@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Applied Science, Faculty of"@en, "Civil Engineering, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Leclair, Donna Gail"@en ; dcterms:issued "2010-09-16T19:51:05Z"@en, "1988"@en ; vivo:relatedDegree "Master of Applied Science - MASc"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """In-situ piezocone, flat dilatometer, and screw plate tests were carried out adjacent to the site of several large earth embankments, founded on a deep deposit of compressible soil. Settlement records since construction were available for two of the embankments. Geotechnical parameters were not back analyzed from the case record, rather, embankment performance was predicted on the basis of parameters interpreted from the in-situ tests alone. Consolidation characteristics were interpreted from the measurement of dissipation of excess pore pressures using the piezocone and dilatometer. Both devices provided complementary results in terms of an appropriate coefficient of consolidation. The excellent stratigraphic profile furnished by the piezocone (CPTU) tests proved to be a most valuable feature. The stratigraphic detail provided by the CPTU tests performed across the site identified continuous, free (framing soil layers which would generally be missed in a conventional geotechnical investigation using a drilled borehole with discrete sampling. The identification of these layers was of paramount importance in the prediction of settlement rate. A one-dimensional analysis formed the basis for the settlement predictions, and was found to be satisfactory. Settlement magnitudes were predicted within 10% of the observed measurements, parallelling the observed rate of settlement throughout the embankment construction period in the early 1970's and to the present date. Key words: settlement, deltaic soils, embankment, in-situ testing, piezocone, flat dilatometer, screw plate, coefficient of consolidation, compressibility, pore pressure dissipation."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/28495?expand=metadata"@en ; skos:note "PREDICTION OF E M B A N K M E N T PERFORMANCE USING IN-SITU TESTS By DONNA GAIL LECLALR B. A . Sc., The University of British Columbia, 1986 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF T H E REQUIREMENTS FOR T H E DEGREE OF MASTER OF APPLIED SCIENCE in T H E F A C U L T Y OF G R A D U A T E STUDIES DEPARTMENT OF CIVIL ENGINEERING We accept this thesis as conforming to the required standard T H E UNIVERSITY OF BRITISH COLUMBIA September, 1988 © Donna G. LeClair, 1988 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of £WlL - & 1 o r g a n i c m a t e r i a l 3) 1 c l a y 4) 1. 5 e l l t y c l a y t o c l a y 5> 2 c l a y e y e l l t t o e l l t y c l o y 6) 2. 5 eandy e l l t t o c l a y e y e l l t 7) 3 e l l t y eond t o eandy e l l t 8) 4 eand t o e l l t y . e a n d 0) 5 • a n d ID) 6 g r a v e l l y eand t o eand 11) 1 v e r y e t l f f f i n e g r a i n e d <•> 12) 2 eand t o c l o y e y eand (•> (•) overconeo 1 1 d a t e d o r cemented Fig. 5.3 Soil Behaviour Type Classification Chart (after Robertson et aln 1986) - 3 2 -i j i Layered fine SAND and SILT j \\ Grey, soft, layered fine SAND and SILT Grey loose clean fine to medium SAND Grey, compact fine clean SAND Grey, soft sensitive sandy slightly clayey SILT Depth (m) 0 — 30 End of hole 10 20 40 — 50 60 _ S Z . Sandv to clayey SILT Hill iiliiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiliiiiiiiiiiiiiiiilllll Silty SAND to sandy SILT Hill iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiilllll ijlji i;•;•;•;•;•:•:•;•!•: ! SAND to silty SAND. !!!! S! i ! i ! i!!! S! i ! i ! i ! i ! II : • 1 i 1 \\'\\ • * i •. i ! \\ \" Medium dense to dense .• cleanSAND •'. • ;• T T T T T . ; . ; . • SAND to silty; SAND •: i :• Dense to very dense •: cleanSAND IIIIIIIIIIKsTJ i^iidlllllllllll Sensitive fine-grained i i i i i m i i i i i i i i i i i i i i i i i m i m m i : layers of SAND, silty SAND: '; or sandy SILT to 0.7m thick • SandyS to sensitive clayey SILT Further penetration refused (a) Log of test hole no. 68-18 R. C. Thurber & Associates (September, 1968) (b) Interpreted profile from cone penetration test CPTU-1 (August, 1987) Fig. 5.4 Comparisonof Son Profuesmterpreted from IM1 Hole and Piezocone Pen^ - 33-5.2 Deformation Characteristics 5.2.1 Shear Modulus. G The low-strain shear modulus, G m a x was determined from shear wave velocities, using measurements from the seismic piezocone test CPTU-6 and equation 4.1. Previous research (Campanellaera/., 1985) has shown that Gmax generally tracks well with cone bearing, inferring a relationship of the form Gmax = mq c [5.2] where m is a correlation coefficient depending on soil type. Fig. 5.5 shows shear modulus and cone bearing measurements obtained at the embankment site. While there is some scatter exhibited in the modulus values, an average ratio of Gmax to q c of 40 is found. That is, the m value in equation 5.2 is equal to 40 for the clay silt deposit, with Gmax and q c in bars. The shear modulus determined from shear wave velocity is a low strain (v<10\"3%) modulus. Therefore at the level of strain induced by the embankments, the shear modulus mobilized may be considerably smaller than G m a x . However, knowledge of the shear modulus in the clay silt stratum is still pertinent to the present study since the interpretation of consolidation characteristics is dependent on rigidity index, which is the ratio of shear modulus to undrained shear strength. Also, for seismic design, G m a x is an important parameter and one which is easily determined from the seismic cone penetration test. 5.2.2 Young's Modulus. E Penetration testing is generally assumed to be a drained penetration in cohesionless soils and an undrained penetration in cohesive soils. Consequently, discussion of Young's - 34 -Cone bearing, q c (bar) Shear Modulus, G m a x (MPa) Fig. 5.5 Cone Bearing from SCPTU and Interpreted Shear Modulus -35-modulus will be divided into two sections, one for the drained parameter and one for the undrained. 5.2.2.1 Equivalent (Drained) Young's Modulus. Es An equivalent Young's modulus may be determined from the screw plate test and, using empirical correlations, from the cone penetration test. Schmertmann(1970) presented a method for detennining Es, assuming a homogeneous, elastic half space, from the results of the screw plate test. Assuming a constant modulus withm the strain area beneath the plate, Schmertmann(1970) showed that Es = 1 . 2 C i ^ B [5.3] P where: Ap = applied screw plate stress p = measured plate deflection B = screw plate diameter, C i is a correction factor to incorporate the effect of strain relief due to embedment, and C i - 1 - 0 . 5 ^ [5.4] where: a ' 0 = effective vertical overburden pressure. A complete derivation of equation 5.3 may be found in Berzins (1983). The method is basically a back calculation of modulus using elastic settlement theory, where the known load-settlement behaviour of the screw plate replaces the foundation pressure and unknown foundation settlement The method has application over a stress range of 100 to 300 kPa, and cannot be confidently applied atlow stresses, particularly at depth. - 36-From a review of calibration chamber results (Baldi era/., 1981), Robertson and Campanella (1985) providedthe relationship Es = nq c [5.5] between the drained secant Young's modulus and cone bearing. The Young's modulus was defined at stress levels commonly induced by shallow foundations, that is, 25% to 50% of failure stress levels. For normally consolidated, uncemented quartz sands, n was found to vary between 1.5 and 3.0, which is in agreement with the value of 2 recommended by Schmertmann (1970). An equivalent Young's Modulus for the sand layer at the embankment site was calculated using equation 5.3. The values of Ap and p were obtained from the initial portion of the load-deflection curve measured in the field SPLT. Test data may be found in Appendix B. Using the values of Es calculated from the screw plate test and average cone bearing values from the five cone penetration tests, the value of n in equation 5.5 was found to be 1.2 ± 0.9. This value may appear to agree with recommended values given above, however, this average masks the actual results, which varied from n = 4 near the surface to n = 0.4 at a depth of 19 m. While the values of Es from the screw plate test remained within a fairly narrow range, q c increased markedly with depth, as the sand increased in density. 5.2.2.2 Undrained Young's Modulus. Eu. and Undrained Strength. Su Undrained Young's modulus is usually estimated from C P T U data using empirical correlations with undrained shear strength, Su, of the form Eu = kSu [5.6] where the constant k is dependent upon stress level, stress history, sensitivity, and other factors. Therefore an estimate of undrained strength is first required. -37-The undrained shear strength is not a unique value for a given cohesive soil, but is a function of the type of test used. Estimates of undrained strength from cone penetration test data were made using the equation S ^ 9 ^ ^ [5.7] where: a v o is the total in-situ vertical overburden pressure Nk is the cone factor, obtained from empirical correlations. A value of Nk = 15 was used for the present analysis. Lunne and Kleven (1981), using field vane strength as a reference, showed that, for normally consolidated marine clays, the cone factor generally falls between 11 and 19, with an average of 15. Undrained shear strength may be determined from the screw plate test using a method similar to that for the cone, where c _ Pult - °VQ R , 8 1 u Nk 1 J where: pult is the ultimate average plate stress Nk depends on boundary conditions, the soil-plate interface, and plate stiffness. Selvadurai et al. (1980), after reviewing classic theoretical and empirical solutions, concluded that 9 for partial bonding Nk(screw plate) = [5.9] 11.35 for full bonding A value of Nk = 10 was assumed for the present analysis. The data reduction program accompanying the flat dilatometer test apparatus also provides an empirical correlation for S^. Fig. 5.6 shows the values of undrained strength predicted by each of the three in-situ test methods, CPTU, DMT, and SPLT, along with the values determined from laboratory vane tests on undisturbed field samples which were obtained -38--10 -20 Depth . 3 0 (m) -40 -50 -60 1 i 1 1 ; 1 r 1 i • — i — i — i — i — — i — i — i — i — 1 1 1 r i . Cone Penetration Test (Nk = 15) A HatMatometerTest(Marchetti, 1980) • Screw Plate Test (Nk = 10) B Laboratory Vane Test (Thurber & Assoc. Ltd.) i i H B -S B B o • • • -B • • m m • -• CB • • • • i 0 50 100 150 Undrained Strength (kPa) 200 250 Fig. 5.6 Summary of Undrained Strength Values -39-in the 1968 investigation program. The three in-situ test methods show good agreement and consistently predict higher undrained strengths that the laboratory tests, which may reflect the influence of scale and sample disturbance on the lab-determined strength. Test interpretations show a generally linear increase in S u with depth, a trend commonly seen in normally consolidated deposits, with a departure from this trend between the depths of 26 m and 35 m, where considerable sandy silt layering is encountered. The scatter in values interpreted from the C P T U in the mixed clay and silt soil is due to internal averaging of the actual values over 0.25 m increments by a computerized interpretation program. The undrained Young's modulus for the clayey silt at the embankment site was interpreted directly from screw plate test results, using the expression (Selvadurai and Nicholas, 1979) 5 pa/Bj = X [5.10] where: 6=plate displacement p = average stress on screw plate a = screw plate radius X = a modulus factor which falls in the range of 0.60 to 0.75. The upper limit applies when the plate is partially bonded to the soil, which may be the case in a sensitive soil, and was used for the present analysis. The values of p and 5 were obtained from the initial portion of the load-deflection curve measured in the field SPLT. Test data may be found in Appendix B. Since the screw plate test provided estimates of both Eu and Su, these were used to determine the constant, k, in equation 5.6. The value of k was found to range between 200 and 300 at the depths tested, between 20 m and 24.5 m. -40-Fig. 5.7 presents the values of Young's modulus, both drained and undrained, interpreted from the screw plate test and piezocone penetration test. The figure shows Eg and Eu as calculated directly from screw plate test results, using equations 5.3 and 5.10, respectively. From piezocone penetration tests, Eg was calculated using equation 5.5, with n = 2; Eu was calculated using equation 5.6 with k = 250 and Su from Fig. 5.6 (Nk = 15). As Fig. 5.7 reveals, good agreement is found in the clayey silt where the two sets of data coincide; therefore, some confidence can be placed in the values correlated from the C P T U Su values at depth. Good agreement is also manifest in the upper 10 m of sand, however from depths of 10 m to 20 m, there is significant difference in Eg. As mentioned previously, Eg correlated from cone bearing reflects the increasing density and increase in confining pressure, with depth, of the sand, whereas Eg calculated from the screw plate test remains within a narrow range. This may be due to increased friction along the rod lengths when testing within the dense sand deposit. 