You are on page 1of 57

Proc. Instn Ciu.

Engrs,

Part 1, 1985,78, Dec., 1325-1381 GROUND ENGINEERING GROUP

8917

Settlement of foundations on sand and gravel


J. B. BURLAND, PhD, DSc(Eng), FEng, FICE, MIStructE
M. C . BURBIDGE, BSc,MSc,DIC, FGSt
The Paper describes the analysis over 200 records of settlement of foundations, tanks and of embankments on sands and gravels. A remarkably simple picture has emerged relating the settlement to the bearing pressure, the breadth of loaded area and the average SPT blow count or cone resistance over the depth of influence. The influence of a number of factors such as shape and depth of foundation, depth of water table, grain size and time have been investigated. The Paper first briefly describes the application the results to the prediction of of settlement with particular emphasis on the limits of accuracy. Paragraphs 6 2 4 are self contained and may be used on their own for design purposes. The Paper follows this with a detailed account of the analysis of the case records.

Notation radius of loaded area foundation subgrade compressibility (ApJAq),mm/(kN/m2) most probable value of a, width of loaded area, m depth of founding level effective Youngs modulus correction factor for thickness sand layer of correction factor for shape correction factor for time thickness of sand layer depth of water table below founding level index of compressibility (ar/BO) rate of increase of Youngs modulus with depth length of loaded area volume compressibility from oedometer test average SPT blow count over the depth of influence corrected valueof SPT blow count average bearing pressure, kN/m2 cone resistance, MN/m2 time-dependent settlement (expressed as a proportion of pi) occurring during first 3 years after construction pi) time-dependent settlement (expressed as a proportion of occurring each log cycle of time after3 years defined in equation (14) most probable value ofT
Ordinary meeting, 5.30 pm., 25 February 1986. Written discussion closes 14 March 1986. For further details see (ii). p. *Imperial College Science and Technology. of ?E. J. Wilson, Consulting Engineering Geologist.
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

l325

BURLANDANDBURBIDGE
t
ZI
V

Pr
Pi PI
U

time in years depth of influence of loaded area effective Poissons ratio final measurement of settlement settlement at the end of construction or completion loading of settlement at time t after completion of loading
standard deviation

maximum previous effective overburden pressure, kN/mZ

Introduction
Numerous methods of predicting settlement of foundations on sands and gravels have been published-many more methods than for clays. The reason lies in the extreme difliculty of obtaining undisturbed samples for the laboratory determination of compressibility under appropriate conditionsof stress and stress history. Hence resort has been made to the interpretation of field in situ tests such as the standard penetration test (SPT), cone penetration test and plate loading test, and much of the literature has been devoted to such interpretations. This extensive literature will not be reviewed here as it has been adequately covered by Sutherland, Simons and Menzies and N i ~ o n . ~ 2. The practical importance of the problem was perhaps put in perspective by Terzaghi4 when he stated that all buildings resting on sand which were known to him had settled less than 75 mm (3 in) whereas the settlement of buildings on clay foundations quite oftenexceeded 500mm (20 in). This statement providedthe impetus for the study described in the present Paper in which a large number of case records of settlement on sands and gravels been have assembled by Burbidge. 3. The essential details of most of these case records are tabulated in Appendix 1 of the present Paper and the associated references are given in Appendix 2. The case record numberingused by Burbidge has been retained ease of reference. for 4 The prime objective of the study was to check whether the above statement . of Terzaghis still held true and reference to Appendix 1 shows that, with a few exceptions, it does for buildings. However, settlements well in excess of 75mm have been recorded for tanks and embankments on very loose sands. In view of the small settlements usually experienced with sands and gravels the second objective of the study was to analyse the data on actual observations settlement employof ing a minimum of interpretation to see if a simple and useful picture emerged. A preliminary study of this type was undertaken by Burland et aL6 and a similar but more detailed approach is described here. 5. The picture that has emerged from the statistical analysis of over 200 cases is remarkably simple and gives a range of settlements which is generally less than the range of predictions offered by the current commonly accepted methods. A brief description of the method and its application is given first, followed by a detailed account of the analysis of the settlement records.

Settlement prediction
6. Theoutcome of theanalysis of thelarge number of settlementrecords summarized in Appendix 1 is presented first, inthe form of a simple direct method of settlement prediction. Paragraphs 6 2 4 are self contained and can be used on their own for design purposes. However, frequent cross-referencesare made to the work described later,so that thebasis of the various assumptions canbe studied.
1326 Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONSONSANDANDGRAVEL

Determination o the foundation subgrade compressibility f 7. The nub of the method is the empirical relationship which has been established between the slope the pressure-settlement relationship for thefoundation of (ApJAq'),the breadth of the foundation B and the average SPT blow count over the depth of influence of the foundation. The quantity ApJAq' is the foundation subgrade compressibility, denoted by a , , andthe units are mm/(kN/m*). The relationship is shown in Fig. 1, where u , / B " ~is plotted against N on double log axes. The quantity u , / B " ~is denoted as I,, the compressibility index. The full line in Fig. 1 has been derived from a regression analysis of over 200 settlement records on sand and gravel. The chain dotted lines approximate to two standard deviations above and below the mean line. Mathematically the regression line is given by

with a coefficient of correlation of 0.848. 8. The following features should be noted about Fig. 1.
(a) a, is the subgrade compressibility for a normally consolidated sand or gravel. In 8 6 4 7 2 it is shown that the relationship between bearing pressure and settlement is approximately linear for normally consolidated granular materials for factors of safety against bearing capacity failure of 3 or more. When the material is overconsolidated or loaded at the base of an excavation, the values of a, and I , are reduced by a factor of 3 for pressure changes below the effective preconsolidation pressure
4 0 '

(b) The SPTblow count is not corrected for effective overburden pressure and the horizontalaxis is therefore not strictly a measure of relative density. Instead a new classification is proposed in which ranges of uncorrected N values are assigned to compressibility grades. The correlation between N and compressibility grade is given in Table 1 and in Fig. 1. The concept of compressibility grades proved particularly valuable in the analysis described in 26 et seq. Q (c) Although the N values are not corrected for overburden pressure it is necessary to make certain other corrections. It is shown in Section 8.5 that for very fine and silty sand below the water table the correction proposed by Terzaghi and Peck' gives improved results, i.e. when N is greater than 15 N' = 15 + 0.5(N - 15)
(2) where N' is the corrected value of N . When the material consists of gravel or sandy gravel it is shown in $9 103-106 that a correction should be applied such that

N'

1.25 X N

(3)

( d ) The results of cone penetration tests may be converted to equivalent N values using Fig. where q,/N is related to grain size (qcis in MN/m2). 2, (e) The results of plate loading tests may be related to compressibility grade using themethods described in 47-50. Care must be in used the application of equation ( l ) in conjunction with plate loading test results
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1327

BURLANDANDBURBIDGE

a,

AP

4mm/(kN/m2)

B in metres

Compresslbllity grades

0.1

I I I I I

I I I I I

10 SPT

100

Fig. 1. Relationship between compressibility ( I , ) and mean SPT blow count over depth of influence. Chain dotted linesshow upper and lower limits (see Figs 22 and 2 3 )

(m)

Table 1. Classijcation o compressibility of normally f tonsolidated sandsand gravels withS P T blow count
Compressibility grade No. of blows N*
<4

Interval
3 5
7 10

IV 111 I1

r
1328

4-8 9-15 1C 2 5 2640 41-60 > 60


~

15 20

* Uncorrected for overburden pressure.


Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONSONSANDANDGRAVEL
200

/-

10:
-

/-

/I

oh2
Coarse

0.06

0.2 0.6 Partlcle sue: mm Fine

20
Coarse

6.0 Fme IMediurn Gravel

Medium]
Silt

Medlum

Sand

Fig. 2. Relationship between q J N and grain size. Values o N are not correctedfor f overburden pressure since as B increases the value of N will often increase as well due to the associated increase in the depth influence. of
Depth o influence and the derivationf I f o ? 9. An important feature of the method is the assessment of the depth of influence z, of the foundation. This is discussed in detail in & 51-63 where it is shown that, when N increases with depth, the relative depth of influence (z/B), decreases significantly as the breadth of the foundation increases. Although the depth of influence depends on manyfactors, for present purposes it is assumed to be given by the full line in Fig. 3 for cases where N increases or is constant with depth. Where N shows a consistent decrease with depth the depthof influence is taken as 2B or the bottomof the soft layer, whichever is the lesser. The value of N for use in Fig. 1 or equation (1) is given by the arithmetic mean of the measured N values over the depthof influence. Calculation of settlement 10. For a normally consolidated sand the immediate average settlement pi at the endof construction, correspondingto the average effective foundation pressure q, is given by
pi = q
X

I,

(4)

where pi is in mm, q in kN/m2 andB in metres. Values of I , corresponding to the best estimate and the upper and lower limits are obtained from Fig. 1. 11. For an over consolidated sand, or for loading at the base of an excavation, for which the maximum previous effective overburden pressureis U , the average : , end of construction settlement pi corresponding to the average gross effective pressure q (where q > aka) is made upof two components as follows
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29
1329

BURLANDANDBURBIDGE

O f

B: m

Fig. 3. Relationship between breadth o loaded area B and depth o influence z, f f (within which75% of the settlement takes place)

= (q - &,)B07

l , mm

(54

When q is less than otothe aboveexpression becomes


pi = q
X

- mm 1, 3

(5b)

Corrections for depth offounding, depth o water table, shape thicknessof layer f and 12. In # 91-106 a statistical analysis of the influence of the above factors is described for foundations with depth ratios D / B < 3. It is shown that, within the limits of accuracy of the analysis, there is no obvious correlation between D/B and settlement. This result agrees with the results of DAppolonia et al. who found from the analysis of a number of results on one site that only a 12% reduction in settlement occurred when D / B increased from 0.5 to 1.0. 13. It is also concluded that the level of the water table beneath the founding level does not have a statistically significant influenceon thesettlement. This result appears to support Meyerhofsg that theeffect of the water tableis reflected in view the measured blow count. Thus water table changes subsequent to the determi? nation of I may have some influence on settlement. 14. Thestatisticalanalysisindicatesthatthere is a significant correlation
1330
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONS O N SANDANDGRAVEL

between settlement and LIB (the length-to-breadth ratio of the foundation). The correction factor is quite small and be expressed as can

where p i ( L / B > 1) =f,X pi(L/B = 1). It can be seen thatf, tends to 1.56 as LIB tends to infinity. 15. There were insufficient data to study the influence of the thickness of the sand or gravel layer beneath the foundation (H,) but it is recommended that when H , is less than z, (the depthof influence) a correctionf, should be applied suchthat

ZI

Time-dependent settlement 16. The case records referred to in $9 107-115 indicate quite clearly that foundationsonsandsandgravelsexhibittime-dependentsettlement.However,no distinct pattern emerges. In some cases the time-dependent process appears to be ,more orless continuous, with the settlement followingan approximately linear log time relationship (after an initial transition period). In other cases the process appears to be stepwise with quiescent periods of up to 3 years interspersed with periods of significant ratesof settlement. 17. The records show very clearly that foundations subject to fluctuating loads suchastallchimneys,bridges, silos and turbinesexhibitmuchlargertimedependent settlements than those subject only staticloads. to 18. The results suggest that the time correction factor for the settlement (p,) at of any time t , when t is 3 years or moreafter the end construction, is given by

Pi

where f, is the correction factor for time, t > 3 years, R , is the time-dependent settlement (expressed as a proportion of pi) that takesplace during thefirst 3 years after construction and R, is the time-dependent settlement (expressedas a proportion of pi) that takesplace each log cycle of time after3 years. 19. For static loads conservative values of R , and R, are 0.3 and 0.2 respectively. Thus at t = 30 years,f, = 1.5. For fluctuating loads conservative values of R , and R, are 0.7 and 0.8 respectively so that att = 30 years,f, = 2.5. Summary 20. In summary the average settlement of a foundation at the end of construction and then at anytime t, 3 or more years after the end of construction, may be expressed by the following equations:
pi =f,x f i
X

[(q - fo:,)

I,] mm

(94

and

where q is the average gross

effective applied pressure

(kN/mZ), c, isthe r :
1331

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

B U R L A N DA N DB U R B I D G E

maximum previous effective overburden pressure (kN/m2), B is the breadth in metres, I, is the compressibility index obtained from Fig. 1 or equation (l),f, is a shape correction factor given by equation (6), f , is a correction factor for the thickness of the sand layer given by equation (7) andf; is a time factor given by equation (8). 21. The probable limits of accuracy of equation (9a) can be assessed from the upper and lower limits of I, given in Fig. 1 and it may be necessary to take these into accountin the design. 22. It must be emphasized that the factor of safety against bearing capacity failure should always be checked in addition to the settlement. If the factor of safety is less than about 3 the pressure settlement curve maybe non-linear and the method will underestimate thesettlement. 23. Furthermore, the method has been based on case studies with quartzitic sandand graveldeposits.Siteswherecoral(calcite)orothermineralogically unusual sand and gravel deposits are encountered should not be analysed by this method unless the deformation properties of these deposits can be demonstrated to be similar to quartzitic deposits. 24. Themethod is well suited for routinedesignpurpases.However,it is suggested that, for major projects, or those where the proposed structurehas strict permissible total ordifferential settlements, other well-established methods of estimating the settlement are also used as a check. On such projects it may prove valuable to refer to the case studies listed Appendix 2 in which similar structures in or ground conditions are involved. In general it seems unlikely that the limits of accuracy can be significantly improved unless resort is made to the direct determination of in situ compressibility. 25. In conclusion itis appropriate to bear in mind the following remarks by Sutherland Before a designer becomes entangled in the details of predicting settlement (in sand) he must satisfy himself whether a real problem actually exists and ascertainwhatadvantagesandeconomiescanresultfromrefinements in settlement prediction.

