-

379NOTES

(ii) proposed buildings are cluse to slope sboulders, and providiug attractive accommodation iu mixed uni" of(iii) groundwater level changes occur to depths greater high-rise and low-rise buildings.than twice the depth of climatic influence. Table 3 usesthe damage risk classification from Table 2 to outline 2. Large chemicalfertilizer plantconstroction considemtions 10 regions of expaosive The modemfertili1<rplant at Hubeiprod- 300000 tsoils. of synthetic ammonia and 450000' of urea annually.

.' Strictlimits-5 mm, and in places less-were placed onSuccessful examples of l?dust~la! and ~rban absolute settlements. Site investigation and testing

structures on expansive sOilsm Chma showed the foundation soil to be Quaternary alluvial-

1. New city devdopment diluvialexp""ive snil with expansion pressure 3(}-80One-third of the buildings in Yunxian County, Hubei kPa, expansionrate 4-8%, and shrinkageratio 20-30%

Province has suffered damage from expansive soil, and Construction methods included deepeningth f

.

it was decided to relocate the city near the Danjiang ings, increasing the contactpre

lie e Oot

River Reservoir. Soil deformations were accentuated by sand cushions . SSuressoilr 1

.

a complex, uudulatiug topogropby. C' pIers,andlar .' epacement,Site iuvestigatinns examinedthe d' . arefulattentionwasp'd ge dlamelerbored piles

expansioo-shrinkage eharac\erif ISlnhulion, soilmois\nreconserv:t to controllingsurfacewatec'chemicalpropertiesofthesoils ASlCS, and physico- constructionmanage~'on,s;r"tructure stiffness,andcal and subsoilfeaturescouldb ~~res~lt, topographi- building after many ye;~t. f efo~atlons of the mainsiting new buildings usin g the ~t . en l~tOa~count in have met all operational r0 ~peratlon are 1-5 mm, and

2 and 3 S't

.. cn ena outlIned10Tables eqUlrements.. I e preparatIOnmclud d' .

and construction of cuts and ~l c~ttm~steps 10slopesMultiftoor buildings wer 1 sd or highway access.neered foundations in are;s ~;~~ hon car~fully e~gi-age potential while g expanslOn-shrink-for smaller ~oneng~~as oJ low potential were reservedtion was p~id to cons:e f apartments. Careful atten-piles, retaining waIls an~Ion on slopes. Piers, shortcontrol down-slope ~ovem:~:oncslopes ~ere used tobeen in progress for som s. ons~ctlon has nowe years and IS successfully

Hou, S., ANDWANG,S. 1980. Primary consideration ofprinciples for classifying areas of expansive-shrinkablesoils according to engineering geology. Engineering In-vestigation and Exploration (Internal Journal), No.5,Institute of Geodetic, Geotechnical and HydrogeologicalInvestigation, Hubei, China.

Hou, S. 1980. Types of expansive-shrinkable soils in Chinaand their engineering geological characteristics. Bulletin ofthe International Association of Engineering Geology, No.21, pp. 5-10.

Heave of coal shale fill

JACK A. CALDWELL AND ADRIAN SMITH

Steffen Robertson and Kirsten, 801-1030 W. Georgia Street, Vancouver, B.C., Canada V6E 3J7

AND

JOHAN WAGNER

Rand Mines Limited, Rand Mines Building, Simmonds Street. Johannesburg, South Africa

Received August 8, 1983

Accepted January 18, 1984

Washing plant discard, predominantly laminated and micaceous carbonaceous shale, was used as fill beneath a thick concretefloor slab. This paper discusses the cause of heave of the fill and the severe cracking of the floor slab-up to 70 mm-about a yearafter construction.

Observation of the fill and the nature of the fill indicate that the heave is caused by absorption of water (but not clay mineralswelling) and predominantly by alteration of the sulphides in the fill to sulphates.

These observations are confirmed by measurement of the chemistry of fresh washing plant discard, weathered fill, and freshdiscard subjected to accelerated weathering. Oedometer tests confirm the swelling of the rocks as a result of water absorption. Asthese processes were still occurring, further heave was anticipated; accordingly, the fill was removed.

Keywords: heave, shale, sulphide alteration, weathering, clay mineral swelling.

Can. Geotech. J. 21, 379-383 (1984)

-_..--

-~..