5.2.3 Constrained Modulus. M The constrained modulus relates stress and strain where strain is assumed to occur only in one direction, usually vertically. Therefore the constrained modulus is often referred to as a one-dimensional modulus, and can be used to compute vertical settlements. The data reduction program accompanying the flat dilatometer apparatus provides an empirical correlation for M , based on Marchetti (1980). As well, numerous empirical correlations have been developed between cone resistance and constrained modulus, these having the form M = ocqc [5.11] In order to calculate the factor a for the normally consolidated clayey silt deposits of the Fraser River delta region, the M values obtained from flat dilatometer tests were compared to values of cone bearing at the same depth. The result, as shown on Fig. 5.8, is a factor of -41 -0 -20 Depth (m) -40 -60 ) | ! i , , , • a 0 a a \\a i B • a | • 1 B O B C T 1 1 1 i T 1 1 1 1 B B a i ' • i B : B H t • a • • • • o • - o • o o o • -o o 0 o o o o o o B Es — SPLT • Eu — S P L T • Es — C P T U o Eu — C P T U -o o o o o o o o --o o o o o o o ' 1 ' 1 1 ' 1 • 1 1 1 i 1 1 1 : 1 i i I 0 100 200 300 400 500 Young's Modulus (MPa) Fig. 5.7 Summary of Young's Modulus Values -42-Fig. 5.8 CorrelationBetwe^Cme-EHmensionalC^ -43-a = 2.37 ± 1 . 0 8 [5.12] which agrees very Well with the range of a = 1 to 3, given by Mitchell and Gardner (1975) for silts of low plasticity where q c is less than 20 bars. Since dilatometer testing was carried out only to a maximum depth of 35.6 m, this correlation was utilized with q c data throughout the entire depth of clay silt to obtain a profile of M with depth. This profile is shown on Fig. 5.9, with values for the upper 20 m of sand taken directly from the dilatometer data reduction program. Dilatometer testing was conducted sufficiently far away from the existing embankments so as to test virgin soil unaffected by the embankment load. The modulus values, as shown on Fig. 5.9, were used to calculate settlement during the embankment construction and preloading phase. The determination of an appropriate modulus to use in settlement analysis requires considerable judgment. In-situ testing offers the advantage of testing soil at its existing level of stress, however, as modulus is stress-level dependent, decisions must be made on an appropriate level of stress at which to compute the modulus. The determination becomes even more complicated when the same soil is subject to changes in in-situ stresses, as in the case of preloading. The in-situ tests used in this analysis provided reasonably complementary results for theparameters required in settlementcalculation. Where values diverged among test methods, the parameter often had minimal effect on the end result, that is, the calculated settlement was not highly sensitive to large variations in the estimated parameter value. 5.3 Consolidation Characteristics Laboratory consolidation tests on field samples have traditionally been performed in order to measure properties for use in geotechnical settlement analyses. The ability to evaluate flow and consolidation characteristics from the time rate of pore pressure dissipation using the piezocone -44-0 -10 -20 Depth (m) -40 -60 1 1 1 H 1 1 1 El El E El El ' 1 1 1 : i El 4 , 1 1 1 r i 1 -El El El El El 1 3 -- H H • B EI - a a H -H • H H - H S - H El - S -B - El - S SI - S - H El -- H El SI - B S El El 0 20 40 60 80 100 120 Constrained Modulus, M (MPa) Fig. 5.9 Summary of Constrained Modulus Values -45-and, recently, the dilatometer, has provided an alternative approach to decerning the consolidation characteristics of a soil. 5.3.1 Pore Pressure Dissipation Tests A piezocone dissipation test is conducted during a pause in penetration at any depth where consolidation characteristics are required. The decay of excess pore pressure is monitored with time. In the case of the present research, pore pressure measurements were taken at five-second intervals. Fig. 5.10 shows data collected at the embankment site during a dissipation test at a depth of 54.8 m, with pore pressure sensing elements at two locations on the cone, U2 behind the cone tip, and U3 behind the friction sleeve. The push rods were not clamped during dissipation tests since recent research (Campanella and Robertson, 1988) has shown that changes in pore pressure measurements caused by movement or creep of the rods is generally not significant when the piezometric element is located behind the cone tip. Fig. 5.10 indicates that while only three minutes are required for half of the initial excess pore pressure to dissipate, as measured at the U2 position, times approaching one hour are required to re-establish equilibrium (hydrostatic) pore pressure. Since the closing pressure (C reading) closely represents the pore pressure on the flat dilatometer membrane in soft clays, Robertson era/. (1988) imply it should be possible to record the C reading with time and obtain a dissipation curve using a standard Marchetti dilatometer. Fig. 5.11 shows D M T dissipation readings obtained by two procedures, one where membrane lift-off was achieved and C reading taken (A-C reading procedure), and another where membrane lift-off was followed by 1 mm expansion, then the C reading taken (A-B-C reading procedure). Fig. 5.11 illustrates that slightly more stable readings are obtained using the A - C reading procedure than with the A-B-C reading procedure. The difference in response is probably related to small changes in effective and total stresses around the membrane during dissipation, depending upon whether or not the soil is pushed out by expanding the membrane once more. 160 140 120 100 Pore Pressure, u 80 (m of water) 60 40 20 i— i — i 1—i—i 1 r IB • ft • V i 1 — i — I 1 — i 1 1 — I — i 1 — i — i 1 — i 1 1 — i — i 1 — i 1 — i 1 1 — i — r CPTU-6 Dissipations at 54.8 m UBC Cone #7 22 October 1987 • U2 • u 3 • • • • Hydrostatic pressure-• • » 3 Q . . . . f i . . . ( ? ! . . . f l . . . ^ . . & . . . f i . . . f r . . f t . . ^ . . d . . . . f i . . . . a . . . f t . . . . a . 1 1 1 i i i i i I ' ' ' , I 1 1 1 1 1 1 I I 10 15 20 25 30 Time (minutes since penetration stopped) 35 40 Fig. 5.10 Pore Pressure Dissipations Measured at Different Locations on Piezocone ~ i 1 r— Q • • I • • • • • • a i • ••• D M T C Reading (bars) — i 1 i — i — i — i — i i 1— • • • • _. • • • • • • • — i i 1 j 1 i 1 1 I 1 i 1 1 ; 1 i r -• A - C sequence at 21 m n A-B-C sequence at 23 m • a • • • - • • • • E B - -Hydrostatic pressure~ 1 0 10 20 30 40 50 Time (minutes since penetrationstopped) Fig. 5.11 Record of DMT Dissipations Using Two Procedures -48-Levadoux and Baligh( 1986) suggested that normalized pore pressure, U , provides a good measure of degree of consolidation, with u = H < 5- I 3J where Au = ut - u D AUi = Uj - U Q Ut = measured pore pressure at time t Uo = equilibrium (often assumed to be hydrostatic) pore pressure Uj = initial pore pressure at the commencement of the dissipation test. Extending the same idea to dilatometer C reading data, an equivalent normalized pore pressure may be obtained from: u - c f ^ t 1 5 1 4 1 where C t = C reading at time t Q = initial C reading at time t = 0. While a C reading is never obtained at time = 0, this value may be obtained from a plot of C reading versus square root of time, and the initial, straight-line portion of the curve back extrapolated to zero. When dissipation data is normalized as described above, it is possible to compare C P T U and DMT data. Fig. 5.12 shows that, at a depth of 23 m, it takes approximately eight minutes to reach 50% consolidation (tso = 8 minutes) in a CPTU dissipation test using a 10 cm 2 cone, and approximately twice as long, (tso = 16 minutes) in a D M T dissipation test. 1.0 0.9 0.8 0.7 0.6 Normalized Pore 0.5 Pressure 0.4 0.3 0.2 0.1 0.0 A Test DMT-2 at 23.0 m A Test CPTU-6 at 22.9 m A A A 1 10 Time (minutes since penetration stopped) 100 Fig. 5.12 Piezocone and Dilatometer Dissipation Record -50-Although a similar trend is evident in the C P T U and D M T dissipation curves shown in Fig. 5.12, Fig. 5.13 shows a different trend obtained from two dissipation tests at a depth of 35 m. At this depth, the clayey silt stratum contains numerous silty and sandy silt layers. This increased permeability is indicated by the CPTU, but not by the DMT. Therefore, it appears that piezocone measurements are more sensitive to thin drainage layers than dilatometer test measurements. The dissipation rate of excess pore pressure is controlled by the consolidation and permeability characteristics of the soil. The coefficient of consolidation in the horizontal direction, Ch, may be calculated from a dissipation test using one of several theoretical solutions. 5.3.2 Theoretical solutions Gillespie (1981) provides a comprehensive discussion of the theoretical solutions available for obtainmgconsohdation characteristic Two of these methods were utilized in the present analysis. Torstensson (1977) theorized that pore pressures caused by steady cone penetration could be estimated by one-dimensional solutions corresponding to the expansion of spherical and cylindrical cavities. His analysis assumed an isotropic, elastic, perfectly-plastic soil with isotropic initial state of stress, and used linear, uncoupled, one-dimensional finite difference consolidation theory to estimate consolidation rates. He proposed matohing theoretical predictions and measured values at 50% consolidation, U = 0.5, to find the coefficient of consolidation in the horizontal direction, ch, using Ch = ^ R 2 [5.15] where T50 is the dimensionless time factor at 50% consolidation and is a function of E/Su, 1.0 0.9 0.8 0.7 0.6 Normalized Pore o.5 Pressure 0.4 0.3 0.2 0.1 0.0 . Test DMT-2 at 35.6 m 0 Test CPTU-6 at 35.8 m ' Q D Q ( , 1 10 Time (minutes since penetration stopped) 100 Fig. 5.13 CPTU Dissipation Exhibits Greater Sensitivity to Drainage Layers than D M T -52-tso is the measured time to achieve 50% consolidation, and R is an equivalent cavity radius. Levadouxand Baligh (1986) noted that the above method for determming Ch does not account for non-linearities during consolidation, soil remoulding, or creep effects, and found no acceptable argument for curve fitting about tso- Rather, Baligh and Levadoux (1986) recommend a method based on predictions obtained from linear, uncoupled consolidation analyses and initial pore pressure distributions calculated by the strain path method for undrained penetration in Boston blue clay, using the normalized excess pore pressure distribution, U , and tabulated values of the time factor, T. 5.3.3 Coefficient of Consolidation Levadoux and Baligh (1986) report that, in clays, dissipation is principally controlled by the horizontal coefficient of consolidation, especially in the early stages of consolidation, since permeability in the horizontal direction is generally greater than in the vertical. Local research (Gillespie, 1981) has shown that consolidation in the clayey silt underlying the Fraser Delta is also controlled by horizontal drainage, therefore use of Ch in a consolidation analysis is appropriate. For the present analysis, the horizontal coefficient of consolidation was determined by curve-fitting