Analysis of case recordsof settlement on sands and gravels


26. The object of thestudy described in thissection of thePaper was to assemble as muchdata as possible on actualfield observations of settlement with a minimum of interpretation to see if a simple picture emerged.The most important factors controlling settlement p are the effective bearing pressure q, the breadth B of the loaded area and the compressibility of the ground withinthe depth of influence of the loaded area. There are many other factors influencing settlement such as depth of founding, geometry of the loaded area, depth water table, time of etc. These factors were felt to be secondary compared with the above three principal factors and could be examined separately after the main trends had been established. 27. For any case record the quantities p, q and B are well defined. Thus, in their preliminary study Burland er al.6 chose to correlate p / q with B. The compressibility characteristics of the ground are much more difficult to define and Burland et al. only distinguished between three categories of granular material: loose, medium dense and dense. In the present study the same basic approach is adopted but a more refined method of classifying the compressibility of sands and
1332
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONSONSANDANDGRAVEL

gravels has been found to be justified. Moreover,it has proved necessary to consider in some detail the depth within which the compressibility significantly influences the settlement (i.e. the depth of influence z,) and also the validity of the assumption of a linear pressuresettlement relationship. These matters are discussed in the following paragraphs as a preliminary to the presentation of the analysis of the caserecords.
The standard penetration test ( S P T ) as a measure of compressibility 28. For the majority of the case records assembled for this study the ground conditions were investigated using the standard penetration test (SPT). For this reason, and because it is a test which is widely used, it was decided to use the SPT blow count asa measure of the compressibility of granular soils. Nevertheless, it is of the utmost importance to appreciate the limitations both of the test itself and the correlationof its results with compressibility. 29. The standard penetration test. At present the two most widely used standards are BS 1377: 197511 and ASTM D1586-67. The testing procedures are broadly similar, and outside the UK and the USA one of these two standards is normally used. An important exception to the general SPT procedureis in Brazil where the Mohr-Geotecnica sampler extensively used. is 30. There are numerous details of the test and its operation which are not standard.13 For example, there are considerabledifferences in the dimensions and lengths of drilling rod used in the test.Also, thedrivingtechniquecan vary significantly. The British and European standardsspecify the use of a trip hammer whereas American practice is to operate the driving weight manually using a cathead. Other factorswhich can influence the N value are the diameterof casing, the conditionof the driving shoe, the typeboring rig and the method of of cleaning the base of the borehole.According to Schmertmann14 almost all samplers used in the USA haveenlargedinternaldiameters to hold a liner. However, they are frequently used without a liner, which leads to a significant reduction in the N value. Over and above all these factors the crucial importance of maintaining an adequate level of water in the borehole must course be emphasized. of 31. It has always been recognized that the SPT is an empirical test. It is a test which will have to become completely standardized if its use as a yardstick for judging in situ properties, suchas compressibility, is to be enhanced. The need for standardization has been emphasized by Nixon3 who calls for the international use of the 1977 ISSMFE Report of the Sub-Committee on Penetration Test for use in Europe.15 Any future changes or standardization in the test that do take place should not deviate significantly from present procedures, so that experience already gained from the is not lost. test 32. Influence ofgrain size. The effects of grain properties, such as angularity and uniformity coefficient, on SPT resistance have not been adequately studied. Holubec and DAppolonia16 suggest that the SPT is influenced by the angularity of granular soil. Gibbs and Holtz found that the grain size had some influence. Tests on dryloose sands showed that the value for coarse sand was marginally N higher than for fine sand at the same relative density and overburden pressure. However, for dense sand there was no appreciable difference between fine and coarse sands. DAppolonia and DAppolonia concluded that the particle size does not appear to have a major influence provided gravel sizes are notpresent. ~ 33. Influence of submergence. Schultze and M e n ~ e n b a c h and Bazaraa have shown that the SPT resistance for coarsesandand gravel is not affectedby
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1333

BURLANDANDBURBIDGE

submergence. Terzaghi and Peck recommend that for dense ( N > 15), fine or silty sands beneath the water table, the measured N values should be reduced, and put forward thefollowing procedure

= 15

+ 0.5(N

15)

(2)

34. This proposal appears to be contradicted by the results of some laboratory tests of Gibbs and Holtz and Schultz and Melzer.* However, Bazaraa concluded from analysis of a large number of results of SPT tests within l m above and below the water table that theeffect of submergence on penetration resistance on very fine or silty sand is generallyto increase the blow count.On the basis of his results he suggested thatthe measured N values should be corrected by the formula N = 0.6 X N ( 10) 35. InJuence o overburden pressure. Although SPT resistance for a granular f soil is likely to be dependent on a number of factors it appears that the twomost important ones are the relativedensity and the effective overburden pressure. Thus, in order to assess the relative density, numerous methods have been proposed for correcting the SPT blow count to a standard overburden pressure (e.g. those of Gibbs and Holtz,Bazaraa2 and Thorburn2*). 36. Turningnowto compressibility, laboratory tests by D a r a m ~ l ashow ~~ that, for a given K O stress history, the two most important factors influencing the vertical compressibility are relative density and stress level-the same as for SPT resistance. 37. It therefore appears that, in attempting to correlate compressibility with SPT blow count, the effect of overburden pressure should not be eliminated since it has an importantinfluence on both. Hence no correction for overburden pressure was usedin this study. However, is recognized that the SPT it blow count does not, on its own, reflect the previous consolidation history of a deposit to any significant extentand theeffect of this hasto be accounted for separately. 38. Cornpressibility grade in terms of S P T . TerzaghisZ4 descriptive correlation between the relative density and N value was originally based on the Terzaghi and Peck allowable bearing pressure chart and theterms were therefore originally used as qualitative measures of compressibility. Since their original introduction the influence of overburden pressure on blow count has been recognized, as discussed under the previous sub-heading. Moreover, when an attempt was made to correlate foundation compressibility a, from the case records given in Appendix 1 with relative density , it was found that the rangeof compressibilities associated with each density zone was very uneven. It will become apparent that any descriptive classification of compressibility based on SPT requires a scale in which the range of N values associated with each zone or grade increases approximately exponentially to give an even spread of a, values. 39. A new descriptivecorrelationbetween SPT and compressibility of normally consolidated granularmaterials has been introduced and is given in Table 1. The opportunity has been taken to dispense with the terminology of relative density and replace it witha number of compressibility grades which, since they relate to uncorrected blow count, are a function both of relativedensity and overburden pressure. 40. It must be emphasized that the SPT blow count can never be anything more than a crude indicator of compressibility, even when restricted to normally
1334 Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

S E T T L E M E N TO FF O U N D A T I O N SO NS A N DA N DG R A V E L

consolidated sands (as in Table 1) and when the procedures are perfectly standardized.
The relationship betweenS P T and other tests 41. As discussed in Q$ 2 8 4 0 , the SPT resistance is used as a measure of compressibility in the present study. However, for many of the case records given in Appendix 1, no standard penetration tests were carried out. In order to make use of these case recordsit has been necessary to attempt to correlate SPT count blow with other tests, the three main ones being cone penetration tests, oedometer tests and plate loading tests. It is recognized that the correlations are only approximate. Nevertheless, is it important to relate these widely used tests to the compressibility grade if the resultsof the present study are to generally useful. be 42. Cone penetrationtests. MeyerhoP5 investigatedtherelationshipbetween N value and staticcone resistance qc for a number of sites, mainlyfor fine and silty sands andsuggested that
q, = 4.4N

where qc is in kgf/cm2. This relationship was found to be independent of density. 43. Meigh and Nixon,26Rodin and Sutherland have shown that the above relationship is restricted to fine and silty sands and that the ratio q c / N increases with grain size. Burbidge collected together the original data used by the above workers,together with other results,includingthoseassociated with thecase records referred to in this Paper. Following the work of T h ~ r b u r n , these data ~ have been correlated with average grain size and were found to be within the zones shown in Fig. 2. For the case records from Brazil the measured blow counts were reduced by afactor of 0.7.30*31 The resultsconfirm that thedensityhaslittle influence on q J N although there is a slight trend for loose sands to lie towards the upper limit of the scatter of the results. 44. Oedometer tests. The oedometertestisthemostcommonly used laboratory test for estimating settlements on sands. In Poland and Russia it is fairly frequently used but outside these countriesit has been less popular. 45. The major difficulty with oedometer tests, as with other laboratory tests on granular soil, is obtaining undisturbed samples. It has been found that fine sands are generally not as prone to mechanical disturbance as coarse sand and where representative samples have been carefully hand cut from excavations oedometer tests have been reasonably successful. 46. About a quarter of the case studies found in the literature contain oedometer results. The opportunity was taken to compare m, values from such tests with N values for sites where the two testswere made. Average N values were obtained over a depth range of 5-15 m and were compared with values ofm, evaluated for a loading intensity of 100 kN/m2. Fig. 4 shows the results of the comparison for eight sites. The number against each point refers to the appropriate case number given in Appendix 1. Seven of the cases are for fine sand. Many more data are required before any firm conclusions can be drawn about the relationship between m, and N . For the purposes of this study the full line in Fig. 4 was used to assess the compressibilitygrade from oedometer results. 47. Plateloadingtests. Terzaghi and Peck7 publisheda diagram showing a collective pressure-settlement chart for 0.3 m (1 ft) square plates bearing on loose to very dense sand strata above the groundwater table. At the time when the chart was constructed only limited plate bearing test data were available and, moreover,
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1335

BURLANDANDBURBIDGE
OF

Fine sand 8Medlum sand

, 1 1 1 ,

I l l , ,

05

0.1

0-05 m, mz/MN (at U"'

0 01 0.005 100 kN/mz)

0,001

Fig. 4. Relationship of N with m, (numbers referto case records in Appendix I )

the corresponding SPT values were probably from two different sized spoons. It was thereforedecided to construct an updated version of the chart using the compressibility grading classification. Data were collected from the case records examined by B ~ r b i d g e Bazaraa," Meigh and Nixon26 and ~ d i n . ~ ~ ,~ R 48. The resulting charts are shown in Fig. 5 and they relate to three size ranges of plate: 0.254.4 m, 04-0.7 m and 0.7-1.2 m. The scatter of results on which these charts are based is large but tends to decrease with larger plates. In compiling the charts no difference was apparent between tests on dry and moist sands. A few test results were availablefor which the water table was at a depthof less than B and these showed considerably larger settlements than for dry sand. The influence of the depth of water table is discussed in $9 98-101 where it is shown that the results of plate tests with H J B > 1 correlate well with the data for larger loaded areas andhigh water tables. 49. Despite the approximate nature of Fig. 5 several interesting features emerge. (a) The larger the plates the greater the linear range and the lower the curvature of the pressure-settlement curves. (b) For any given pressure and compressibility grade, settlement increases with platesize. (c) The initial tangent slopes to the pressure-settlement curves become more distinct with increasing plate size. Thus, for the smallest size of plate, bedding errorsandminor densityvariationswouldleadtoserious errors in interpretation. 50. Meigh32 has suggested that the grain size and grading of sands could be important factors influencing their compressibility under a test plate. While constructing the charts in Fig. 5 no discernible relationship with grain size could be found, a conclusion also reached by Terzaghi and Peck.7 In the case of gravels and gravelly sands the datawere found to be only in the zones for grade IV and better with little correlation withSPT blow count.
1336
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

S E T T L E M E N T O F FOUNDATIONSONSANDAND

GRAVEL

Bearlng pressure: kN/m2

0.25 m-0 4 m

(a)

(C)

Fig. 5. Charts for assessing the compressibility grade from plate loading tests of sand f carried out at shallow depth in the baseo wide excavations or
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1337

BURLANDANDBURBIDGE

Depth o influence f 51. Guidance varies on the depth of influence beneath a loaded area on sand. Terzaghi and Peck' recommend taking the average blow count over a depth equal to the breadth B. Parry33 takes the depth of influence equal to 2B but places greater emphasis on the values immediately below the foundation. N 52. S ~ h r n e r t m a n n ~ ~ takes the depth of influence equal to 2 B and uses a also simpleinfluence diagram to obtain the distribution of verticalstrain.Intheir statistical analysis of a number of settlement observations Schultze and Sherir5 took the depth influence equal to2B. of 53. For a uniformly distributed circular load on an isotropic elastic half space the depth of influence is usually taken as 2B. The settlement at this depth is about 25% of the surface settlement. Hence,for practical purposes, the depth ofinfluence may be assumed to be the depth at which the settlement is 25% of the surface settlement and is denoted by zI (or the relative depth of influence (z/B),). 54. There are not many experimental data for assessing the depth of influence for foundations on sand and muchof the data are from model tests. Morgan and G e ~ ~ a plot ~ ~ distributions of vertical displacement with depth for a number r d the of tests on model footings ranging from 0.2 m to 0.9 m in diameter. The 25% settlementpointscorrespondto ( z / B ) , varyingfrom 1.8 to 1-13. Breth et d 3 ' measured the settlement distributions beneath 1.0 mdia.footings on carefully prepared beds of very loose medium to coarsedried sand. An approximately linear distribution of settlement with depth for all the tests was observed with 25% of the surface settlement occurring (z/B),equal to about at 1.5. 55. Turning now to the field measurements,Shvets a n d K u l c h i t ~ k i i ~ ~ measured the settlementdistributionbeneath l msquare plates ontwo alluvial soils-a slightly silty sandy gravel and a very silty slightly gravelly sand. The results are given in Fig. 6 and it can be seen that the 25% settlement point occurs between ( z / B ) ,equal to0.8 and 0.6. 56. Figure 7 shows the normalized distribution of measured settlement with z / B beneath five buildings on deeplayers of sand. Theresults from Nikitin et al.39 are from a 61 m dia. ring foundation for a television tower (Case 63). Within the main ring were footings for a service tower along with a second foundation ring. The whole foundation complex occupied most of the area and may treated as a be single entity. The soil profile consisted20 m of dense to very dense sand, 15 m of of stiff to very stiff clay, followedby rock. Reference points were located at depthsof 6 m, 12 m and 25 m below the foundation. The maximum observed settlement the of foundation was 37.8mm, of which at least 19.5 mmtook place inthe clay. Eighty-five per cent of the compression of the sand tookplace in the top 12 m, i.e. for z / B equal to 0.2. The results plotted in Fig. 7 relate only to the immediate compression of the sand. 57. The resultsfromBreth and Chambosse4' arefor a reactorbuilding at Biblis, Germany (Case 27). The settlement distribution was measured down a borehole 1.8 m to one side of the 60 m dia. circular raft foundation. The ground conditions beneath the foundations consisted of 7 m of dense gravelly sand, 48 m of dense tovery dense fineand medium sand, followed by a great depth of very stiff clayey silt. When the settlement of the raft had reached 40 mm the surface settlement of the instrument was 1 3 3 mm. Near the surface the settlement decreased very rapidly with depth, decreasing to 5 mm at z / B equal to 0.25. Thereafter it decreased more slowly, becoming about 2.5mm at a depth of 57 m at the top of the silt layer. The normalized settlement distribution shown by the curve labelled
1338
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONS O N S A N D A N D

