in Figure 18.

Figure 19 illustrates the performance of the modified coring tool in preventing the aboveproblem at the interface between rock and soil by coring through to the required socketdepth together with the casing. Conventional method of construction, where the tempo-rary casing is installed using a vibro-hammer, is unable to penetrate into the rock layerand thus causes situations such as those shown in Figure 18 and also loss of concrete dur-ing concreting of the pile.


Figure 17. Modified rock coring tool for bored pile construction in limestone formation.

Figure 18. Collapse of loose soil (slime) surrounding the bored hole.

Figure 19. Performance of modified coring tool.


3.3. Definition of Rock Coring for Design Verification at Site During Construction andfor Payment Purposes

One of the common disputes during bored piles construction is the definition of rock coringeither for technical reasons or payment purposes. Usually design consultant either didnot specify clearly in the specification or the definition is not clear and cannot be reliablyassessed or quantified at site and thus subject to argument/contractual dispute later. Boredpiles contractor will usually consider payment for rock coring to be counted once thereis change of tools from soil boring tools to rock coring tools. This is very subjective andinconsistent as each contractor will have different types of boring machine with differentcapacity including different level of skills and patience of the bored pile rig operator andas such, it is at the sole discretion of the contractor. Therefore, it will be difficult to comparerock coring on the same basis when two different types of boring machines (or differentoperator) are being used. The situation becomes more complicated in metasedimentaryformation where very hard soils are commonly encountered with SPT’N’ more than 50blows/300 mm (e.g. with extrapolated SPT’N’ values of more than 100) as the transitionfrom soil to “rock” is not clear.

The Authors recommend a systematic approach to define rock coring for design veri-fication at site during construction and payment purposes. The recommendations can beincluded in technical specifications, drawings and bills of quantities during tender so thatall tenderers are clear on the requirements and adequate provisions were made duringtender to prevent dispute after award and during construction. Under the recommenda-tions of the Authors, similar method with some small modification has been adopted forlarge scale infrastructure projects such as the Klang Valley MRT and some big developmentprojects in Malaysia. This systematic approach is originally developed for Kuala LumpurLimestone but has since, been modified for other types of rock formations.

The recommended definition of rock coring shall fulfill all three (3) criteria below:

– Change of tools to rock coring tools, and– The rock materials shall be verified by carrying out point load test on at least three (3)

rock samples to achieve minimum index strength, Is(50) of 3 MPa subject to site confir-mation on typical rock lump sizes based on the size correction factor, F = (De/50)0.45

where De is the equivalent core diameter in mm (Is(50) = F*Is), and (1).– Recovered rock coring materials of more than 50% subject to site calibration upon start

work unless otherwise agreed by the engineer.(2)

Note:

(1) The value of Is(50) varies for different rock formation and site conditions. Therefore, theGeotechnical Engineer designing the bored piles have to determine the proper valueto be used as it is also related to rock socket capacity adopted for the bored piles.

(2) The criteria of recovered rock coring materials can be relaxed by the Consultant basedon site conditions and rock types as it is dependent on machine capabilities.

Any coring/boring in rock like materials that do not fulfill the definition of rock coringmentioned above shall be considered as partial rock coring where the payment will bemade based on the equivalent rock length as tabulated in Table 3.


Table 3. Sample of equivalent rock socket length for every 1 m of partial rockcoring not fulfilling Is(50) of 3 MPa.

Average Values Is(50) Equivalent Rock Socket Lengthper layer (Mpa) to Is(50) = 3 MPa (m)

< 0.5 00.5 to < 1.0 0.201.0 to < 2.0 0.402.0 to < 3.0 0.60

≥ 3.0 1.00

Figure 20. Point load test carried out at site for Klang Valley MRT project.

For example, for a 2 m rock coring with Is(50) = 2 MPa, the equivalent rock socket lengthto Is(50) = 3 MPa is 1.2 m where the rock coring payment is solely for this 1.2 m rock socketlength only and the remaining 0.8 m length to be paid as soil boring.

Minimum two (2) sets of point load tests shall be conducted for each pile and one (1) setof point load test for every 1 m rock coring. Each set of point load test shall consist of min-imum three (3) rock fragments to be selected by Engineer’s representative from the mostrepresentative rock samples recovered for rock core at every 1 m depth. The provision ofpoint load test equipment at site for the above-mentioned testing shall be by the Contractorand the equipment is simple and can be easily operated at site.

Figure 20 shows the Point Load Test (PLT) being carried out at site on rock samplesobtained from bored pile construction.