about tso, using values of T recommended by Torstensson (1977) and by Baligh and Levadoux (1986), applying equation 5.15 to dissipation test data from both the piezocone and dilatometer. The various theoretical values of the dimensionless time factor may be found tabulated in Appendix A. In the case of the piezocone, a value of R = 0.561 cm, the radius of a standard 10-cm2 cone shaft, was used in equation 5.15 to estimate ch- Fig. 5.14 provides an example of Normalized Pore Pressure 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 • • 1 H i \\ H U M ! ! i l l • Measuredat 54.8 m, CPTU-6, u2 position • Levadoux and Baligh, 1986 • Torstensson Spherical, E/Cu= 100 A Torstensson Cylindrical, E/Cu= 100 t. • • A— -• • • • -• _ •\"• i ] • i -I ^ » ] • • c \\ \\ _ J ... . . . . -I D -1 10 Time (minutes since penetration stopped) 100 Fig. 5.14 Theoretical Curve Fitting About Time for 50% Pore Pressure Dissipation -54-theoretical curve fitting, presenting measured C P T U dissipation data along with data points predicted by theoretical analyses atU = 0.8, 0.6, 0.5, 0.4, and 0.2. At all stages of consolidation, the spherical solution of Torstensson (1977), with a soil rigidity index of 100, provides the best fit to the measured data. Further deviation from the measured data was observed as rigidity index was increased from 100 to 500. The interpretation of a spherical cavity is potentially the most rational, as Levadoux and Baligh (1986) indicate that contours of Au/a ' v o around a 60° cone are spherical in shape. In the case of the dilatometer, an equivalent radius of the standard Marchetti blade, of R = 2.057 cm, was used, as recommended in Robertson et al. (1988). To date, this procedure for the D M T has only been verified for soft, normally consolidated to lightly overconsolidated soils. As with theoretical curve fitting for C P T U data, the spherical solution of Torstensson (1977), with a soil rigidity index of 100, provided the best fit to the measured data. A horizontal coefficient of consoUdationwas determined at tso, using Torstensson's (1977) spherical solution, for each dissipation test conducted with the piezocone and dilatometer. Fig. 5.15 presents a summary of the Ch values determined in this manner alongside laboratory-determined values for comparison purposes. Points representing the average values obtained by R. C. Thurber & Associates Ltd. (1968), are shown. l^bxjratory-determinedcv values and in-situ-determined Ch values show close agreement. The majority of values fall within a narrow band averaging 4 x 10\"3 cm2/s. Potential drainage layers of sandy silt, where Ch is considerably greater, are clearly identified by the piezocone dissipation tests. Levadoux and Baligh (1986) suggest that pore pressure dissipation, during early stages of consolidation of the soil in the vicinity of the piezocone, takes place in a recompression mode for both normally consolidated and overconsolidated soils, and suggest c n in the normally consolidated (NC) range may be evaluated by: -55-0 * C P T U dissipations in clayey silt -10 a C P T U dissipations in sandy silt DMT C reading dissipations -20 o 2 O O: Laboratory consolidation test (UBC) m Laboratory consolidation test (Thurber & Assoc. Ltd.) -30 Depth (m) -40 -50 -60 -70 .001 .01 /cur Coefficient of Consolidation f— Fig. 5.15 Profileof Coeffident of CtonsoUdation with Depth -56-/ X T _ . RR(piezocone) . . . I C , , , Ch(NC) = — CR Ch(piezocone) [5.16] where RR (piezocone), the recompression ratio, represents the strain per log cycle of effective stress during recompression, and CR, the compression ratio, is the average slope of the strain versus log effective stress plot in the effective stress range expected during consolidation. The slope of the compression portion of a consolidation curve is generally five to ten times that of the recompression portion, therefore it would be expected that Ch(NC) would be approximately one-fifth to one-tenth the value determined from the early stages of a piezocone dissipation test. However, for the coefficient of consolidation determined by curve fitting about tso, it is likely that consolidation is once more occurring in the compression mode, therefore the values of Ch determined in this analysis were not adjusted by the RR/CR ratio. Local experience has shown that consolidation in the Fraser River delta deposits occurs more quickly than predicted on the basis of laboratory consolidation test parameters. This is likely due, in part, to the influence of undetected drainage layers. Judgment must be exercised when determining a coefficient of consolidation from penetration test pore pressure dissipation data. When measured pore pressures are plotted with the common logarithm of time during a pause in penetration, as in Fig. S.14, theoretical solutions predict slower consolidation than observed for U less than 30%, that is, nearing the completion of consolidation, or a degree of consolidation greater than 70%. Conversely, theoretical solutions predict faster than observed consolidation early in the dissipation process. Therefore, different coefficients of consolidation are predicted at different degrees of consolidation, that is Ch determined at tso may not be the same as that determined independently at tso or t,20- However, for a parameter where order of magnitude estimates are frequently made, the variations in Ch found in this analysis were relatively small. -57-As noted by Baligh and Levadoux (1986), laboratory measurements of coefficient of consolidation or permeability in fine-grained soils can underpredict in-situ values by several orders of magnitude. Profiles determined from tests conducted on samples obtained from discrete depths typically show significant scatter and can easily miss drainage layers essential in a field scale consolidation analysis. In-situ tests offer strong advantages in identifying these important layers for subsequent specific tests. - 5 8 -6. PREDICTION OF PERFORMANCE Embankment performance, in terms of rate and magnitude of settlement, was predicted using the results of piezocone, flat dilatometer, and screw plate tests. For the predictions, no new methodology was introduced. Analyses were performed using current, accepted practice. 6.1 Simplified Approach; Crie-Dimensional Analysis With, perhaps, the exception of high risk structures, most settlement calculations are based on one-dimensional analyses mvolving the estimation of vertical displacements induced by a design load. The total settlement, S, is calculated as the sum of three components S = Sd+S c + Ss [6.1] where Sdis the distortion settlement, S c is the consolidation settlement, and S s is the secondary compression settlement In order to evaluate each of these components, it is necessary to quantify the load applied, the resulting increase in stress, the distribution, with depth, of this stress increase, as well as the relevant soil properties. 6.1.1 Stress Increase To evaluate the distribution of stresses within a soil mass, the theory of elasticity is invariably used. Although soil is a non-linear material, and inherently anisotropic, the assumption is often made that the soil is isotropic elastic. Rigorous solutions for more complex non-linear constitutive relationships are only possible in very few cases, and for most applications, the use of elastic theory results in an acceptable degree of accuracy for the evaluation of stress distribution. -59-The values of vertical stress increase at the ground surface, from each component of loading, are identified in Table 6.1. These values formed the basis for computing the distribution of stress increase with depth. Table6.1 Stress Increase Induced by Placement of Embankments, Surcharge, and Abutments Location Stress Increase (kPa) due to: Embankment Surcharge Abutment Arthur Laing Bridge south approach 173 50 44 McConachie Way Overpass 151 35 106 The solution for the distribution of stresses within a semi-mfinite, homogeneous, isotropic mass, with a linear stress-strain relationship, due to a point load on the surface, is first credited to Boussinesq (1885). Because the solution is a linear function of applied load, the principle of superposition may be applied to account for variations in loading conditions, from point to line loads and from strip to circular areas. In an elastic analysis, an embankment is considered to be an infinite strip area carrying a combination of uniform pressure and linearly increasing pressure. Fig. 6.1 illustrates how the increase in vertical stress, Ao\"z, due to an embankment, may be computed by means of elastic theory. The equations of Fig. 6.1 are commonly expressed in the form A(JZ = Iq [6.2] -60-where I is an influence factor which takes into account the geometry of the loaded area. Numerous charts have been compiled (Harr, 1966; Scott, 1963; Foster and AMvin, 1954; Fadum, 1948) from which values of I may be obtained for various geometric loading configurations. In an analysis using elastic theory, the value of q, the uniform pressure, is evaluated by the \"normal loading approximation'' q = Y H [6.3] where y is the unit weight of the embankment fill, and H is the embankment height (a) Stress increase due to uniform load, q oz \" ^ {a + sin(a)cos(a+2p)} (b) Stress increase due to linearly increasing load fx 1 B a sin(2p) Fig. 6.1 Determination of Vertical Stress Increase by Elastic Theory -61 -Perloff (1975) comments that the above approach neglects the shear stresses which develop between an embankment and its foundation, and proposes an alternate approach (Perloff et al., 1967), which considers the embankment and foundation as a single body loaded only by self weight This is called the \"elastic embankment'' approach and may be more realistic because it considers the effect of the material itself on the distribution of stress, allows for shear distortions at the embankment-foundation interface, and produces a result found to be consistent with field measurements of pore pressures beneath an embankment (Bozozuk and Leonards, 1972). Fig. 6.2 presents the distribution of vertical stress increase, with depth, for the normal loading approximation and elastic embankment methods, for the combined embankment and abutment loads of the McConachie Way Overpass. At a depth of 20 m, where the clayey silt stratum is first encountered, the elastic embankment method predicts a vertical stress increase 20% less than the normal loading approximation. The presence of 20 m of sand may reduce the stresses distributed to the underlying compressible clay silt layer. When the stiffness of a load-bearing stratum is larger than that of an underlying soft soil, the load distributing effect can be approximately accounted for by calculating stresses in the lower layer assuming the upper stiff layer to be increased in thickness. Perloff (1975) suggests an increase of 15% in the thickness of the upper layer has been used successfully. Therefore, in calculating the stress increase in the clayey silt, for both the normal loading approximation and elastic embankment methods, an additional 3 m of sand was assumed to exist, or A0\"z = A(7 z +3 in the clay silt. For example, 21m became 24 m for determination of the stress increase at that depth. This is evident as the break in the curves occurring at the 20 m depth in Fig. 6.3, which illustrates the distribution of stress increase due to embankment and preload but, unlike Fig. 6.2, does not include the abutment load. -62-Embankment % Vertical Stress Increase (kPa) Fig. 6.2 Profile of Vertical Stress Increase Predicted by Normal leading Approximation and Elastic Embankment Methods—McConachie Way Overpass -63--10 -20 h -30 El E) E) E) ..EL El SI -S-. El El El El El El El 63 B -E EB 63 -EB El El S - ~ El El El El El El El 63 63 63 63 63 -40 El SEl El 8 -50 -60 El El El El El El El El El S E) El El El El El El El El 63 63 B 63 63 63 63 63 EB 63 63 63 SB EB E) I 1 f— EB EB Normal leading Approximation EI Elastic Embankment Method Adjusted to account for stiffness of upper 20 m (Perloff, 1975) 50 100 150 200 Vertical Stress Increase (kPa) Fig. 6.3 Profile of Vertical Stress Increase Due to Embankment and Preload—McConachie Way Overpass Embankments -64-6.1.2 Undrained Deformation Initial distortion settlement is an immediate deformation which takes place, in cohesive deposits, under undrained