TRANSPORT AND ROAD RESEARCH LABORATORY

Department of Transport

CONTRACTOR REPORT 44

THE EFFECTS OF SIMULULATED TRAFFIC VIBRATIONS ON A DWELLING HOUSE

by

Dr R A Hood, MSc, PhD and C P Marsha l l , BSc, CEnK, MICE

The authors of this report are employed by Travers Morgan PlanninK. The work reported herein was carried out under a contract placed on them by the Secretary of State for Transport.

The views expressed are not necessarily those of the Department of Transport.

This report, llke others in the series, is reproduced with the authors own text and illustrations. No attempt has been made to prepare a sCandardised format or style of presentation.

Veh i c l e and Envi ronment D i v i s i o n Veh i c l es Group Transport and Road Research Laboratory Old Wokingham Road Crowchorne, Berkshire RGll 6AU 1987

ISSN 0266-7045

Ownership of the Transport Research Laboratory was transferred from the Department of Transport to a subsidiary of the Transport Research Foundation on 1st April 1996:

CONTENTS

.

2.

3.

4.

.

~ CF ST~Crt~E

4.1 Gemral 4.2 R)urda tiore 4.3 Wa11~ 4. 4 Floors and Ceilings 4.5 Roof

SETS ~G~IOg

5.1 5.2

(]:)ject: of Invescigaclon nesccipClon of Stage I Invescigaclo.

5.2.1 5.2.2

m-sire Test/r

5.3 Results of Stage 1 InvesClgaCion

5.3.1 5.3.2 5.3.3 5.3.4 5.3.5 5.3.6 5.3.7

Descriptlon of Geologi,~1 S~=am Desm=iptlon of Soil ~-ading Charac~m-istics Mols~e Contm.nt: 8o11 Om~.si~ S e t - - t : PenetromeCer Test: Results

. SIM[EATIONEXP~

6.1 S£mulation

6.1.1 6.1.2

Grouncbcrne Vibrations Airborne Vibrations

6.2 Vibration and Noise Monitz~ring

6.2.1 6.2.2 6.2.3

~t~ing ~uilmmnt Monitoring Positions Measurements Under takon

6.3 YmnitDring of Movement

6.3. i Level Survey 6.3.2 ~ire Photography 6.3,3 Crack Survey 6.3. 4 Soil Mmve~ts

6.4 Results - Vibration and Noise Monitoring

6.4.1 6.4.2 6.4.3 6.4.4 6.4.5 6.4.6 6.4.7 6.4.8

Response of Structure to Groun~bcrne Vibration Response of Floors and Ceilings ~o G~oun~xEne Vibration Response of Rafter to Groundbcrne Vibration Response of Structure to "Nurmal Use" Response of Structure to Airborne Vibration Existing Noise and Vibration Levels Noise Levels Induced by Ground~rne Vibrations Fenestration Attenua ticn

6.5 Results - Monitoring of Movement

6.5.1 Level Survey 6.5.2 ~ire Phot~/raphy 6.5.3 Crack Survey 6.5.4 Soil ~%ovements

6.6 Final Site Investigation

.

7.1

7.2

7.3

TEST FOUND~ION STRIPS

Objectives

[~scriptlon of Experiment

Results

7.3.1 7.3.2 7.3.3 7.3.4

Vibration Monitoring Level Survey Electrolytic Levels Final Site Investigation

. D~gCUSSIGN OF S;/TuT/tTICN EXPERIMENT

8.1 Simulation

8.2 Response of Structure to Groundborne Vibrations

8.2.1 Response of Foundations 8.2.2 Response of Structural Elements 8.2.3 Response of Floors, Ceilings and Roof

8.3 Response of Structure to Airborne Vlbraticns

8.4 Response of Structure to "No~mal Use"

8.5 Statuary of Possible Damage Mechanisms

8.6 Movement and Damage Actually Recorded

8.6.1 Soil Mcuements 8.6.2 Movements within the Structure

. ~LL~IGNS OF SLMUIATICN

10. EXPERIMENT

10.1 Objectives

10.2 ~he~ry

10.3 Choice of Material

10.4 Description of Experiment

i0.5 Results

10.5.1 Response to PuLse 10.5.2 .Response to Frequency Sweep

ll.

12.

RBCO~kTICNS FOR F ~ WORK

ii.i

ii. 2

11.3

Sin~lation Experiment

Test Foundation Strips

Trench Experiment

;K~ENC~KE~S

13.

APPR~IX i -

LIST OF FI~

Reproduced by permission of the Controller of HMSO. Extracts from the text may be reproduced, except for commercial purposes, provided the source is acknowledged.

ABSTRACT

This report describes an experiment to investigate damage caused to buildings by traffic-induced vibrations. A vacant property founded on loose to medium dense sand was located and subjected to separately generated groundborne and airborne vibrations, designed to simulate the effects of heavy traffic. Gro~rne vibrations were simulated using a geophysical vibrator, airborne vibrations with a loudspeaker system connected to a mi~uter.

The dynamic response of the structure was monitDred using a ccmputer conr~olled data collection and analysis system. Throughout the period of the experiment the condition of the strucuze, mavemn~ within the structure and movaments within the underlying soil were monitored, using a range of precise measuring te~niques.

In order to widen the scope of the study, six sections of foundation strip were constructed close to the test house. These were and s~jected various static loads and an~litudes of vibration. The behaviour of the strips was also monitored throughout the experiment.

The report describes the experimen~l work and results in detail. Possible damage mechanisms are pos~lated and discussed in the context of the very small amounts of movement and damage actually recorded. The distinction iS made bet~.n the effects of airborne and groundhorne vibrations. Conclusions are drawn ~ a r d i n g the overal l r i sks of damage from t r a f f i c vibrations and circumstances which increase this risk are highlighted.

I. DERDDUC~ON

Recent studies by the TRRL on the subject of traffic vibration have concentrated on examinino the causes of building vibration and the disturbance resultingl,2. From the results of these studies it is apparent that of the people who experience traffic vibration in their homes, a large proportion consider that such vibrations cause damage to their property. The steady increase in c~,i,ercial traffic 3 and recent increases in ~ermitted vehicle weights have tended to exacerbate this concern ~.

There is little firm evidence linking traffic vibration building damage, and since traffic vibrations are generally of

low amplitude it has become common practice to place the blame fmr building damage on other factors. Although this argument has some force it ignores the possibility that damaqe could result from long term exposure to low level stress reversals, or from the c~,~ined effect of traffic vibrations and existing stresses in a structure. It is also possible that traffic vibration could be a cause of differential settlement of a structure.

AS the knowledge of the subject is limited at present, it is important to establish whether research can lead to a better understanding of the problem.

A pro~me of research was designed by ~ and Travers Morgan Planning to establish the feasibility of different approaches to studying the problem. As a first st-?. a full scale similation of the effects of traffic vibration on a conventional building structure has been carried out by TEarers Morgan Planning. It was hoped that the results obtained would give general insight into the mechanim~s of vibration transmission into buildings and enable observations to be made of soil mmmment and building darage under controlled conditions over an extended period of exposure.

During this project the opporUmity was taken to study the effects on vibration propagation in soil of a 2m deep trench with and without a polystyrene filler.

This report describes the experiment and the results obtained.

i-

. CSJECrT~S

The experiment was designed to determine whether simulated traffic vibration caused damage to a specimen house and if so, what damage mechanisms were in~Dlved. ~he structure selected for the study ~as a pair of semi-detached houses, approximately 90 years old, founded on a loose to medium dense sand. Vibration simulation was split intm two parts. Airbo~,e vibration was simulated by a loudspeaker sys~mu, mounted in the side of a lorry and connected to a om~puter. Grouncborne vibration was achieved using a geophysical vibrawr which impacted the ground.

The rain aim of the work was ~ assess whether road traffic vibration oaused damage tm houses. In order tm make this assesm~.nt, four types of n~nitmring w e r e ~rried out.

t} Site Investigation. ~:ial pits were dug on the site before and after the simulation, to assess ground conditions and tm evaluate soil densities. This work was conducted in order to assist in interpreting any settlement of the structure which occurred.

ii) Structural survey. A full survey of the structure was carried out befmre and after the simulation. As well as this, visual inspection of cracks was carried out at regular intervals dm:ing the w=k.

ill) Vibration monitoring. An extensive series of vibration mmsurements was undertakon, covering foundations, waLLs, flours and ceilings. The purpose of this work was tm learn as much as possible about the response of the structure, so that possible damage mechanisms could be postulated.

iv) FcmitDring of nDvement. A variety of precise measuring techniques were en~loyed to monitor any .Dvement which occurred in the structure and the underlying soil. r~asurements were repeated at intervals throughout the simulation period.

-2-

FREVIOUS WORK

A great deal of research work has been undertaken in the fields of geo~ical engineering, vibrations, and damage to structures. However, very little specific study of damage to buildings caused by road traffic vibrations has taken place.

For the p~poaes of this study, a review ~s ~arried out of work in associated fields which was relevant to this experiment. Partic,,1~r subject fields covered inclucMd levels of vibration generated by road traffic, the nature of building response to vibrations, criteria frr vibration induced damage, behaviour of soils undergoing vibrations, and soil-st:cucture interaction.

~,]] details of the literature review are given in Appendix i.

3-

4

4.1

4.2

4.3

4.4

4.5

EESCR/Y~ON OF

G~@mAL

The structure comprised two semi-detached houses, two storeys in height, with a basement over part of the plan area. Construction details are illustrated on Figs.4.1. and 4.2. and a general view of the site is shown on Plate I.

An initial s~vey of the house revealed extensive cracking of plasterwcrk in both walls and ceilings. However the amount of crackir~ was no more than is typical of houses of this age and was not indicative of any structural defects.

R~Ct~

The foundations of the cellar were not e~mined. The reminder of the house was founded on mass concrete strip foundations 600ram wide and 280ram deep. The underside of the mass concrete was 68~m below ground level and rested on approximately 10@t,, of ballast.

~rx.~

All ~earing, party and external walls were constructed of brirlazrk in lime mortar. The external walls had battens fi~ed to the internal faces to which laths and plas~-~ were attached. The internal brick walls had plaster applied directly to both faces of the brickwork.

Non-loacb~ing partition walls were constructed of 100ram x 50ram studding with 100ram x 62ram sill and head pieces, raths and plaster were attached to both faces.

FIEORS AND CEI~

The ground flour joists spanned front to rear on sleeper walls. The joists were 12.~m x 50ram or ll2mm x 50ram. There was herring bone strutting at intervals between the joists. Flou~boards were 22ram thick, nailed ~m the joists.

The first floor joists also spanned front to rear and ~re 175ram x 62ram at 375mm cen=es. There were trimmers around the staircase (175mm x 100ram) and the fireplace (175ram x 7~mm).

The first floor ceiling joists were 125m x 50mm at 350ram centres.

All ceilings were constructed of laths and plaster.

ROOF

~e roof construct ion cm~rise4 rafters, boarding and slates.

-4-

.

5.1

5.2

5.2.1

SITE INVESTIGATION

A ~ stage investigation ccmprising trial pits, hand augered bore~les and in-situ penetrQmeter tests ~as carried out by Southern Testing r~hcra~ories. The first stage took place in the first week of December 1985, before the vibration experiment began. The second stage took place in March 1986, at the end of the experiment.

OBJB~T OF INVESTIGATION

The object of the first stage of the investigation was to define the geological strata at the proposed site. Of primary importance was the measurement of the in-situ soil densities. The object of the second, stage was to measure changes, if any, in the soil densities which had occurred during the vibration experiment.

DESCRIPtiON OF S~%GE I INVESTIGATION

This part of the investigation comprised 3 trial pits and two hand augerec boreholes, the locations of which are shown on Fig 5.1. In each trial pit in-situ density tests were performed at depth intervals of about 0.Sin. Samples were removed at the same elevations for labora~y testing.

In-situ Testing

Methods of measuring in-situ density were as follows:

(i) Sand replacement tests performed in the base of each pit

(ii) Density measurements from core samples

(iii) Calibrated penetrometer readings

The sand replacement tests are described in BS 5930 (Test 15). In this test a sample of soil is extracted and is replaced with sand of known bulk density. By this method the bulk density of the excavated sand can be determined.

Three types of penetrcmeter were used. They were

(i) (ii) (iii)

The Borros Penetrometer The Perth Penetrometer The ~&~ckintosh Penetrometer

In all three devices, the nun~er of blows required to move the probe tip through a set distance is recorded. However, the probe shapes differ, as does the energy input per blow. These differences in detail mean that each of the penetrometers is best suited to particular soil condition and test requirements. Since these were not known precisely in advance, the three different instruments were employed in order to provide wL=cboration of the results.

-5-

5.2.2

5.3

5.3.1

5.3.2

5.3.3

~"his type of tecnnique is useful for detecting changes in strata. In granular materials, the blow count can be related to the in-situ density of the mterial. To achieve this the readings were calibrated by carrying out pene~rometer tests on samples of soil from the site which had been reccmpacted in the l~hcra=y to a variety of densities.

Labora ~cry Testing

In the labcra~ry, tests were carried out ~ detarmme particle size distribution, moisture content and specific gravity. ProctEr ~ction tests were also per£oL,~ed with a view to determining the moisture content at which the maximum dry density of the soil oocurs.

P~--dLTS OF SI~(~ 1 INVESTIGATION

Description of Geological Strata

The investigation confirmed that the houses were built on the "Folkes~ beds, comprising sand which was probably deposited in shallow sea conditions. In the Sevencaks area, these beds are lightly cemented and may contain discontinuous pebble beds. They exhibit sate local variations in the i r physic1 properties, notably in particle orientatio~ though grain size is generally uniform.