3.3.1. Use of Point Load Test Index Strength (Is(50)) as Construction Control Criteria for RockSocket Length of Bored Piles in Kuala Lumpur Limestone

Geotechnical engineers involved in bored pile design and construction need a systematicway to verify the actual rock socket length cored at site fulfill the requirements of rocksocket capacity adopted in design. Fully instrumented preliminary test piles subjected tomaintained load of up to three (3) times working load or up to failure is the best wayto confirm the rock socket capacity ( f srock) of bored pile. However, this method is bothcostly and time consuming. Therefore, only few preliminary tests piles can be carried out


at site (e.g. usually about 1 to 2 preliminary test piles or about 0.2% to 0.3% of the totalnumbers of working piles) and it is not practical to solely use this method as quality controlduring construction for all the working piles. Thus, a simpler, faster, cheaper and reliablealternative of using point load test on rock samples obtained from coring during bored pileinstallation was developed.

In order to use point load test index strength (Is(50)) as a parameter to determine theactual rock socket length at site that fulfills the design criteria, the following steps shall becarried out:

1) Sufficient numbers of Unconfined Compression Tests (UCT) and Point Load Tests (PLT)shall be carried out on rock cores obtained from boreholes during subsurface investiga-tion (SI) in the design phase. Both tests are carried out on the same sample with UCTcarried out first and followed by PLT.

2) The site/project specific correlations of point load test index strength (Is(50)) toUniaxial Compressive Strength (UCS) are determined based on results from Step 1)above.

3) Maintained load tests on fully instrumented preliminary piles shall be carried out toobtain the ultimate rock socket capacity of the rock. The ultimate rock socket capacityshall be correlated to the UCS and Is(50) of the same rock (e.g. carry out borehole firstand necessary UCT and PLT before the construction of the preliminary pile on the exactsame location).

4) Use the Is(50) correlated from the above steps as construction control criteria on rockcoring depth that fulfilled the design criteria for the working bored piles.

5) Calibrate and improve the correlations of Is(50) with rock socket capacity through testsresults of maintained load tests on instrumented working piles.

To illustrate the above approach, a sample project which the Authors are involved isused as an example.

A total of sixty one (61) numbers of Unconfined Compression Tests (UCT) and PointLoad Tests (PLT) were carried out on Kuala Lumpur Limestone samples obtained fromboreholes to determine the correlation between Point Load Test Index Strength (Is(50)) toUniaxial Compressive Strength (UCS). Table 4 summarises the results obtained.

Table 4. Summary of uniaxial compressive strength (UCS) and point load test (PLT) index strength(Is(50)).

No. of Tests Uniaxial Compressive Strength (UCS) (in MPa)

StandardMinimum Maximum Mean Deviation

10.2 97.9 58.1 19.6

61 Point Load Test Index Strength (Is(50))

StandardMinimum Maximum Mean Deviation

1.1 6.6 3.6 1.3


Figure 21. Frequency chart of uniaxial compressive strength (UCS) of Kuala Lumpur Limestonesamples tested.

Figure 22. Frequency chart of point load test index strength (Is(50)) of Kuala Lumpur Limestonesamples tested.

Figures 21 and 22 show the frequency chart of the UCS and Is(50) respectively. Figure 23shows the relationship between UCS and Is(50). The ratio of UCS/Is(50) ranges from 8 to ashigh as 33 with average value of about 15.

From the test results, it is also important to take note that the commonly used correla-tions of UCS = 20–25* Is(50) cannot be assumed to be valid for all rock types as demon-strated above where the average correlations for this particular project is UCS = 15* Is(50).Therefore, it is important that site specific correlations are established for each project dueto the variable nature of rocks and also limitations of the testing method.


0 1 2 3 4 5 6 7 8 9 10Is(50) (MPa)

0

10

20

30

40

50

60

70

80

90

100

UC

T (M

Pa)

0 1 2 3 4 5 6 7 8 9 10

0

10

20

30

40

50

60

70

80

90

100

Lower BoundUCS/IS(50) = 8

Mean valueUCS/IS(50) = 15

Upper BoundUCS/IS(50) = 33

Figure 23. Relationship of point load test index strength (Is(50)) to UCS of Kuala Lumpur Limestonesamples tested.

4. SUMMARY

The design and construction of foundations in limestone formation pose tremendous chal-lenges to geotechnical engineers due to the highly irregular karstic features of limestoneformation. Understanding of the potential difficulties arising from these karstic featuresis essential in order to provide safe, construction friendly and cost effective foundationsolutions.

Some typical piling problems associated with limestone formations were presented anddesign and construction recommendations for driven piles, jacked-in piles and bored pileswere also discussed. The design and construction of foundations in limestone formationrequires careful planning from the design stage up to the construction stage wherecontinuous input from the construction team to the design team is very important to ensuresuccessful construction and satisfactory performance of foundations in limestoneformation.