GRAVEL

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1339

BURLANDANDBURBIDGE

(27) in Fig. 7 is for the sand layer only. Although no measurements were made beneath the centre of the raft it is clear that most of the settlement takes place above zjB equal to0.25. 58. The results presented by Dunn41 are for a nuclear power station founded on a raft 55 m wide and 101 m long (Case 32). The underlying ground consists of 31 m of very dense fine beach sand overlying stiff silty clays and dense silt. Five settlement plates were located at various levels in a borehole beneath the raft, with the deepest plate being at a depth of about 12 m. Curve (32) shows the observed settlement distribution. Undoubtedly some settlementwill have taken place in the underlying clays and silts, inwhich case the settlementsin the sandwould diminish more rapidly than shown by the curve. 59. Curves A and B in Fig. 7 show the settlement distribution beneath two buildings in Berlin Kriegeland W e i ~ n e r The .detailed normalized distribution of ~~ settlement varied with the magnitude of the loading. The points show the extreme values and the curveshave been drawn through the mean values. It must be emphasized that very little information is given by Kriegel and Weisner about the ground conditions for these two buildings except that the sand is 4&50 m deep and is of medium density. Clay layersare frequently encountered in Berlin and the possibility of a deep clay layer this site should not be ruled out thereby reducing at the valuesof (zjB),. 60. InFig. 8 thevalues of ( Z / B ) correspondingto p/po = 25% have been ~ plotted against breadth for the various model studies and field measurements. It can be seen that, although the scatter large, there is a marked tendency for (z/B), is to decrease as thebreadth increases. 61. It should be emphasized that the depth of influence corresponding to a given value of B will not be unique and will depend on the variation of stiffness with depth.Neverthelesstheresults given inFig. 8 indicate atrend which is
0.1 0 1

B: m
1

10

100

I
Shvets and Non-homogeneous elastic Breth er a / 37

.Case 63 *Case 27

Kriegel and Weisner4*

Melbourne Series II
2-OL

Fig. 8. Relationship between measured depth ofinJuence z, andfoundation breadth. Full line is takenfrom Fig. 9 and isjitted at B = 0-2m assuming that Ebjak = 10
1340
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONSONSANDANDGRAVEL

broadly in accordance with theory for a non-homogeneouslayerhaving an increasing Young's modulus with depth. Fig. 9 shows the normalized distribution of settlement with z/2a beneath the centreof a rigid rough circular load of radius a on a Gibsonsolid for various values of Eb/ak (the results were obtained by means of a finite element analysis). For a given value of EL and k it is clear that as a increases the relativedepth of influence (z/2a),decreases. The full line in Fig.8 was
PlPO' %

0.

m 1.

a
E 1.

1,' = 13 /

2.

Fig. 9. Distribution oj'settlement with depth for a circular rough rigid foundation resting on an isotropic non-homogeneous elastic soil
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1341

BURLANDANDBURBIDGE

derived from Fig. 9 and was fitted at a breadth of 0.2 m corresponding to avalue of Eb/ak equal to 10. Garga and Q ~ i give~a similar relationshipfor the depth of n ~ strain influence fora non-homogeneouslayer. 62. Many more measurementsareneeded of thedistribution of settlement with depth beneath foundations on granular soils both from the point of view of establishing the depth of influence and, of more importance, for studying the in situ deformation properties. For the purposes of this study the full line in Fig. 8 was used as a rough guide to the depth of influence when N is constant or increases with depth. In a very few cases N decreased with depth and in these instances the best fit to the general trends of the data was obtained by taking the depth of influence equal to2B. 63. The arithmetic mean of the SPT blow count ( N ) over the depthof influence was used to obtain the compressibility grade of the foundation subgrade. The full line in Fig. 8 when plotted as B against zI on double log axes forms a straight line as given in Fig. 3.
Pressure-settlement relationship 64. Most of the current methods settlement prediction on sandsassume that of the relationship between bearing pressure settlement linear the and is over working range of stresses. S ~ h u l t z e ~ ~Shultze and SheriP5 conclude from the and study of a number of case records that the pressuresettlement relationship is linear over the periodof construction. It has already been noted from Fig. 5 that as the size of test plates is increased the initial portion of the pressure-settlement curve becomes more linear. 65. A number of the caserecordscollected by Burbidge contain complete pressuresettlement data and make possible a study covering a range of ground conditions, foundation dimensions and bearing pressures. Five examples will be given. 66. Case 27 is a nuclear reactor founded at a depthof 5 m on a 3 m thick m 60 dia. raft.The underlying ground consists of 60m of dense sand andgravel, assessed as grade 11, overlying Tertiary sands and clays. The net pressure against average settlement relationship for the reactor is shown in Fig. 10 and is, for all practical purposes, linear. 67. Case 51 consists of two 12 storey towers each founded on four 5 m deep footings 4 m wide and 7 m long. The underlying ground consists of 7 m of dense sandy gravel over weathered sandstone. SPT tests on the gravel indicate that it is of grade 111. The net pressuresettlement curves for the four outermost footings are shown in Fig. 11. Settlement observations only began once the pressure had reached 134 kN/mZ.Three of thefootingsexhibitedlittlesettlement uptoa bearing pressure of 223 kN/mZ but thereafter the relationships are again linear for all practical purposes. This case is of interest since the bearing pressures exceed 500 kN/mZ. 68. Case 60A is an 18 storey reinforced concrete building founded on a 1-2 m thick raft. The raft is 22.9 m X 32.6 m and is founded at a depth of 3 m. The underlying ground consists of fine to medium sands to great depth with the top 7 m compacted by vibroflotation to grade IV. Settlement observations were started at a gross pressure of 55 kN/m2. It can seen from Fig. 12 that the initial portion be of the pressure-settlement curve islinear butat higherpressuresthecurve steepens. This is thought to be due to time dependent settlements occurring during the slow application load near the end of construction. of
1342
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONS O N SANDANDGRAVEL


0
100
Net bearing pressure: kN/m2 200 300 400

500

10-

F
-

20

30 -

40

50

Net bearing pressure: kN/m2

Fig. IO. Case 27. Nuclear reactor founded at 5 m depth on dense sand and gravel assessed as grade I1 (N = 41-60)

'B--&

U r - & m
I

B-:

1 4 m

I F - P :HF o o t l n g s mFig. II. Case 51. T w o tower blocks each founded on four 5 m deep footings resting on sandy gravel assessed as grade ( N = 2 6 4 0 ) 111
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1343

BURLANDANDBURBIDGE
Gross bearlng pressure: kN/mz

* 2

"A

Fig. 12. Case 60A. An 18 storey building on a raft founded at a depth o 3 m onjine f to medium sands assessed on grade IV ( N = 16-25)

69. Case 41 is a 10 storey building founded on a raft at the centre surrounded by pad footings. The raft is 11.0 m wide and 33.4 m long, and is founded at a depth of 5 m. The footings are also founded at a depth of 5 m. The underlying ground consists of 12.4 m of sand, assessed from SPT values as grade IV, over stiff clay. Settlement observations were begun when the gross pressure on the raft was 38 kN/m2. The pressure-settlement relationship for the raft is shown by the full line in Fig. 13. Once the gross pressure exceeds the initial vertical effective pressure a : , the curve becomes significantly steeper. The broken line is for a 4.1 m square footingadjacenttotheraft.Settlementreadingsonlybegan when thegross bearingpressure was largerthanthe effective overburden pressure, andthe pressure-settlement relationship can be seen to be linear. 70. Case 69 is for a building in north-west Berlin which is founded on a 5.5 m X 6.5 m raft on sand of grade V. Although the precise depth of founding is not known it is presumed to be between 2 and 3 m. The pressure-settlement relationship is plottedinFig. 14 and itis evident thatthere is amarked change of curvature over the initial portion of the pressuresettlement curve, after which it is linear. 71. The casesdiscussedin $5 64-72 include sandsand gravelswithgrades ranging from I1 to V, foundation widths ranging from 4 m to 60 m and bearing pressures up to 500 kN/m2. All the deposits are believed to be normally consolidated. It can be concluded that for pressuresin excess of theinitial effective overburden pressurethe pressuresettlementrelationship is, for practicalpurposes, linear. For pressures less than the initial effective overburden pressure the compressibility is reduced by a factor of 2 to4. Observations presented by Dunn4' on the settlement of theDungeness B nuclearpowerstation(Case 32) are in agreement with this conclusion. DAppolonia et al.45 deduced that the modulus
1344
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F F O U N D A T I O N S ONSANDANDGRAVEL
Gross bearlng pressure. k N / m 2

150

1000

50
I

200

2\

4-

E
g
m

8-

\
\

10-

\ \
\ \ \

m F 1214\

161818 5 l

\
\

Rafl
-7 74 .

Footing m +

l
v
m

wE i

I
33.4 m X 11-0 m

I
Footlng 4.1 m square

Fig. 13. Case 41. A 10 storey building founded at a depth 5 m, partly on a raji and of = 16-25) partly onfootings, on sand assessed as gradeIV

(m

[ M = E/(l - v)] for a preloaded sand was approximately twice that of a normally consolidated sand. 72. The conclusion that compressibility is reduced at pressuresbelowthe maximum previous overburden pressure is at variance with the conclusions of S ~ h u l t z e Sherif,46 and Schultze and Sherif3* who conclude that the pressure,~~ settlement curve is uninfluenced by the removal of overburden pressure. A study of the observations presented by S c h ~ l t z e ~ revealsthe following. First, rather ~.~ large time corrections have been applied to the settlement observations, and secondly, in many instances excavation for the raft foundations took place below the water table. If the uncorrected settlement observations are used and the gross total pressures towards the end of construction are reduced by the hydrostatic uplift of the groundwater then the pressure-settlement relationships reveal small but discerniblepreconsolidationpressures. For example, the results for Case 83 are plotted in Fig. 15 and a kink in the vicinity of the effective overburden pressure is apparent, giving a change of slope of about 2.
Relationship between foundation subgrade compressibility breadth and 73. In $8 6 4 7 2 it was shown that the slope of the pressure-settlement curve Ap/Aq (equal to a,, the foundation subgrade compressibility) is approximately
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1345

BURLANDANDBURBIDGE
Bearing pressure: kN/mz
0

100

200
I

L 5 . 5 m-+

10E E
C

m
+ .-. 3

30-

40-

Fig. 14. Case 69. Building founded on raft at a depth o 2-3 m on sand o grade V f f = 9-15)

(m

0
I

,.l 00

Gross bearmg pressure. kN/m* 200


I

E E
C

E =
m
W

10-

m m

20

Fig. 1.5. Case 83. Building founded on 1716 m X 84.0 m raft at a depth o 10.7 m in a f I V (N = 16-25). Water table at 8.5 m depth sand and gravel assessed as grade below ground level Delivered by ICEVirtualLibrary.com to: 1346 IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONS O N SANDANDGRAVEL

constant for normally consolidated sands and gravels. In Appendix 1 the important details of each of the case records referred to in the present Paper are summarized. Where the detailed pressure-settlement relationship was available the virgin portion of the curve was used to determine Ap/Aq. Where the pressuresettlement relationship was not available and only the immediate settlement and the gross effective pressures are given the assumptionwas made that thereloading curve up to theeffective overburden pressure a has a slope equal to one-third of : , the virgin curve. Thus thevalue o Ap/Aq is given by f

74. In many cases (e.g. for most footings) only the net bearing pressure and immediate settlement were known, in which case values of pJqh,, are given. 75. Relationshipbetweena, and B. As stated in 26 and 27, the approach to analysing the case records is similar to that adopted by Burland et namely, to correlate the values foundation subgradecompressibility a, given in Appendix 1 of with the breadthB for each compressibility grade. 76. In Figs 16 to 20 the measured values of foundation subgrade compressibility a, (mm/(kN/m2) are plottedagainst B (metres) on double logaxes for compressibility grades I1 to VI. It can be seen that for each grade the majority of the observations give a well-defined linear correlation between log and log B. A a, few of the cases lie outside the general spread of the results. In Fig. 16 (grade 11) Cases 29 and 32 lie well above the scatter the results. Case 29 is the 93 m dia. oil of storage tank in the Ekofisk Field of the North Sea. The soil profile showsthat the sand is underlain at a depthof 26 m by a 50 m thick stratum of hard clay. It seems very probable thatsignificant settlementstook place within this clay stratum. 77. Case 32 is the Dungeness B nuclear power station in Kent, England, which is founded on fine sand. It will be shown later that there is some evidence fromthe present study to suggest that SPT blow counts on submerged fine sand give N values which are too high and should be reduced in accordance with the recommendations of Terzaghi and Peck (see 33 and 34). On this basis the value of N for Case 32 decreases from 60 to 36; this falls within compressibility grade 1 1 1 and is plotted in Fig. 17, where itlies within the spread of the results. 78. In Fig. 19 (compressibility grade V) Case 85 lies well below the spread of the results. This case is a chimney for a power station at Cologne, Germany, reported by S c h ~ l t z e . ~ The SPT results were consistently less than 10 below ~.~ the water table and, in view of the small settlement of the structure, it can only be concluded that the SPT results are unreliable. Parry33 also noted the anomalous results from this case. 79. In Fig. 20 (compressibility grade VI) Case 79B lies above the spread of the results. Cases 79A and B consist of two preload embankments next to each other, being 8 m and 11 m high respectively. A study of the pressure-settlement curves for each embankment shows that Case 79B had a steeperpressure-settlement curvefromthe start of loading and that the much largervalue of a, cannot therefore be attributed to local yielding under the higher embankment. It appears that thecompressibility of the soil for Case 79B corresponds to grade which is VII, consistent with some of the cone test results the site. on 80. In Figs 1 6 2 0 the broken lines drawn through the points have been fitted using linear regression of log a, on log B. The cases discussed previously which lie
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1347