(H)proposed buildings are close to slope shoulders, and(Hi)groundwater level changes occur to depths greaterthan twice the depth of climatic influence. Table 3 usesthe damage risk classification from Table 2 to outlineconstruction considerations in regions of expansivesoils.

Successful examples of industrial and urbanstructures on expansive soils in China

1. New city developmentOne-third of the buildings in Yunxian County, Hubei

Province has suffered damage from expansive soil, andit was decided to relocate the city near the DanjiangRiver Reservoir. Soil deformations were accentuated bya complex, undulating topography.

Site investigations examined the distribution,expansion-shrinkage characteristics, and physico-chemical properties of the soils. As a result, topographi-cal and subsoil features could be taken into account insiting new buildings using the criteria outlined in Tables2 and 3. Site preparation included cutting steps in slopesand construction of cuts and fills for highway access.Multifloor buildings were placed on carefully engi-neered foundations in areas of high expansion-shrink-age potential, while areas of low potential were reservedfor smaller, nonengineered apartments. Careful atten-tion was paid to construction on slopes. Piers, shortpiles, retaining walls, and apron slopes were used tocontrol down-slope movements. Construction has nowbeen in progress for some years and is successfully

---

NOTES 379

providing attractive accommodation in mixed units ofhigh-rise and low-rise buildings.

2. Large chemicalfertilizer plantThe modernfertilizerplantat Hubeiproduces300 ()()()t

of synthetic ammonia and 450000 t of urea annually.Strict limits-5 mm, and in places less-were placed onabsolute settlements. Site investigation and testingshowed the foundation soil to be Quaternary alluvial-diluvial expansive soil with expansion pressure 30-80kPa, expansionrate 4-8%, and shrinkageratio 20-30%.

Construction methods included deepening the foot-ings, increasing the contact pressures, soil replacement,sand cushions, piers, and large diameter bored piles.Careful attention was paid to controlling surface water,soil moisture conservation, superstructure stiffness, andconstruction management. Deformations of the mainbuilding after many years of operation are 1-5 mm, andhave met all operational requirements.

Hou, S., ANDWANG, S. 1980. Primary consideration ofprinciples for classifying areas of expansive-shrinkablesoils according to engineering geology. Engineering In-vestigation and Exploration (Internal Journal), No.5,Institute of Geodetic, Geotechnical and HydrogeologicalInvestigation, Hubei, China.

Hou, S. 1980. Types of expansive-shrinkable soils in Chinaand their engineering geological characteristics. Bulletin ofthe International Association of Engineering Geology, No.21, pp. 5-10.

Heave of coal shale fill

JACK A. CALDWELL AND ADRIAN SMITH

Steffen Robertson and Kirsten, 801-1030 W. Georgia Street, Vancouver, B.C., Canada V6E 3J7

AND

JOHAN WAGNER

Rand Mines Limited, Rand Mines Building, Simmonds Street, Johannesburg, South Africa

Received August 8, 1983

Accepted January 18, 1984

Washing plant discard, predominantly laminated and micaceous carbonaceous shale, was used as fill beneath a thick concretefloor slab. This paper discusses the cause of heave of the fill and the severe cracking of the floor slab-up to 70 mm-about a yearafter construction.

Observation of the fill and the nature of the fill indicate that the heave is caused by absorption of water (but not clay mineralswelling) and predominantly by alteration of the sulphides in the fill to sulphates.

These observations are confirmed by measurement of the chemistry offresh washing plant discard, weathered fill, and freshdiscard subjected to accelerated weathering. Oedometer tests confirm the swelling of the rocks as a result of water absorption. Asthese processes were still occurring, further heave was anticipated; accordingly, the fill was removed.

Keywords: heave, shale, sulphide alteration, weathering, clay mineral swelling.

Can. Geotech. J. 21, 379-383 (1984)

380 CAN. GEOTECH. J. VOL. 21, 1984

Des residus d'usine de lavage, comportant essentiellement des schistes carbonates et micaces lamines, ont ete utilises commeremblai sous une epaisse dalle de beton. Cet article discute des causes du gonflement du remblai et de la fissuration importante dela dalle (jusqu' it 70 men) environ un an apres la construction.

L'examen du remblai et la nature du remblai indiquent que Ie gonflement est cause par I'absorption d'eau (mais pas par Iegonflement de mineraux argileux) et, de fa<ronpredominante, par I'alteration des sulfures en sulfates dans Ie remblai.