Description of Soil

The =ial pits and hand augered boreholes revealed the soil at the site as-

0 - 200ram ~psoil

2 0 0 - 500ram hrse brown slightly clayey SAND with organic material and s~me gravel

500- 150ram Loose brown slightly silty SAND with traces of gravel

1500- 3300ram Medium dense lightly cemented orange-bruwn and yelluw-br~n banded medium ~%ND with some coarse sand and traces of gravel

Trial pit No.l revealed sand which was so loose that the pit became unstable. It %~s not passible to i~stall shoring to prevent collapse of the walls and the pit was abandoned at 2.7m.

G~ding Characteristics

Grading analyses of soil from trial pits 2 and 3 indicate a very uniform soil profile with depth (see Fig.5.2 and 5.3). The grading profile of soil ~ trial pit 2 hardly ctmnges at all with depth. The results fro trial pit 3 indicate a slightly more gravelly topsoil, but are otherwise (~nsis~ent with trial pit 2. In general, the material is a uniformly graded medium sand.

-6-

5.3.4

5.3.5

5.3.6

r%ois~re Content

The results of the dry density/moisture content tests (Proctor cm~action) are shown on Fig. 5.4. The optimum m~isD~e content (i.e. that at which the soil achieves a maximum dry density) is about 11%. Despite the scatter in the data, a slight tendency for the optimum moisture content to increase with depr.h can be inferred.

The natural nDis~re content at the site %as found to vary as shown on Fig. 5.5. The moisture content was higher at the ground surfmce than at depth. In the ~9 metre the natural moisture conten~ %~s at, or above, the optimum moisture convent. Below 2.0 metres it was at leas~ 5% less. It will be shown later that the moisture contenn IEofile at the end of the experiment was almost the same as at the

Soil Density

Fig. 5.6 s~m the optimum dry density of the soil. This tends to decrease with depth. There is some separation between the results from trial pit 2 and those from trial pit 3, which could be due to the slight difference in silt content be~en the two pits. However, the grading ceres f~r the two pits (see Figs. 5.2 and 5.3) are almost identical. It is therefare more likely that the separa~.ion is coincidental. In these circumstances, the trend line shown should be taken as representative of the site as a whole.

The in-situ densities measured by the sand replacement tests and frcm (rre samples are presented on Fig. 5.7. There is clear scope fur densification, particularly in the trap two m~tres.

Set~t

The poten~a] for settlement within the soil can be calculated as follows:

~H - ~e

H i~o

where H = original thickness of a volume of soil AH = change in thickness (settlement) ~e = change in voids ratio eo = original voids ratio

The voids ratio is the ratio of the volum of ~ids ~ the ~lun~ of solids in the soil.

Accordir~ to the above fr~m,,Im, if the ~:~ i. 0m of sand were to densify from measured ~ maximum density, this would result in approximately l~mm of settlement. If densiflcation extended ~ a depth of 1.5m, this would increase to approximately 21ram.

-7-

5.3.7 Penetrcmeter Test Results

The Burros Penetrcmeter results are presented on Fig.5.8, and an average of the three sets of results is shown on Fig.5.9 The results show a clear change in soil properties at about l.Sm depth, where blowcount increases sharply. This trend is least marked in trial pit I, which gives lowe~ blowcounts than the soil elsewhere on the site.

The insitu densities of the soil as calculated frum the Borros proDe results agree well with measured densities up ~ a depth of l.Sm, but exceed the maximum possible values below this depth. This is because the sand beccmes cen~nted at this depth which invalidates the calibration of the penetrcmeter, as this ~as performed cn reccmpacted (and thus uncmnented) sand.

The Perth Pene~ter test results, which are shown in Fig.5.10 and 5.11 show a pattern which is similar to the Rrros Penetrometer results, though less distinct.

The blowummt of each individual Perth Penetrmneter rose s~dily fur the first 300ram of penetration. This prct~bly reflects the effect of stress relief due tD excavation. The effective stress ~as lower because soil had been removed from the pit. H~ever, beyond 300ram of penetration the effective stresses were prctmbly little changed from the insitu values, due ~ the soil arching be~en the restrained edges of the trial pit.

To allow fur the effect described above, a bounding envelope has been cK-awn around all the results. The envelope thus reflects the "insitu" stress s~te more fairly, because the profile is governed by the deeper, less disturbed, part of each individual result.

The results of the Mackintosh Penetrcmeter tests showed considerably more scatter than the other tests, though the general trend agreed with that described above.

The fluctuations in the blowco~t profile of the Perth and Borros penetru~eters ~ l y provide a true reflection of local variations in the degree of cementation of the sand particles, and possibly also indicate minor variations in in-situ density. It is fair ~ say, though, that the top l.Sm of sand at the site consisted of loose uncemented sand. Below this it was denser, and was lightly cemented.

8-

.

6.1

6.1.1

SIMULATIC~ EXPERIMENT

SIMULATION

Building vibrations are generated by road traffic by two separate mechanisms. In simple terms, tyre contact with irregularities in the road surface leads to groundborne vibrations; engine and exhaust noise produces airborne vibrations. For this study, these two vibration sources were simulated independently of each other.

Groundborne Vibrations

Monitoring of building foundations close to roads with uneven surfaces has indicated that the dominant ccmponent of groundborne, traffic-generated vibration is a response to '%lheel hop" in which tyre-road contact excites vibrations of wheels and suspension. It typically has the following characteristics:

Frequency - in the range 5 tO 30 Hz. The predominant frequency in any particular case depends upon the nature of the vehicle suspension and its loading.

Pulse duration - Up tO several tenths of a second, depending upon vehicle speed and number of axles.

Maxirmum a~litude - the vertical component of vibration in building foundations adjacent to uneven road surfaces scmetimes exceeds i mm/s. (See Appendix i, Para 2). Figures of 2-3 mm/s have ~:n recorded in a few instances.

Simulation of the dominant cc~nent of groundborne vibration was achieved using a geophysical surface vibrator(see Plate 2) This equipment is capable of vibrating the ground surface with vertical impulses of a selected duration and frequency. A suitable interface was required between the vibrator plate and the soil, both to prevent local soil failure, and to ensure that the passage of vibrations into the ground realistically modelled the effect of a lorry on a road pavement. It was decided that this would be best achieved by the construction of a small area (Sin x 5m) of flexible road pavement close to the structure under test, on which the vibrator would be mounted. The road pavement was built to withstand 7 million axles in accordance with the DI~ Specification for Road & Bridge Works.

The arrangement of the road pavement and vibrator is shown in Fig. 6.1

The way.form generated by the vibrator had the following c~aracteristics (measured on the adjacent house foundations):

Fundamental Frequency - 12 - 13 Hz

Pulse duration approximately I second, including rise and fall. The waveform typically included 8 reversals at peak amglitude which is broadly equivalent to 4 heavy goods vehicle axles.

-9-

6.1.2

Maxim~ ampli rude - 2.5 ~m/s (vertical component)

Figure 6.2 compares the wavefarm produced by the vibrator with a typical roadside measurement of a heavy goods vehicle.

In addition to the generation of pulses as described above, the vibra~ur was used on a nun~er of occasions to generate a waveform of continuously varying frequency, such that the dependence of response on frequency could be determined. These "frequency sweeps" ranged from 80Hz to 10Hz in a period of about 15 seconds.

Airborne Vibrations

Noise is generated by traffic over a wide frequency range. Bswever, significant structural excitation only results from low frequency noise (up to about 200 Hz). Noise throughout this range is generated by vehicle engines and exhausts. For the large diesel engines fitted tD heavy c~mercial vehicles vibration levels often peak in the 50-80 Hz region.

Simulation of airborne vibration was achieved by mounting four 18" Celestion loudspeakers in the wall of a high sided refrigeration lorry, parked adjacent to the house facade. These were powered by a 500 watt amplifier. Reasons for selecting a high sided vehicle in which to house the loudspeakers were twofold:

(1) to provide infinite baffle for the loudspeaker enclosure and provide a ~od distribution of noise.

(ii) to isolam the vibration source from the ground. The vehicle tyres achieved satisfactory damping in this respect.

The signal to the amplifier ~as provided by a CED 1401 Laboratory Interface connected to a ~C microcomputer. This system is camhle of generating a waveform with any frequency content, by means of an inverse Fourier transform technique.

Inie~lly the system was used to generate a broad band signal, ranging in frequency from 0 to 200 Hz, and peaking in the 50-80Hz range characteristic of heavy goods vehicles. Signal amplitude ~Is monitored in the space between the vehicle housing the loudspeakers and the house facade. The signal was pulsed such as to produce a noise level of ll0dB during pulses and 100dB between pulses.

The resulting vibration levels were lower than had been expected even in the window adjacent to the loudspeaker system. A decision was therefore taken to replace the broad band signal with a single frequency source at the resonant frequency of the window (27 Hz).

-i0-

6.2

6.2.1

6.2.2

6.2.3.

In this ~ay an "upper bound" experiment was carried out. The window vibration levels achieved were the highest possible for the particular noise level (ll0dB). Therefore it could reasonably be assumed that they would equal or exceed any produced by traffic.

The pulse length used for the majority of the experiment was 2 seconds and the passage of approximately 500,000 vehicles was simulated.

~ION AND NOISE MCNI~ORING

Mmni toting Equipment

Vibrations w~e ,Dnitored using five 3-dimensional arrays of geoph~nes. Signals from the geophones were fed into a CED 1401 labora~ory Interface. ~his equipment is cable of analogue to digital conversion of up ~ 16 channels of data simultaneously. For most measurements, conversion rates of about ikHz per channel were utilised.

Analysis of the data was performed using the same equipment. Facilities avai ]able included :

- screen display of selected channels of data - printing of selected data - fast Fourier analysis of any chosen part of

a selected signal - =~ss correlation of wavefmrms - production of amplitude envelopes for

selected wavefmrms.

NDise was monitored using three systems: a B & K Sound rovel Metre Type 2209; a CEL Environmental Noise Analyser Type 162 and CEL Level and Wa~fcrm Pecorder Type 160. ~he systems were aalibrated using a B & K Piston Phone Type No 4220 which had recently been re-~mlibrated by the National Physical rahoratcry. Noise levels were analysed using third octave band filters, and a Fourier analysis undertaken using the CED 1401.

Moni ~ocing Positions

Brackets to which geoohcnes could be attached were mounted on a tmtal of 82 positions on the structure. Locations monitored included fo,,Idations, walls (externally and internally), floors, ceiling joists and roof ti~ers. Further details of specific locations are given below for particular experiments.

Measurements Undertaken

During the course of the work, measurements concentrated on a nu~er of aspectm. These are described below.

-11-

(i)

(ii)

(iii)

(iv)

(v)

(vi)

Response of structure to groundborne vibration.

With the vibrator operating normally, measurements were taken at 20 locations on the outside of the structure, and 9 locations internally. The purpose of this experiment was to aSsess the dynamic response of the main structural elements to the vibration. By assessing how the grourd- bo~,a pulse moved through the struc~re, likely areas of stress concentration were identified.

Response of flours to grcundbcrne vibrations.

Subjective response of house occupants to vibration is often associated with the fact that suspended flours can amplify vibrations. In the structure under test, first floor romm had simple suspended floors, whereas in ground floor rooms, floors were supported on foundation sleeper walls. A series of measurements were carried out to assess the following:

- response of floors to n~,Lal vibrator pulse - frequency dependence of floor response - dependence of response on support conditions - natural frequency

Respmse of rafter to g r ~ e vibration.

The response of one of the rafters to a frequency sweep %Bs measured.

Vihrations due to normal use. ~n order to help put the vibration levels in the structure in response to traffic simulation into context, a series of measurements were undertaken during which "normal use" of the house was simulated Activities measured included slamming doors, jumping on floors and running up and down stairs. In one case, as well as monitoring walls and floo:s, the level of vibration reaching the foundations was measured.

Response of structure to airborne vibration

Noise and vibration levels within the structure in response to the airbo~,,e vibration source were monitored.

Existing noise and v ib ra t ion levels prior to experiment

The test structure was only 8 metres from the ;%225 Sevenoaks to ~rtford Road, which has formed a link in the orbital road network around Imndon in recent years. 24 hour • easurements were undertaken prior to the beginning of the simulation experiment to assess existing noise and vibration levels resulting from the heavy traffic load.

-12-

6.3

6.3.1

(vii) Noise levels in response to groundborne vibration

The excitation of floors and walls by the groundborr~ vibrations resulted in low-frequency noise within the houses. Noise levels were measured in all the front rooms with the vibrator alone operating, and a fine frequency analysis was performed.

(viii) Fenestration Attenuation

The attenuation afforded by the window closest to the noise source was measured in response to a random noise and a third octave band analysis was perfDrmed. An attempt to obtain a fine frequency analysis of the attenuation by sweeping through the frequency range was unsuccessful because interference f~. passing traffic noise adversely affected the c:mpressor signal.

~ G O F ~

As well as monitoring the dynamic behaviour of the structure as outlined in section 6.2, a number of techniques were employed to monitor changes in the condition of the structure during the aourse of the experiment. Methods were used to assess mm,m,ents of the structure, movements within the soil underlying the struc~re and changes in the pattern of cracking. Details of individual ~iques are given below.