BURLANDANDBURBIDGE

Regresslon line of log a, on log B

Compressiblllty grade II

(m

41-60)

Breadth: m

Fig. 16. Relationship between a,and B for compressibility grade I I (S = 4 1 6 0 )

outside the spread of the results have not been included in the analysis. For the grade I1 results in Fig. 16 the regression line is heavily weighted by the relatively large number of cases for B less than 3 m and theparallel chain dotted line is felt to be more realistic. 81. Table 2 lists the slope, correlation coeficient and standard error for each regression line in Figs 1&20. In all cases the correlation coefficient exceeds 0.8. The standard error for a, varies from x1.46 to x1.9 with a tendency to increase as the compressibility increases. 82. A particularly significant feature to note in Table 2 is the similarity in the slopes of the regression lines. The weighted average of the slopes is 0.704 (the weighting takes account of the correlation coefficient and the number of cases
1348 Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONSONSANDANDGRAVEL

_ _ _ Regression lme of log a, on log B


Compressibllity grade 1 1 1

(m=26-40)

>

01'

'

10

100

Breadth: m

Fig. 17. Relationship between a, and B for compressibility grade 111

(m

= 2640)

associated with eachregressionline) and the greatestdeviationsfromthis are +20% and - 14%. It therefore appears that the slopeisindependent of the compressibility grade-an observation which leads to considerable simplifications in the subsequentanalysis. 83. In Fig. 21 the regression lines for each compressibility grade are shown as broken lines. The full lines all have slopes 0.7 and their locations havebeen fixed of by a least-squares analysis of the deviations of log a,. They are termed ' adjusted mean lines '. 84. Comparison of the adjusted mean lines in Fig. 21 with the observations reveals an interesting result. It transpires that the adjusted mean line forany given grade forms a reasonable upper bound for the next grade up and a reasonable
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1349

BURLAND AND BURBIDGE

- - - Regression line of log a, on log B


Compressiblllty grade IV

(r

16-25)

0.l1

7
10 100

Breadth: m

Fig. 18. Relationship between a, and B for compressibility grade I V

(N = 16-25)

lower bound for the next grade down. This is illustrated in Fig. 17 for grade 111 compressibility. The adjusted mean lines for grades IV and I1 are shown chain dotted and are seen to form very reasonable upper and lower limits to the spread of the results. Upper and lower limit lines for the other compressibility grades have been obtained in the same way and are shown by chain dotted lines in the appropriatefigures. In general the limit lines shown in Figs 16-20 correspond to a spread of rather less than plus or minus two standard errors from the regression line for each grade. Very few of the results lie above the upper limit lines whereas rather more lie below the lower limit lines, particularly for values of B less than about 3 m. Thus Fig. 21 forms a convenient summary of the data and could be
1350
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENTOFFOUNDATIONSONSANDANDGRAVEL

-- - Regression line of log a, on log E Cornpressibility grade V ( N = 9-1 5)

O1

d
1

Breadth: m

10

100

Fig. 19. Relationship between a, and B for compressibility grade V Table 2. Analysis o r,egression lines in Figs to 20 f 16

(N = 9-15)

Grade

No. of cases
19 45 68 39 27

Slope m

Correlation coefficient
0.89 0.91 0.82 0.84 0.86
X

I1 111 IV V
VI

Weighted average

0.669 0.710 0.592 0.833 0.805 = 0.704

1.60
1.46

X X X X

1.60
1.90 1.79

0.620 1.279 2.168

X
X X

10- 10-
10- 10- 10-*

5.585
14.656

* at is in mm/(kN/m);

B is in metres.

used for design purposes. 85. It is of theutmost importance to recognize that the regression lines or adjusted mean lines cannot be used for extrapolating the settlement for a small footing to a larger one. By increasing the breadth of the footing the depth of influence is increased and this may well result in an upgrading of the compressibility grade. 86. Relationship between a , , B and N.The equation for the adjusted mean lines
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29
1351

BURLANDANDBURBIDGE
1 ooop
L

179'B , / ,

1-

O1

Regression line of log a, on log B Compressibility grade VI ( N = 4-8)

_--

I l l 1

' l ;

10 Breadth: m

100

Fig. 20. Relationship between a,and B for compressibility grade VI

(N

=4

3)

1352

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONSONSANDANDGRAVEL

10 Breadth. m

100

Fig. 21. Relationship between a , , B and compressibility grade showing mean lines and upper and lower limits
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1353

BURLANDANDBURBIDGE

in Fig. 21 is given by log a, = m


X

log B

+ C(#)

where the slope m = 0.7. The term C ( N )is a function of the compressibility grade and hence of the mean SPT blow count N . The value of C corresponding to a particular compressibility grade is given by the intercept of the appropriate adjusted mean line with a, axis (i.e. when B = 1 m). Hence the

a, a, C ( N ) = log - = log p(= log I,) B"


where I, is termed thecompressibility index. 87. In Table 2 the values of u,/B"~( I , ) for each compressibility grade are = listed and in Fig. 22 they are plotted as open points against on double log axes. The upper and lower limits for each compressibility grade are also shown. It cay be seen that there is an approximatelylinear relationship between I, and log N l_og and the reason for having a progressively increasing interval in N for successive compressibility grades nowbecomes apparent. The spread I , between the upper of and lower limits increases from a factor of about four for grade I1 to about eight for grade VI. 88. In view of the apparently linear relationship between log I, and log R an independent regression analysis was carried out on all the cases in Appendix 1 for which SPT or cone test data are available. The results of the analysis are given in Fig. 23. The regression line for log( u J B " ~ on log is shown as afull line. It has a ) slope of - 1.43 and an intercept on the N = 1 axis of I , = 1.7. The coefficient of correlation is 04348. 89. The regression line in Fig. 23 has been plotted in Fig. 22 and is seen to agree well with the points for each compressibility grade. Mean upper and lower limit lines have also been drawn in as chain dottedlines and arereproduced in Fig. 23. It can be seen that most of the individual cases lie between these limitlines and the majority of those that do not arefor B less than 3 m, for which the scatter is somewhat larger. 90. Thus Fig. 22 is a more compact form of Fig. 21 and can be used for design purposes-see @ 7 and8.

The influence of various factors on settlement 91. The collection of a relatively large set of data such as is given in Appendix 1 makes it possible to study statistically the influence of various factors on the immediate settlementof loaded areas on granular materials. 92. The regression line in Fig. 23 can be represented by the expression

where brackets the in figure represents one standard Denoting error. I O ~ ( N " ~ / B ' ' ~ )a, X 10' as T , the value of T for each case given in Appendix 1 X may be calculated. Fig. 24 shows a histogram of T for the complete data set which can be seen to be approximately normally distributed. The mean value of T is denoted by and is equal to 2.232. For any given foundation the best estimate of the foundationcompressibility ii, is given by
1354 Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONS ON SAND AND GRAVEL

t
1
I

Cornpressibility grades

I , ,

10SPT N

100

Fig. 22. Relationship between compressibility indexI , and compressibility gradederivedfrom Fig,21

94. The deviation of any measured value of a, from the best estimate may be expressed as aJ2, or G,/a,. The values of aJ2, and 2,/a, corresponding to one
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

7 1 ,

1355

BURLANDANDBURBIDGE
lOOr

---

Upper and lower limlt llnes from Fig. 22


D

B33 m

oB<3rn
0

0.1

I 1

10 SPT N

I l l

100

Fig. 23. Relationship between compressibility index I , and N for all cases in which SPT or cone tests were carried out

standard deviation from i, are both 1.82. Hence any statistical analysis will only i detect major influences on a, and care must be taken not to read too much into minor trends. 95. It is assumed that the majorfactors influencing a, (apart from N and B) are the length L, the depth D,the depth of the water table H , and the thickness of the
1356
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENTOFFOUNDATIONSONSANDANDGRAVEL
30r

Fig. 24. Frequency distribution of the settlement observations all the cases for

sand layer beneath the foundation H , . Each of these parameters will be studied separately. 96. Influence of LIB. In Fig. 25 values of a,/ii, have been plotted for various values of LIB for those cases in which DIB < 0.25, H,/B < 0.2 and H J B > 2. The chain dotted lines correspond to one standard deviation either side of the mean. It can be seen that the measured values of a, are larger than 2, for the majority of cases, indicating a positive correlationbetween LIB and a , . However, the influence of LIB is not large and it would appear that the average valueof a,/ii, is unlikely to exceed about 1.5 at LIB equal to 5. Although there are very few observations for larger values of LIB the results do not point to any further increase in the average value of a,/ii, beyond about 1.6. For comparison the relationship for homogeneous elastic theory is shown as a broken line and isseen to give significantly larger averagevalues of a,& than observed. The full line in Fig. 25 isgivenby the empirical expression

which appears to represent the observed trend reasonably well and tends to 1.56 as LIB tends to infinity. 97. Influence o DIB. InFig. 26 observed values of a,& have been plotted f against DIB for the cases in which LIB < 1.5, H,/B < 0 . 2 and H J B > 2. There are a number of observations for D / B < 0.3 but relatively fewfor larger values. It appears that there is no obvious correlation between a,/ii, and DIB for DIB 3. In their analysis of a large number of observations on one site DAppolonia et aL8 report only a 12% reduction in settlement as DIB increases from 0.5 to 1.0. Such a variation is too small to detect in relation to the scatter of the results presented
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1357

B U R L A N DA N DB U R B I D G E
a,/%,

,I'

Z,/a,

Fig. 25. Influence o LIB on foundation subgrade compressibility f a,

.-

0
0

I
Zf/al

OB 2 3 m

Fig. 26. Influence o DIB onfoundation subgrade compressibilityaf f

here and supports the conclusion that theinfluence of DIB is small. 98. InJluence of depth o water table H,. There are differing opinions about the f influence of the depth of the water table on settlement. For example Terzaghi and Peck' assume that for a deepwater table the settlementof a foundationis half that for a water table at founding level. However, MeyerhoP recommends that the presence of the groundwater table shouldbe ignored on the basis that its effect is already reflected inthe SPTblow count. 99. In the following analysis a distinction is drawn between results from building foundations andresults from plate loading tests. It can be seen from Appendix
1358
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

GRAVEL

30H,/B 0s 20i ;

>

I I

c
a ,
3

No of values = 24 T(mean = 2.103 U = 0.215 ($af) mean = 1.34

10-

I l
I

1.0
T
=

2.0
n1.4

3-0

log

-7X a, X 10 ~ 0 .
(b)

Fig. 27. ( a ) Frequency distribution o settlement observationsfor buildings with H , f greater than 5 m; ( b ) frequency distribution o settlement observations for plate f loading tests withH,/B greater than unity

1 that, forthemajority of building foundations, embankments and tanks, the water table is close to founding level. Hence it is possible to analyse the results of those cases where the water table deep (taken as greater than m) and compare is 5 them with the whole data set. There are 15 cases which fall into this category and Fig. 27(a) shows the frequency distributionof T . The mean valueof T for this data set (given by the full line) is 2.176 and the standard deviation 0 is 0.187. These values may be compared with the corresponding values for the complete data set, which are = 2.232 (shown by the broken line) and U = 0263. The average value of aJa, = 1.13, i.e. the settlements of the foundations with deep water tables are, on average, only 13% less than the best estimates from the wholedata set. It must
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1359

BURLANDANDBURBIDGE

be concluded from the above analysis that the level of the water table has no significant influence on the valuea, for building foundations. of 100. Almost all the plate loading tests listed in Appendix 1 have H J B > 1. Fig. 27(b) shows the frequency distribution of T for 24 plate tests with H,/B > 1. The mean valueof T for this data set is 2.103 and the standard deviationis 0.215. The average value of ii,/a, is 1.34. Thus the settlements of the plates are, on average, 25% less than the best estimates. However, the loading of the plates was carried out in less than a day, whereas the loading of the building foundations usually took place over a year or more. It will become evident later in the Paper that significant time-dependent settlements occur on sand. Hence thefact that a, from the plate tests is less than i, is not surprising and can be attributed prini cipally to time effects. 101. It appears from this study that the position of the water table has only a small influence on the value T . It is important to emphasize that of this conclusion must not be taken to imply that the positionof the water table does notinfluence the settlement. Whatit does do is to confirm Meyerhofsview that the effect of the watertable is probably reflected in thevalue of N. If awatertablechanges subsequent to the determination of the N values the settlements maydiffer appreciably from the predicted values. 102. Influence of thickness of sand layer. A few of the cases given in Appendix 1 have a thicknessof sand layer H , beneath the foundationof less than 2B. As might be expected there is a tendency for the values of a, to lie below iir but there are insufficient cases for any useful trend to emerge. It is suggested that, for design purposes, the predicted value a, should be reduced by the factor of

when H, z, where zI is the depth of influence given by Fig. 3. 103. Influence ofgrain size on N. It is generally accepted that grain size does not have a major influence on the number of blows in an SPT. Two soil types where uncertainties exist are fine sands or silty sands below the water table, and gravelly soils, as discussed @ 33 and 34, and 9 32 respectively. in 104. Finesandsandsiltysands.InFig. 28 theblackpoints refer tothe observed settlements of foundations on fine sands andsilty sands for which N was evaluated from SPT tests below the water table. In $ 33 and 34, two methods of correcting for submergence for fine sands were mentioned, Terzaghi and Peck' and Bazaraa." The open points in Fig. refer to corrected blow counts N' where 28 N' = 15 0.5(N - 15), as proposed by Terzaghi and Peck. It can be seen that only five cases are affected and only two of them (32 and 64) significantlyso. The effect of the correction is to bring these two cases closer to the mean regression line taken from Fig. 23. Application of the Bazaraa correction (W = 0.6N) translates all the points significantlyto the left, which results in a poorer overall correlation. Therefore on the basis of the limited evidence available it appears that the SPT correctionproposed by TerzaghiandPeckforsubmerged fine or siltysands results in an improved assessmentof compressibility. 105. Gravel and gravelly sands. Fig. 29 shows a plot of I , against N for all the cases involving gravel, sandy gravel and gravel/sand. inspection it can be seen By that the mean the points tends lie to theleft of the meanregression line for the of to complete data set. This is confirmed by a statistical analysis of the results, which