conditions. The following equation for vertical distortion settlement, Sfj, due to a distributed load acting on a rectangular area near the surface of a relatively deep stratum, was first given by Schleicher (1926)0 S d = C d p B ( ^ ) [6.4] where Cd is a parameter to account for the shape of the loaded area and the depth of the layer for which the settlement is being calculated, p is the magnitude of the uniformly distributed load, B is a characteristic dimension of the loaded area, u is Poisson's ratio, and Eu is the undrained Young's modulus. When a soft, compressible stratum is underlain by rock or very hard or dense soils, as in the case of Sea Island, where the compressible clayey silt is underlain by dense glacial till, the effect of layering may have an appreciable influence on the magnitude of (calculated immediate settlement. The factor Cdof equation 6.4 was replaced, in the present analysis, with the factor Cd* to account for the presence of the rigid base. Harr(1966) cites values for C£ which depend upon the shape of the loaded area and thickness of the compressible stratum relative to the width of the loaded area. To account for the load distributing effect of the overlying sand, it was assumed that the entire 61 m of soil beneath the embankment consisted of compressible clayey silt, and the distortion settlement, Sd6i, was calculated using equation 6.4 and Cd\\ The distortion settlement in the upper 20 m of sand, Sd20> was then calculated by the same method, and this value subtracted -65-from Sd6i- The distortion settlement in the clayey silt stratum alone, arising from the combined embankment and preload pressure of the McConachie Way Overpass embankment, was thus found to be 18.4 cm. A similar calculation performed for the south approach embankment of the Arthur Laing Bridge yielded a distortion settlement in the clayey silt stratum of 25.1 cm. It is evident when considering equation 6.4, that calculated distortion settlements depend directly on the assumed values of Young's modulus and Poisson's ratio. For saturated clayey soils, which are thought to deform at constant volume duimg the initial time in which elastic distortion settlements develop, a value of Poisson's ratio of u = 0.5 was assumed, and the undrained Young's Modulus as determined from the screw plate test, Eu = 28 MPa, was used. As discussed previously, use of the screw plate to determine Young's modulus may underestimate Eu due to the influence of inhomogeneities and the small size of the plate with respect to embankment size. Comparing the initial modulus value, Ej = 28 MPa, with the unload-reload modulus, Eu. r= 40 MPa, indicates the Sea Island clayey silt is a strain hardening material. If the value of Young's modulus were allowed to vary between 28 MPa and 40 MPa, and the value of Poisson's ratio to vary between 0.35 and 0.5, the calculated distortion settlement in the clayey silt stratum would range, for the McConachie Way Overpass embankments, between 14.5 cm and 22 cm. The value of Sd = 18.4 cm used in prediction lies in the middle of this range, and the potential variation of ±4 cm is insignificant in comparison to the settlement which occurs due to consolidation of this stratum. Because of the high permeability of sands, the distortion settlements occur at the same time as consolidation settlements. The prediction of settlements of cohesionless soils is often based on semiempirical methods, correlated for compatibility with field observations. The method proposed by Schmertmann (1970) uses the fouowing equation n S d =CiC 2 Ap^T [ |).AZi [6.5] - 6 6 -where Ap is the net load intensity at the foundation depth, I z is a strain influence factor, E is the Young's modulus for the centre of the ft1 layer, Azj is the thickness of the 1 t h layer, and Ci and C2 are correction factors. and was used in the present analysis. To incorporate the effect of strain relief due to embedment, the correction factor C1 is defined as follows, with a rnidting lower bound of 0.5: where a' 0 is the effective vertical overburden pressure at the depth of interest. The correction factor C2 accounts for the time-dependent increase in settlement due to creep which is observed to occur. where t is time, in years. To use equation 6.5, the upper 20 m of sand at the site was divided into layers, the first layer being 3 m thick and encompassing the mixed sandy, silty, clayey soil, and me remaining layers of primarily clean sand each being 1 m thick. An influence factor was calculated using both the normal loading approximation and elastic embankment method, and Young's modulus as determined from the screw plate test, for each layer, was used in equation 6.5. As a result, Srj calculated for the upper 20 m of cohesionless soil was found to be 2.4 cm and 1.7 cm by the normal loading approximation and the elastic embankment method, respectively, in the case of the [6.6] [6.7] -67-McConachie Way Overpass embankments, and 2.9 cm and 2.1 cm in the case of the south approach embankmentof the ArthurLaing Bridge. As was the case for the compressible clay silt, distortion settlement of the cohesionless stratum is dependent upon Young's modulus. While correlations from in-situ tests to deformation moduli are empirical in nature and may not be considered highly reliable (Jamiolkowski et al., 1985), wide variations in this parameter have a negligible effect on the overall results of the present analysis, as the magnitude of this component of settlement, approximately 2 cm, is extremely small in comparison to the total settlement. 6.1.3 E)ramed(Corisolidation)Deformation The effect of foundation loads applied rapidly to cohesive soils is manifest by increased pore water pressure. With time, water flows out of soil voids, and pore pressures dissipate. This process is known as primary consolidation. If boundary conditions in the field are such that volumetric strains and accompanying settlements are only vertical, for instance, when the dimensions of the loaded area are large relative to the thickness of the compressible stratum, or when the compressible material lies between two stiffer soils whose presence tends to reduce the magnitude of horizontal strains, a one-dimensional (vertical) consolidation analysis is appropriate, and may be conducted in two steps: 1. Evaluationof ultimate consolidation settlement (amount, ormagnitude), 2. Estimationof time-setflementhistory(rate). This comprises the simplified approach of the present analysis. 6.1.4 Amount of Settlement One-dimensional consolidation analysis assumes zero lateral strain. In reality, the condition of zero lateral strain is not often met, especially where deep compressible strata are -68-involved. In practice, however, except in the case of high risk structures, generally a one-dimensional settlement analysis is carried out To estimate the magnitude of settlement, a soil profile is divided into layers and the increment of consolidation settlement, dS c, for mat layer is computed from dS c = myAOzdz [6.8] where A O z is the stress increase at the centre of the layer, dz is the thickness of the layer, and m v is the coefficient of volume compressibility, equal to 1/M, the reciprocal of constrained modulus at the centre of the layer. Hence, equation 6.8 can be rewritten as dS c = - j^dz [6.8a] The total settlement for the entire soil profile is the summation over all the layers, i. e. SC = I dS c = Z ^ d z [6.9] To determine the consolidation settlement due to embankment load and surcharge load for the McConachie Way Overpass embankments and Arthur Laing Bridge south approach embankment the 61m thick soil profile was divided into 56 layers. The majority of the layers were taken as 1 m in thickness, with the exception of the top layer, which was 2 m in thickness, and the seven sand layers from the 10m through the 20.5 m depths, which were taken as 1.5 m in thickness. The constrained modulus values shown in Fig. 5.9 were used in equation 6.9. A vertical stress increase was determined for the centre of each layer from the profile shown in Fig. 6.3, using both the normal loading approximation and the elastic embankment methods. - 6 9 -In-situ tests for this study were performed at sufficient distance from the embankments to ensure foundation soils were not influenced by the added loads. The values of constrained modulus determined from those tests are valid, therefore, for the initial loading conditions. Once the foundation soils had been preloaded, however, the soil would possess different properties, as the subsequent removal of preload would leave the soil with a stress history, that is, in a lightly overconsolidated state. The M values used to compute settlement due to embankment and surcharge, therefore, required some modification for the calculation of settlement due to abutmentloading. Schmertmann (1986) proposed a special method for computing foundation settlement, from dilatometer test results, which accounts for the variation in M with varying stress level. This method recognizes that the effective stress at the time of structure loading may not be the same as at the time the D M T was conducted, whether due to excavation, surcharge, dewatering, or other circumstances. This special method involves construction of an appropriate modulus-effective stress curve for both normally consolidated and overconsolidated soil conditions, using the values of preconsolidation pressure, p' c, determined from the DMT, effective stress at the time of the DMT, p'd effective stress at the time of structure loading, o'Q, and structure load, AO\"' Z . The Schmertmann (1986) method uses the tangent modulus relation of Janbu (1967) to compute the adjusted modulus at the revised stress level, a' = a' 0 + Ao z M = k m p a ( £ ) 1 \" a [6.10] where k m is a dimensionless modulus number, Pm is atmospheric pressure, a reference stress, o' is the appropriate level of effective stress, and a is a stress exponent, approximately equal to 0.5 for sands and silts and 0 for clays. -70-For the present analysis, the stress exponent was taken equal to 0.5, and values of k m were back calculated for each soil layer from the original M values shown in Fig. 5.9. The average values of back calculated modulus numbers, 400 in the upper dense sand, 24 in the clayey silt, and 52 in the sandy silt, are within the range of typical values cited by Janbu (1967). Equation 6.10 defines the curve for normally consolidated conditions. In order to reconstruct the overconsolidated portion of the curve, the point (M, p' [ 6 - 1 1 ] where S c is the ultimate magnitude of consolidation settlement and Sc(t) is the magnitude of consolidation settlement at time t. One-dimensional consolidation theory relates the quantities excess pore water pressure, u, depth, z, and time, t, through the partial differential equation at - c az2 [ 6 A 2 ] where c is the coefficient of consolidation, Cy in the vertical direction or c n in the horizontal direction. The initial distribution of excess pore water pressure, however, depends on in-situ stress conditions, which may vary from a simple linear distribution with depth to very complicated distributions. While the solution to equation 6.12 involves integration, analytical solutions to the integrals have been developed for several distributions of excess pore water pressure. These solutions relate U(t) and the dimensionless factor, T, where T - | [6.13] and d is the length of the longest path by which pore water may escape. The drainage path, d, is equal to the full thickness of the compressible stratum if drainage can occur at only one boundary, or half the thickness of the stratum if drainage can occur at both top and bottom. One of the greatest sources of error in predicting time rate of settlement is the definition of drainage boundary conditions. For predicting rate of settlement at the embankment site, the compressible clayey silt was divided into two strata, separated by a more freely draining sandy silt layer at a depth of about 28 m. Fig. 6.5 presents a schematic diagram of the boundary conditions assumed for calculating the rate of settlement. Each of the two clayey silt strata was Fig. 6.5 Profile Assumed for Settlement Rate Analysis -75-assumed to have double drainage, the upper stratum draining to the overlying sand and to the sandy silt at 28 m, and the lower stratum draining to the sandy