All m-~murmmnts were carried out simultanecusly on a series of selected dates. These were as follows

Date Number of vibrator Percentage of pulses oompleted total

i0.12.85 0 0 19.12.85 20,000 2 23.12.85 42,000 5 6.1.86 85,000 I0

15. i. 86 164,000 19 i. 2.86 317,000 36 22.2.86 558,000 63 20.3.86 888,000 I00

Level Survey

In otter to assess whether any heave or settlement was occurring, levelling stations were installed at 36 locations on the structure (see Fig 6.3). Thesewere levelled using a Wild N3 level and an Invar staff, capable of resolving movements of 0.I mm over a distance of 20 metres.

-13-

6.3.2

Since the extent of the vibra~r's zone of influence could not be accurately determined before the work began, it was necessary to establish reference stations remote from the vibrator. Due to the constraints of the site, the only available location was on the far side of Otford Road from the house. This was not ideal, since it necessitated levelling with the instln~ment close to the busy road. Huwever, no more satisfactory arrangement was possible in the circumstances and efforts were made to minimise the disruptive effect of passing traffic.

1oire Pho aphy

To dete~ine whether or not any differential movement was occurring within the structure, high resolution Moire photography was e~91aYed with the collaboration of the National Physical Laboratory. This is a technique in which paper printed with a fine grid of lines is attached to the house facade and photographed with a specially modified 35ram camera. When two negatives produced on different dates are overlaid, interference fringes are produced which can be related to movements of the structure. By this technique, differential movements in the plane of the facade can be resolved to 0.1ram.

For this experiment, the p a t ~ e d paper was affixed to the entire front facade of the structure. A sample of the paper is shown on Fig 6.4.

6.3.3 Crack Survey

During the preparation of the houses for the experiment, all wallpaper and ceiling finishes were favored. This revealed very extensive existing cracking in the plaster. Due to the extent of the cracking, it was not possible to accurately measure and monitor every crack. Instead, 40 existing cracks in various locations were selected and monitored with a Demec gauge. At each location 3 or more studs were affixed to the wall or ceiling, straddling the crack. The gaps be~een the studs were measured at intervals during the experiment, including in each case one which did not strac~le the crack, as a reference. In addition, the location of every significant crack was re~orded on a plan or elevation of the relevant facade. On each measurement day, an inspection was carried out, and any further crauking =ecoz~L~.

-14-

6.3.4 Soil ~vements

The level survey described in section 6.3.1 was intended to measure settlement of the structure. Such settlement could develop by three mechanisms:

(i) Settlement directly under the vibrator pad extending to the soil beneath the struc~re.

(li) Soil densification below the house footings.

(iii) Soil particle movements, perhaps without change in density, within a "log-spiral" zone under each froting. A log spiral is the curve along which a failure surface would typically develop within the soil. Slight shearing along such a surface could lead to settlement.

In order to assess which mechanism was responsible for any observed settlement and tm ccrrcDorate the re~Its of the level survey, two teohniques were employed for monitoring movements within the soil.

(i) Electrolytic levels

Electrolytic levels are transducers which measure change in inclination. FOr this s~dy, six such levels were positioned at in~rvals along ea~ of t~) horizontal bore~mles drilled outwards from the house cellar. The boreholes were between 1.0m and 1.2m below ground level. Bore/Die locations are shown on Fig.6.1. By integrating the measured inclination changes, ~tical soil movements were detected.

This technique was developed at the Building Research Establishment 40 but has never previously been used in a sand deposit. Installation was achieved by jacking a steel casing into the soil and removing the soil from within the tube. A "Gryflex" plastic hose was then positioned within the casing and held under tension. ~e casing was removed and the hose relaxed, allowing it to expand t~ fill the hole. The electrolevels were then pushed intD position. We assess the accuracy of the system to be better than + 0.1ram.

(ii) Magnetic Extenscmeters

Magnetic Ext~nsometers are commercially available. They consist of a series of magnets embedded in the soil, a plastic guide tube, and a probe. The probe is passed up and down the guide tube in the soil and detects null points in the mgnetic fields. This enables relative vertical movements within the soil to be deduced. For this s~udy, systems were installed and operated by Soil Instrt~ents Limited, Uckfield, Sussex. The accuracy claimed for the system,s + 0.1mm.

Four extensometers were installed (see Fig.6.1). At each loaation, magnets were installed at 500ram intervals to a depth of 4m.

-15-

6.4

6.4.1

RESULTS - VI~ATICN AND NOISE MCNITIP/NG

Response of Structure to Grouncborne Vibration.

This experiment studied the vibrations of the front facade of the s~ucture in vertical and radial directions, with respect to the vibrator. (Vibrations out of the p/ane of the house facade were also measured. Levels recorded were generally luwer than for vertical or radial campanents. These out of plane vibrations

have had relatively little effect on the overall behaviour of the structure since they primarily consisted of localised mo~ments of individual structural elements. They were therefore not considered further. ) The peak levels measured are shown on Figs.6.5 & 6.6. These are plotted as graphs of peak vibration level against distance from the vibra~m: on Fig. 6.7.

Wavef~ms measured at each geophone station at ground level have been correlated, to assess the time taken fur the waveform to pass through the foundation. Plots of distance from the vibrator against time are shown on Fig. 6.8.

The following points of interest ~Terge fram the results:-

The ~tical cm~[rnent of vibration at foundation level attenuates with distance approxima~.ly in accordance with an inverse square root function, as would be expected f~ a strface wave. ~he behaviour a t eaves level mirrors that at the foundations, with an amplification which is typically of the order 30-50%.

The radial cm~ponent of vibration a t foundation level sh~ws no significant attenuation. At eaves level, radial vibrations appear to peak at either end of the s~n/c~/re with a drop in level t~wards the middle. At the party wall between the t~) houses, only 15% amplification of the level on the foundation is evident, whereas at the end closest ~ the vibrant, the amplification is between 4 and 5 times.

At eaves level, there is a very large change in radial vibration level between the end of the structure and the adjacent measuring position (4.20mm/s reduces ~o 1.73mm/s) in a space of less than

metres. This is probably attributable to the restraint afforded by the stiff and massive chimney. At the f~r end of the s~ructure frm the vibra=r, radial vibration in,eases once again beyond the second chimney. This suggests that the energy is reaching the fa r end through the ground or structural foundations rather than through the houses at eaves level. ~he central partion of the struc-~e is restrained be~een the ~o chimneys, whereas the t~o end portions are freer to vibrate.

-16-

6.4.2

The speed at which the radial component of the wave passes through the foundation is significantly greater than that at which the vertical component travels. Approximate speeds are 450 m/s (radial), 360 m/s (vertical).

Very high local vibration levels occurred in individual structural elements. The most extreme case was in the wall adjacent to the vibrator where the radial component of vibration exceeded I/.5 mm/s - an amplification of ii times the level on the foundation immediately below.

Response of Floors and Ceilings to Grouncborne Vibration

(i) Response to normal vibra tot pulse.

The peak levels recorded in various rocms are shown on Fig. 6.9. In the first floor rooms, measurements were taken on joists in the middle of each floor. On the ground floor, positions directly over a sleeper ~all were selected. There is a marked difference in peak levels between grotmd and first floors, which is prctmbly associated with the differing support conditions.

(ii) Frequency dependence of response.

A wide range of peak vibration levels were recorded on different floors. This is bemuse floor response is dependent on na-ral frequency, which in turn varies with ~ size, floor stiffness and floor mass.

The natural frequen~ of ~ of the floors was assessed by applying an impact to each floor and reuo~ding the response. For one of these the response to a groundborne frequency sweep was also recorded. These measurements are shown on Fig. 6.10.

The results of this experiment show that floor and ceiling response is highly frequency dependent. Where the driving vibration contains a significant cauponent at the natural frequency of the floor or ceiling, a large amplification of foundation vibration can result. Since the normal vibrator pulse contained energy at very few frequencies, care must be taken in drawing any conclusions regarding typical relationships between foundation response and joist behaviotu: from this experiment.

(iii) Dependence of response on local support conditions.

In one of the large ground floor roa,s, the response ~o a pulse was measure

i} ii)

iii)

on a joist directly over a sleeper wall on a joist midway between sleeper walls on a floorDoard midway between joists.

These results are presented on Fig. 6.11. The sleeper wall modifies the wavefcrm, and reduces the peak vibration level by up ~o 25%.

-17-

6.4.3

6.4.4

6.4.5

.Pasponse of Rafter to Grcundborne Vibration.

The response of a rafter to a groundborne frequency sweep was measured. This once again showed a highly frequency dependent behaviour with a maximum amplification of the foundation vibration of nearly 9 times (at a frequency of 17Hz).

Response of Structure to "Normal Use"

The vibration levels produced in the structure by such activities as door slamming will depend on how energetically they are carried out. Attampts were made to represent events which would be likely to occur in an occupied house, from time to time.

Specific activities were as folluws:-

m Slam internal door Slam front door Jump on floor (ground floor) Jump on floor (upper floor) Run down stairs.

Lmzations at which these events occurred and at which measurements were taken are shown on Fig. 6.12. Peak vibration levels recorded are tabulated beneath Fig. 6.12.

The results shawed that peak vibration levels associated with occupation of a house were higher than those generated by the vibrator, which was simulating the effects of traffic at high exp3sure sites. ~:wever when making a cumparison, consideration must be given to how often the two types of of vibration occur. The large impulses ~easured wauld be relatively infrequent events during the normal occupation of a house, whereas in same circumstances, traffic vibrations can recur frequently over long periods of time.

Another difference between the two sources is that groundborne traffic vibration enters the house through the foundations, rather than being generated within the structure. The vibration level measured on the foundation in response to an impulse on the upstairs floor was found to be less than 1% of the peak level generated within the structure.

Response of Structure to Airborne Vibration

The airborne vibration source gave a noise level of ll0dB outside the window. The corresponding level inside the roan was 99 an, 3d8 lower than that generated by the response to groundborne vibrations. A peak vibration level of 131 mm/s was measured on the windaw pane adjacent to the noise source. However, the highest vibration level r~sulting elsewhere within the structure was less than 2.5 mm/s (in a suspended floor). This was lower than the cooL,spending response to grcundborne vibration.

-18-

,~.4.6

6.4.7

6.4.8

6.5

6.5.1

Existing Noise and Vibration Levels

Fig.6.13 shows the distribution of noise throughout the 24 hr period in terms of dB fin. and dB(A) for the hourly values of L I, LI0, Leq and L90 indices. The 18 hr LI0 for the property was calc/la£ed to be 71 dB(A). The peak linear noise levels were of the order of 95 dB. The frequency analysis of this noise indicated that the energy was preck~inantly in the frequency range 40 - 80 Hz. All measur~nents were taken im fr~n the first floor win~w on the front of the facade of the house at a height of approximately 6m a~ove the road surface.

Vibration levels generated by passing vehicles were monitored on the house fmundation. Highest levels rt~orded were of the order of O. 7~s.

NOise Levels Induced by Groundborne Vibrations

Notwithstanding that the vibrator engine could not generally he heard inside the house, high noise levels were generated due to the vibrating floors and walls. Noise measurements throughout the two buildings are shown on Fig.6.14. A third octave band analysis of the noise is also given on Fig.6.14

By examining the vibrations of the floor and ceiling it was concluded that although these two surfaces were vibrating in phase, they were doin~ so at significantly different amglitudes and thus generating high levels of noise at the lower frequencies.

Fenestration Attenuation

The attenuation of the window closest ~o the noise source (i.e. the small 9~und floor front room of 56 Otford Read), was measured by generating random noise through the loudspeaker system and measuring the third octave noise levels inside and outside the room. Fig.6.15 shows the attenuation.

RESULTS - ~3NITORING OF MO%~MENT

Level Survey

There is sane scatter in the results of the level survey, but no trends of either settlement or heave appear at any location on the structure.

[M~r ideal conditions the equ mt used can resolve levels to _+ 0.1ram. However in this experiment the scatter in the results is such as to limit their accuracy to approximately + 0.3 ram. This error range is attributable to ~ independent factors. The actual precision of the levelling operation may have been affected detrimentally by site conditions. The cramped nature of the site led to problams with the use of consistent sight lines. Furthermore, the need to complete each set of measurements on a particular date resulted in much of the work being carried out in adverse weather conditions.

-19-

6.6

String B indicates and upward mc~ement close to the test foundation strips and is discussed in Section 7.

( ii)Magnetic Extensometers

Unfortunately, a technical problem developed in the magnetic extenscmeter meamming system and the soil movements in the four vertical boreholes drilled around the house could only be measured to an accuracy of 2nm. The readings were subject to scatter which could not be corrected.

The data, notwithstanding the scatter, indicated that no mm~mant developed in the soil immediately adjacent to the house. This observation was supported by a precise level survey on the top plates of the extensumeter systems.

FINAL SITE INVESTIGATION

The purpose of this second investigation, which ~ place at the end of the simulation experiment, was to determine the insitu density of the soil beneath the house foundation and the test foundation strips. Density was measured by sand replacement tests and by removal of m~isUEbed core samples for laboratory analysis. Results for the test foundation strips are diso/ssed in Section 7.

Under the house footing, and the vibrator pad, both methods of measurement indicated densities very close to the average for the site measured before the experiment began. ~he moisture content was also very similar to that at the start of the experiment, which suggests little seasonal effect on the general moisture content of the soil. These results are presented on Figs. 6.20 and 6.21.

The optim~ density of the soil and the moisn~re content at which it was obtained both fell within the range for the site in general.

-21-

7

7.1

7.2

TEST KX~N[I%TIC~ SIRIPS

CSJ~CIT~ES

The main simulation experiment was limited to a study of the response of a single sUmcUme ~ a particular amplitude of vibration. In order ~ broaden the scope of the w~rk it was considered desiraDle to have scm~ indication of the effects of vibration amplitude and fmundation loading on response.