A relatively cost effective method to verify rock socket quality for bored piles construc-tion in limestone formation using Point Load Test was introduced by the Authors. Themethod is found to be effective in removing some of the uncertainties associated with def-inition of rock for both design and payment purposes as the method allows quantitativeassessment to be carried out at site in a relatively simple manner.

ACKNOWLEDGEMENTS

The contributions of the Authors’ colleagues, Ir. Darryl Fong who painstakingly compiledthe point load test data is gratefully acknowledged.


REFERENCES

1. Barton, N., Lien, R. and Lunde, J., “Engineering classification of rock masses for the design oftunnel support”. Rock Mech., 6 (1974), pp. 189–236.

2. Bieniawski, Z. T., “Engineering classification of jointed rock masses”. Transactions of the SouthAfrican Institute of Civil Engineers, 15 (1973), pp. 335–343.

3. Bjerrum, L., “Norwegian experiences with steel piles to rock”. Geotechnique, Vol. 7 (1957),pp. 73–96.

4. Chow, C. M. and Tan, Y. C., Jack-in Pile Design – Malaysian Experience and Design Approachto EC7. IEM Course on Eurocode 7 – Geotechnical Design, 16-17 July 2009, Petaling Jaya, Selangor,Malaysia (2009).

5. Chow, C. M. and Tan, Y. C., Performance of jack-in pile foundation in weathered granite. 17th

SEAGC, Taipei, Taiwan (2010).6. Culshaw, M. G. and Waltham, A. C., “Natural and artificial cavities as ground engineering

hazards”. Quarterly Journal of Engineering Geology, 20 (1987), pp. 139-150.7. Deeks, A. D., White, D. J. and Bolton, M. D., A comparison of jacked, driven and bored piles in

sand. Proc. 16th ICSMGE, Osaka, Japan (2005), 2103–2106.8. Gue, S. S., “Foundations in limestone areas of peninsular Malaysia”. Civil and Environmental

Engineering Conference (C&EEC) New Frontiers & Challenges, Bangkok, Thailand (1999).9. Gue, S. S., Private communication (2013).

10. Horvath, R. G., “Field load test data on concrete to rock bond strength”. University of Toronto,Publication No. 78–07 (1978).

11. Jakucs, L. D. Morphogenetics of karst regions. Translated by Balkay, Adam Hilger, Bristol (1977),p. 283.

12. Jennings, J. N., Principles and problems in reconstructing karst history. Helicite 20(2) (1982).pp. 37–51.

13. Neoh, C. A., “Design & construction of pile foundations in limestone formation”. Journal –Institution of Engineers, Malaysia. Vol. 59, No. 1 (1998), pp. 23–38.

14. Omar B. Ibrahim and Hon, T. F., “Piled foundation in limestone formation”. Proc. 8th SEAGC,Kuala Lumpur, Malaysia (1985).

15. Rosenberg, P. and Journeaux, N. L., “Friction and end bearing tests on bedrock for high capacitysocket design”. Canadian Geotechnical Journal, 13 (1976), pp. 324–333.

16. Tan, B. K. and Ch’ng, S. C., “Weak soil above limestone bedrock”. Newsletter of the GeologicalSociety of Malaysia, No. 2, Vol. 12 (1986).

17. Tan, Y. C., Chen, C. S. and Liew, S. S., “Load transfer behaviour of cast-in-place bored piles intropical residual soils”. Proc. 13th SEAGC, Taipei, Taiwan, ROC (1998), pp. 563–571.

18. Tomlinson, M. J., Foundation design and construction. 6th edn. Longman (1995).19. Waltham, A. C. and Fookes, P. G., “Engineering classification of karst ground conditions”.

Quarterly Journal of Engineering Geology and Hydrogeology, 36 (2003), pp. 101–118.20. White, D. J. and Lehane, B. M., Friction fatigue on displacement piles in sand. Geotechnique,

54, No. 10 (2004), 645–658.21. Williams, A. F. and Pells, P. J. N., “Side resistance rock sockets in sandstone, mudstone, and shale.

Canadian Geotechnical Journal, 18 (1981), pp. 502–513.22. Yeap, E. B., “Irregular topography of the subsurface carbonate bedrock in the Kuala Lumpur

area”. Proc. 8th SEAGC, Kuala Lumpur, Malaysia (1985).23. Yee, Y. W. and Yap, L. C., “Hand-dug caisson foundations for tall building over cavernous

limestone”. Proc. 13th SEAGC, Taipei, Taiwan, ROC (1998), pp. 681–688.

foundation design and construction in · pdf filefoundation design and construction in...

Documents