silt at a depth of 28 m and to the basalttilatapproximately 61 m. While dense glacial till would normally be assumed to act as an impermeable boundary, rapid dissipation of pore pressure was manifest in a C P T U dissipation test at a depth of 60.9 m, leading to the conclusion that drainage would take place at the interface of the clay silt and till. Local experience has shown that considerable weathering exists at the surface of the till, again supporting the idea of drainage at the base of the clay silt A drainage layer at a depth of 28 m was assumed based on the stratigraphy as defined by profiles of cone bearing with depth and on the coefficients of consolidation calculated from in-situ dissipation tests. Profiles of cone bearing 30 to 40 m apart at the embankment site gave evidence of considerable sandy silt layering surrounding the 30 m depth, therefore the areal extent of this more freely draining layer was judged sufficient for it to act as an effective drainage path. Furthermore, horizontal coefficients of consolidation interpreted from C P T U dissipation tests atthis depth were approximately one order of magnitude greater than those in the remainder of the clay silt, agam mdicating preferential drainage. Ahorizontalcoefficientof consolidation of c n = 0.004 cm2/s was assumed for the two compressible clay silt strata. Fig. 5.15 shows only a narrow scatter band around this value, for dissipation and consolidation tests conducted throughout the deposit The drainage layer at the 28 m depth was assumed to be of insufficient thickness to contribute to consolidation settlement merely acting as a drainage interface. Generally it is found that consolidation settlement in cohesionless soils occurs so rapidly it is virtually impossible to distinguish between settlement occurring as a result of consolidation and that occurring as a result of distortion. Dissipation tests conducted in the upper 20 m of sand revealed that any excess pore pressures generated during cone penetration dissipated within one minute. Given this information, it was assumed that consolidation settlement in the sand took place almost immediately with each new construction loading phase. -76-With the above system so defined, settlement at any time, t, was calculated using the following equations for the time factor, T, in each compressible stratum: TUpper = ^ t = 0.066t, with t in months [6.14a] Tiower = ^ t = 0.004t, with t in months [6.14b] and a rearranged version of equation 6.11: Sc(t) = U(t)Sc [6.15] Values of U(t) were interpolated from Case 2 (half sine curve) of the table relating U and T, which is given in Appendix A. A half-sine curve was assumed as the initial distribution of excess pore pressure, as it was expected that drainage would occur at a slightly slower rate at the 28 m depth than at either the boundaries between the clayey silt and upper sand or the clayey silt and basal till. Consequently pore pressures would remain higher at the 28 m boundary than at the ground surface or at the 61m depth. Fig. 6.5 shows the assumed initial distribution of excess pore pressure. Following surcharge removal from each embankment, the last value of settlement computed due to the surcharge load was maintained constant, continuing to act as a fixed contribution to the overall settlement throughout the remainder of the analysis. No rebound, or negative settlement, was assumed to occur. The determination of settlement during unloading and reloading requires assumptions as to how and when the foundation soil will respond. While the ground will likely continue to settle for some time after a portion of the surface load has been removed, this settlementmaybecounterbalanced by the potential rebound which may occur due to loadremoval. -77-6.1.6 Correction for Construction Period As embankments, or other structures, in their final form, are not placed immediately, but require time for construction, time-settlement curves must be corrected to allow for the construction period. Terzaghi (1943) proposed an empirical correction method whereby it is assumed the net foundation load is applied at a uniform rate during the construction period, tc, and that the degree of consolidation at the end of this time is the same as if the load had been acting for half that time, tJ2. In other words, settlement during the construction period is the same as that which would be calculated assuming instantaneous loading at half the construction time, with the load reduced proportionally to account for the fact that the total load is not acting during this time. For the example case of the McConachie Way Overpass embankments, 8.3 m of compacted fill and 2.1 m of uncompacted surcharge were placed in five months, that is, tc = 5 months. The fill was not placed uniformly during this period, however, with approximately 2 mof fill being placed in three months, and the remaining 8.4 m being placed in two months. In order to match the construction sequence more closely when predicting settlement rate, the load due to an embankment 1 m high (one half the load due to placement of 2 m of fill) was assumed to act at 1.5 months (one half the time for placement of 2 m of fill), following which the load due to an additional embankment height of 3.15 m (one half of 6.3 m) was assumed to act at 3.5 months, and the surcharge load due to 1.05 m of uncompacted fill (one half of 2.1 m) assumed to act at 4.5 months. In this manner, at the end of the construction period of five months, surface settlements due to consolidation in the clayey silt were found to be 6 cm and 4 cm for the normal loadingapproximationand elastic embankment method, respectively. Consolidation settlement in the sand, 12 cm and 9.3 cm, was assumed to have been completed by the end of the construction period, and the contribution of this component of settlement was added in linear increments from construction start to finish. -78-Qjnstruction of the south approach embankment for the Arthur Laing Bridge proceeded uniformly to completion within two months. At the end of the construction period, 2.5 cm and 2.6 cm of settlement were computed to have occurred, by the normal loading approximation and elastic embankment method, respectively, due to consolidation of the clayey silt, and the contribution of consolidation settlement from the 20 m of sand, 14.4 cm and 12.5 cm, was added in linear increments over the two months. 6.1.7 Secondary Compression Observations both in the laboratory and m me field mdicate mat settlements continue under conditions of constant effective stress, that is, even after excess pore pressures have dissipated, or the process of primary consolidation is complete. This settlement, known as secondary compression, is believed to continue at a very slow rate for an mdefinite period of time. It is often assumed that secondary compression takes place after primary consolidation is complete. This is based on the observation that curves of settlement versus logarithm of time show a distinct inflection point at the time when primary consolidation is essentially complete, for laboratory consolidation tests run under a load increment ratio of unity. Secondary compression has been related to water content, consolidation pressure, clay mineralogy, temperature, and other effects (Mesri, 1973), but is recognized to be a function of time. The coefficient of secondary compression, C a , is usually calculated from laboratory measurements, using the equation Ae = -Calog(g] [6.16] where Ae is the change in void ratio as measured in a laboratory consolidation test between the times ti and t2-While much work has been done in quahtatively considering the concept, very little can be said about parameters which quantitatively describe the magnitude of secondary - 7 9 -compression. Mesri and Castro ( 1987) suggest that, for a given soil, the ratio of C a to Q , the slope of the virgin compression portion of a laboratory consolidation curve, remains essentially constant, and that for a majority of inorganic, soft clays the ratio ^ • = 0 . 0 4 ± 0 . 0 1 [6.17] Use of this relationship to detennine C a , however, requires an estimate of the laboratory-determined parameter, C c . Obviously, application of this concept to the full-scale field case requires many assumptions. Secondary compression is often ignored in practice. For the embankment site, the calculated time for completion of primary consolidation is approximately 3 5 years. Although it is generally assumed that secondary consolidation will occur after this time, it is likely that some secondary compression has occurred at the embankment site. In the present analysis, however, the component of settlement due to secondary compression has not been included. 6 .2 Modified Approach Where the thickness of a compressible stratum is large relative to the loaded area, the three-dimensional nature of the problem influences the magnitude and rate of settlement. Semi-empirical approaches are used in order to modify settlement magnitude to account for these effects. When there is an axis of symmetry in a field loading case, as in the centreline of the subject embankments, Skempton and Bjerrum (1957) give an expression whereby consolidation settlement beneath the centreline, incorporating 3 - D effects, can be expressed in terms of the settlement predicted from a 1 - D test: S S - D ^ S L D [6.19] where the correction factor, X , is a function of stress history, and X = A + P(l-A) [6.20] -80-where d I AO^C I Z P—5 [6-21] and A is the Skempton pore pressure parameter expressing the proportion of the principal stress difference which is responsible for the increase in pore water pressure. For normally consolidated soils, the factor X approaches unity, therefore no correction was required in the present analysis. For overconsolidated soils, however, an adjustment should be made. Skempton and Bjerrum (1957) provide a chart for determining X based on overconsolidation ratio and the foundation size relative to the thickness of the compressible stratum. It should be noted that some account has been taken for foundation size with respect to the thickness of the substrata by the correction factors C i and C2 used in the 1-D analysis for calculating distortion settlement. 6.3 Complex Approach The advent of the personal computer has spawned increased interest in the use of numerical methods and finite element methods (FEM) to compute foundation settlements. While the prospect of handling non-linear constitutive relationships and complicated boundary conditions in a settlement analysis is certainly appealing, the material parameters required as input are not generally obtainable with the degree of accuracy required to justify a sophisticated analysis. Often, the approximations required to fit the real problem to that for which a F E M solution is available are inconsistent with the precision of the solution procedure (Perloff, 1975). Therefore, approximate analysis of settlements, or the one-dimensional, simplified, approach often remains appropriate and is still the most common method used in geotechnical practice. - 81 -Burland (1987), in reviewing the most commonly-used methods of settlement prediction for clay soils, demonstrated that traditional settlement calculations are usually adequateforpractical purposes, provided the appropriate in-situ soil properties have been obtained. The settlement magnitudes predicted by conventional 1 -D analysis, the Skempton and Bjerrum method, the stress path method, and me finite element method were evaluated as ratios to the exact theoretical values in the cases of a homogeneous, isotropic, elastic material, a homogeneous, anisotropic, elastic material, and an anisotropic material with increasing stiffness with depth. In most cases, the simple 1-D method gave the best predictions of total settlement, raising the questions of whether the sophistication of F E M analysis is necessary, and whether greater accuracy is, in fact, achieved. The findings of Burland (1987) emphasize the importance of assuring a high quality of geotechnical data, obtaining as accurately as possible values such as m v (=1/M), E and G, whereafter a simple 1-D settlement analysis should prove sufficient for most practical applications. A finite element settlement analysis was not conducted as part of the present research. 