An experiment was devised in which stall foundation strips were consUmcted. These were placed at different distances from the vibratmr and loaded with varying dead loads, the intention being

relate behaviour, particularly settlement, ~ these two variables.

[ES~CN OF EXPERIM~E

Slx '}{' shaped foundation strips were cmns~-ucted as shown on Fig. In each set of three, one was loaded such as to produce a ground bearing press~e equal to the estimated value under the house foundation. The other two were loaded to 2 and 0.5 times this value. The arrangement of the strips in r-lation ~ the vibrator is shown on Fig. 7.2.

7.1

~ration levels on each ~-t strip were measured in response ~e nomal vibrant pulse and to a frequency sweep. ~li~es rec~z~a and apparent resonant frequencies of vibration were rela~a tD the location of each strip and its loading.

L~elling stations were es~hlished on ~ch test strip (~qo per strip) and these were monitored before the strips were loaded, after leading and immediate settlement, and at intervals during the simulation experiment:.

One string of electrolytic levels (see Section 6.3.4) ran close tD Row B of test foundations. These were monitored throughout the experiment.

After the simulation was co.feted, trial pits were dug adjacent each row of test strips such that the soil density under each

strip ccu;a be detsrmined. ~hese densities were compared with values found elsewhere on the site at the start of the experiment (see section 5 - Site Investigation). ~ indirect comparison ~as necessary since the soil in the test area had t~ be undis~ed during the experiment, thus precluding density measurements in the area ~ i ~ ~ the exper~ent .

-22-

7.3

7.3.1

7.3.2

RESV~TS

Vibration Monitoring

Vertical vibration ampli~/des in response to a pulse are shown below. Levels are all in the range 2.9 - 3.6 mm/s and appear bear little relation to test strip dead loads. The strips close tm the vibrator exhibit vibration levels which, when compared with corresponding blocks further away, are in the range 5% tD 20% higher.

~%BLE 1 - TEST KX~KTICN STRIP V~RATION LEVIS IN RESPONSE TO NCR~L PULSE

Block reference 50ad Vertical vibration amplitude

A1 2.0w 2.90 A2 LOw 3.56 A3 0.5w 3.58 B1 2.0w 3.06 B2 1.0w 2.99 B3 0.5w 3.24

(W = estimtmc] load on house foundation)

A factor which might af~ct vibration levels is slight variations in ground stiffness be~wean the different strips which, Wether with the varying dead loads, resulted in different na~al frequencies of vibration fur the various blocks.

In urder to test this theory, the blocks ~e subjected t~ a frequency sweep, and the responses recorded. These are shown on Fig. 7.3. As can be seen, the predominant frequencies vary from block to block.

level S ~-vey

The results of the level survey are presented on Fig.7.4. In contrast tm the behaviour of the house (which showed no significant settlement), settlements in the range 1-11 mm %~re n~asured on the test strips.

The pattern of settlements corresponds with the loading and location pat~rns~ i.e. settlement inureases with proximity ~ the vibra~ur and with increasing dead load. However, care must be taken in ass~ a

direc~ ~use and effect relationship. If for example, the ~ most heavily loaded strips are compared, the ratio of vibration levels is 0.95, whereas the ratio of final settlements is 4.7. This once again suggests that some addltional fac~ur is influencing ~ results.

It is interesting ~ note that in all six test foundations, some tilting occurred. In each case, two levelling stations were m~nitmred. The ratios of higher settlement : lower settiement fur the six blocks in the final condition are tabulated below,

-23-

~nr;. 2 - TILTING OF TEST FOUNDATION STRIPS

Block reference Hax settlement Hin settlement Ratio (m) (m)

A1 14.0 7.0 2.0 A2 2.4 2.1 i.i A3 1.6 0.6 2.7 Sl 2.9 1.5 1.9 B2 1.7 1.4 1.2 B3 1.2 0.6 2.0

7.3.3 Electrolytic Levels

The electrolytic levels ran close to Row B of test foundation strips. The changes in level r z-~rded are plotted on Fig.6.19. These are somewhat surprising. Despite the fact that the test fourdations settled, the adjacent soil appears t~ have heaved. This could have resulted frcm various mechanisms, which are discussed below.

(i) General heave due to removal of 2.5m thickness of ~xic ~aste pit immediately pricr to the experiment. In n~mal circ.mmtances any such heave would have occurred very rapidly. Pmwever, on this site, the sand was lightly ammnt~d and could have d i l a t e d slowly under the influence of vibration.

(ii) Direct dilation of the sand due ~ vibration. Such dilation could have occurred in the area between the house and the ~-t foundations, where no loading was present, but is less likely under the test foundations themselves.

(iii) Heave associated with adjacent settlemmt of the test foundations. This mechanism could again explain part, but not all of the observed results.

It is likely that in fact =re than one of these mechanisms was operating.

7.3.4 Final Site Investigation

At the end of the simulation period, trial pits were dug adjacent t~ all six test foundation strips. Core samples were taken beneath the test strips for 1~hcra~-y analysis, and in-situ density measurement by sand replacement were perfom,~.

-24-

Sand properties under the test foundation strips were found to be very similar to those es~hlished elsewhere on the site prior to the simulation. In particu/ar, measurements of soil density, both by sand replacament and from core samples indicated no significant difference from the pre-vibration norm. (see Fig. 7.5).

Since the results indicate that no soil densification occurred beneath the test foundation strips, the settlements observed can pm:~ably be attributed to rotational mechanisms within the soil. Slight sb~ring along a typical log-spiral failure plane would accord with this theory and explain the observed tilting of the blocks.

-25-

8

8.1

8.2

8.2.1

8.2.2

DLg:USSI(~{ OF SIMULATION EXPERIMENT

SIMUIATI~

Simulation of groundborne vibration was successful. A total of 888,000 pulses were recorded, sire/fating the effect of over 3.5 million heavy goods vehicle axles. The response of the house foundation accorded well with messuremen~s of traffic induced vibrations. The fundameneA| frequency of the vibrant pulse was selec~d fru~ within the range m~st commonly generated by heavy goods vehicles. This generated resonances within some, but by no means all, structural elements within the test house. When interpreting the results it is important to remamber that response is dependent upon driving frequency. At any real site, passing traffic will generate vibrations with a rarge of frequencies whereas the simulation was carried out without m/~ variation.

Problems were encountered with the simulation of airborne vibration. The initial attempt to produce a waveform which simulated the entire range of low frequency airborne vibration resulted in very low vibration levels within the struc~re. Any particular vehicle would produce noise with a na~r frequency spectrum, so, during the latter part of the experiment, a single frequency source was selected. This generated noise at 27Hz, - the frequency which produced maximum response in the adjacent wL-z]ow. Airborne vibration simulation can t he re fo re be r~arded as a "worst case" for the particular noise level generated.

R~'I~T~E OF STRIETURE ~D (;~L~6C~ VIBRATIDNS.

Response of Foundations

The m~st marked characteristic of the foundation response is the difference ~ in behaviour between the vertical and radial amponents of the vibration (see Fig. 6.7). The results suggest that some considerable part of the energy contained in the vertical component is dissipated into the underlying soil.

Response of Structural Elements

The amplification of the vertical ccm~onent of vibration which occurs between foundation level and eaves level is probably associated with the smaller dead load carried at the higher level. This smooth response, with no obvious local peaks in level will be ~likely to result in severe stress ccmcentrations within the struck/re.

The radial comE~nent of vibration behaves quite differently. In this case amplification is mue2~ larger at either end of the structure than elsewhere. It appears that the two chimneys which run up through the structure afford sane restraint against radial vibration, so that the high amplitudes oco~r in localisad areas. This could result in concentrations of stress at locations where the chimneys join other structural members, despite the low vibration amplitude at these locations.

-26-

8.2.3

8.3

8.4

High stress lewis would also be expected in areas of high local vibration. The most extreme case of this was the end wall of the structure adjacent to the vibrator, where additional cracking of plaster was observed during the experiment in the first floor rock.

Response of Floors, Ceilings and Roof

The measurements undertaken show that vibration of wooden joists and rafters is dependent on the frequency of the vibration source, and the support conditions of the tLmbers. Although vibration levels on floors were normally higher than on adjacent walls, there is no simple relationship between vibration levels in the foundations and those in the floors and ceilings.

This is im~orrmnt, as human perception o£ vibration is often associated with vibrations in floors, whereas stress concentrations, which might lead to damage to a structure, are associated with vibration levels in the main structural elements. Vibrations in floors, ceilings and roofs could be associated with superficial damage but are likely to have little relation to any risk of structural damage.

I~SPONSE OF ~ TO ~ VTIg~TI(~S

Airborne vibration generated significant vibration levels in • windc~ and window frames, and caused some excitation of floors. Major struc~ral elements appeared to be unaffected, and no damage, even to windows, was recorded. Since the s~.,,l~tion of airborne vibration was devised as a "worst case", it would be reasonable to conclude that in general any damage resulting from this source ~m/Id be superficial. For cracks to develop in winduw glass or surrounds, unusual stress concentrations would probably have to exist therein prior to exposure.

RESPONSE OF S ~ TO " ~ USE"

The simulation of "normal use" of the house resulted in some very high, though transient, peak vibration levels in local elements of the struc~n~. These might lead to superficial damage of plaster, but are unlikely to lead to structural damage. Any damage associated with soil o-,~action and associated settl~ment would be highly unlikely due to vibrations generated within the structure, since levels reaching the foundation were found to be so low.

-27-

8.5

8.6

8.6.1

OF POSSIBLE [I~MAGE MBCHANISMS

The following mechanisms for vibration-related damage to Duildings are suggested:

(i) Compaction or migration of soil resulting in differential settlement and structural and/or architectural damage.

(ii) Damage to structural elements due to high local vibration levels. This could occur either as "trigger damage" in which stresses associated with an increasing vibration level combine with static stresses to exceed a damage threshold, or alternatively, repeating vibration could result in fatigue damage.

(iii)~mmge to structural elements and/or plamter due to changes in structural stiffness or mass. In the structure tested, the chimneys represent members of mass and stiffness greatly in excess of those possessed by adjacent and connected structural ram,hers. This resulted in relatively large changes in vibration level occurring within very small areas of brickwork. Suc~ differential movement could result in damage.

(iv) Damage to plaster due to vibration of floor and ceiling joists. High vibrat/on amplitudes in floors and ceilings could result in cra~king of ceiling plaster, or cracks at wall/ceiling joints.

(v) Roof damage, due to high vibration amplitude in rafters.

(vi) Dmmqe to windows and surrounding plaster due to airborne vibration.

AND DAMAGE AC1'U~r,r.v RBODRIED

Soil ~y~.mnts

One of the most significant results of this experiment was that ve~ I little movement of the test structure or the underlying soil was receded. Be site investigation indicated that the soil conrmined the potential for up to 20ram of settlement by densification. No settlement of the structure, by this or any other mechanism, was reoorded.

In contrast to the main structure all the test foundation strips did settle, by varying amounts. Any discussion of the test house behaviour n~st therefore take account of this fact.

-28-

8.6.2

Since the house did not settle, and no soil densification was recorded under the test foundation strips, it would appear reasonable to conclude that the vibration levels being generated were insufficiently high to densify this particular soil.

The settlement under the test foundation strips appeared to be associated with localised rotational movements within the soil. The main structure was unable to undergo simi1~r movements due to the difference of scale. A shearing plane beneath one of the test strips would be contained within the top layer of loose uncemented sand. For part of the house structure to rotate in a similar fashion, soil nDbilisatlon would he necessary to a greater depth. Under those parts of the structure built on top of the cellar, the deeper foundations probably contributed additional restraint against this mechanism. Huwever, the cellar would not have provided significant restraint elsewhere in the structure.

It is important to identify other circumstances in which settlement of a full sized structure might have been recorded. These are summarised below:

(i) Different soil type. Larger settlements would be likely in saturated sand (especially if loose) or normally consolidated soft clay or peat.

(ii) Other ground conditions. The rotational settlement mechanism o b s e t ' ~ under the test foundation strips could possibly have developed under the main structure had the loose unc~rented sand extended to a greater c~th.

(ill)Different structural conditions. Unlike the test foundation strips, individual footings of the test house were unable to rotate due to the restraint afforded by the structure. Had any single facade been poorly tied to the rest of the structure, a similar localised rotational settlement might have occurred.

Movements Within the Structure

Some small movements were recorded within the structure. ~ditionsl cracking of piaster did occur, though the extent was very small when compared with the cracking which already existed prior to the experiment.

The most significant m~ents within the plaster (new cracks or mu~sments in existing ones) ce.curred in the end wall of the house facing the vibrator, ana in ceilings close to the chimneys. These locations correspond with damage mechanisms (ii) and (iii) described in section 8.5.

The overall level of damage was very slight. It is probable that in a normally decorated house, none of the damage rt~orded would have bec~e evident.

The slight n~vements which did occur appear to have been as a result of a fatigue effect since they appeared relatively late in the simulati~ period. NO "trigger damage" was recorded in the experiment. This may have been due to the extensive cracking which already existed, which would have relieved any stress concentrations within the plaster.

-29-

. CC~.LUSIONS OF SIMUIATION EXPF2/MENT

This study involved an extensive and wide ranging series of experiments. ~ r it is important to recognise that the work was carried cut on a single site under particular conditions. The soil at the test site was a uniform, dry, loose to medium sand close to the surface, becoming lightly cemented at a depth of about two metres. The test structure was about ninety years old and constructed of brickwork in lime mortar , which is considerably more flexible than modern cement mor~r. In drawing conclusions frc~ the work these particular characteristics must be borne in mind as they will hm~e had a considerable effect upon the results recorded.