1360

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONSONSANDANDGRAVEL

.N
O N'
=

15

+ %(N-15)

\
l , ,
I
I

I I I J

10 SPT i? and

100

Fig. 28. Relationship between compressibility indexI , and N for submergedfine and silty sands

gives a mean value of T = 2.085 and U = 0.246 compared with = 2.232 and 3 = 0.263 for the whole data set. It is a simple matter to show that T',,,, can be made equal to by correcting the N values such that N' = 1.25 X N . This is a fairly small correction and in many cases could perhaps be neglected. It is worth noting that the lowest value of N for the case records involving gravel is 13 and more data are required for lower valuesof N . 106. Inherent variability. Natural deposits of granular soils are inherently variable, both laterally and vertically:thevariability will differ from one site to
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29
1361

BURLANDANDBURBIDGE

t t
0.1
1

0 6 2 3 m o 6 < 3 m

10

? 00

SPT

Fig. 29. Relationship between compressibility index I , and gravels

A for gravels and sandy

another. Effects of different foundation geometries and loadings have made it difficult to isolate the effects of inherent variability on the settlement of separate foundationsata given site. Thepresentstudyindicatesthatthe influence of geometry and load can largely eliminatedby expressing the measurements from be a given site as the quantity T . The variation in T for a given site is then a measure of the inherent variabilityof the settlement characteristics the site. Fig.30 shows of a plot of TIT mean for all the cases where more than one foundation was observed at a given site. The measured settlements generally between about 50% of the lie average, although on some sites it is less than f20%. Therefore, given perfect
1362
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONS O N SANDANDGRAVEL

Fig. 30. Investigations into the inherent variability of settlement characteristics at a number of sites

methods of measuring the compression characteristics of granular deposits and predicting settlement, one could normally expect differences of up to a factor of about 3 in the actual settlements, depending on the site conditions. The method of correlating foundation compressibilitywith compressibility grade given in Fig. 22 gives differences varying between factors of 4 to 8. Thus there is still room for considerable improvement in predictive methods but the limitations of inherent variability should always be borne in mind.
Time-dependent settlement 107. As pointed out by S ~ h m e r t m a n n , ~ is notcommonto consider the it ~ time-dependent settlement of sand. However, all the case records reported here which have measurements subsequent to completion of construction show timedependent settlement, as can be seen from Appendix 1. However, of the 27 cases given in Appendix 1, 14 have to be treated with caution owing to the presence of clay or silt layers beneath the loaded area or owing to the fact that the sand has been recently placed. Moreover, for a number of the remaining cases the periods over which time-dependentsettlementmeasurementshavebeenmeasured are relatively short anddo not give a clear pattern of behaviour. 108. Fortunately,a very complete set of settlementobservationshas been published by Bolenski4* for ten structures founded on sand in Warsaw (Cases 16
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1363

BURLAND AND BURBIDGE

to 25). Bolenski, who was not in fact an engineer, while working for the Polish Building Research Institute, and later, as a hobby, collected settlement data on structures over a period of about 20 years. In view of the length of the records (some over 16 years) they are invaluable for studying the time-dependent settlement of foundations on sand. 100. In Fig. 31 the measurements on four buildings in Warsaw (Cases to 25) 22 are plotted in terms pJpi against log time after completion construction (pi is of of t the settlement at completionof construction and pt is the settlement at time after construction). In spite of the length of the settlement records no clear pattern emerges. Cases 22 and 25 appear to show continuing settlement which is approximately linearwith the logarithmof time. However, Cases23 and 24 show stepwise behaviour with long periods of little settlement followed by sudden downward movement. 110. Bolenski also presented long case records of settlement of chimneys on sand and these show markedly more time-dependent behaviour than buildings. In Fig. 32 the results for three chimneys (Cases16 and 19) are plotted aspJpi against log time. Case19 shows linear settlementwith the logarithm of time but cases 16A and B again show stepwise behaviour. comparing Figs3 1 and 32 it can be seen By that the chimneys reach much larger values of pJpi than the buildings and also show a more rapid rate of settlement. The reasonfor this is thought to be due to the action of wind inducing fluctuating bearing pressures on the sand. Bolenski also reports some records settlement of turbine foundationswhich show similar of characteristics to the chimneys and which were undoubtedly subjected to fluctuating loads. The stepwise nature the timesettlement characteristic of some of the of structures may be due to perturbationsin loading (possibly minor seismic events) triggering grain slip within the mass of the sand. 1 1 1. If it is assumed that the settlement pI at times greater than 3 years after construction increases linearly log time then with
PI

= Pi

+ A P + Apt log(t/3) ~

100 1 .o

Time after completion of construction: days 500 1000


I ,

5000
I ,

1.3

Fig. 31. Timesettlement characteristics of four buildings in Warsaw observed by


Bolenski4'
1364
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENTOFFOUNDATIONSONSANDANDGRAVEL

where t is the time in years after completion of construction and is 3, A p , is the increase in settlement during the first 3 years and Apt is the increase in settlement per logcycle of time after 3 years. Dividing by pi
-= 1
Pi
PI

+ R , + R , log(t/3)

where R , is the propprtional increase in settlement during the first 3 years, and R, is the proportionalincrease in settlement per cycle of time after3 years. log 112. In Table 3 the values of R , and R, are listed for the relevantcases. There is no obvious correlation with soil type. Case 51 consists of eight large footings on gravel and it can be seen that the values of R , and R, have a wide range even on the same site.
Time after completion of construction: days
1
I
I
I
I 1

1000
1 1 1 1 I
I
1

10 000
l

I l l

\
Fig. 32. Time-settlement characteristics o three chimneys in Warsaw observed by f Bolenski4*
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1365

B U R L A N DA N DB U R B I D G E 113. A conservative interpretation of Table 3 leads to the following expression for buildings

_ - 1.3 + 0.2 log(t/3) p' Pi

and for chimneys


=

1.7

Pi

+ 0.8 log(t/3)

114. The latter expression may be appropriate for other foundations subject to fluctuating loads such as bridge abutments and silos. Another way of interpreting the above expressions is that after 30 years p' = 1 . 5 for ~ ~ buildings and p' = 2 . 5 ~ ~ for chimneys. 115. The above expressionshavebeenderived from limited data which are mainly restricted to grade 111 sand and gravel. Clearly, there is a need for more of post-construction settlement observations over a period years.

Discussion and conclusions 116. No attempt is made here tocomparethe results of thecorrelations derived in this Paper with predictions of other methods. To doso would require a case-by-case comparison. The confidence limits of the correlation summarized in Fig. 1 are large and are believed principally to reflect the limitations of the SPT, cone penetration test and other indirect methods for assessing the compressibility of granular materials. Most other methods of settlementprediction, although more analytically based, rely on such indirect methods of assessing compressibility and will therefore also have wide confidence limits. Recognition of this and of the variability inherent on anysite, is important in the design process.

Table 3. Time-dependent settlement offoundations on sand and gravel*


~~ ~

Case Grade

Principal soil type

I
22 23 24 25 51 83 84

(loading period in days)


751 334 355 894 880 822 488 1208 542 61

ti

R3

R,

Buildings 111 111 111

I1 I I1 I IV V
Chimneys

Fine/medium sand Clayey silty sand Silty fine sand Fine sand Gravel Sand/gravel Sand-gravel Medium sand Medium sand Fine silty sand

0.11
0.30 0.26 0.14 0.37t 0.13

0.23

0 0
0.21 0.17f 0.13 0.07 0.85 0.53 0.67

0.05
0.55 0.35 1.oo

19

* Note: pJpj = 1 + R , + R , log(@) with t in years-see t mean of range O.OM.62. 1Mean of range 0.02-0.4.
1366

equation (8).

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENTOFFOUNDATIONS

O N SANDANDGRAVEL

117. Schultze and carried out a similar correlation the one to described here usinga multi-correlation technique. Their data base was very much more limited than in this study and was largely dominated by cases with N equal to 20 and N equal to 30. Nevertheless itis of interest to compare theresults of their correlation with the one derived here. For a square footing resting on the surface of a deep sand layer Schultzeand Sherifsresults maybe expressed as
a, =

0.364

El

N0.87

which may be compared with equation (1 5)


a, =

1.706 X
N . 1 4

118. The two equations are compared in Fig. 33 for E = 3 m and E = 30 m and agree reasonably well for footings up to about 5 m wide. For larger foundations equation (15) gives significantly larger settlements. The difference in slope between the two relationships is probably due to thefact that in equation (18) is determined over a much greater depth (2E) than for equation (15). It should also be noted that Schultze and Sherif arrived at much larger shape and depthcorrection factors than were obtained in the present study. 119. The following conclusions can be drawn from the study described in this Paper. The results of a statistical analysis of over 200 case records of settlement on sands and gravels has resultedin a simple correlation between a, (the foundation subgrade compressibility), E and A, the average SPT blow count over the depth of influence. The standard error of a, varies from about (x/+)1.5 for N greater than 25 to (x/+)1.8 for N less than about 10. Thus the accuracy of the correlation is not particularly high but, in view of the small settlements that are usually involved, it is good enough for most practical purposes. However, it is recommended that other widely accepted methods are also used as acheck. (b) If more precise predictions of settlement on granular soils are required they must be based on direct methods of determining in situ compressibility and not on indirect methods such as the cone and SPT. It is hoped that the results of this study will serve to stimulate the development of such methods while at thesametimeprovidingasimple approach for routine design purposes. (c) The available experimental evidence suggests that the relative depth of influence ( z / E ) ,beneath a foundationdecreases significantly as the value of B increases. There is an urgent need for field measurements of settlement at various depths beneath loaded areas to establish not only the depth of influence but also the in situ compressibility of granular soils with depth. ( d ) For normally consolidated sands the relationship between the effective foundation pressure and settlement is approximately linear up to about one-third of the bearing capacity. The effect of overconsolidation and loading at the base of excavations is to reduce the foundation subgrade compressibility for bearing pressures below the effective preconsolidation pressure.
(a)

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1367

BURLANDANDBURBIDGE

\
\

t
1-

Equation (1 5)

- - - Schultze and
Sherif35

I 1
I l l 1 1

10

100

Fig. 33. Comparison between the correlation between a,and N derived in the Paper with that obtained by Schultze Sherif3 and ( e ) Using the complete data set as a basis for comparison it appears that for DIE 3 the depth of founding and the level of the water table do not have a significant influence on a , . However, the effect of increasing L/B is to increase a, by up to about 50%. cf) Ithas been shownthatthe Terzaghi and Peck recommendations for correcting the SPT blow count for submerged dense finesands andsilty sands give an improved correlation. Similarly, an analysis of all the case records involving gravelsand sandygravels indicates that the SPT blow count should be increased by a factor of about 1.25 for the purpose of assessing the compressibility. ( g ) An analysis of the results from a number of sites where the settlement of two or more foundations hasbeen observed indicates that the inherent
1368
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT OF FOUNDATIONS ON SAND AND GRAVEL variability of the groundfrequently gives rise to settlements which differ from the mean by a factor of 1.5 or more. This finding gives support to Terzaghi's4 statement that the settlements of uniformly loaded areas on natural sand strata vary erratically. (h) The field measurements show that time-dependent settlements take place on sands and gravels. For static loads this additional settlement is fairly small and may reach 50% of the end of construction settlement after about 30 years. For fluctuating loads the time-dependent settlements are muchlarger.

Appendix 1. Details of case records and measurements 120. Table 4 gives details of the various case records referred in the Paper. For ease of to reference the numbering of the records is the same as that used by B~rbidge,~ although for various reasons not all his cases have been used.An explanation of some of the columns is given below: Column 3 R, mean of N over depth of influence values the (see 51-63). Column 4 grade, Table see 1. Column 5 method, standard SPT, penetration C, cone test; static penetration test; Oed, oedometer; P, plate loading test. Columns 6 8 B = breadth; L = length; D = depth of founding. Column 9 H , , depth of table water beneath founding level. H , , thickness of sand or gravel stratum. Column 10 Columns 11-13 qsrossis gross the bearing pressure at founding level; qbe, is net the effective bearing pressureat founding level;A&, is the known change in Qbd p i , observed average settlement the end of construction.* at Column 14 Api , observed increaseaverage in settlement to due Aqbe,. Column 15 Column 17 ApJAq', obtained columns 13 and 15 or from of from slope pressuresettlement curve from equation(1 1). or Column 18 t i , length of construction or loading period. Columns 19 and 20 total final settlement p, and corresponding time t, since start ofconstruction or loading.
'

*When only the edge settlement of a tank has been measured a factor of been applied 1.1 has for tanks up 40 m in diameter and a factor of 1.2 for diameters above to 40 m.5

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1369

BURLANDANDBURBIDGE

Table 4. Details of case records

(g

(2) Principal soil type

14) Grade

(5) Method

T
B

I 2 3lA 318 61P 6lR 71A 71p 8lB 81P 9 1 ~ 96 l21A


13/A

D
0 I .2 I I 2.8 3.6 2.85 2.85 2.5 2.0 3.0 3.0 2.6
0

T
1.5

(11)

(1.2) (I! Foundatlon press.

kN/m'
q,roll

_ .