6.4 Comparison of Predictions with Observed Settlement Fig. 6.6 and Fig. 6.7 present the combined results of the first and second steps of the one-dimensional settlement analyses, showing the time rate of settlement predicted using in-situ test results with the observed settlements recorded by Transport Canada for the McConachie Way Overpass embankments and the south approach embankment of the Arthur Laing Bridge, respectively. 6.4.1 McConachie Way Overpass Embankments Both rate and magnitude of settlements were predicted with a high degree of accuracy by the elastic embankment method, when compared to the observed settlement, until the time of abutment loading in 1975. Following construction of the abutments, settlements predicted by in-situ test methods was greater than those observed. The rate of settlement, however, Settlement (cm) -175 6 8 10 12 14 Time (years from January 1, 1970) 16 18 20 Fig. 6.6 Predicted and Observed Settlements — McConachie Way Overpass Embankments TXJ*—i—i—i—i—i—:—i—i—r—r—i—i—i—;—r~i—r—i—i—i—i—i—I—i—i—i—i—i—i—i—i—r— Settlement (cm) -25 -50 -75 •100 •125 •150 •175 • • ' • 1 • ' • • ' ' • ~I—I—I—I—I—i—I—i—I—I—I—I—i—•—I—t— Arthur Laing Bridge South ApproachEmbankment • Predicted settlement— \"Normal Load\" method • Predicted settlement— \"Elastic Embankment\"method Observed settlement • f i . i ° ° g . B . A — -.^ .^ .•.•.p-, u -..A JA.A. T — i — i — i — r ~ 6 8 10 12 14 Time (years from January 1, 1970) 16 18 20 Fig. 6.7 Predictedand Observed Settlements — Arthur Laing Bridge South ApproachEmbankment -84-continued to closely model the actual rate, as all settlement curves on Fig. 6.6 are essentially parallel. The normal loading approximation, overpredicts observed settlement from the outset of construction, however this result is not unexpected, as Boussinesq solutions, while widely used, are generally found to be conservative. At the time of the most recent Transport Canada survey (December, 1986), total settlement had reached 106 cm. Using in-situ test parameters, the settlement predicted by the elastic embankment method was 116 cm, an overprediction of 10 cm (9%), and the settlement predicted by the normal loading approximation was 143 cm, an overprediction of 37 cm (35%). This represents a significant improvement over the original predictions given by Bertok (1987), based on laboratory data, where the predicted settlement due to abutment construction alone was 20 cm compared to the 33 cm observed, a difference of 13 cm (40%). 6.4.2 Arthur Laing Bridge South Approach Embankment Rate and magnitude of settlements as predicted from in-situ test methods match observed settlements with remarkable precision. Fig 6.7 indicates that throughout the construction, preloading, and abutment construction phases, predicted settlements are very similar to the observed settlements. Following abutment construction, settlement is slightly overpredicted by in-situ test methods. The predicted rate of settlement, however, continued to closely, model the actual rate through to 1987. 1 Differing from the McConachie Way Overpass embankments, the settlements predicted by the normalloading approximation matched the observed settlements more closely than that predicted by the elastic embankment method, although the difference is small. This may result from the assumption inherent in the Boussinesq analysis, that the loaded area is an infinite strip. The size and shape of the south approach embankment is closer to the assumed infinite strip than the embankments for the McConachie Way Overpass. -85-At the time of the most recent Transport Canada survey (November, 1987), total settlement had reached 126 cm. Using in-situ test parameters, the settlement predicted by the elastic embankment method was 143 cm, an overprediction of 17 cm (13%), and the settlement predicted by the normal loading approximation was 135 cm, an overprediction of 9 cm (7%). Again, this represents some improvement over the original predictions, where the predicted settlement due to abutment construction alone was 23 cm compared to the 20 cm observed, a difference of 3 cm (15%). In general, the elastic embankment method1 proposed by Perloff et al. (1967) provided better predictions of settlement. This case record is not sufficient to state which is the better approach, however, since both the elastic embankment and the normal loading approximation methods yielded reasonable results. i - 86-7. DISCUSSION The present analysis cannot be considered as a Class A estimate, since actual settlement data have been published, neither can it be considered as a Class C estimate, as presently-available data were not back analysed in an effort to refine estimated parameters. Back analysis of parameters from field monitoring, in fact, was not attempted, due to lack of monitoring data in this published case history. It is unfortunate that the lack of monitoring data, such as pore pressure and surface and deep settlement measurements due to embankment loads, prohibited the evaluation of interpreted geotechnical parameters. The present study would have been improved if detailed monitoring data were available. The good prediction of settlements using in-situ test data may have been due to counterbalancing errors. Nevertheless, the results of this research, in its present form, have a twofold significance. Firstly, it has been shown that, for this embankment case history, settlement magnitudes can be predicted with reasonable confidence based on parameters interpreted from in-situ tests. Also, it has been shown that the detailed stratigraphic information gathered using in-situ tests provides a solid basis for accurate prediction of the rate of settlement by increased precision in the identification of potential drainage layers within the soil profile. Secondly, it has been demonstrated, for this embankment case history, that a simple, one-dimensional analysis can adequately predict settlements. For the south approach embankment of the Arthur Laing Bridge, predicted performance paralleled the observed performance with a degree of accuracy not often found in the prediction of settlement for large structures founded on compressible soils. For the McConachie Way Overpass embankment, performance predicted by in-situ test methods proved to be an improvement over that predicted by conventional methods, however, the predictions did not parallel observations as closely as in the case of the Arthur Laing Bridge south approach embankment. The original -87-findings outlined in Bertok (1987) also were indicative of poorer performance predictions for the McConachie Way Overpass embankments. The less precise performance prediction for the McConachie Way Overpass embankments may stem from several causes. However, a most obvious problem unique to these embankments is evident in Table 6.1 where a breakdown of the surface load imposed by each component of construction shows that the surcharge was insufficient to account for the final load imposed by the abutments. This problem may derive from two sources: one, that the preload itself was of inadequate thickness, or two, that placement of the footing was such that a larger portion of the abutmentload than anticipated was placed on the original ground surface. Bertok (1987) states that abutments of the McConachie Way Overpass were founded on spread footings located in the compacted sand fill approximately 1 m above the original ground surface. A review of the original design drawings (Phillips, Barratt, Hillier, Jones and Partners, 1974) confirmed that the underside of the spread footing was to rest 0.99 m (3.25 ft.) above original grade. Being placed at this elevation, full use was not made of me load-spreading capacity of the compacted embankment sand fill, and indeed, large loads would be applied to the original ground surface. Therefore, although the ground had been consolidating under the embankment and surcharge loads, once the final phase of construction began, the loads induced by the abutments soon exceeded the maximum past pressure, and virgin compression conditions once more came into effect. Aside from the above complications due to this load-unload-reload sequence, considerable uncertainty exists in the distribution of excess pore pressure at this stage of construction. The excess pore pressures induced by the embankments and surcharge had not reached equilibrium before the abutment load was applied, imposing another, unknown, distribution of excess pore water pressure. As a result, prediction of settlement following abutment construction becomes a task laden with added uncertainty, and the results obtained appear to be far above what might be expected, even for the McConachie Way Overpass embankments. -88-In an effort to further refine the prediction of performance, perhaps the most influential parameter is the deformation characteristic, or modulus. The portion of settlement due to distortion, approximately 30% of the total, is directly dependent upon Young's, modulus, E , whereas the consolidation settlement, in a one-dimensional analysis, is directly dependentuponthe constrained modulus, M . For the original analysis performed in 1968, the values of Young's modulus assumed in order to compute initial elastic settlements were: (48 MPa in cohesionless silty sand E = [7.1] .1000SU in cohesive soils For this study, based on the in-situ screw plate test, rl05 MPa in cohesionless silty sand E = [7.2] 1(200 to 300)Su in cohesive soils Comparing these values of modulus, however, does not automatically account for the differences in settlements predicted in 1968 and by the present analysis. While more settlement in cohesionless soils would have been predicted in the original analysis, due to a modulus roughly half as large as that interpreted from the screw plate test, it is not clear whether or not this would be counterbalanced by the fact that a larger modulus may have been predicted for the cohesive deposit, based on the correlations with undrained strength. The constant of proportionality between E and S u is much larger in the case of the original analysis, however, Su determined by the 1968 laboratory testing program was consistently lower than that interpreted from in-situ testing, as indicated on Fig. 5.6. The empirical nature of the interpretation of constrained modulus makes its value highly questionable. An increase or decrease in M of 25% would result in a decrease or increase, respectively, in consolidation settlement of approximately 20%. Fortified by local experience, however, and given the consistency in correlations found in the present research and reported in -89-the literature, a high degree of confidence may be expressed in the values used for this analysis. While some inadequacy may be suggested in the adjustment of M to account for changes in stress history, based on the performance of the south approach embankment of the Arthur Laing Bridge, it appears that this methodology produces reasonable results. Changes in the assumed vertical stress increase, Aa z , have approximately the same effect on the predicted settlement as changes in M , that is, a 25% increase in A a z results in a 20% increase in computed settlement. In the original analysis, as reported in Bertok (1987), the Osterberg method, an elastic theory method utilizing influence factors and superposition, was used to compute stress increase. Of the methods used in the present analysis, both appear to be conservative, but distinctly adequate. The importance of the two-dimensional nature (shape) of the loaded area is evidenced here, as deviation from the assumed condition of an infinite strip appears to result in increasing conservatism in the estimation of vertical stress increase. The value of Poisson's ratio, u, affects the calculated distortion settlement, but only as a second order term, and it is less influential on the final result than the previously-discussed parameters. Use of u= 0.5 in the present analysis assumes compressible soil deforms at constant volume. Researchers (Bozozuk and Leonards, 1972) have suggested values of u = 0.3 to 0.35 may be more appropriate due to the inherent anisotropy of many cohesive soils, however, in-situ tests do not provide the means for determining the appropriate in-situ value. No indication has been given of the original prediction of settlement rate. Bertok (1987) stated only that