Ces observations sont confirmees par les mesures de la composition chimique des residus frais de I'usine, du remblai altere etde residus frais soumis it une alteration acceleree. Des essais oedometriques confirment que Ie gonflement du remblai resulte deI'absorption d'eau. Comme ces processus etaient toujours en cours, une continuation du gonflement etait attendue; enconsequence Ie remblai a ete enleve.

Mots-eMs: soulevement, schiste, alteration des sulfures, alteration, gonflement des mineraux argileux.[Traduit par la revue]

Introduction

The fill used at a new kitchen and dining roomcomplex in Utrecht, Natal, South Africa, was thediscard from the mine's coal washing plant. Approxi-matelya year after being placed, the fillhad heaved up to70 rom, and in so doing, had caused extensive crackingof a thick floor slab and outward movement of walls.

This paper describes the nature of the fill and thepossible causes of heaving of the fill. Heave is shown tobe caused by (l) absorption of water (but not claymineral swelling) and (2) alteration of sulphides tosulphates in the fill.

Description of the problemThe Utrecht operations mine coal from the Vryheid

sequence of the Karoo Series. Underground mining ofthe coal seam inevitably involves some removal of theoverlying and underlying shales and mudstones. Mate-rial removed fromthe mine is crushed to less than 75 mmand passed through a washing plant. The light coal isfloatedout and the heavier coals and rocks discarded onsurface waste dumps.

Washing plant discard was used as fill beneath thefloor slab of the new complex. Between 400 and1500romof washing plant discard were compacted intoplace to bring the floor to grade.

Figure 1 shows the pattern of floorheave observed inthe dining room approximately 2 years after placing thefill. The greatest rate of change of the heave profileappears to be within approximately 2 m of the outerwall. The concrete is continuously but irregularlycracked around the dining room approximately 2 m fromthe wall.

The north end wall of the dining room has movedapproximately 25-35 mm outwards at the top. Approxi-mately 10-15 rom of movement of the brick wallbeneath the damp-proof course is observed.

The fill in the kitchen is approximately 400 mm thick.Floor profiles were not measured in the kitchen, butextensive cracking of the floor slab has occurred.Single-brick division walls approximately 1.5 m highhave cracked and opened at the top approximately15-25 rom.

Flu.. DEPTH ABOuT l400mm

EE

~

20

7 _- ~Crock ;.::..\\

o....oo.,.o:::on...i~x......i!:

Crack/'

FilL DEPTH ABOuT 400m

FIG. 1. Contours of dining room floor heave. (Heave con-tours in men.)

Description of the fiUThe fill, as observed in test holes in the dining room,

is slightly moist, very dark gray, mottled yellow, andash gray, generally medium dense to dense, mediumcoarse (up to 250 rom), and composed of angular to platypieces of laminated and micaceous carbonaceous shalemixed with occasional sandstone and low-grade coal.By comparison, the fresh material from the washingplant is black and gray and has no yellow mottle.

The yellow mottle is observed not only on thesurfaces of the individual particles but also on thebedding planes of the shale, which may be exposed bybreaking. The flatstone and shaleparticles in the filltend

- --

100

90

80

70.

60.

.0.

40

:JO

201

IOi

0.0.01 0.1 It, 10 100

P"'''TlCLE SIZE I"".'

FIG. 2. Particle size analysis.

to lie in the same orientation, and the general impressionis that the fill was adequately compacted.

The temperature of the fill when examined at the endof summer was 26°C. The room temperature was 22°Cand the outside temperature in the shade was 24°C. Thefill beneath the kitchen floor is saturated, mainly as aresult of water leaking from broken pipes below the floorslab.

Figure 2 is the grading curve of the fill. Table I givesthe moisture content of various fractions of a samplefrom beneath the dining room floor.

The table shows that the average moisture content ofthe fill is approximately 4-5%. The moisture content ofthe fines(arbitrarily defined as the 10%and less fraction)is 7-8%, while the individual, larger particles have amoisture content of 3-4%.

The major minerals of the fill are organic material,quartz, feldspar, and mica. Minor constituents arepyrite, gypsum, water soluble salts, hydrated ironoxides, and possibly sulphur. X-ray diffraction indicatesthe clay minerals illite and kaolinite only; the clays arenonswelling. Pyrite is randomly distributed through thesamples; some are devoid of pyrite, while others containconcentrations of up to 10%. Gypsum rosettes and othersecondaryproducts are observed on the weakly cohesivebedding planes.