4.d

I3jB I 3c j 14
1, 5"

1517 15/17 l5/8-l8 I6/A I6jB 19 20;A 20/B 21 22 23 24 25 27 29 3011-7 3018

Fine 10 coarse sand Fine sand Sand Sand Sand Sand Sand Sand Silty sand Silty sand Sand Sand Silty sand Sdty fine sand Silty fine sand Silty fine sand Fme sand Sand Sand Sand Sand Medium sand Medium sand Fine silty sand Finelmedium sand Finetmedium sand Fine/medium sand Finelmedium sand Clayey silty sand Silty fine sand Fme sand Gravelly sand Finelmedium sand Finelrnedium sand Fine/medlum sand Fine/medium sand Finelmedium sand Fine/medium sand Fine/medium sand Fine/mediurn sand Finelmedium sand Finclmedium sand Finelmcdium sand Finelmedium sand Fine/medlum sand FlncJmedium sand Fme/mcdiurn sand Flneimedlum sand Fine/medium sand F!ne/medium sand Flne sand Fme sand Fine sand Flnc/mcd!um sand Sandigravel Medium sand Medium sand Sand Gravelly sand Medium sand Medlum sand Medium sand Medlum sand Medium sand

28 17 8
8

111

SPT
SPT C C C C C C C C C C C C C C C C C C C Oed Oed OCd Ocd Oed OCd Oed Od C Oed OCd C
-

IV VI VI
111 111 111 111 V V

30 30 35 38 10 1 0 60 60 17
15 15

7.3 I .6 I -6 - 1.5 - 2.3 - 1.6 - 1.6


0.5 0
-

150

I1
11

I5
7 6 6 6 6
~

IV V V V VI VI VI VI VI 111 111 111 111


111 111 111 111 111 111 11 11

0
0

0 I .o I .o I .o 1-2.6 4.0 3.7 6 22 2.3 4.0 2.7


3-0-5.0

I .o I .o I .o 0 2 2 I .6 2 - 1.0 - 0.7

- 0.5

193 I30 52 52 I62 162 93 140 93 147 284 284 121


80

60 78 I64 74

64
196 196 245 151
151

0.5
0.7 0.6 2.2 - 1.7 3.2 0.7 2.6 - 3.7 0 4 4 4 4 4
4

47
-

2.2 2-8 5.2 0


1.5

245 140 I34 I47 I80 417

75 70-86 I l8 I26 I43 I l4 I12 I77 77 66 I10 I32 23 I 247 139-290 97-225 102-161 113-166 97-199 I39 161 I50 113 I77 215 I66 I56 I54 24 I

20
20 20 20 20 20 20 20 20 20 21 22
-

IV IV IV IV

SPT SFT S PT
S PT

1-7
1.8

30p- I5
30/16-18 30119-30 30/3 l -32 3013343

2.0 2. I 2.3 2-5 2.6 3.0 32 3.4 3.5 0.6 0 0


0

IV
IV
IV

SPT

SFT SPT
SPT SPT SPT SPT SPT P S PT S PT S PT S PT S PT
P
P

30144
30145 30146 30/4748 30/49-50

3o/c
3I/A 3118 3 I/C 3 I/D 32 33lA 3318 33ic 34 35!A 35/B 36 37 38 39/0 39JC 39iP 39/P

21 19 17 20 60
-

IV IV IV IV IV I1 IV IV

4 4 4 4 4 4
-~

IV I1
IV IV IV 111 V V

0 9.7
5.0 45

S 6 LL6 0-6 0-6 - 7.2 0


0.5 10

II II 25

IV
I V

60
I2 21 21 16 16

P P 5 PT S PT FPT SPT SPT SPT SPT j PT


$PT

4-0
0

289 294 147 186

0 01 . 20.9
1.5 5.0
5.0 0

06 06 3.7 -9 2
0 10.0 10-0 10.0

209 70 I l8 490 I82 I58 I80

500

IV IV
IV

254 245

IV

1 0 0

1370

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

S E T T L E M E N T OF F O U N D A T I O N S O N S A N D A N D G R A V E L

(14)
P,*

(18)
I, ,

Remarks

mm
18 22 20 35 10.5 11.0 6.5

days 9.3 16.9 Centre settlement Edge settlement X 1 . 1

38.5
67.3 6. 5 S. I 7.0 2.1 8.6 8.2 0.35

loo0
loo0

m
1800

30 . 8
12 I 3 6-12 52 80 7 143 74

I800 500
500 1100 I 100

0.88
5.0-16.5 77.5 I33 9 87.2 100 78-141 123 112-173

Bridge piers founded in base of 3-8 m deep cuttmg Bridge piers founded in base of 3.8 m deep cuttlng

300
400
-

I 120
-

50-90
75 84121 7.2
10.1

5.4 17.3 14.5 21.1 10.1 3.3 8-6 12.5 45


81 (mean) 12.2 10-4 (mean) 75

1208 542 61 522 500


1400

751 334

355
894 880 3.4 mean) 4-9 4.6 mean) 5. I mean) 4.9 mean) 4.54.9 6. I 6.8 6.3 9.7 3.6-5.1 4.3-5.0 0.8 40.2 57.7 w.9 54 6 1460 I460 I460 I460 I460 I460 I460 1460 I460 I460
1460

Measured edge settlement 130mm Poorly graded sand, some thin layers of clay and silt Poorly graded sand, some thin layers of clay and silt Poorly graded sand, some thin layers of clay and silt Poorly graded sand, some thin layers of clay and silt ApplAq' (corrected for 0.b. pressure) AplAq' (corrected for 0.b. pressure) ApplAq' (corrected for 0.b. pressure) ApplAq' (corrected for 0.b. pressure), clay underlying sand ApjAq' (corrected for 0.b. pressure), clay underlying sand AplAq' (corrected for 0.b. pressure), clay underlying sand ApplAq' (corrected for 0.b. pressure) ApplAq' (corrected for 0.b. pressure) AplAq'(correc1ed for 0.b. pressure), clayey silly sand below 12.1 m WAq' ApjAq' from slope of pressure-settlement curve. Stiff to very stiff sandy clay below 60m Hard clay below 26m Range of p , = 6 . 6 1 I . 2 m m p/q' = 2 . 8 4 7 ;

(mean) 7.I
(mean) 5- 1-8. I 8.7 9.4 10.2 14.5 4-1-5.1 7.6-8.5 I .7 80 90
100

1460
I
~

131 65 4s 24 25 14 232 I96 39 4 15

2821
1500

I820 I820 2.9 I28 I24 21.9 52 30 761

Occasional clay bands Occasional clay bands Occasional clay bands Occasional clay bands Stiff silly clays below 21.3 m. ApplAq' from slope of q' v. p curve Clay ll-14m. mar1 below 14m. AplAq' corrected for 0.b. pressure clay below 1 I m. AplAq' corrected for 0.b. pressure ( p , = 3 mm at depth = 14.5 m) Clay below 12.5m. AplAq' corrected for 0.b. pressure.
9.6-1 1 . 1 m limestone; l I 1-16.1 m very stiffsilty clay 9 . 6 1 1 . 1 m Imestone: 11.1-16.1 m very st~ffsilty clay 7.&9.1 m stiBclay. Below 9.1 m Gneiss bedrock hpJAq' from pressure-seltlemenl curve

11.0

(mean) 99 (mean)

4.5 man) 4-0 nean)

I I

Boreholes 3 and 6 used for SPT Boreholes 3 and 6 used for SPT I2 plate tests: p = 6.3-20.5mm: plq' = 2.68.4
i plate tests: p = 7.0-14-Omm: pplq' = 2 9-57

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1371

Table 4. Continued
(1) (2)

( >)

(4)

(5)

Case no.

Princlpal soil tYF

Grade

Method

40 41 43/A 4318

T
B

T
Ysru... -

q"d -

Aq' 55 7

44/PI
44/P2 44/MI 44/M 3
45/A 45jB 45JC 47/A 4718 47/c 48 49 5O/A 50/B 51/A to 52lC 52/H 52/A3 52/D3 52/J 53 56jB 5610 57 58/A 5818 %/C 59/A 59/B 59jC 59/D 59lE

Fine sand Medium sand Sand Sand Medium sand Medium sand Medium sand Medium sand Fine to coarse sand Fine to coarse sand Fine to coarse sand Sand with gravel Sand with gravel Sand wlth gravel Medium sand Medium sand Silty fine sand Silty fine sand Gravel Sandlgravel Sand/gravcl Sandlgravcl Sand/gravel Sand/graveI Sllty sand Fine sand Fine sand Fine sand Sandy gravel Sandy gravel Sandy gravel Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine 10 medium sand Fine to medium sand Fine to medium sand Fine to medium sand Fine sand Compacted moist sand Compacted moist sand Compacted moist sand Sand and loam Fine sand Sandlgravel Fine sand Fine sand Coarse sand Coarse sand Fine to medium sand

12 22
-

35 50 28 45 18

V IV 11 11
111 11 111 11

220 l44 144 I50 I50

S PT Oed Oed

S PT SPT SPT S PT
CISPT
c/sPT CjSPT 193 193 193

I50 I50

IV

I8 I8
29 26 18 30 6 20 20 37
50 50 30 20 20 I2
-

IV IV
111 111 V 111

S PT SPT SPT SPT


C

215 215 215


270 75
41

VI IV IV
111 11 11 111 IV

SPT SPT SPT


SPT SPT SPT S PT SPT
C

33 SI8 300 300 300 300 300 91 171 171 I23

6 13 13 13 35 25 25 25 35 35 25 25 25 25 25

IV V V V VI
V V V Ill IV

Oed Oed
C

SPT SPT SPT


C C C
I67

78 77 77

230
230 284 195

IV IV
111 111

C
C C

59/F
59jG 59lH 5911 5911 59/K

RI
226 2 50
250

1v
IV
IV IV
IV

!94 206 !94 104

59/M
59/N 59/0

40
40

111 111

40

111
111 111 111

!04
104
104

59/P
W Q

4 0
40
40

59/R
W A 6018 W C 61/A 61lB 61/CI 61jC2 63 W C 65 66/A 66/B 69lA 69/B 70

I04

30
30

IV
IV

I65 I48 96

25 34
15 15

IV
111

!20 64
l39
>

11

$5

I1 IV 111 V V Delivered VI VI
2 3 25 I2 I2

184
200 140
'20 68 88 44 41 37 I60

IPT

, by ICEVirtualLibrary.com to: , IP: 86.154.85.54 > V On: Thu, 15 Sep 2011 11:00:29

;PT ;PT ;PT

1372

(18)
l,,

(20)
lr
3

days -

days -

Remarks Firm silly clay below 26.7111. Grade determmed over (:/BJl = I Stiff clay below 12.5m Treated by vibroflotation to 7.5m depth. Limestone below l5 2 m Treated by vibroflotation to 7.5m depth. Limestone below 15.2m Compacted by vibroflotation

600

488 580 488 I


1 1

853 853 853

Compacted by vibroflotatlon AplAq' from slope of q' v. p curve. Firm clay below 19.2 m AplAq' from slope of q' v. p curve. Flrm clay below 19.2 m AplAq' lrom slope 01 q' v. p curve. Flrm clay below 19.2 m

7 30 15 207 207 880 l I


1 1

Ap/Aq' corrected for 0.b. pressure Moraine below 8.7 m Low values of SPT below W.T. due to boiling

1838

AplAq' from slope of q v. p curve. ApIAq'

1.542

I Sandstone below 7.1 m

loo0
125 125 I25 Pressure Increased to 91 kNlm' without lurther settlement Pressure increased to 107kN/m2 without further settlement ApIAq' corrected for o.b. pressure

AplAq' corrected for 0.b. pressure

304

AplAq' from slope of pressure-settlement curve AplAq' from slope of pressuresettlement curve Model tests Model tests Model tests Model tests 2020 257 257 Model tests Settlement measured over depth of sand layer AplAq' from slope of pressure-settlement curve StiR clay bclow 5.3m Stiff clay below 5.3 m

366
I 1

I
1
1

1220 520 l 47 47 I

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1373

B U R L A N D AND B U R B I D G E

Table 4. Continued
(1)

(2)
Prlnclpal soil type

Case no.

) ! (

Grade

(5) (4) Method B

(7) (8) (6) Dimensions, m L D

(9)

4
m

(10) H,

(11)

(12) (13) Foundation press, k N/m'


*:er

4.ro..