reliable prediction of the rate of settlement to be expected at this site was very difficult, adding that the reliability of predictions was tenuous due to sampling and testing problems in the soft, cohesive soil, and to lack of understanding of drainage effects from only one deep test hole. By contrast, the simplicity and versatility of in-situ tests such as the C P T U and DMT are clear advantages. These in-situ tests provided excellent stratigraphic detail and definition of -90-potential drainage seams, enabling rational decisions to be made on site geometry and drainage. As a result, when combined with pore pressure dissipation test information, rate of settlement was predicted with precision. In relating soil properties determined in the laboratory with those determined from field measurements, Olson (1985) comments that the ultimate check on the usefulness of laboratory data is a comparison between predictions and field measurements. The same may be said for soil properties determined from in-situ tests. The ultimate check, in the case of the present performance prediction, provided a favourable endorsement for using in-situ tests to predict both rate and magnitude of settlement. Recognizing this, it would appear that in-situ testing promises to be a preferential method of conducting geotechnical investigations. Still, the fact remains that there is some resistance to relying on in-situ testing. Despite the cost effectiveness of the C P T U and DMT, a substantial investment of both capital and technical expertise is required by the consultant, contractor, or agency committed to the successful use of in-situ testing. Although laboratory testing of field samples will likely continue to be the basis for numerous geotechnicalinvestigalions and analyses for some time, in-situ testing offers the ability to enhance the quality of such investigations and will increasingly be recognized as a viable alternative or addition. -91 -8. CONCLUSIONS A N D RECOMMENDATIONS In-situ testing is not without its drawbacks, and many areas remain where the complex soil behaviour duringpenetration is not well understood, however, given the limitations of alternative techniques, it becomes evident that practical problems can be handled with confidence when parameters have been interpreted from in-situ test results based on local correlations, experience, and sound judgment Based on the present research, the in-situ testing program undertaken provided adequate information concerning the settlement properties of the sensitive clay silt stratum which figured prominently in predicting embankment performance. Hence, the fouowing conclusions can be made. The soil profile beneath the areas covered by the south approach embankment of the Arthur Laing Bridge and McConachie Way Overpass embankments is reasonably uniform. The subsoil conditions consist of approximately 20 m of predominantly sand, underlain by an approximately normally consolidated clay silt to a depth of approximately 61m at the north end of the site, and to a greater depth to the south. Underlying the clay silt is the Pleistocene till. More free draining sandy silt layers occur at a depth of about 30 m and were clearly identified in several tests. Thus these layers may be assumed to exist in sufficient areal extent to act as effective drainage layers for excess pore pressures generated from surface loads. The undrained shear strength in the clay silt appears to increase linearly with depth from 50 kPa at a depth of 20 m to 90 kPa at 60 m. The undrained Young's modulus, Eu, of r (200 to 300)SU by SPLT [8.1] 100SU by C P T U dissipation test correlations appears to provide a useful correlation, since Su is often more readily and consistently obtained by a variety of methods than Eu. The range of rigidity index obtained by the two test methods, -92-E/Su = 100 to 300, shows good agreement in view of the overall differences in the cone penetration and screwplate tests. The correlation found between constrained modulus, M , and cone bearing, q c , of M = 2.4 q c [8.2] is in general agreement with previously published correlations given in the literature. The Special Method of Schmertmann (1986) provides a rational method by which to modify M when design stress conditions are other than those which existed at the time of in-situ testing. Rezocone dissipation tests yielded a detailed record from which to assess the consolidation characteristics of the clay silt. Such dissipation tests are sensitive to changes in the behaviour type of soil in the vicinity of the porous element, and thereby provide a means of identifying potential drainage paths. Dissipation tests conducted with the flat dilatometer did not show the same sensitivity to potential drainage paths. The spherical solution of Torstensson (1977) with E/Su = 100 appeared to provide the best model for predicting consolidation characteristics of the Sea Island clay silt, and a horizontal coefficient of consolidation, Ch = 0.004 cm2/s, was effective in predicting the rate of consolidation settlement. Based on the results of the present analysis, it appears that horizontal drainage dominates, andthatthe consolidation was controlled by ch. It was not found necessary to adjust the value of ch(piezocone) to ch(NC) from dissipation tests conducted for a length of time sufficient to determine tso. The piezocone penetration test (CPTU) was capable of providing all the required information pertinent to a settlement analysis, especially when correlations are enhanced by local experience. The flat dilatometer test (DMT) was found to be adequate as a stand-alone test for determining the parameters required to predict settlement. Dissipation tests with the dilatometer -93-show promise in determining consoidation characteristics, however, longer times to reach tso are required, as compared to the CPTU. The effective radius, R = 2.057 cm, used in this analysis to determinech from the dilatometer, provided good results, but requires local experience in a variety of soils before its general use may be recommended. The appeal of the screw plate test lies in the correct orientation of the test for predicting vertical settlement. Site conditions, however, such as the dense sand encountered at the embankment site, may render the test difficult and cumbersome to conduct and may influence the interpretedparameters. Additionally, the screw plate test alone did not furnish all the required information for a settlement analysis. Researchers have reported success in determining consolidation characteristics from the SPLT, however, these tests have been in stiff clays. At Sea Island, in the soft, sensitive, compressible soils at depth, it was not possible to simulate small, constant stress increments and thereby ascertain displacement response. The normal load approximation, or Boussinesq method, appears to provide a conservative prediction of the distribution of vertical stress increase with depth. When the embankment shape more closely resembles an infinite strip, the Boussinesq method more accurately predicts stress increase. However, the elastic embankment method outlined by Perloff (1967) appeared to give a more realistic stress distribution for the present analysis. It would be instructive if settlement monitoring of the subject embankments could be continued into the year 2000. This would provide some assessment of the effects of secondary compression. At the present time, there is uncertainty as to when secondary compression begins andhowto quantitatively account for it in practice. The process is not well understood, and it is not known if parameters interpreted from in-situ tests inherently include some measure of the secondary compression which may be on-going from the time of initial soil deposition. The analytical techniques existing in the geotechnical community today appear to be adequateforconducting a one-dimensional settlement analysis. Complex finite element methods -94-ornumerical analyses generally are notrequiredfor settlement calculations, and do not necessarily provide results superior to the 1 -D analysis. More important than the type of analysis performed is the accurate determination of soil parameters, in particular deformation and consolidation characteristics. The greatest effort should be expended in deterniining these characteristics as accurately as possible. From this analysis, in-situ testing emerges as a viable alternative to the traditional approach of obtaining geotechnical parameters required in the prediction of settlement. While interpretation of in-situ test data is, by and large, empirical in nature, the large amount and diversity of the data obtained enables the engineer to obtain a better sense of site conditions and variability, leading to a generally more reliable geotechnical solution. -95-9. BIBLIOGRAPHY Baldi, G., Bellotti, R., Ghionna, V. , Jamiolkowski, M . and Pasqualini, E. (1981). \"Cone Resistance of a Dry Medium Sand\", Proceedings, 10th International Conference on Soil Mechanics and Foundation Engineering, Stockholm, Vol. 2. Baligh, M . M . and Levadoux, J.-N. (1986). \"Consolidation after Undrained Piezocone Penetration. II: Interpretation\", Journal of Geotechnical Engmeering, Vol. 112, No. 7, pp. 727 - 745. Bertok, J. (1987). \"Settlement of embankments and structures at Vancouver International Airport\", Canadian Geotechnical Journal, Vol. 24, pp. 72 - 80. Berzins, W. E . (1983). \"Determination of Drained and Undrained Soil Parameters Using the Screw Plate Test\", M . A. Sc. Thesis, The University of British Columbia. Blunden, R. H . (1973). \"The Urban Geology of Richmond, British Columbia\", Adventures in Earth Sciences Series No. 15, Department of Geological Sciences, The University of British Columbia. Boussinesq, M . J. (1885). ApplicationDesPotentiels, a L'Etude del'Equihbre etdu Mouvement Des Sondes Elastiques. Paris: Gauthier-Villars. Bozozuk, M . , and Leonards, G. A. (1972). \"The Gloucester Test-Fill\", Proceedings, ASCE Specialty Conference, Performance of Earth and Earth-Supported Structures, Purdue University, Lafayette, Vol. 1, Part 1, pp. 299 - 317. Burland, J. B. (1987). \"Analysis of Settlement on Clays\", Short Course on Recent Developments in Foundation Analysis and Design, Imperial College, April, 1988. Campanella, R. G. and Robertson, P. K. (1988). \"Current Status of the Piezocone Test\", Proceedings of the First International Symposium on Penetration Testing, ISOPT-1, Orlando. Campanella.R. G., Robertson, P. K. and Gillespie, D. (1985). \"A Seismic Cone Penetrometer for Offshore Applications\", Soil Mechanics Series No. 93, The University of British Columbia. Campanella, R. G. and Robertson, P. K. (1981). \"Applied Cone Research\", Soil Mechanics Series No. 46, The University of British Columbia. Douglas, B. J. and Olsen, R. S. (1981). \"Soil Classification Using Electric Cone Penetrometer\", Symposium on Cone Penetration Testing and Experience, Geotechnical Engineering Division, ASCE St. Louis, pp. 209 - 227. Fadum, R. E. (1948). \"Influence values for estimating stresses in elastic foundations\", Proceedings of the Second International Conference on Soil Mechanics and Foundation Engineering Vol. 2. Foster, C. R. and Ahlvin, R. G. (1954). \"Stresses and deflections induced by a uniform circular load\", Proceedings, Highway Research Board, Vol. 34, p. 467. -96-Gillespie, D. G. (1981). \"The Piezometer Cone Penetration Test\", M . A. Sc. Thesis, The University of British Columbia. Harr, M . E . (1966). Foundations of Theoretical Soil Mechanics. New York: McGraw-Hill. Jamiolkowski, M . , Ladd, C. C , Germaine, J. T. and Lancellotta, R. (1985). \"New developments in field and laboratory testing of soils\",Proceedings ofthe 11th International Conference on SoilMechanics andFoundation Engineering, San Francisco, pp. 57 - 153. Janbu, N. (1967). \"Settlement calculations based on the tangent modulus concept\", University of Trondheim, Norwegian Institute of Technology, ButtetinNo. 2. Kay, J. N. and Avalle, D. L . (1982). \"Application of Screw Plate to Stiff Clays\", Journalofthe GeotechnicalEngineeringDivision, ASCE Vol. 108(GT1), pp. 145 - 154. Kummeneje, O. (1956). \"Foundation of an Oil Tank in Drammen\", Norwegian Geotechnical Institute Publication No. 12. Levadoux, J.