TABLE I. Dining room fill moisture content

Moisture content

SampleSpecific

Average density

Total sampleFines only (i.e., particles smaller than

4.75 mm)Particles larger than 4.75 mm onlySelected coarse particles (uncrushed)Selected coarse particles (crushed)

4.47 0.45

7.55 0.754.57 0.403.88 0.393.69 0.38

NOTES 381

1000

The fresh material from the washing plant does nothave gypsum rosettes or other oxidation products alongthe bedding planes. Fresh particles of material from thewashing plant readily adsorbed water and tended tobreak up and disintegrate.

Possible causes of heave of the fillPossible causes of heave of the fill or structural

distress of the floor considered but discounted are thefollowing:

(1) Formation of gases in the fill: the open nature ofthe fill and the opportunities for escape of gas makes thismost unlikely.

(2) Heating of the fill and concrete slab with subse-quent expansion: the temperature increase of the fill isinsufficient to account for the volume change observed.

(3) Swelling of a clay fraction: as noted above thereare no swelling minerals in the fill.

(4) Stress relief of the particles: stress relief is likelyto occur soon after mining and thus is not a likelyexplanation.

(5) Settlement ofthe foundations: the foundations areon competent sandstone; also, the horizontal movementof the brickwork discounts foundation settlement.

There are, however, two apparently viable causes ofheave which might be considered: (I) absorption ofwater by the shales with a subsequent increase in volumeand (2) alteration of the sulphides in the fresh washingplant discard to sulphates and secondary oxidationproducts.

As noted above, the fill beneath the kitchen floor andthe south end of the dining room is only approximately400 mm thick. The fill in these areas is saturated. Bothwater absorption and sulphide alteration may accountfor the increase of volume of the fill. Conversely, the fillbeneath most of the dining room floor is only partiallysaturated. The predominant cause of heave in this area isbelieved to be sulphide alteration.

The following sections discuss the nature of the twophenomena and the testing done to confirm that heave isdue to water absorption and sulphide alteration.

Sulphide alterationThe pH, sulphide, and sulphate content of both the

discard from the washing plant and the fill from beneaththe floor were measured. In addition, these parameterswere measured on samples of the washing plant discardafter different degrees of simulated accelerated weather-ing in the laboratory, using methods described by Smithetal. (1982). This essentially involves repeated wettingand dryingat elevatedtemperatures.Figure3 summarizesthe results.

Figure 3a indicates that the fresh washing plantdiscard has a high pH and a low sulphate content. Thosesamples subjected to accelerated weathering display a

//

/HOLE H5 @ O.7m I

II

I/

/.-'.- -

382

0./

CAN. GEOTECH. J. VOL. 21, 1984

OCll"

o

pH

0./00

I

i 0_050

002'

o

SUI."PHlDE . ~ 5

FIG. 3. Accelerated weathering test results.

reduced pH and a slight increase of sulphate content.The samples from the fill have a considerably lower pHand a considerably greater sulphate content than thefreshmaterial. Figure 3b shows that the sulphidecontentof the fresh material is greater than that of the weatheredmaterial. The sulphate content of the fresh material is,however, lower than that of the weathered fill.

The observation of gypsum rosettes, calcium sul-phate, in oxidised fill implies a source of calcium.Although no calcium was identified as a mineral phaseof the shale, this does not mean that there is no availablecalcium in the shale. Analysis of such shales normallyshows a significant proportion of calcium that can bemobilized in the saturated fill.

Thus, it is apparent that, with weathering, thesulphides alter to sulphates. As described by Coveneyand Parizek (1977), this leads to an increase in thevolume of shales. Depending on the sulphide oxidationprocess perceived and on simplifying assumptions onthe nature of the oxidation product, a volume increase inexcess of 60% can be postulated. In the case describedby Coveney and Parizek (1977), a floor cast on a shale,with a pyrite content up to 3%, heaved between 100and200 mm.

Heave of black shale has also been reported in Canadaby Quigley and Vogan (1970). In the case history

described by those authors 76 mm of differential heavecaused severe structural deformation of a lightly loadedbuilding founded directly on drained black bedrock inOttawa. Heave was attributed to oxidation of dissemin-ated iron sulphide in the shale by autotrophic bacteria toproduce secondaryhydrous sulphatesof greater volume.