4'

71 72 73 74 75 76 77 78.4 79/A 7918 80 8ljC 81/D 8I/E 81/F B3 84 85 86 B7 89/A 91 92/A 92iB 92iC 92fD 92/E 93/A 93/B 94/A 94/B 95iA 9SjB 96/A 96iB 96JC 96/D 97/P 97/E 98:A 98/B
I0 0

Fine sand Fine sand Fme/medlum sand Fine/medium sand Fine sand Fine/coarse sand Medlum sand Silly fine sand sand 788 Fine sand/silt Sandy gravel Flne sand Fme sand Fine sand 7 Fine sand Sandigravel Sand'gravel Gravelisand Sandigravel Sandigravel Sandigravel Sand Sand Sand Sand Sand Sand Sand 5 Sand 5 Flneicoarse sand Sandigravel Silty fine sand Sllty fine sand Silly fine sand Silly fine sand Medlum Qnd Medium sand Fme sand Flne sand Fine sand Fine sand Fine medlum sand

v1
V I V
II

V I

20
M)

P P P
.3

1.5 3.0 2.8 4.8 22.5 10.04 20.0

IV 111

SPT SPT
C C

5 V I 5 SdtyV fine I 5 V I 5 FineVI sandisill 36 111 V I 5 6 VI

c
C

c
C

VI
8

VI

IV 20 SPT V SPT 14 V 10 SPT 26 111 SPT 111 34 SPT 111 37 SPT 27 111 SPT 50 2.4II 2.4 2.1SPT II SPT 50 11 50 SPT

27.5 41.2 41.2 0.9 0.9 0.9 0-9 I2 1.2 1.2 17.6 84 0 16.0 43.0 20.5q5 14.5 14.5 3.5 33.06 10.7 24.44

2.3 Strip 3.5 3.5 3.0 3.0 56 65.0 10.0 1.5 20.0 3.0 20.0 20.03.0 20.0 27-S 0 27.5 0
10.0

Shallow
-

Shallow
~-

~12.8 $B ~B S I5
2 30

78 187 I57 216 I96 245-295 240 85 85 130 I76 158 I33 I l3 199 268 240 228 173 255 216 293 I20 584 697 575 584 347 35 45 386 383 17.0 220 220 I20
I10 I10

-2.5
10

-I -I
0 0 -5.5 Deep Deep Deep Deep -2.2 -I 8 25
1.5

12.0 32
45

0.3 0.9 0.2 1.2 0.9 10.7 7.3 3.5 5.3 2.6 1.0
0

--

40 40 13

34

-2.5
~

237 223 226 21 5 8.2 5. l


.-

I1 I1 50 3.0
50

V 1
VI
18

IV
II

50
~

SPT SPT SPT SPT SPT SPT


Oed Oed Oed Oed 2 3 Oed Oed

I V IV V V
V
V

7 7 4 4
-

V I
VI

SPT SPT

VI VI
IV

c
Oed

2.1 2-1 I 5 1.5 2.8 I8 2-1 2.4 3-0 4.1 2-1 8.2 61-0 I 8 Strip 27 30.2 308 3.8 7.0 Strip 30 2-5 37 2.5 3.0 2.3 2-5 3-4 2.7 2.5 23 3.4 1.5 2.8 1-5 33 6.0 6.0 0 20-0 0 20.0 14.0 I0 28 14.5 3.3 1.0 18.0 26.0 1-5-3.5

__
~

6.5 22 3 6.0 0.4 3-7 10.0 0.4 10.0 I 6 13.0 16 130 2-6 13.0 2.6 13.0 0-9 18.0 0.9 18.0
-

v0
190

10

10.0

10.0

145 I42 99 140

> 18

1374

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

S E T T L E M E N T O F F O U N D A T I O N S O N S A N DA N DG R A V E L

(18)
l,.

days
~

Remarks AplAq' from slope of pressure-settlement curve AplAq' from pressure-settlement curve ApplAq' from pressure-settlement curve

9.5 4.9 16.2


5.8 5-0 15-18

15.2 26.2
10-3 4.7

7-0

26 . 10-5-19.7 2.9 231 I65 107-283 5W796


11.8

Il 6
81

139-368 993-1401
10

700 696 854 752 12 40


I I I I

AplAq' corrected for 0.b. pressure Limestone below 12m. Sand compacted to 5 m depth ApjAq' Corrected for 0.h. pressure AplAq' corrected for o.h. pressure

219487 12.35 2515

100 160

ApjAq' from pressure-settlement curve

76 6.4
13-0

I? 7 21.2 179
X0

5.7 57 6.5 4.7 11.4 7.5 46 6. I 20.5 3.7 11.9 0.7 0.33 0 47 0.79 0.52 54.3 55 6 25.9 1.6 x.9 7.7
10.0

822 488 195 532 7


I
1

23.3 18.6

I462
I l58

l5 5

434 10.9 1 3 4 4.4 2.3 27 46


18

AplAq' from pressurC-seltlemenl curve AplAq' from pressure-settlement curve Settlements suggest grade I I . Suspect SPT results ApfAq' from pressure-settlement curve. Stiff clay below 21.5rn ApjAq' from pressure-settlement curve. Clayey sllt 8-2-15.7 m

I I

1 9 25 91 6 4.x
15 10

350

I2
IX

164
15.5

1 7
17

18.9 59 83 68 37
20-8

790 I594 790 546


500 500

18

I020

22 24

I594 1594

74
I20

AplAq' corrected for 0.b. pressure AplAq' corrected for 0.h. pressure AplAq' from pressuresettlement curve. Hydraulic sand fill AplAq' from pressure-settlement curve. Hydraulic sand fill Hydraullc sand fill Hydraulic sand fill Hydraulic sand fill Hydraulic sand fill ApjAq' from pressure settlement curve Overconsolidaled clay below IOm Overconsol~datedclay below IOm A d A a from Dressure settlement curve. Comoacted sand fill

97 37
27 2

39I

36.9

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1375

BURLANDANDBURBIDGE

Appendix 2. Case recordsstructures, geology and references Reference Structure Bearing strata Case Recent and Pleistocene sands Steel tank Baker49 1 Baker4 Recent and Pleistocene sand 2 Steel tank de BeerSo Recent and Eocene sands Bridge 3 de Beerso Recent alluvial/Eocene sands Bridge 6 de BeerSo Eocene sand Bridge 7 de BeerSo Eocene sand Bridge 8 Bridge de Beer5 Eocene sand 9 Bridge de Beer Recent river/Eocene sands 12 Embankments Recent river sand Bjerrum 13 Bjerrum Steel tank Recent river sand 14 Bogdanovic et al. 14 storey Quaternary river sand 15 building Bolenski4 Recent river sand 16 lOOm high chimneys Chimneys Bolenski4 Pleistocene lacustrine sands 19 Recent river sands 20 11 storey Bolenski4 buildings Bolenski4 Pleistocene river sands 21 22 storey building 11 storey Bolenski4 Recent river sands 22 building Bolenski4 Pleistocene river sands 23 9 storey building Pleistocene lake sands 24 11 storey Bolenski4 building 12 storey Bolenski4 Pleistocene river sands 25 buildings Nuclear Breth and C h a m b ~ s s e ~ ~ Pleistocene river sands 27 reactor 29 Concrete tank Pleistocene sands Clausen et Steel mill DAppolonia et aL8 Recent beachldune sands 30 Steel tank Davisson and Salley5 Recent river sand and 31 gravel Nuclear Dunn41 Recent beach sand 32 reactor Chimneys Egorov and PopovaS6 Quaternary sand 33 120/250 m Pleistocene river sand/gravel 34 Bjerrum and Eggestad5 Silo Farrent Steel tanks Quaternary beach sand 35 Silo Quaternary beach sand 36 FarrentSE Fischer et aLS9 37 Nuclear Paleocene/Cretaceous sand reactor Quaternary beachldune sand 38 Building Frost6 Recent river sand Steel mill Garga and Q ~ i n ~ ~ 39 complex Recent marine sand Geilly et 40 22 storey building 41 10 storey Recent dune sand Glick6 building Compacted Cainozoic sand 43 Greenwood and T a i P Steel tanks 44 Greenwood and T a i P Test footings Pleistocene river/dune sand Recent beach and lagoon sands Grimes and C a n t l a ~ ~ ~ 20 storey 45 buildings Bazaraa Plate tests 47 Pleistocene river sand
1376
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT O F FOUNDATIONS ON SAND AND GRAVEL


48 49 50 51 52 53 56 57 58 59 60 61 63
64 65 66

13 storey buildings Embankment Concrete tank 12 storey buildings Plate tests

Tertiary sand Hydraulic sand fill Quaternary marine sands Pleistocene river sand/ gravel Quaternary river sand/ gravel Recent river/Eocene sand Recent river/Eocene sands Recent coastal sands Pleistocene river gravel Pleistocene river sand Pleistocene river sand Pleistocene river sand Pleistocene river/ Jurassic sand Recent marine sand
-

K~rngold~~ Lagging and Eresund66 Langfelder and J ~ h n s t o n ~ ~ Levy and Morton6*


Levy and Morton6'

Bridge Bridge 10 storey building Factory building Various 18 storey buildings Test footings 533 m tower Silos Plate test 2 storey building Buildings 25 storey buildings Concrete tank
10 storey buildings Embankments Building Plate tests

Marivoet6' Marivoet6' Martins et al.70 Meigh and NixonZ6 Muhs and Kah17' Muhs and Weiss7' Muhs7' Nikitin et al." N~nveiler~~ Oweis7' Bazaraa" pre~s~~.~' Rios and Silva7' Ronan7' Sanderat et
al.''

Pleistocene river sand Quaternary river sand Tertiary sand Recent aeolian sand (compacted) Quaternary river sands

6%75 76
~~

77 78 79 80 81 83 84 85 86 87 89 91 92 93 94
95

Quaternary river sands Recent river gravel Recent coastal/river sand Pleistocene river/ 30 storey Tertiary sand building Pleistocene river/ 20 storey Tertiary sand building 120 m chimney Recent/Pleistocene/ Tertiary sand 120 m chimney Pleistocene river sand/ gravel Pleistocene river/ Nuclear Tertiary sand reactor Quaternary deposits Buildings Quaternary coastal sand Steel tank Quaternary coastal sand Machine hall Recent dune sand 3 storey buildings Quaternary deposit 5 storey buildings Recent river sand 1 storey frame building

Sanderat et al." Sanglerat et al."' S~hmertrnann'~ Sch~ltze~~.~~ Sch~ltze~~.~~ S~hultze~~.~~ S~hultze~~ S~hultze~~ Schultze and Sherif" Thorne" Thorne'' Tomlinsona2 Tschebotario@' V~tjakov'~

Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1377

BURLAND AND BURBIDGE


1 storey Recent sand/ river frame building Footingisand Quaternary 97 marine embankment Bridge 98 100 storey 9 sands building aeolian

96

Votjakov Webba6 Recent sand Pleistocene river/ Wennerstrand Zakharenkov

fill

References 1.SUTHERLANDH. Granularmaterials(reviewpaper), B. Proc.ConJ the Settlement of Structures, Cambridge,Pentech Press, 1974,473-499. 2. SIMONS, N. E. MENZIES K. A short course in foundation engineering.IPC Science and B. and Technology Press, 1976. 3. NIXON I.K. Standard penetration test state-of-the-art-report. 2nd European Symposium on Penetration Testing, Amsterdam, 24-27 May 1982, Balkema. 4. TERZAGHI Discussion on paper by Skempton and MacDonald: Theallowable settleK. ments of buildings.Proc. Znstn Civ. Engrs, 1956, Part 3,5775. 5. BURBIDGE,M. case study reviewof settlementson granularsoil. MSc/Dissertation, A Imperial College, Universityof London, 1982. J. State of ArtReport, 6.BURLAND B. et al.Behaviouroffoundationsandstructures: session 2. Proc. 9th int. ConJ Soil Mech. Foundation Engng,Tokyo, 1977,2,495-546. 7. TERZAGHI and PECK B. Soilmechanicsinengineering practice, lst/2nd ed. John K. R. Wiley, New York, 1948/1967. 8. DAPPOLONIA D. J. et al. Settlement of spread footings in sand. J . Soil Mech. Fdns. Diu. Am. Soc. Civ. Engrs, 1968,94, SM3,735-760. 9. MEYERHOF, G.Shallow foundations,J . Soil Mech. Fdns Diu. Am. Soc. Ciu. Engrs, 1965, G. 91, SM2,21-31. 10. CLAYTON R. I. et al. Dynamic penetration resistance and the prediction of the comC. pressibility of a fine-grained sand-a laboratory study. Ghotechnique, 1985, 35, No. 1, 19-31. 11. BRITISH STANDARDS INSTITUTION. Methods oftests for for civil engineering purposes: soils Test 18, Determination of the penetration resistance using the split barrel sampler.British Standards Institution, London, 1975, 1377,103-104. BS 12. AMERICAN SOCIETY FOR TESTINGAND MATERIALS. 158667 Standard method for penetration test and split barrel sampling of soil. of Standards, ASTM, Philadelphia, Book 1967,part 19. G. F. uses and abuses J . Soil Mech. Fdns Diu. 13. FLETCHER A. Standard penetration test: its Am. Soc. Civ. Engrs, 1965.91, SM4,67-75. J. H. 14. SCHMERTMANN 1979, Statics of SPT. J . Geotech. Engng Diu. Am. Soc. Civ. Engrs, 1979,105, GT5,655-670. 15. INTERNATIONALSOCIETY FOR SOIL MECHANICS FOUNDATION ENGINEERING.of AND Report theSub-Committeeon the Penetration Test for Use inEurope. ISSMFE,1977, 3, 95-120. 16. HOLUBEC, and DAPPOLONIA Effect of partical shape on the engineering properties I. E. of granularsoils. Proc.Symp.Eval. Rel. Dens. AmericanSocietyforTestingand Materials, Philadelphia, 1972, Special Technical Publication 523,314-318. 17. GIBBSH. J. and HOLTZ W.G. Research on determing the density of sands by spoon penetration testing. proc. 4th Znt. Con& Soil Mech. Foundation Engng, London, 1957, 1, 35-39. D. E. 18. DAPPOLONIA J. and DAPFQLONIA Useof the SPT to estimatesettlement of footings on sand. Symp. on Foundations onInterbedded Sands, CSIRO, Perth, 1970. 19. SCHULTZE E. MENZENBACHJ. Standard penetration test and compressibility of and K. soils. Proc. 5th Znt ConJ Soil Mech. Foundation Engng, Paris, 1961,1,52?. 20. BAZARAA A. R. S. S. Use of the standard penetration test for estimating settlements of shallowfoundations on sand. PhD thesis, Universityof Illinois, Urbana, 1967.
1378
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENT OF FOUNDATIONS O N SANDANDGRAVEL