Damage to buildings caused by road traffic vibrations will, in our opinion, result from: differential settlement in the underlying soil; trigger damage, (in which existing critical stress concentrations are "topped up" by dynamic stresses from vibration); or fatigue effects due to repeated application of vibrations.

In this experiment only the last of these three mechanim~s was observed, and the resulting damage was limited to slight cracking of plaster.

~ring in mind the particular site conditions and the limited damage ac-,~lly observed the following conclusions can be drawn.

(1) Airborne vibration from traffic is unlikely to lead to structural damage. Windows and near~ plaster under stress could be damaged by trigger mechanisms though none were generated during this su~y.

(ii) The test structure was subjected to a considerable period of groundborne excitation during which vibration levels were at the high end of the range that traffic would be likely to induce in buildings close to an uneven road surface. Damage was nevertheless very slight. It would therefore appear unl~ely that struc~2ral damage would result directly from traffic vibration except in cases where settlement or trigger damage occurred.

The slight fatigue damage recorded might have been wu~&e if the plaster had not been extensively cracked already, or had been of a more brittle type. A newer house could be more susceptible to such a fatigue effect, as could a building which had suffered structural deterioration or major alterations.

(iii) The test structure suffered no "trigger" damage. Trigger damage results when existing high stress levels are "t~oed up" by an additional factor, such as vibration.

-30-

(iv)

(v)

In the test structure, extensive cracking of plaster had already taken place (though not to a degree which is unus,a~ for a structure of this age). This cracking would have gone a long way towards relieving any stress concentrations within the plaster, thus reducing the risk of trigger darage. Similarly, the flexibility of the lime mortar would, over the years, have relieved stress concentrations in the brickwork. In a house with less extensive plaster cracking the risk of vibration induced trigger damage to plaster would be greater. We consider that structural damage is highly unlikely to result from this mechanimu.

Differential settlement had been predicted prior to the experiment, but none was actually recorded. This appears to have been a function of the particular conditions existing on the site. Other soil types and gecmetries exist in which differential settlement would be more likely, and under such circumstances some damage might result (See section 8.6.1). Even on a soil such as that at the test site, settlement might have been observed due to localised rotational movements had the facades of the structure been less well tied together.

Previous work in the field of vibrations and buildings suggests that maxLmum nuisance to occupants is caused by ra t t l i ng windows and vibrating floors. ~he vibration monitoring work undertaken during this experiment showed that these effects are not simply related to vibration levels in foundations and walls. An increase in traffic related nuisance does not necessarily therefore imply an increase in the risk of damage. In our view, for any risk of damage to exist at all due to traffic induced vibration, considerable nuisance would have to be present.

-31-

i0

I0.i

i0.2

TRENCH LXPE~

OBJECTIVES

During the course of the sim,lation experiment, a problem was enoountered with disturDance to residents in a house some 45 metres from the vibrator. To reduce the disturbance, a 2m deep, partially retained trench was dug close to the vibrator between it and the house in question. This action succeeded in reducing vibration amplitudes by s~ne 20% sufficient to alleviate the problem.

Subsequently, it was decided that the site =~uld be used to assess the degree to which attenuation was afforded by such a trench if it was hackfilled with a material which provided a suitable impea~nce ~tCh. The purpose of the experiment was to evaluate the practical usefulness of such a vibration barrier in situations where ba~filling would be essential.

TIEOR~

To select a suitable fill material, the theory of wave transmission through three media was considered 41. For the case in which the first and third media are the same, this defines a sound power transmission coefficient ~tin terms of the densities of the two materials, the wave speed in the ~ materials, the wave frequency and the fill thickness. The relationship is shuwn below:

OCt =

4

4 2 + [PS2 I

+ p1¢1 2sin 2 k÷

!o.3

Where Pl = density in outer material P2 = density in filler material C 1 = speed of wave in outer material C2 = speed of wave in filler material I = thickness of filler material k 2 = (2 Tr ) x frequency of wave

C2

This relationship is only strictly correct for fluid media with infinite boundaries and for steady state conditions. ~:wever, it gives a good indication of the fac~rs which influence wave transmission.

CHOICE C~ MATERIAL

Since the wave speed in the soil and the soil density were known, a filler naterial was sought which would minimise tr~ission in accorclance with the above theory.

-32-

10.4

For this particular soil and filler thickness, transmission is minimised by reducing both filler density and wave speed in filler. The filler material must also be strong enough to withstand soil pressures within the trench. To match these requirements at a realistic cost, expanded polystyrene was selected as a filler material.

Various grades of expanded polystyrene were tested. The grade selected had a density of ll.lkg/ml-3. The wave speed in this material was 523 m/s. The corresponding figures for the soil were approximately 1800 kg/m 3 and 175 m/s, giving an impedance m~=match of over 50 to I.

The thickness of material used in the experiment was 1.22m. This rather large figure was considered necessary because the tranm, issien coefficient is highly frequency dependent. At the low frequencies in which we are primarily interested (10-20 Hz) the coefficient increases sharply. Figure I0.i shows the theoretical values of ~tfor the particular soil and filler in question for a range of filler thicknesses. For the limited trench dimensions in the experiment, values of ~t will be higher at all frequencies than those shown, due to passage of waves under and around the trench, but the overall trend could be expected to be similar to that slxz~.

~ F ~ Q q ~ E~:JERINENT

A ~ench was excavated to a depth of 3m, close to the vibrator as shown in Fig. 10.2 Vibration monitoring stations were es~ahlished on the vibrator p~a and in four positions at various distances beyond t he trench. Geophones were attached to a drain catchpit, the foundation of a house under construction and an existing house foundation. Measurements were taken at each monitoring position in each of the following conditions:-

i) ii)

iii)

open trench trench filled with polystyrene and back- filled around polystyrene re,Dyed and trench backfilled.

Pot each trench condition, response to a normal vibrator pulse and to a frequency sweep was recorded.

It was not possible to measure vibrations ~ough the undisturbed ~Luund at this stage due to the existence of the original trench. Measurements which had been taken prior to excavation of the original trench were not meaningful in this context as substantial construction work had taken place on the site in the meantime, which would have altered the passage of vibrations through the ground.

-33-

10.5

10.5.1

I:~SULTS

Response to pulse

These results are ~hulated below. In all cases the polystyrene barrier appears to perform better emn either the backfilled trench or the open trench, although in s~,e instances, the figures are narginal.

~BLE 3 - RESPONSES TO NCR~L ~ R POISE (RATIO OF H~%SORED P~K PARTICLE VELOCITY TO REFERI2rE PEAK pARTI~R VE[ECITY)

TRENCH CONDITION

Location Open trench Polystyrene Backfilled trench

A 0.276 0.223 0.307 B 0.132 0.130 0.159 C 0.082 0.074 0.087

10.5.2 Response to Frequency Sweep

~he results are plott~a on Fig.10.3. R~ve 40 IIz no significant differences are apparent he,men the different trench co~itions. Below this frequency, however, considerable variation occurs. ~ e fol~wing points, in particular, emerge:

Results for all trench condi~ are highly frequency dependent. The different responses at each measuring position suggest that aon- ditions local to each geophone array (i.e. weight, size and shape of foundations to which the array is fixed) domira~e the response. Comparisons are only made, therefore, between different measurements at a single measuring position.

In all cases, the polystyrene fill attenuates the vibration m~e effectively than the back- filled trench ~L~ough most of the frequency ~ange. The effect is m~st harked in the range 10-20 HZ, where the vibration levels passing through the backfilled trench are reduced by up to 50% by the polystyrene.

~mhe effect of the polystyrene ( ~ a r u ~ ] with the backfilled trench) is biggest close ~o the trench, a t most frequencies.

-34-

10.6

The open trench appears to provide bet~ attenuation than the polystyrene in the range 20-40 Hz but poorer attenuation in the range 10-20 Hz.

DISCtESION AND CGNCI/JSICNS OF TR~CH

The results of this experiment appear to indicate the grouncbcrne vibrations can be attenuated, at least to some degree, by a solid barrier. The most encouraging result was that the in~rovement in the atten-~tion affmrded by the polystyrene applied to peak at the low frequencies typically generated by traffic.

However, the results must be treated with some caution. The esperiment was under severe constraints of time and space. As a result it was only possible to experiment with one filler material in the tzench. Similarly, only one thickness of filler could be teated. The sits circumstances prevented measurements through undisUmbed ground. (It was no t possible to fully cc~act the the soil when the trench was finally backfilled). Finally, it was not possible to co,act the soil against the polystyrene when the trenc~ was bac~fille~ around it. The barrier therefore cc~rised two narrow bands of relatively loose soil, and a layer of polystyrene.

Making al/owance for these points it is only really poseible to conclude that sane additional attentuatlcn can be achieved by a solid barrier, but not precisely how much. To mere usefully quantify the results a more thorough experiment would have to be carried out.

-35-

ii

Ii.i

11.2

11.3

REX~Da.TIONS FOR ~ WORK i

SD~u%TION EXPER~4~T

In this study we identified several mechanis~s which could give rise vibration damage. Hswever, not all of these occurred on this

particular site and hence t h e i r effects upon buildings are yet to be established.

A second simulation exp=iment would provide an opp~rUmit7 to invesUiga~e these p~enmmma m~re fully.

We ree~.nd that an isola~ sUmcU=e is Io~, which is at least 20 years old. This should be founded on loose saturated sand cr soft clay. A new house could then be built adjacent

the existing structure and the ~o subjected tm vibration. The scope of the research would thus be ex1~.nded to cover variations in ground conditions and in type and age of structure.

T~T ~ O N STRIPS

This aspect of the current project yielded sane inter esning results. It would appear that a great deal more could be learnt about soil structure in~raCcion by this method. By controlled experima~ts on test fmm~ations, the relative i,%~rmnce of su~ variables as soil

and configuration, vibration level, fzmn~ticn loading and fmoUing gem~ry could be investigated in great.- derail.

~ E X P ~

The work carried out during this project to investigate at~_nua~ion of grouncbsrne vibrations was of a llmi~ nacre. The results indicated that a polystTrene barrier could afford saue at~nmtion, though the amount recorded was relatively m~all for a 1~rge (and hence expensive) barrier. It would appear unlikely that su~ a technique could be commercially viable except in a few sensitive locations.

This walk does, hswever, su~]est that attentuaLion of traffic vibration by a buried barrier is possible, and the effect of locating such a barrier directly ber~ath a road w~,la be worthy of investigation. A simple comparison could be achieved relatively cheaply by the cons~uction of a series of test sections of road, founded on different barrier layers.

-36-

12 ~ C ~ . R ~ S

The work described in this r e~ort was carried out under contract to TRRL. The Project Managers for TRRL were Dr Nelson and Dr Watts who provided valuable support and assistance throughout the study. The contributions of other memDers of TRRL are also gratefully acknowledged.

In addition, we gratefully acknowledge the help and assistance given us during this study by.

The National Physical TahnratDry who were responsible for the Moire Photography S~dy.

The Building Research Establishment, who ins~mlled the Electrolytic Levels

South Eastern Gas Board (Estates Division) - the original owners of the site.

Alfred McA1pine ~mes Limited -who acquired the site frcm the SEG8 for development and provided us with resources for a number of aspects of the project.

-37-

13

i.

.

.

.

.

.

.

REFERENCES

Martin D J, Nelson P M and Hill R C (1978)

Measurement and analysis of traffic-induced vibrations in buildings

Transport and Read Research r~hora~ory Supplementary Report 402

Watts G R (1984)

Vibration nuisance frcm road traffic-results of a 50 site survey

~ ~ t of ~:ans~ort

Transport Statistics Great Britain 1972-1982

Civic Trust (1970)

Heavy lorries. A Civic ~ust Report

Watts G R (1986]

Traffic- induced 9coundborne v~brations in dwellings

Proceedings of Institute of Acoustics Conference. Salfcrd 1986

Steffens R J (1974)

Structural vibration and damage

Building Res~rch Establishment Report

Whlffin A C and Leonard D R (1971)

A survey of traffic-induced vibrations

Road Research r=horatzzy Report LR418

-38-

.

.

10.

11.

2.

13.

14.

Martin DJ (1978)

Low frequency traffic noise and building vibrations

TRRL Supplementary Report 429

(1983)

Vibrations : building and human response

Building Pes~rch Es~hlishment Digest 278

Studer J and Suesstrunk A (1981)

Swiss standard frr vibrational damage to buildings

Proceedings of 10th ICSMFE 1981 Vol. 3 pp 307-312

(1966)

C~acking in Buildings

Building Research Station Digest 75

Stavinitser L R, Farpen~ V P (1977)

A lahora~ry study of the vibration s~hility of a sandy foundation bed

Translated from- O~1ovaniya Fundamenty i Mekhanika Grt~Itmv, No 2, 26-28 March 77

Goldshtein M N, Khain V Ya, ~olyubchik V S (1974)

Experimental l~ratcry investigations of vibration creep of sandy ioundation beds

Translated from Osnovaniya ~menty i Mekhanika Gruntov No I, pp 33-35, Feb 1974

IIyichev VA, Kerch~sn V I, P~bin S I, Platetsky V M

P~perimental s-~y of sand soil vibrocreeping

1 S~mp soils under Cyclic and Transient leading, S~nsea. Ed Pande & Zienkiewicz pp 239-245 (Balkema - Rotterdam)

-39-

5a. Eastwcod W (1953)

The factDrs which affect the Natural Frequency of Vibration of Foundations, and the effect of Vibrations on the Bearing Power of Foundations on Sand.