In a later paper Quigley et ai. (1973)describe in detailobservations and interpretations of the alteration andheaveof black shale. Their observations of the alterationproducts of sulphide oxidation in black shale are basedon photomagnifications (up to 7000 times), X-raydiffraction analysis, and X-ray fluorescence. Quigley etai. (1973) show that the growth of gypsum crystalsimbedded in the shale appears to be responsible for theheave of the two buildings in Ottawa founded on theshale. Jarosite and natrojarosite alteration products,probably produced by bacterial action, filled opencavities and cracks. The authors concluded that whilethe jarosite is probably critical to the overall alterationprocess it is probably of secondary importance ingenerating heave pressures.

Gillott et ai. (1974) describe a microstructural analy-sis of altered and unaltered black shale from a site wherefoundation heave occurred. They examined a range ofsamples from relatively fresh to severely rotted speci-mens that were coated with secondary alteration pro-ducts. In unaltered specimens examined by them theprincipal minerals are illite, chlorite, quartz, calcite, andpyrite. In altered specimens, gypsum andjarosite are themost abundant new minerals. Pyritic sulphur is presentin both altered and unaltered shale. Sulphate sulphur ispresent in the altered but not the unaltered shale andthere is a marked decrease in carbonate content betweenthe unaltered and the altered shale. Gillot et ai. (1974)present a detaileddiscussionbased on their observations,as summarized above, of whether it is the gypsum or thejarosite that is responsible for the heave of the blackshales, and on thebasis of their discussion conclude that:heave of black shales results from the formation ofsecondary minerals of increased volume formed inassociation with the bacterial oxidation of pyrite and theproduction of sulphuric acid; the main reaction productsare gypsum and jarosite; and it is not certain which ofthese minerals generates heave pressure.

Grattan-Bellew and Eden (1975) describe an investi-gation of the heave of the basement floor of a church inthe New Edinburgh area of Ottawa. They discoveredthat the concrete under the floor had been reduced to amushy consistency due to attack by sulphate solutionformed by the oxidation of pyrite in the underlyingshale. Gypsum was the main product in the weatheringzone of the shale, although small amounts of jarositewere observed. They showed that heave resulted fromthe growth of gypsum crystals between the lamellae inthe shale.

---

TIME IN MIHIITU

FIG. 4. Results of swell tests on carbonaceous shale.

Water absorptionThe finer fraction of the fill was placed in an

oedometer and saturated. The change in volume withtime is shown in Fig. 4. This confirms that wateradsorption can account for a volume change of the fill ofthe order of magnitude observed.

The phenomenon of swelling of rocks by waterabsorption can account for a volume change of the fill of(1968). They performed an extensive series of tests onvarious rock types and showed that the greatest swellingdue to water absorption occurred for clays, marls,shales, and mudstones. None of the materials tested bythem contained clay minerals with an expanding claylattice structure.

Conclusions

On the basis of limited physical, mineralogical, andchemical testing, the cause of heave of a floor slab in a

NOTES 383

new dining room and kitchen complex is shown to be(I) absorption of water and (2) alteration by weatheringand oxidation of sulphides in the fill to sulphates.

As the majority of the fill was still only partiallysaturated and as there is still abundant unweatheredpyrite in the fill, continued heave is possible. Accord-ingly, the fill was removed and replaced with an inertmaterial.

000

COVENEY, R. M., andPARIZEK,E. J. 1977.Defonnationofmine floors by sulphide alteration. Bulletin of the Associa-tion of Engineering Geologists, 14(3), pp. 131-156.

DUNCAN,N., DANNE,M., and PETTY,S. 1968. Swellingcharacteristics of rocks. Water Power, May, 1968, pp.185-192.

GILLOTT,J. E., PENNER,E., and EDEN, W. J. 1974.Microstructure of Billings shale and biochemical alterationproducts, Ottawa, Canada. Canadian GeotechnicalJournal,11(4), pp. 482-489.

GRATTAN-BELLEW,P. E., and EDEN,W. J. 1975. Concretedeterioration and floor heave due to biogeochemicalweathering of underlying shale. Canadian GeotechnicalJournal, 12(3), pp. 372-378.

QUIGLEY,R. M., and VOGAN,R. W. 1970. Black shaleheaving at Ottawa, Canada. Canadian Geotechnical Jour-nal, 7, pp. 106-112.

QUIGLEY,R. M., ZAJIC,J. E., McKYES,E., and YONG,R. N.1973. Biochemical alteration and heave of black shale;detailed observations and interpretations. Canadian Journalof Earth Sciences, 10,1005-1015.