21. SCHULTZE E. MELZER K. The determination of the density and the modulus of and J. compressibilityofnon-cohesivesoilbysoundings. proc. 6th Int. Conf Soil Mech. Foundation Engng, Montreal,1965,1,354-358. S. 22. THORBURN Tentativecorrectionchartforthestandardpenetrationtestinnoncohesive soils. Ciu. Engng Publ. W k s Reu., 1963, June, 752-753. 0. 23. DARAMOLA The inyuence of stress history on the deformation of sand. PhD thesis, 1978. Imperial College, University of London, Proc. 24. TERZAGHI K. Recent trends in subsoil exploration. 7th ConJ Soil Mech. Foundation Engng, Texas, 1947,l-15. G. J . Soil 25. MEYERHOF G. Penetration tests and bearing capacity of cohesionless soils. Mech. FdnsDiu. Am. Soc. Ciu. Engrs, 1956,82, SMI, Paper 866,l-19. and I. Proc. 5th Int. 26. MEIGH A. C. NIXONK. Comparison of in-situ tests for granular soils Conf Soil Mech. Foundation Engng, Paris, 1969,1,499-507. 27. RODINS. Experienceswithpenetrometerswithparticularreferencetothestandard 1961, 1, penetrationtest. Proc. 5th Int. Conf Soil Mech.FoundationEngng,Paris, 517-521. 28. SUTHERLANDB. The use of in-situ tests to estimate the allowable bearing pressure of H. cohesionless soils. Struct. Engr, 1963,41,85-92. 29. THORBURN, Discussion on paper by Thorburn and Macvicar. Proc. Conf on BehauS. iour o Piles, Institution of Civil Engineers, London, f 1971,53-54. 30. VARGASM. Foundations ofla11 buildings on sand in Sao Paulo, Brazil. Proc. 5th Int. Con6 Soil Mech. Foundation Engng, Paris, 1961,1,841-843. 31. SANGLERATThe penetrometer and soil exploration. Elsevier, Amsterdam,1972. A. 32. MEIGH, A. C. Discussionon: Settlements on granular soils, observations, and comparisonwithprediction. Proc.European Conf Soil Wiesbaden, 1963, Mech.Foundation Engng, 2, ?l-72. 33. PARRY H. G. A direct method of estimatine settlements in sand from R. SPT values. Proc. Symp. on the Interaction of Structure andFoundation,Birmingham, 1971, Midland Soil Mechanicsand Engineering Society,29-37. Fdns Diu. Am. Soc. Ciu. Engrs, 1970, W, SM3,lOll-1043. 35. SCHULTZE and SHERIF Prediction of settlements from evaluated settlement observaE. G. tions for sand. Proc. 8th Int. Conf Soil Mech. Foundation Engng, Moscow, 1973, 1.3, 225230. 36. MORGAN R. and GERRARDM. Behaviour of sands under surface loads. Soil Mech. J. C. J. Fdns Diu. Am. Soc.Ciu. Engrs, 1971,97, SM6. 37. BRETH et al. Settlement of Shallow Footings (Zur Setzung von Flachgrundungen). H., Vort. Baugrundtag. Dt. Ges. Erd-u. Grundb. 1976, Sept.,Nurnberg, F.R.Germany, 603-626. 38. SHVETSV.B. and KULCHITSKII B. Experimental investigation of the depth of comG. pressed soil foundation stratum under a plate. Osnou. fund. mekh. Grunt., 1970, JanFeb, No. 1, 1C~12. N. osadok fundameta televizionnoj basni v Ostakino 39. NIKITIN V. Issledovanie (Investigation of foundation settlement of a television tower at Ostakino). Osnou. Fund. Mekh. Grunt., 1970, Mar.-Apr., No. 2,32-38. 40. BRETH and CHAMBOSSE(1974) Settlement distribution in the subsoil underneath a H. G. nuclear reactor. 1974, Conf Settlement of Structures, Cambridge, Pentech Press, 1974, 1&13. 41. DUNN C. Settlement of a large raft foundation on sand. S. Con$ Settlement of Structures, Cambridge, Pentech Press, 1974,1421, H. H. 42. KRIEGEL J. and WEISNER H. (1973):Problems of stress-strain conditions in subsoil. Proc. 8th Int.conf soil Mech. FoundationEngineering, Moscow, 1973,1.3,133-141. 43. GARGA K. and QUIN J. An investigation on settlements of direct foundations on V. T. sand. Conf Settlement of Structures, Cambridge,1974,22-35. E. fur in und 44. SCHULTZE Beispiele Setzungsbeobachtungen bindigen nichtbindigen Boden. (Examples of settlement observations in cohesive and non-cohesiveProc. soils). 3rd European Conf Soil Mech. Foundation Engng, Wiesbaden, 1963,1,143-162.
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1379

BURLANDANDBURBIDGE
45. DAPFQLONIA D. et al. Discussion on settlement of spread footings on sand. J . Soil J., Mech. Fdns Diu. Am. Soc. Ciu. Engrs, 1970,960, SM2,754-761. G. 46. SHERIF Setzungsmessungen an Industrie und Hochbauten und ihre Auswertung. Bull Inst. Soil Mech. Foundation Engng, Tech. Uniu., Aachen, 1973. 47. SCHULTZE E. Probleme bei der Auswertung von Setzungsmessunger (Problems in the analysis of settlement measurements).Vortrage Der Baugrundtagung, Essen, 1962,343381. 48. BOLENSKI M. Osiadania now0 wznoszonych budowliW zaleznosci of podloza grunto-

wego:Wyniki20-letnichBadan W InstytucieTechnikiBudowlanej,(Settlement of constructions newly erected and typeof subsoil: The results 20 years studies carried of out in the Building Research Institute). Prace Instytutu Techniki Budowlanej, Warszawa, 1973. N. 49. BAKER C. (1965) Discussion. Shallow foundations. J . Soil Mech. Fdns Diu. Am. Soc.
Ciu. Engrs, 1965,91, SM6, 119-121.
50. DE BEERE.Settlementrecordsofbridgesfoundedonsand.

Proc. 2nd Int. Con$ Soil

Mech. Foundation Engng, Rotterdam,1948,2, 1 1 1-121.

51. BJERRUM L. Development of an industrial site on a silty sand deposit. Authors notes from a series of two lectures presentedMassachusetts Institute ofTechnology,1962. at to large variations in live load. 52. BJERRUM L. Secondary settlements of structures subjected Symp. on Rheology and Soil Mechanics, Grenoble, Springer-Verlag, Berlin, 1964, 4 6 G 471. et al. 53. BOGDANOVIC L. Comparison of the calculated and measured settlements of buildings in New Belgrade. Proc. European Con$ Soil Mech. Foundation Enyng, Wiesbaden, 1963,1,205-2 13. 54. CLAUSEN J. F. et al. (1975) ObservedbehaviouroftheEkofiskoilstoragetank C. foundation. 7th Ann.Offshore Technol. Con/., Houston, Texas, May1975,3, OTC 2373, 399413. 55. DAVISSON, M. and SALLEYR. Settlement histories of four large tanks on sand. Proc. T. J. Conf: onPerformance of Earth and Earth-supported Structures, Purdue University, June 1972, American Society of Civil Engineers special conf., Part 2,981-996. 1, and POPOVA V. Comparison of computed and factual settlements of 0. 56. EGOROV K. E. high smoke stack foundations. Proc. 4th Asian Con$ Soil Mech. Foundation Engng, 1971,1,9-15. 57. BJERRUM L. EGGESTAD Interpretation of a loading test on sand. Proc. European and A. Con$ Soil Mech. Foundation Engng, Wiesbaden,1963,1,199-204. 58. FARRENT A. The prediction andfield verifications of settlements on cohesionless soils. T. Proc. 4th Aust.-N.Z. Con$ Soil. Mech. Foundation Engng, Adelaide, 1963,ll-17. J. 59. FISCHER A. et al. Settlement of a large mat on sand. Proc. Conf: on Performance of Earth and Earth-supported Structures, Purdue University, June1972, American Society of Civil Engineers special conf., No. 2,997-1018. 1, R. 60. FROST, J. (1970) Discussion to SessionNo. I, Symp.onFounds.onInterbedded 45. Sands, Perth, p. J. 61. GEILLY et al. Settlements of two buildings supported on rafts: comparison with predicted settlement calculated from static cone penetrometer data. Con/. Settlement of Structures, Cambridge, 1974,3741. G. 62. CLICK L. Settlement study of main road department building Symp. on Foundations on Interbedded Sands, CSIRO, Perth, 1970,87-93. D. J. 63. GREENWOODA. and TAIT B. Prediction of foundations on sands.Symp. on Foundations on Interbedded Sands,CSIRO, Perth, 1970,23-33. A. W. 64. CRIMES S. and CANTLAY G. A twenty storey ofice block in Nigeria founded on loose sand.Struct. Engr,1965,43,45-57. Sao Paulo. Lhrchitecture daujourdhui, No. 65. KORNGOLD L. Immeuble CBI Esplanada a 21,7542. 66. LAGGING B. and ERESUND B. L. (1974) Provbelastning av sandfyllning (Test loading of a hydraulic sand fill).Swedish Nat. Res.Rep., 1974, No. R46,221-227. 1380
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

SETTLEMENTOFFOUNDATIONSONSANDANDGRAVEL
67. LANCFELDER J. JOHNSTON D. W. Settlement of two tauks on loose cohesionless soil. and Proc. 4th Pan Am.Con$ Soil Mech. Foundation Engng, 2,15-25. 68. LEVY F. and MORTON Loading tests and settlement observations on granular soils. J. K. con5 Settlement of Structures, Cambridge, 1974,43-52. de ponts a fondation directe (Settlement obser69. MARIVOET L. Observation des tassements vation on bridges built on shallow foundations). Proc. 3rd Int. Conf: Soil Mech. Foundation Engng, Zurich,1,418-430. J. Proc. European Con$ Soil 70. MARTINS B. et al. Settlement of a ten-storeyed building. Mech. Foundation Engng, 1963,1,313-317. and KAHL H. Tragfahigkeit und setzungen sandiger Boden. Berichte Bauforsch. 71. MUHS H. 1961, 18. 72. MUHS H. and WEIS K. Die Berechnung der Bauwekssteifigkeit von Hochhausern aus

den Ergebnissen von Setzungsbeocachtungen (The calculation of building stability of tower houses from the results of settlement observation). Bautechnik, 1963, 11, 377382. 73. MUHSH. et al. DiezulassigeBelastungvonSandaufgrandneuererVersucheund Erkenntnisse. Bautechnik, 1963,10-11,0ct.-Nov., 130-147. E. 74. NONVEILLERSettlement of a grain silo on fine sand. Proc. European Con$ Soil Mech. Foundation Engng, Wiesbaden,1963,1285-299. I. 75. OWEIS S. Equivalent linear model for predicting settlements of sand bases. J . Geotech. Engng Diu.Am. Soc. Ciu. Engrs,1979,105, GT12,1525-1544. H. 76. PRESS Baugrundprobebelastungen, ihre Auswertung und die den Bauwerken gerness-

enensetzungen(Buildinggroundtestloads.Theirevaluation andthesettlements measured on the buildings). Bautechnik, 1932,10, No. 30,391-395. H. 77. PRESS Setzungsbeobachtungen (Settlement observations). Bautechnik, 1938,16, No. 2,
2627. F. 78. RIOS,L. and SILVA P. (1948) Foundations in downtown, Brazil. proc. 2nd. Int. Conf: Soil Mech. Foundation Engng, Rotterdam, 1948,4,69-72. S 79. RONAN . R. Heavy structures founded in aeolian soils. Proc. 3rd Aust-N.Z. Con$ on Geomechanics, Wellington,1980,1,3944. G. 80. SANCLERATet al. Controle in situ des previsions de tassements basees sur les essais de penetration statique (pour 79 ouvrages sur 17 sites differents) (In situ checkof settlement predictions based on static penetration tests). Ann. Inst. Techn. Bat. Trav.Publ., 1979, No. 369, Feb., 31-50. 81. THORNEC.P. Discussion to Session 1. Symp. on Foundations on InterbeddedSands, CSIRO, Perth, 1970,47-50. Foundation design and construction, Pitman, 1969, London, 212. 82. TOMLINKIN M. J. G. P. McGraw-Hill, 83. TSCHEBOTARIOFF Soil mechanics, foundations and earth structures. New York, 1951,44and 155. I. 84. VOTJAKOVIz opyta stroitelstva proizvodstvennogo korpusa na manyvnych gruntach (From the experience of constructing an industrial building on hydraulic fill). Osnou. Fund. Mekh. Grunt.,1974, No. 5, Sept.-Oct., 7-9. I. Iz 85. VOTJAKOV, opyta stroitelstva proizvodstvennogo korpusa na namyvynch gruntach. Osnov. Fund. Mekh. Grunt, 1977,19, No. 5,29-30. D. 86. WEBB L. Settlement of structures on deep alluvial sandy sediments in Durban, South Africa. Conf: on In-situ Investigations in Soils and Rocks, British Geotechnical Society, London, 1969,181-188. Proc. 87. WENNERSTRAND J. Comparison of predicted settlements for a fine sand. 7th European Con$ Soil Mech. Foundation Engng, Brighton, 1979,2,295-298. V. 88. WHITMAN R. et al. Performance of foundation for ALTAIR radar.J . Soil Mech. Fdns Diu. Am.Soc. Ciu. Engrs, 1971,97, SM1, 1-18. 89. ZAKHARENKOVM.M.Opytstroitel'stvadevjatietamogodomonapescanojpoduske (Experience with the construction of a nine-storey building on a sand fill). Osnou. Fund. Mekh. Grunt.,1969, No. 4, July-Aug., 34-35.
Delivered by ICEVirtualLibrary.com to: IP: 86.154.85.54 On: Thu, 15 Sep 2011 11:00:29

1381

You might also like