Proc 3 ICSMFE Switzerland, Volume 1 pp 118-122

Eas~md W (1953)

Vlbrations in Foundations

be Structural Bngineer March 1953 pp 82-98

16. Mencl V, Kazda J (1957)

Strength of sand during vibration

Proc 4 IC~ Vol I, pp 382-383

7. R~dy A S, Srinivasan R J, Nm~iar P (1970)

Influence of g~oundwatmr on the dynamic c~sract~-istics of ~i.~s on sand.

I E (I) Journal of Civil ~ineering Vol 50 ~larch 1970

18. Bmmund W F, Leonards G A (1972)

Subsidence of sand due to surface vibration

ASCE, 98, SMI, pp 27-42

8. Kolbuszewski J, Alyanak I (1964)

Effect of vibrations on the shear strength and pcEosity of sands

Be Surveyor and Municipal Engineer Volume 123, ICE, June 1964, pp 31-34

0. Stavinitser L R, Karpushina Z S (1973}

D~mic =iaxial tests of sandy soils.

O~sll 'uct ion Properties of Soils. ~Tom: Osnovaniya, gundamenty i Mekhanika Gruntmv, No i, pp 23-25

-40-

21.

22.

3.

24.

5.

6.

7.

Cisek T (1981)

Strength of cohesionless sand under vibra~cry load.

Int Sym 9 Soils under Cyclic and Transient loading, Swansea. Volume I, pp 143-155

Cundall P A and Strack 0 D L (1979)

A discrete num~ical model for granular ass~lies.

Geotechnique, 29, i, pp 48-65

~ee K L, Seed H B, D~nlcp P

Effect of transient loading on the strength of sand.

Proc 7 ICS~$FE, Mexico 1969, pp 239-247

T ~ n D H and Wu T H (1969)

Behaviour of dry sand under cyclic loading

ASCE, 95, S~4, pp 1097-11/2

lqmqa,4 T (1978)

On the v i b r a t i o n a l c h a r a c t e r i s t i c s o f a sand l ayer as a fou,dation model.

JSCE No 275, Vol I0 pp 225-233

Borcherdt R D

Rayleigh-type surface wave on a linear visco-elastic half-space.

J Acoust, Soc, Am. Vol 55, No 1 1974 pp 13-15

G~zetes G (1980)

Static and dynamic displacements of foundations in heterogeneous multilayered soils

Geotec~nigue 30, NO 2, pp 159-177

-41-

28 Gazetas G, Roesset J M (1979)

Vertical Vibration of ~achine Foundations

~SCE 105, 6T12 pp 1435-1454

29. Gazetas G (1981)

Strip fDtmdations on a cross-anisotrcpic soil layer subjected tX~ dynamic loading

Geo~::l~5"Fle 31, No 2, pp 161-179

30. Gazetas G C, Noesset J M (1976)

Methods of Structural analysis Vol 1 Ed W Saul, A Peyrot. ASCE Spec Conf, Univ Wisconsin.

"Forced Vibrations of Strip Footings on Layered Soils"

31. Gazems G, Yegian M K (1979)

S~r and Raleigh Waves in Soil D ~

ASCE 105 Grl2 Dec 79, pp 1455-1470

32. Gazetas G (1981)

Machine Foundations on deposits of soft clay overlain by a ~sa thered crust.

Geotechnique 31, No 3, pp 387-398

33.

34.

Lapin S K (1979)

Experimental determination of the coefficient of apparent additional mass of soil for verti~l vibrations of foundations.

• :anslated from Osnc~aniya, ~menty i Mechanika GruntDv No 3, pp 9-10, 1979

Anderson K H, Brown S F, Foss I, Pool J H and Rose,rand W F (1976)

Effects of cyclic loading on clay behaviour.

Proc. Conf on Design and Construction of off-shore structures. London pp 75-79

-42-

35.

36.

37.

38.

39.

40.

41.

Castro G (1969)

Liquefaction of sands

PhD thesis, Harvard University, Massachusetts, USA

Koutsoftas D C (1978)

Effect of cyclic loads on undrained strength of two marine clays

American Society of Civil Engineers, ASCEI04, GT5 pp 609-620

~kahashi M, Hight D W and Vaughan P R (1980)

Effective stress changes observed during undrained cyclic triaxial tests on clay.

Proc. Int. Syrup on Soils trader Cyclic and Transient Loading, Swansea, i, 201-209

Smith I M, Molsnkamp F (1980)

DFnamic displacements of offshore structures due t~ low frequency sinusoidal testing

Geotechnique 30, No 2, pp 179-205

Awojcbi AO (1982)

Vertical vibration on a compound halfspace of a homogeneous stra~xn with Gibson subsoil

Geotechnique 32, No 4, pp 305-313

Cooke R W and Price G (1974)

Horizontal inclincmeters for the measurement of vertical displacement in the soil around experimental foundations.

Field Instrumentation in geotechnical engineering. p~ i12-125. London : Butterworths

Kinsler L E and Frey A R

Fundamentals of Acoustics pp 136-142

-43-

APPENDIX

i.

.

.

i. LITERATURE REVI~

GENERAL

A mawr review by Steffens 6 was published in 1974. This examines factors affecting structural response, natural frequencies of buildings and building elements, human sensitivity to vibrations, criteria for damage and ccnmDnly measured vibration levels. Further details of some aspects of this review are given under the relevant headings below.

VIBRATION LEVELS GENERATED BY ROAD TRAFFIC

The generation of groundborne vibrations is discussed by Whiffin and Leonard 7. They report that vibration levels are highly dependent on the quality of the road surface. Measuring 3.65m from the edge of the road, they recorded levels of 0.i - 0.25 n~n/s on a smooth surface, rising to 0.7 - 1.47 mm/s for the same vehicle passing over an artificial wedge 21ram high. Steffens I reports measurements of vibration amplitude close to roads. These are mostly in the range 0-10 microns (approximately 0-i ~m/s) though some are as high as 60 microns (approximately 5,m/s at 15 Hz). Both authors report vibration frequencies in the range 5-30 Hz.

Recently Watts 5 has carried out measurements in buildings close to relatively large road surface irregularities. At one site vibration levels of 130mm/s 2 at approx. 10-13Hz (approx.2mm/s) were recorded close to a poorly backfilled trench.

Martin, Nelson and Hill I report vibration levels in floors of I0 mm/s 2 RMS acceleration at i00 Hz (approximately 0.02 ram/s), reducing to 1 mm/s 2 at 25 Hz. The corresponding values on the foundation were 0.5 n~u/s 2 and 0.3 mm/s 2. The vibration source was a lorry travelling on a smooth road surface generating predominately airborne vibrations.

RESPONSE OF BUILDINGS TO VIBRATION AND CRITERIA FOR [I~%%GE

Martin 8 examines building response at four sites, once again concentrating on airborne vibration. He notes floor response in two frequency ranges: 63 to 125 Hz - corresponding to the driving frequency of the airborne vibration, and i0 to 25 Hz - corresponding to the natural frequencies of the floors. His report investigates the amount of nuisance reported by house occupants from traffic vibrations.

The subject of nuisance is further investigated by Watts 2. He reports that rattling doors and windows, and vibrating floors are the main causes of vibration disturbance. He also observes that the fear of vibration induced damage expressed by occupiers is disproportionate to actual vibration levels cc~Dnly recorded.

.

4.1.

Steffens 6 considers various scales which attempt to provide a numerical measure of intensity of vibration which can be directly related to damage. These include the vibrar and Zeller scales, both of which predict a damage threshold (i.e. possibility of small cracks in plaster) of about 3-4 mm/s for vibration at 13Hz. The German standard DIN 4150, 1970 is also reported. This sets a limit of I0 mm/s resultant velocity for transient vibrations in sound buildings (i.e. buildings previously undamaged apart from minor cracks in plaster).

BRE Digest 2789 reports that cases where vibration has caused damage to buildings are extremely rare. It Suggests that many complaints of architectural damage due to vibration are unfounded and result from the extreme sensitivity of human beings to vibration. Some advice is given on the design of structures to resist vibration.

Studer 10 describes the Swiss standard SN 640 312 "Effects of vibrations on structures". This proposes maximum permissible vibration levels in structural foundations. Different levels are given for different structural types, frequency ranges and vibratioral sources. For a masonry structure founded on concrete, excited by traffic vibrations in the range 10-30 Hz, the maximum level permitted is 5 mm/s.

It should be borne in mind when assessing vibration damage to buildings that cracking can occur for a large number of other reasons and is cx~monplace in buildings. BRE digest 7511 summarises a number of l causes of cracking in buildings and discusses their relative importance.

SOIL/S~R~ INTERACTION

The behaviour of soils is generally well understood in both static and cyclic loading conditions. However, the effect of vibrations on soil already subject to a static load has been considered very little.

Behaviour of Footings on Sand

The stability of footings on sand has been investigated by Stavinitser and Karpenko 12, who found that vibrations reduced bearing capacity. Goldshtein 13 examined the effect of "vibrocreep" - continuing settlement of a statically loaded footing due to vibrations. In some cases, vertical settlement continued indefinitely, although at a steadily decreasing rate. In others, bearing capacity failure was observed.

Ilyichev 14 also investigated vibrocreep in sand. He tested a series of fotmdation blocks of various sizes and weights. One block was seen to rise, as a result of settlement in nearby blocks. The na~)ral water ~hle was i. 6m below ground level. After initial tests, the sand was saturated up to ground level, which resulted in greater settlements. Relationships were established between settlement, stress level and proximity to the vibration source.

4, 2.

4.3

Fastwood 15 investigated the response of footings on sand to vibration. He observed a non linear relationship between footing load and natural frequency. Damping characteristics of the sand were examined and related to cbserved settlements. He also found that saturated sand had lower natural frequency and lower bearing capacity than dry sand. The observations regarding bearing capacity were confirmed by Mencl and Kazda 16. Reddy et a117 corroborated the results regarding resonant frequency.

BrumLmd and Leonards 18 observed continuing displacement with time during vibrating footing tests and related this to nett energy input.

Behaviour of Sand under Cyclic Loading

Some laboratory work on the dynamic behaviour of sand has been carried out. Kolbuszewski 19 found that with increasing vibration level sand densifies down to a minimum before starting to dilate again. S~vinitser and Karpushina 20 vibrated sand samples in a triaxial apparatus. They found that at frequencies above 10 Hz, shear strength decreased. Cisek 21 observed strength reductions even at lower frequencies.

Lee et a123 report that the shear strength of sand is increased with higher rates of loading. This could be explained in the light of results presented by Cundell and Strack 22. They found that "columns" of sand within a deposit carry the load. As the load is increased the col~uns buckle and other favourably positioned grains form columns and in turn carry the load. At high loading rates there is less time for this process to take p/ace, and hence mcbilised shear strengths are higher.

Timmerman and Wu 24 subjected sand to c~,bined static and cyclic loading in a triaxial apparatus. They found marked increases in strain when cyclic loading was applied, at frequencies between 2.5 and 25Hz.

Kagawa 25 investigated resonance characteristics of a sand layer on a shaking table, finding several harmonic peaks. However, significant boundary effects were observed in the particular test arrangement.

Theoretical Work

Theoretical studies intD wave propogation in soils are naturally limited by the idealisations of the soil employed. However, we feel it is relevant to mention some of the more important analytical papers in this short review.

The propogation of Rayleigh waves hasbeen examined by Hartan and by Borcherdt 26. C~zetas and others 27-32 have investigated machine foundations, usually modelling the soil as a multilayered semi-infinite half space, lapin 33 has addressed the problem of the volume of soil which moves with a footing.

Low frequency loading of soil has been examined by many authors 34-38. In particular, Awojobi 39 considered vertical vibrations transmitted through a homogeneous stratum in to a "Gibson '' subsoil (one in which shear modulus increases linearly with depth). He concluded somewhat surprisingly, that the strata behave as if they were homogeneous throughout. This has practical implications, as many soils exhibit constant properties in a layer near the surface, which then increase at depth.

LIST OF FIGURES

Plate 1 Plate 2

General View of Test House The Geophysical Vibrator and Loaded Test Foundation Strips

4.1 4.2

Construction Details - Ground Floor Construction Details - First Floor

5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11

Location of Trial Pits and Borehole Sieve Analysis at 5 Depths in Trial Pit 2 Sieve Anlysis at 3 Depths in Trial Pit 3 Variation of Optimum Moisture Content with Depth Variation of Natural Moisture Content with Depth Variation of Maximum Dry Density with Depth Variation of Dry Density with Depth before Vibration Borros Penetrometer Results Average of 3 Borros Soundings Perth Penetrometer Tests - Trial Pit 2 Perth Penetrometer Tests -Trial Pit 3

6.1 6.2

6.3 6.4 6.5 6.6 6.7

6.8

6.9 6 .i0 6.11

6.12 6.13 6.14a 6.14b 6.15 6.16 6.17 6.18

6.19 6.20

6.21

Site Layout Comparison of Vibrations Produced by Heavy Lorry & Vibra tot Location Plan - Levelling Stations Sample of Moire Photography Paper Peak Vibration Amplitudes - Vertical Component Peak Vibration Amplitudes - Radial Component Variation of Vibration Amplitude in Structure with Distance from Vibrator Variation of Wave Arrival Time with-Distance from Vibrator Peak Vibration Levels in Floor and Ceiling Joists Frequency Dependence of the Response of a Floor Joist Comparison of Floor Responses under Differing Support Conditions Location Plan - Normal Use Experiment Existing Noise Levels at Site Prior to Experiment Noise Levels in Front Rooms due to Vibrator Alone Frequency Analysis of Noise in Room 4 Variation of Facade Attenuation with Frequency Movement of Observed by Moire Photography Visible Cracks Prior to Simulation Experiment New Cracks and Crack Movements Observed during Simulation Experiment Ground Movements Measured by Electrolevels Variation of Natural Moisture Content with Depth after Experiment Variation of Dry Density with Depth after Experiment

7.1 7.2 7.3 7.4 7.5

Test Foundation S trips Arrangement of Test Foundation Strips Response of Test Foundation Strips to Frequency Sweeps Settlement of Test Foundation Strips Insi~u Soil Densities beneath Foundation Strips at End of Experiment

i0.i

10.2 i0.3

Theoretical Values of Coefficient of Sound Power Transmission for Various Thicnesses of Polystyrene Trench Experiment Monitoring Positions Vibration Levels Resulting from Different Trench Arrangements

PLATE 1 GENERAL VIEW OF TEST HOUSE

PLATE 2 THE GEOPHYSICAL VIBRATOR AND LOADED TEST FOUNDATION STRIPS

m ~ m ~ v o ~

: . . . . i ~ . . . . . . . . 7 " " ~ . 1

1 ~ L : ...... ] i i - ;z : I : . '-<.