SMITH,A., DEHRMANN,A. A., and HARPLEY,D. P. 1982.Laboratory procedures manual. Steffen Robertson andKirsten internal report, Johannesburg, South Africa.

/./ -/0 - -,. V

/'8

/"'.6'.' .-1--

>--- ..-'.4

/ !16 - >--- - l- .-'I

1--1--'/

,"769D ZO "" 100 ZOO .,., """ ""'" .

384

DISCUSSIONS

Some experimental studies concerning the contact stresses beneath interfering rigid stripfoundations resting on a granular stratum:! Discussion

J. GRAHAM

Department of Civil Engineering, University of Manitoba, Winnipeg, Man., Canada R3T 2N2AND

G.P.RAYMOND

Department of Civil Engineering, Queen's University, Kingston, Ont., Canada K7L 3N6Received December 12, 1983

Accepted January 13, 1984

Can. Geotech. J. 21, 384-385 (1984)

The paper is a welcome addition to the number ofexperimental projects in sand reported from Canada inrecent years. The authors are to be complimented on thefundamental nature of the question they have chosen toexamine, on the elegance of the design of their testingprogram, and on the quality of their results for twointerfering footings.

Consideration of a closely similar interference prob-lem between three parallel footings (Graham et al.1984) permits comments to be made on several pointsraised by the authors. In these three-footing tests, thestress fields associatedwith the outer footingscaused thestresses under the central footing to be symmetrical, asthey would be, for example, in a line of parallelinterfering footings. This contrasts with the asymmetricstresses under the authors' footings.

Separate series of tests in the writers' laboratorieshave examined how loads are shared between neigh-bouring footings through superstructures of differentstiffnesses. Prefailure and failure conditions have bothbeen studied. In one series (Fig. la), the footings wereforced downwards together by a rigid loading system,and the distribution of loads between them was mea-sured by separate load cells. The ratio Aof the load onthe outer footing to the load on thecentral footing did notvary greatly between the beginning and end of the tests.In the second series (Fig. Ib), the central footing wasindependent of the outer footings, and different valuesof A were held constant during the tests. The resultsprovide general support to the authors' view thatinterference effects are largely restricted to centre linespacing SIB < 4. However, some evidence of interac-tion with the test container is seen in both of our testseries, especially when the sand has a higher coefficientof uniformity than in the authors' tests. It is perhaps

lpaper by A. P. S. Selvadurai and S. A. A. Rabbaa. 1983.Canadian Geotechnical Journal, 20, pp. 406-415.

BhAFFECTEOBYTEST BOX

1.0 ~NOINTERACTION .~._._.-BETWEEN FOOTINGS - _-/.........--PREFAILUREt ./ j """ /'

/ r>../ /

//J/~ClOSE TO FAilURE

1//

\,1

-<Q!<i 0.9a::o<3 O.B...J

0.1

(0) CONTROLLED DISPLACEMENTS

0.6I 3 4

SPACING 51 B

6

A%100 ROUGH .100 SMOOTH0

1. .. +.0 .. t!.

II

°

SPACING SIB

FIG.I. Load sharingand efficienciesbetweensymmetri-callyloadedinterferingfootings.

more common to expect interference to cease at centreline spacings of six to eight footing widths.

The authors tested "smooth" footings (8 < 7°), partlyto correspond with assumptions of the physical model-ling in their testing equipment. At first sight this mayrepresent a limitation on the usefulness of their resultssince most real footings are rough. However, thedifference in bearing capacity between smooth andrough footings is small (Vesic 1973). Recent tests(Graham et al. 1984) on 38 mm wide x 305 mm longisolated footings in crushed quartz sand suggest that thedifferences may be only 5-10%.

The paper directs attention to the important questionof the contact stress distribution beneath footings. The

4r °;:>-uzwU;;:u.w

FIG. 2. Stress characteristics for interfering surface footingwith SIB = 2.02, linear 8-solution, <I>= 35°.

authors show stresses that differ markedly from theunifonn distribution commonly assumed for design, orthe triangular distributions that result from theoreticalsolutions using complete contact friction mobilization(8 = <1»across the full width of the footing base.Friction mobilization is a strain-controlled responseand not a fundamental soil property (Graham andStuart 1971).