• 5 ; . " ' . . . . . . " ~ ' ~

.... !5 ! i = 0 1,1,1 !

m : ....... ~ z ~ "~

. . . . . . . . . . . .

-:- ">" "1 -,- ~ ~= <~,,, I L ]1 ~=./ . == l u . i I I \ , . " - , _ 1 I " ' - i n " ; , ~C - ~ I n i l - . . . . :, O O ~ ; I

/ + . . - ~ 1 , I V ~ 7 , ~ q ~ m ~ i '

,,,., ,. • ~ . . . . ,:,~ , ~ ,, ,:,, ',. ~ : ; : ,~ , ,.,~ . . . . . . . . . . ,, . ~,< ..~.,,, .....

":'t: IL<;. t~'<.."'~'":.. I '.,

"r" ~ 1 , ' . • --- ,,,. ._ ~ ,.. ---4 ~ ~ ~=~ :'~ ~ ,-, / ~ . ~ .

. j " ,~ ,~ ,~ , ,~ : x , ~ x ' , : . , t - : ' ~ t , , . . . ~ , ~ v ~ ,~1 ~'~.~,:~,,. . --,,,,.~, .-..,. ~ _ - -

- - l - - ~ : Z IO0 ( / } 113111

I I

o

I',,,,, o3

04

x

h- I

. - . ]

I

" 1 ' " ' '1 i 1 "

i

I

?,,-. \ .,~

. . J • • . , -

i i

° ~

0 Z ~

Y

0~I-

o.,

Ol

03 -I

-.-_:~l~ff .-~- ~ .

x ---~ CO : ~ "~'I,U ~, 111I u~ : x ; i ! O ~

• ~ "1" ~ ; = , IIII~ _~ .- ~ ,,= '-

I! -,!=- II .A I

| 1 ] , i - I

L. - . . r I i ! , , I ,,~,h~,~,,,!i i~,~ ~',"':~

t-- ' ; i : "r" re

u ' i ca-"'---~- I 1 1 ' =

=: :o~u) (2 ( n z - J /

~u . I 0 0 ° "

i~ I - - - m,¢l: . . . . . 0 ~ ..,I

~ . ,. \

_ _ . , -

~,-I ' --"

=OUJ I-

¢~OD

I-- uJ a

~ e n

Z

#. o

Z

ou) 0 ~ z

(~1

(9 i

[L

Q

TOXIC WASTE TiP

_ ..: • ~i ~

i '

I

7

I

i . , • - :

l

I"-

I

," •._ /

m .2 A2

A3

I

B1 m a

I

L 'L

'\'" I i ~ \

B9

I I I I

I A1

10 METRES

m A1 TRIAL PIT- STAGE 1 KEY m B1 TRIAL PIT" STAGE 2

• HA HAND AUGER BOREHOLE

FIG 5.1 L O C A T I O N OF T R I A L P ITS & BOREHOLE

0 0

0

ONISSVd

0

3E)~I 1N3S)EI3d

EEEEE

o.o<] x

8

o ~

P-

E

"- 8 Z

~ o m m

"" ~ m

N ~

o Z ~ ~, < 6

x ~ uJ

Ixl W ~

z ~

0 0

0 m

IZ.

u

0 0 0

0

8 o o

9NISSVd 3gV~hl~13cl

tLI

u

- - I u

I--

_z

E

J

Z

LU > W m

U~

EEE

0

1"0

~1 2"0

I--

LIJ

3-0

, , / OPTIMUM MOISTURE CONTENT 7 e

5 10 i L

TREND LINE

FIG 5 .4 VARIATION OF O P T I M U M MOISTURE CONTENT.WIdTH DEPTH

0

1"0

2"0

,,n

3: I - (1. IJJ O

3-0

NATURAL MOISTURE CONTENT °/o

I

o

t'44- / /

x x'x / /

/ /

%0" /

;< /

/

/ [] /

( x O

'+'H- Xx O

O

4-

x

÷ O

TREND LINE

+ /

/

/ /

/

/

o

Q

TREND LINE FOR OPTIMUM MOISTURE CONTENT

O HA O TP 1 ;K T P 2 + T P 3

FIG 5.5 VARIATION OF NATURAL MOISTURE CONTENT WITH DEPTH

1'5

I " 0 - -

A

E

, J hl >, I,I - J

o Z

0 {z: L~2.0 -

II1

"I-

n h i ID

3 " O -

MAXIMUM DRY DENSITY 1"6 1 "7 1' 8

I I I

X

O X /

/ /

I /

/ /

/ /

/ o +

/ /

/+ /

/ /

/ /

1"9 I

/ o / /

/

2 "0 Mc j /m3

/ /

/ /

/

TREND LINE

+

-I"

~- HAND AUGERED BOREHOLE

x TRIAL PIT 2

-r TR IAL PIT 3

F I G 5 . 6 V A R I A T I O N OF M A X I M U M D R Y D E N S I T Y W I T H D E P T H

DRY DENSITY

1.5 1. 6 1.7 1.8

0

1'0 - -

~o 2.0 -

D .

3 . 0 -

I I I 1.9 1.10

I i

Mg/.

• o

÷ /k

A

/k x

TREND LINE FOR EXISTING DENSITIES BEFORE VIBRATION

0

D LINE FOR / MAXIMUM DENSITIES

/ (FROM FIG.5.6)

r7 + / /

x

'/ • TP 1 "~ SAND X TP2 I REPLACEMENT + TP 3 TESTS

0 TP 1 ~ DENSITY /~ TP2 ~ CORE TESTS o TP3

FIG 5.7 VARIATION OF DRY DENSITY WITH DEPTH BEFORE VIBRATION

..J IJJ > hl ..I

C]

Z .-~

uJ 1:D

3:: I=-

,,i 2 a

BLOW COUNT 1100mm

2 4 6 8 T T T T

• TP1

Z~ TP2

B T P 3

/x

, o

m

10

FIG 5.8 BORROS PENETROMETER RESULTS

E

~ 1.0 LU :> LU ._1

Q Z

O

8

LU m

- r

~- 2"0 O. LU Q

3'0

0

f

BLOWS PER 100 m m

2 I

6 I I I

>

q .

--L

FIG 5 . 9 AVERAGE OF 3 BORROS SOUNDINGS

BLOWC.OUNT PER 150mm PENETRATION ,,, ,~ ~

1.0-

E v

. J W

Q z 2"0- O

W

W

3 '0 -

\ \ "-- <-,~

\ \ ' , o

~ • • N ",e,,.

, j I

I

\ f " \ r/

,! \ • ~ \ e

/2 ~ . .

. . . . . BOUNDING ENVELOPE TO ALL TEST RESULTS

FIG 5.10 pERTH PENETROMETER TESTS TRIAL PIT 2

0 BLOWCOUNT PER 150mm PENETRATION

101 2~3

1.0-

Z

2.0-

UJ m

I" o

3'0-

o. '%

"I ' %

°,%

e, °'•% "°

" • ". "'O

•" °• ' '" "' ' °o..

- e. "'e.

"'e, ( ( "" /

°•'. •

O . . . . .

" ° ' ° o ° •

°°

I, °°%, ',

°% "1

•% %,,

°-.

'%.

" " ~ll

, . , ° ' •

BOUNDING ENVELOPE TO ALL TEST RESULTS

FIG 5.11 PERTH PENETROMETER TESTS TRIAL PIT 3

Z 0 i ).- 0

0 :

0

o 0 (D

" it a.

(n i

z I 0 m l mm

Z UJ >

"~ W

u. 0 (n i i . . r r ~ I (/) I,- Z '

I - w ~ I I I ..I I -

I - - - - - - "J U)

J H U

|

[ ]

Ill (n

u4 i - u4 :E 0 (n z u4

x uJ

u i.- u4 z c3 <[ :S

I i

[

• i I • i

OOg L : 0 0 0 ~ 0 0 ~ ~ c~ <C Q.

I

I

!

/ /

- / q /

/ / --~

i

0

(n Q.

(n

z o m

< a ; [

0 u.

I - (n w

W z z O

0 < ~ . m n -

=-=-o

0 0

U.I .J

0 U).

0

,,J

W I-- m

U)

(D

m

U.

E E

5

0

- 5 0

1 AXLE 2 AXLES i

P i 3 AXLES _ l

0'.5 ' ' 1"0 ' ' 1;5

TIME (SECONDS)

VERTICAL COMPONENT OF VIBRATION MEASURED ON A WALL lm FROM KERB IN RESPONSE TO FULLY LADEN 6 AXLE VEHICLE TRAVELLING OVER

2Omm DEEP IRREGULARITY IN SURFACE

0 E E

-5 o o.s

TIME (SECONDS)

VERTICAL COMPONENT OF VIBRATION MEASURED ON HOUSE FOUNDATION IN RESPONSE TO VIBRATOR

F IG 6 . 2 COMPARISON OF V I B R A T I O N S P R O D U C E D

BY HEAVY I.ORRY AND V I B R A T O R

27el

30 0

291

281

I 1

12

D3

.4

25 26

=5 69 7 , 8

240

O 23,35 22,34

19,10

11,12, e 36

33 I e20 , -- ... 1 7 ~ 1 6

NOTE: WHERE TWO OR MORE NUMBERS ARE GIVEN, STATIONS ARE AT

THE SAME LOCATION IN PLAN BUT AT DIFFERENT LEVELS.

FIG 6 . 3 LOCATION P L A N - L E V E L L I N G S T A T I O N S

FIG 6.4 SAMPLE OF MOIRE PHOTOGRAPHY PAPER

in

e ~

i n

(9

i

o ~

O)

m

03 m

(D

m

O

rn

:>

I I O

& I I I I I

0~

' O0~/I " o~ (9

A

(n

E E

--I

o m

r r ILl =)

,.J

m a rr

t UJ I z I 0 Q. I :E I 0

°

. J

I - - I--

I © IX ¢ o w I - ~ I I I

I <

CD m m

< W a .

I

e,,~ 6 N

0 IN

I

m

m

0

- I

U

ne i11

IIQ

I_

I t

I0

I I I J I

e l

• / . ~ I -

e l o

I - - I I I I f I

I I I ©

pl .o I - I I I I

16 I

I I

7

0 1 - -

j -

I I I

I

o__.~, ._-

- I

!

a

n .

E E

I - Z I l l Z 0 a.

0

. J

I

U.I a

I - . . I a .

Z 0 I -

U.I a .

m

I.I.

5

I,t.I

4

I -

z~ O E m ~m#

~- 2 <[

m m

0

0

+

0

0

+ + 0

+ + + +

% 0

÷

5 10 15 2 0

HORIZONTAL DISTANCE FROM VIBRATOR (m)

O- EAVES LEVEL

VERTICAL COMPONENT Jr-FOUNDATION LEVEL

5

4

U,I a

I-.. m

3 11. :E.-, < ~ z E OE i

I-- " " 2 n - m

0

0

-I-

0

+ 4 0 +

%

+ + ++

0

+

5 10 15 2 0

HORIZONTAL DISTANCE FROM VIBRATOR ( m )

FIG 6 .7

RADIAL COMPONENT

VARIATION OF VIBRATION AMPLITUDE IN

STRUCTURE WITH DISTANCE FROM VIBRATOR

E w u z

m

20.

15'

10'

.

0 VERTICAL WAVE

t HORIZONTAL WAVE

+

÷

O +

++

+

O O

+ O0

+ +

+ +

+ +

+ O

O

O

Oo

O O

, , , , , ,

o lO 2'0 io io so

TIME LAG msec

FIG 6 . 8 VARIATION OF WAVE ARRIVAL T IME

W I T H DISTANCE FROM VIBRATOR

2'77

0.72

n

3"00

3'57

1"13

6'26

4.83

2.09

17

12.11

1'86

FIG 6 .9 PEAK V I B R A T I O N LEVELS IN F L O O R AND CEIL ING J O I S T S (mm/s)

E E

U.I ¢]

I,- I

, . I Q.

,,¢

Z O I- < rr m m >

20"

15-

10-

5'

0 _ ~o 2'o 3o 4o so 8o 7o

/ 1 J ~ FLOOR JOIST

\~ . . . . . . . REFERENCE ON FOUNDATION

/ /

/11 \,..~)/ \

FREQUENCY Hz

ENVELOPE OF PEAK VIBRATION AMPLITUDES

IN RESPONSE TO A FREQUENCY SWEEP

uJ a

I- " i a . =E <

. J > a n ,

- I

t 0

0,1 10

| , , , . . . . . . .