An advantage of the stress-characteristic method forcalculating bearing capacities is that contact stressdistributions are developed directly from the analysisonce reasonable assumptions have been made regardingthe 8-distribution. Figure 2 shows the stress distributioncalculated for two symmetrically loaded interferingfootings at a spacing of about two footing widths. The

DISCUSSIONS 385

calculations assumed that no shear stresses would occurat the centres of the footings and that 8 varied linearlyfrom <I>at the edge to zero at the centre. Thisassumption was first suggested intuitively by Grahamand Stuart (1971), and has since been supported by theexperimental results of Bauer et at. (1979). The shapesof the theoretical and experimental vertical contactpressure distributions are approximately parabolic, andare similar to those suggested by the authors.

Finally, it is perhaps worth noting that the increasedflexural moments identified by the authors are onlyexperienced when the interference is asymmetric.

BAUER,G. E., SHIELDS,D. H., SCOTT,J. D., and NWA-BUOKEI,S. O. 1979. Normal and shear stress measurementson a strip footing. Canadian Geotechnical Journal, 16, pp.177-189.

GRAHAM,J., and STUART,J. G. 1971. Scale and boundaryeffects in foundation analysis. ASCE Journal of the SoilMechanics and Foundations Division, 97(SMII), pp.1533-1548.

GRAHAM,J. RAYMOND,G. P., and SUPPIAH,A. 1984.Bearing capacity of three closely-spaced footings on sand.Geotechnique, 34, in press.

VESIC,A. S. 1973. Analysis of ultimate loads of shallowfoundations. ASCE Journal of the Soil Mechanics andFoundations Division, 99(SM1), pp. 45-73.

Some experimental studies concerning the contact stresses beneath interfering rigid stripfoundations resting on a granular stratum:} Reply

A. P. S. SELVADURAI

Department of Civil Engineering, Carleton University, Ottawa, Ont., Canada K1S 5B6AND

S. A. A. RABBAA

Department of Civil Engineering, AI-Azhar University, Madinet Nasr, Cairo, EgyptReceived January 11, 1984Accepted January 13, 1984

Can. Geotech. J. 21, 385-386 (1984)

The authors would like to thank Professors J. Grahamand G. P. Raymond for their interest in the paper and forthe stimulating discussion. The problems cited in thediscussion also clearly illustrate the necessity for con-sidering interference effects that can exist betweenclosely spaced foundations. In the three-footing prob-lem, where the plane of symmetry exists at the centre

IDiscussion by J. Graham and G. P. Raymond. 1984.Canadian Geotechnical Journal, 21, this issue.

line of the central footing, the contact stresses at thecentral footing should, of course, exhibit a symmetricalpattern. The outer two footings, however, will exhibitan asymmetric contact stress distribution. Even in thecentral footing the contact stress distribution will be farremoved from the unifonn pattern assumed in conven-tional structural design calculations. The main purposeof the experimental investigation was to ascertain thenature of the contact stress distributions that can exist atthe base of shallow footings resting on granular media.

386 CAN. GEOTECH. J. VOL. 21, 1984

The experimental procedure used determines, relativelyaccurately, such contact stress distributions (see, e.g.,Kempthome 1978; Selvadurai and Kempthome 1980;Rabbaa 1981). The nonuniform contact stress distribu-tions, whether symmetrical or asymmetrical, are typicalof contact stress distributions that can occur at the baseof shallowfoundations resting on loose to medium densegranular soils. It must also be noted that any theoreticalassessment of this class of nonlinear soil-foundationinteraction problem should be capable of accuratelypredicting both the load displacement response of thefootings and the contact stress generated at the interfacefor a complete range of footing settlements. Currentnumerical calculations indicate that while certain nu-merical stress analysis codes duplicate the load displace-

ment response accurately they are deficient in duplicat-ing the particular type of nonuniform contact stressdistribution observed by both the authors and thediscussers.

KEMPTHORNE,R. H. 1978. Contact stress distributions be-neath a rigid strip footing. M.Eng. thesis, Carleton Univer-sity, Ottawa, Ontario, Canada.

SELVADURAI,A. P. S., and KEMPTHORNE,R. H. 1980. Planestrain - contact stress distribution beneath a rigid footingresting on a soft cohesive soil. Canadian GeotechnicalJournal, 17, pp. 114-122.

RABBAA,S. A. A. 1981. Interference between rigid stripfoundations resting on cohesionless soils. Ph.D. thesis,Carleton University, Ottawa, Ontario, Canada.

--