20 30 40 50 60

FREQUENCY Hz

70

SPECTRUM OF RESPONSE TO A SINGLE IMPULSE

FIG 6 . 1 0 FREQUENCY DEPENDENCE OF THE RESPONSE OF A FLOOR JOIST

i

in

Z 0 m

< u4 ~: (/) Z

O 0 3: u.

I

~3

~,= , , . . - o= ~L I= U,I

0 ~ ~ U.I

U31tl tlDl~ I~ I I I

s/~uw s/~uu~ sl~u~u sl~uu~

0

,,o, 8

o ",~

E E E E _~

O

n x w w In z uJ u4 w r r

o a . ~ ,~ w z u 4 /.~ , - I LI.I rn

0 ~ - ~ Q r n n ~

o o o ~ " T '='3 " J I.I- ~

., o •

< .J a. qlm

~m

( 3 m 14.

L

i

11

T

L

I

ii ooo ,, _LI

3

t

I-I

i ,, 1,2 ~ O R ~ O R

iLii ' ,, .L.I 5

4

MEASUREMENT POSITIONS ACTIONS MEASURED

1. GROUND FLOOR (VERTICAL~ A. IMPULSE ADJACENT TO GEOPHONE 1

2. FIRST FLOOR (VERTICAL/ B. IMPULSE ADJACENT TO GEOPHONE 2

3. WAM~GROUND FLOOR (OUT OF PLANE) C. SLAM DOOR i 4. WALL,GROUND FLOOR (OUT OF PLANE) D. SLAM DOOR ii 5. FOUNDATION (,VERTICAL) E. RUN DOWN STAIRS

F I G 6.12 L O C A T I O N PLAN - ' N O R M A L USE" EXPERIMENT

A:tion

Geophone

A B C D E

i i i i

1 2 3 4 5

i

11.8 3.3 7.2 0.6 - 2.3 68.1 3.8 1.4 0.5 1.6 24.7 8.7 4.7 - 4.5 18.3 16.4 6.0 - 3.2 10.2 ii. 6 0.8 -

PEAK VIBRATION LEVELS mmls

A

<

m

,.J I.U > uJ --I

U~

Z

100"

90-

80"

70 ̧

60'

50'

40'

0

18 HOUR Llo=71dB (A)

L1

" ~ / ~ / ~ . . ~ - - ~ . . . ~ ~ ' - ~ . . j . - . . _

~ ~V// f'/ ',,/

,, • , , o ,,

v

5 10 15 20

T I M E IN H O U R S

A

Z l

- I

m "0

..I U.I > U,I ,,,.,I tl.I U~ 0 z

100"

90~

80

70'

60"

50"

40"

0

L1

'~\ /~ / F J . . ~_ ~-"-.. ~' .... "v ~''~--_

^'~.J. ,r ./,y ~"~ "" "'. __'~~.

• J /" / Leo "~ - . . ,

!

5 10 15 2 0

T IME IN H O U R S

FIG 6 . 1 3 E X I S T I N G N O I S E L E V E L S AT S I T E P R I O R TO E X P E R I M E N T

97

q 95 I

I ROOM

94

94

9 4

9 3

103

96

4

FIG 6.14a NOISE LEVELS IN FRONT ROOMS

DUE TO VIBRATOR ALONE (dB)

"O

..J UJ

UJ .J

UJ

(/)

I,g a. a z 0 U~

100-

90-

8 0

70-

60-

50 1C) 2'0 3'0 40 5C) 6'0 70

FREQUENCY Hz

FIG 6.14b FREQUENCY ANALYSIS OF NOISE IN ROOM

8b

rn "o 30" ILl Q

U u.

U. 0 Z 0 m p-

uJ P h. 4

20

, i , ~ , , , .

32 64 125 250 5 0 1 0 0 0 2 0 0 0 4 0 0 0 8 0 0 0

FREQUENCY Hz

FIG 6.15 VARIATION OF FACADE ATTENUATION WITH FREQUENCY

H \

IJ 164 0 0 0 CYCLES

O'_4mm

8 8 8 0 0 0 "CYCLES

FIG 6.16 MOVEMENT OBSERVED BY MOIRE PHOTOGRAPHY

,• , 11

I /

[ ' ,

1

I I ' i l

IL,

. . . . . . . . J. i

I

,-k--'

! I

FIRST FLOOR CEIL ING

• t ! U, i/"~! , , ' II

L

t l f l

( ,,,/V I

I

GROUND FLOOR CEIL ING

FIG 6.17 VISIBLE CRACKS PRIOR TO S IMULATION EXPERIMENT

0.1

E I /

f/J

.... ._ 0-2:_9._2_ " " ~

I,

L o~ I

I i

.0 0'5

1'¢

0"5

' I I

FIRST FLOOR CEILING

. . . . NEW CRACKS /SIZE SHOWN IN mm) WHERE NO SIZE SHOWN,CRACK IS HAIRLINE

0'2 MONITORED MOVEMENTS (MOVEMENT SHOWN IN mm/

_J

r ~ L ~11111 I

I0'2

'0'2 ~ .~ 0 '2/,,

I 1 lr]-J

I J

d

0.1

I

GROUND FLOOR CEILING

FIG 6.18 NEW CRACKS AND CRACK MOVEMENTS OBSERVED DURING SIMULATION EXPERIMENT

0 2

8

tu 10 > 0 12 ZE

14 Z

16 0 0 1 8

20-

METRES FROM INSIDE OF CELLAR WALL

~ 0 2.0 3.0 4.0 s.o 6 . 0 ~ 23.12.85

6 1 8 6

r

ELECTROLEVEL STRING BENEATH FLANK WALL OF TEST HOUSE

15.1.86

1 .2 .86

2 8 . 3 . 8 6

STRING A

E E 6 I-- Z 4 LIJ :E 2 uJ >

:E 2

z

0

0

1'.0 2'.0 3"0

28 .3 .86 6 .1 .86

. . . . 23 .12 .85

4:0 5 .0

METRES FROM INSIDE OF CELLAR WALL

STRING B

FIG 6.19 GROUND MOVEMENTS MEASURED BY ELECTROLEVEL

WATER CONTENT %

E v

.J UJ > uJ .J

Q Z

O ¢

¢ 0 . J

110

I,- eL I, LI a

0

1.0

2.0

3.0

5 10 15 I I I

0

0 •

0

o • 7

TREND FOR SITE DATA BEFORE EXPERIMENT

O BENEATH VIBRATOR PAD CORE TESTS A BENEATH HOUSE FOUNDATION

• BENEATH VIBRATOR PAD ~SAND REPLACEMENT TEST • BENEATH HOUSE FOUNDATION

FIG.6.20 VARIATION OF NATURAL MOISTURE CONTENT WITH DEPTH AFTER EXPERIMENT

A

E

al

> - I

Z

O ~c O

o , .I u l

z k- 0 . u l

a

1.0

1.5

2 . 0 -

3.0-

0

0 I •

0 •

DRY DENSITY Mg/m3

1.7 1.9 I I

FOR SITE

O BENEATH VIBRATOR PAD { BENEATH HOUSE FOUNDATION ~ CORE TESTS

• BENEATH VIBRATOR PAD • BENEATH HOUSE ,FOUNDATION I SAND REPLACEMENT TESTS

FIG 6.21 VARIATION OF DRY DENSITY WITH DEPTH AFTER EXPERIMENT

l l l l

1 000

"/1

0 0 0 l m

PLAN

BRICKWORK~ / / /~ ' / /~

MASS CONCRETE

22s

Z

300 I_ 300 _[ • - r I

0

/ / / ~ / / / '

SECTION A- A

FIG 7.1 TEST FOUNDATION STRIPS

B1 B2

I i I ~ow~ I I IL_

I I I T I

B3

TEST

HOUSE

VIBRATOR

PAD

TEST STRIP LOADINGS

A1,B1 - 9"00 TONS

A2,B2 - 4 ' 50 TONS

A3,B3 - 2"25 TONS

' o w

A3

SCALE 1 : 5 0

FIG 7 . 2 A R R A N G E M E N T OF T E S T F O U N D A T I O N S T R I P S

E E

U,I rt

i - I

,.J a. :! <

Z 0 I

I-- <

m e r a ,

>

10.

.

80

ROW A ROW B

LOAD = 0"5 x HOUSE FOUNDATION

~. --\\ / / / t

i f , , / / ~ ' 7 /

I !

I \ " '~ I \

I ', 1 '.,-,,,,^

1; "J,-? i \ I

\

\ J

I !

FREOUENCY (Hz)

t'o

E t0- W

) . ,

rr m !

>

8 0

ROW A

- - - ROW B A

LOAD = 1.0xHOUSE FOUNDATION h ~ l /

i/ t....I •

~ ~ / V t.j ~ / ' ~ ' /

1 I

FREQUENCY (Hz)

1'0

In

E E 10 tU a

I-- ira, ,.J a. :J

< 5" Z 0 I- ,,¢ n"

I r a ,

>

ROW A A ROW B / /

LOAD = 2.0 x HOUSE FOUNDATION

. ~ i xJ

80 ' ' 1'0 FREQUENCY (Hz)

F I G 7 .3 RESPONSE OF T E S T F O U N D A T I O N STRIPS TO F R E Q U E N C Y S W E E P S

201 10 A1

E E I - z UJ

UJ ..J p.

UJ (n

.

4-

_

O e -

0.4-

J J J /

~ A2

2 4 8 15 25 4o B o No OF PULSES ( x l O 0 0 )

loo 250 400 600 88e

NOTE: SEE FIG.7.2 FOR TEST STRIP LOCATIONS

FIG 7.4 SETTLEMENT OF TEST FOUNDATION STRIPS

A

E 1.5 O

uJ > uJ ,.J

Z

O

0 1 , 0

0 , . J u,I m

Z I- o. 2 . O uJ

DRY DENSITY Mg/m 3

1,6 1,7 1,8 I I I

B3

1.5 • .J 0 uJ > u.I ,_1 Q Z = O m 0 1'O-

O , . I u,I m

"1- I-,

uJ 2,0-

DRY DENSITY Mg /m 3

1.6 1.7 1.8 i i i

t I

B 2

1.5 O

IJJ > uJ ..I

z

0 c- O 1.0

0 , . J lid II)

I " O. w 2 . 0 (2

DRY DENSITY M g / m 3

1.6 1.7 1.8 I i i

I I |

I I

B1

A

E 1.5 0

IJu > tlJ ,,I a Z

0 n- O 1"0.

0 . . I ul 11)

I,-

uJ 2 . 0 e~

DRY DENSITY Mg//m 3

1.6 1.7 1.5 , i O

1 . 8

u,I > uJ .,J O Z

O

1.0-

O

~ w

I -

" 2.0- ~ a

I

DRY DENSITY Mg /m 3

1.6 1.7 1.8 I i I

S

A3 A2

_1 >

- I Q Z = 0 n-

1'0-

0 . J . 1 m

n Lu 2 . 0 - Q

1.5 0

DRY DENSITY Mg/m 3 1.6 1.7 1.8

I I I

I I I

|

t I t

A1

. . . . . AVERAGE SOIL DENSITY AT SITE

BEFORE START OF EXPERIMENT

O CORE TESTS

• .SAND REPLACEMENT TESTS

FOR TESTS STRIP REFERENCES SEE FIG 7.2

FIG 7.5 I N S I T U SOIL DENSIT IES BENEATH

F O U N D A T I O N STRIPS AT END OF EXPERIMENT

Z 0 03 oO

U3 Z <C er" l--

kU

0 a.

(3 Z.,=,

0 U3 LL o I-- Z UJ 0 u. u.. ILl 0

0 ' 6

0"5

0.4"

0 . 3

0.2

0.1

0 0

\ \ \ \ \ \ \ \ \

\ \ \ ~ / - 0 ' 6 m

\ \

\

\ \ \ \ \ ~ , - 0 ' 2 5 m

\ \

X X

\

\

2 . 4 m ~ . ~ ~ ~ . . . . . . . . .

~ ~ " - - - - ~ - - ~ • ~ t - - - - , - - - - - -~ , ' ' - - - -

10 20 30 40 50 • 60 FREQUENCY Hz

70 8 0

FIG 10.1 T H E O R E T I C A L V A L U E S OF ~,t FOR V A R I O U S

T H I C K N E S S E S OF P O L Y S T Y R E N E

L3 m

VIBRATOR - - - ~

REFERENCE GEOPHONE ARRAY

5 m

lm

15m 18m

B c

F IG 10.2 T R E N C H E X P E R I M E N T M O N I T O R I N G P O S I T I O N S

3'0-

i 2.0"

1.0"

0 40 3'0

~ -~ 3.0-

~ 2.~

2 1 ~ 1.0-

N

2o Ib o

FREQUENCY Hz GRAPH 1

r I

I

t/ ~ .; o ~/

40 30 20 1'0 FREQUENCY Hz

GRAPH 4

~ 1'O

¢.

= S

• ~

~.- ~ 0.5

N

40 30 2() 10

FREQUENCY Hz GRAPH 2

40 3'0 2'0 1()

FREQUENCY HZ GRAPH 5

E 1.0

2

Z 0'5

==

40 30 20

FREQUENCY Hz GRAPH 3

FIG 10.3 VIBRATION LEVELS RESULTING

10

BACKFILLED TRENCH (GRAPH 1 -3)

OPEN TRENCH (GRAPH 4 & 5)

POLYSTYRENE FILLED TRENCH

MEASUREMENT POSITIONS (see fig.!O. 2 ) _

GRAPHS 1 & 4 - POSITION A GRAPHS 2 &5 - POSITION B GRAPH 3 - POSITION C

FROM DIFFERENT TRENCH ARRANGEMENTS