-N. and Baligh, M . M . (1986). \"Consolidation after Undrained Piezocone Penetration. I: Prediction\", Journal of Geotechnical Engineering, Vol. 112, No. 7, pp. 707 - 726. Lunne, T. and Kleven, A. (1981). \"Role of CPT in North Sea Foundation Engineering\", Symposium on Cone Penetration Testing and Experience, Geotechnical Engineering Division, ASCE, pp. 49 - 75. Marchetti, S. (1980). \"In Situ Tests by Flat Dilatometer\", Journal of the Geotechnical Engineering Division, ASCE Vol. 106(GT3), pp. 299 - 321. Marchetti, S. and Crapps, D. K. (1981). \"Flat Dilatometer Manual\", GPE Inc., Gainesville, Florida. Mesri, G. (1973). \"Coefficient of Secondary Compression\", Journal of the Soil Mechanics and Foundations Division, ASCE Vol. 99(SM1), pp. 123 - 137. Mesri, G. and Castro, A. (1987). \"C a /Cc concept and K Q during secondary compression\", JoumalofGeotechmcalEngmeering, Vol. 113, No. 3, pp. 230 - 247. Mitchell, J. K. and Gardner, W. S. (1975). \"In-Situ Measurement of Volume Change Characteristics\", State-of-the-Art Report, Proceedings of the Conference on In-situ Measurement of Soil Properties, Specialty Conference of the Geotechnical Division, ASCE North Carolina State University, Raleigh, Vol. U. Mitchell, J. K., Guzikowski, F. and Villet, W. C. B. (1978). \"The Measurement of Soil Properties In-situ\", Report prepared for the U . S. Department of Energy, Contract W-7405-ENG-48, Lawrence Berkeley Laboratory, University of California, Berkeley. Olson, R. E. (1985). \"State of the Art Consolidation Testing\", Consolidation of Soils: Testing and Evaluation, ASTM Special Technical Publication 892, R. N. Yong and F. C. Townsend, Eds., pp. 58 - 69. Perloff, W. H . (1975). \"Pressure Distribution and Settlement\", Foundation Engineering Handbook, H . F. Winterkorn and H.-Y. Fang, Eds., New York: Van Nostrand Reinhold. -97-Perloff, W. H. , Baladi, G. Y. and Harr, M . E . (1967). \"Stress distribution within and under long elastic embankments\", Highway Research Record, No. 181. Phillips, Barratt, Hillier, Jones and Partners, (1974). Grant McConachie Way No. 1 Overpass, Drawings No. 7043-210D, 21 ID. Robertson, P. K . (1988). \"Penetration Testing in the U. K . \" Birmingham, England, July 6 - 8, 1988. Robertson, P. K . , Campanella, R. G , Gillespie, D. and By, T. (1988). \"Excess Pore Pressures and the DMT\", Proceedings of die First International Symposium on Penetration Testing, ISOPT-1, Orlando. Robertson, P. K . and Campanella, R. G. (1986). \"Guidelines for Use, Interpretation and Application of the CPT and CPTU\", Soil Mechanics Series No. 105, Department of Civil Engineering, The University of British Columbia. Robertson, P. K . (1985). \"In-Situ Testing and Its Application to Foundation Engineering\". 1985 Canadian Geotechnical Colloquium, Canadian Geotechnical Journal Vol. 23. Schleicher, F. (1926). \"ZurTheorie Des Baugrundes\", DerBauingenieur, No. 48, 49. Schmertmann & Crapps, Inc. (1988). \"Guidelines for Geotechnical Design Using the Marchetti DMT\", Pennsylvania Department of Transportation, Project 84-24 Manual, Vol. III. Schmertmann, J. H . (1986). \"Dilatometer to compute foundation settlement\", Proceedings of hi Situ 86, Specialty Conference of the Geotechnical Engineering Division, ASCE, Blacksburg, Virginia., pp. 303 - 321. Schmertmann, J. H . (1983). \"Revised Procedure for Calculating K Q and OCR from DMT's with Ir>>1.2 and which Incorporate the Penetration Force Measurement to Permit Calculating the Plane Sixain Friction Angle\", DMT Workshop, Gainesville, Florida. Schmertmann, J. H . (1970). \"Static Cone to Compte Static Settlement Over Sand\", Journal of the Soil Mechanics and Foundations Division, ASCE Vol. 96(SM3), pp. 1011-1043. Scott, R. F. (1963). Principles of Soil Mechanics, Reading, Mass.: Addison-Wesley Pubfohing Co., Inc. Selvadurai, A. P. S. and Gopal, K . R. (1986). \"Consolidation Analysis of the Screw Plate Test\", Proceedings, 39th Canadian Geotechnical Conference, In Situ Testing and Field Behaviour Ottawa, pp. 167 - 178. Selvadurai, A. P. S. and Nicholas, T. J. (1979). \"A Theoretical Assessment of the Screw Plate Test\", Proceedings, 3rd InternationalConference on Numerical Methodsin Geomechanics, Aachen, Germany, Vol. 3. Selvadurai, A. P. S., Bauer, G. E. and Nicholas, T. J. (1980). \"Screw plate testing of a soft clay\", Canadian Geotechnical Journal, Vol. 17, No. 4, pp. 465 - 472. Skempton, A. W. and Bjerrum, L . (1957). \"A contribution to the settlement analysis of foundations on clay\", Geotechniquq Vol. 7, No. 3. Terzaghi, K . (1943). Theoretical Soil Mechanics, New York: John Wiley and Sons. -98-Thurber, R. C. and Associates Ltd. (1968). \"Foundation Soils Investigation, Proposed Hudson Street Bridge Main Piers\", Report No. 1 to Phillips, Barratt, Hillier, Jones and Partners. Torstensson, B.-A. (1977). \"The Pore Pressure Probe\", Nordiske Geotekniske Mete, Oslo, Paper No. 34.1-34.15. \\ -99-APPENDIX A - Tables Relating Degree of Consolidation and Dimensioriless Time Factor - 100-Table A. 1 Values of the Dimensionless Time Factor, T Method Degree of Consolidation (%)* 20 40 50 60 80 Baligh and Levadoux(1986) 0.69 3.0 5.6 10 39 Torstensson Spherical (1977): E/Cu = 500 0.11 0.46 0.81 1.26 3.28 400 0.10 0.40 0.68 1.12 2.85 300 0.085 0.35 0.61 0.98 2.36 200 0.066 0.28 0.47 0.77 1.91 100 0.057 0.20 0.32 0.50 1.16 Torstensson Cylindrical (1977): E/Cu = 500 0.34 2.14 4.29 8.33 23.60 400 0.30 1.75 3.57 6.79 21.00 300 0.24 1.38 2.81 5.37 16.29 200 0.18 1.06 2.32 3.82 10.13 100 0.14 0.83 1.37 2.49 5.03 * Degree of consolidation, in percent = (1 -U)100% - 101 -Table A.2 Average Degree of Consolidation for Various Values of Dimensionless Time Factor, T Distribution of initial excess pore water pressure: 2d Doubly-drained stratum Constant Linear variation Case 1 Half sine Sine Triangular curve curve Case 2 Case 3 Case 4 T U (%) Case 1 Case 2 Case 3 Case 4 0.004 7.14 6.49 0.98 0.80 0.008 10.09 8.62 1.95 1.60 0.012 12.36 10.49 2.92 2.40 0.020 15.96 13.67 4.81 4.00 0.028 18.88 16.38 6.67 5.60 0.036 21.40 18.76 8.50 7.20 0.048 24.72 21.96 11.17 9.60 0.060 27.64 24.81 13.76 11.99 0.072 30.28 27.43 16.28 14.36 0.083 32.51 29.67 18.52 16.51 0.100 35.68 32.88 21.87 19.77 0.125 39.89 36.54 26.54 24.42 0.150 43.70 41.12 30.93 28.86 0.175 47.18 44.73 35.07 33.06 0.200 50.41 48.09 38.95 37.04 0.250 56.22 54.17 46.03 44.32 0.300 61.32 59.5 52.30 50.78 0.350 65.82 64.21 57.83 56.49 0.400 69.79 68.36 62.73 61.54 0.500 76.40 76.28 70.88 69.95 0.600 81.56 80.69 77.25 76.52 0.700 85.59 84.91 82.22 81.65 0.800 88.74 88.21 86.11 85.66 0.900 91.20 90.79 89.15 88.80 1.000 93.13 92.80 91.52 91.25 1.500 98.00 97.90 97.53 97.45 2.000 99.42 99.39 99.28 99.26 - 102 -APPENDIX B - Field Test Data U B C I M S I T U T E S \" r i M G S i t e L o c a t i o n ! L a i n g B r i d g e S. C P T D a t e i 0 8 / 1 2 / 8 7 1 0 i 5 8 P a g e Noi 1 / 2 On S i t e L o c i M c C o n a c h i e 0 / P C o n e U s e d i UBC#8 S t d / B F S PP C o m m e n t s ! C P T U - 1 DIFFERENTIAL P.P. PORE PRESSURE CONE BEARING SLEEVE FRICTION FRICTION RATIO RATIO AU/Ot U <•. of .ater) Qt (bar) (bar) Rf (X) O U B C I M S I T U T EE S r i M G S i t e L o c a t i o n i L a i n g B r i d g e S. CPT D a t e i 0 9 / 1 1 / 8 7 1 2 i 0 2 P a g e Noi 1 / 2 On S i t e L o c i M c C o n a c h i e 0 / P C o n e U s e d i H o g S u p e r C o n e Commente i C P T U - 5 DIFFERENTIAL P.P. PORE PRESSURE CONE BEARING SLEEVE FRICTION FRICTION RATIO RATIO AU/Qt U (m. of water) Qt (bar) (bar) Rf Ctt D E P T H ( m e t Q r s ) CD | \" V ° ? P P o A , , ll | ~ 3 O C3 O 3 p c o D O c IS i > o 1 ' o 3 o r o 0 D c» 0 3 o r n o o M-3 3 Q (Q o — -i o •» a O 10 N e \"D c DO n z in n o o TJ B o H C D a m m m m u r n M i l l n m m n m i l m m i m i m m .60 1421. 1.20 3.90 .00 .000 63. 1.25 14.(2 1.700 .099 1.14 11.50 1.20 46.0 179.2 SANDY SILT .80 954. .92 2.50 .00 .000 22. .52 9.37 1.600 .130 1.45 11.13 1.77 .197 53.2 SILTY CLAY 1.00 331. .61 1.90 .00 .000 11. .35 5.79 1.S00 .1(0 .84 5.25 1.29 .133 21.9 HUD 1.20 186. .69 2.00 .00 .000 12. .35 5.25 1.(00 .131 .86 4.51 1.20 .141 22.1 CLAY 1.40 164. .80 1.75 .00 .015 -1. -.03 P01 » . 1.13 P0 -- 1.13 PI = 1.10 QUESTIONABLE .1.60 130. 1.00 1.80 .00 .034 -6. -.14 P01 = 1.34 P0 = 1.33 PI = 1.15 QUESTIONABLE 1.80 8. .61 2.00 .00 .054 15. .50 3.74 1.(00 .231 .(I 2.(6 .94 .111 22.3 SILTY CLAY 2.00 430. .61 3.52 .00 .074 70. 2.(3 3.14 1.700 .245 .49 1.39 .57 34.4 103.1 SILTY SAND 2.20 564. .95 4.50 .00 .093 94. 2.55 4.09 1.700 .259 .72 2.78 .66 35.5 159.9 SILTY SAND 2.40 542. 1.28 1.70 .00 .113 -19. -.37 P01 = 1.64 P0 = 1.61 PI = 1.05 QUESTIONABLE 2.60 576. 1.19 1.25 .00 .132 -32. -.66 P01 = 1.57 P0 = 1.52 PI = .60 QUESTIONABLE 2.80 593. 1.00 4.70 .00 .152 99. 2.74 3.44 1.800 .303 .(6 2.17 .59 35.4 1S4.1 SILTY SAND 3.00 664. 1.25 (.40 .00 .172 148. 3.2B 4.10 1.900 .319 .91 2.85 .67 35.4 255.7 SILTY SAND 3.20 642. 1.60 6.20 .00 .191 132. 2.44 4.66 1.900 .334 1.23 3.63 .77 34.2 238.6 SILTY SAND 3.40 587. 1.47 5.70 .00 .211 119. 2.39 4.08 1.800 .350 1.08 3.10 .72 33.6 199.5 SILTY SAND 3.60 698. 1.80 8.09 .00 .231 193. 3.41 4.47 1.900 .366 1.26 3.44 .75 34.5 348.3 SAND 3.80 854. 1.B5 B.25 .07 .250 197. 2.43 4.35 1.800 .381 1.17 3.07 .69 36.1 350.9 SAND 4.00 2.02 12.58 .00 .270 349. 6.23 .00 1.300 m m 23S.7 SAND 4.20 2.01 12.31 .00 .289 359. 6.61 .00 1.900 m m 304.1 SAND 4.40 2.25 12.20 .00 .309 327. 5.17 .00 1.900 m m •77.8 SAND 4.60 2.35 13.13 .00 .329 357. 5.53 .00 1.900 m m 303.5 SAND 4.80 0. 2.20 11.50 .00 .348 303. 4.95 .00 1.300 m i l l 257.7 SAND 5.00 2.70 13.00 .00 .368 340. 4.46 .00 1.300 m m 238. ( SAND 5.20 2.80 12.30 .00 .388 329. 4.13 .00 1.300 m m 273.3 SAND ) 2.10 12.30 .00 .407 354. 6.64 .00 1.300 m m 201.0 SAND „.oO 1.48 13.85 .00 .427 415. 14.70 .00 1.800 m m 352.7 SAND 5.80 1299. 2.40 16.(0 .30 .447 492. 8.56 .00 1.900 m m 409.4 SAND 6.00 1344. 4.80 17.(0 .00 .466 431. 3.05 .00 2.000 m m 3(6.1 SILTY SAND « . 2 » 1322. 2.80 15.80 .00 .486 438. 6.18 .00 1.900 m m 372.3 SAND 0 1065. 3.60 13.10 .00 .505 310. 2.98 . .00 1.900 m m 2(3.9 SILTY SAND o.60 1043. 2.30 12.80 .00 .525 347. 6.14 .00 1.900 m m 234.8 SAND 6.80 1599. 3.10 18.50 .11 .545 525. 7.00 .00 1.900 m m 446.6 SAND 7.00 1644. 4.02 4.39 .00 .S64 -21. -.16 P01 = 4.38 P0 = 4.35 PI « 3.74 SUESTI0NA3LE 7.20 3.4B 16.84 .00 .584 451. 4.99 .00 1.900 m m 383.4 SAND 7.40 1410. .65 13.70 .00 .604 419. 32.06 .00 1.900 m m P0I = .39 P0 = .9B PI = 13.05 356.0 SAND 7.60 1322. 1.80 8.40 .00 .623 205. 4.81 .00 1.800 m m 174.0 SAND 7.80 1455. 4.07 16.90 .25 .643 432. 3.33 .00 1.900 I I I I I I 367.0 SAND 8.00 1622. 1.60 2.35 .00 .662 -8. -.18 P01 = 1.94 P0 = 1.93 PI = 1.70 QUESTIONABLE 8.20 1766. 2.90 15.00 .00 .6B2 40S. 5.86 .00 1.900 m m 344.4 SAND 8.40. 1644. 4.00 18.50 .00 .702 493. 4.81 .00 1.900 I I I I I I 418.7 SAND 8.60 1789. 4.70 22.30 .00 .721 606. 5.02 .00 2.000 m m 514.7 SAND 8.80 1900. 3.50 17.80 .26 .741 485. 5.77 .00 1.900 m i n 412.5 SAND 9.00 1455. 4.10 18.00 .00 .761 471. 4.43 .00 1.900 m m 400.! SAND 9.20 1822. 2.80 20.80 .00 .780 620. 11.92 .00 1.900 m m 527.1 SAND 9.40 2401. 3.70 24.60 .00 .800 726. 9.36 .00 1.900 m m 616.3 SAND 9.60 2791. 4.60 30.50 .00 .819 908. 9.13 .00 2.000 m m 771.8 SAND 9.80 2B5B. 5.10 28.20 .30 .839 906. 6.66 .00 2.000 m m 685.1 SAND 10.00 2858. 5.00 29.20 .00 .859 846. 7.37 .00 2.000 m m 719.1 SAND 10.20 2624. 3.80 24.60 .00 .878 722. 9.21 .00 1.900 m m 613.8 SAND 10.40 2958. 3.70 28.20 .00 .898 857. 12.63 .00 1.900 m m 728.4 SAND 10.60 3147. 7.80 30.50 .00 .918 791. 3.72 .00 2.000 m m 672.7 SAND 10.80 2858. 7.90 26.80 .40 .937 653. 2.94 .00 2.000 m m 555.0 SILTY SAND 11.00 2757. 5.50 25.90 .00 .957 704. 5.19 .00 2.000 m m 539.3 SAND '• 20 2512. 4.60 19.50 .06 .976 507. 4.49 .00 1.900 m m 431.1 SAND 10 iBes. 3.90 19.50 .00 .996 533. (.13 .00 1.900 m m 452.8 SAN0_ 11.60 2346. 7.20 27.10 .00 1.016 689. 3.57 .00 2.000 m m 585.9 SAND 11.90 2356. 6.60 25.50 .45 1.035 653. 3.76 .00 2.000 m m 555.0 SAND 12.00 2179. 5.60 21.50 .00 1.055 544. 3.79 .00 2.000 m m 462.1 SAND 12.20 2056. 4.20 19.20 .00 1.075 511. 5.34 .00 1.900 m m 434.2 SAND 12.40 2702. 5.10 27.40 .00 1.094 777. (.85 .00 2.000 m m (60.3 SAND 12.60 2746. 6.10 26.90 .00 1.114 722. 4.81 .00 2.000 m m (13.8 SAND '12.80 2546. 6.70 23.90 .49 1.133 591. 3.35 .00 2.000 m m 502.3 SAND .. 13.00 1912. 4.00 16.80 .00 1.153 431. 4.80 .00 1.900 m m 3(6.1 SAND 13.20 2601. 5.70 27.60 .00 1.173 762. 5.76 .00 2.000 m m (47.9 SAND 13.40 2624. 4.60 21.40 .00 1.192 576. 5.(4 .00 1.900 I I I I I I 439.9 SAND 13.60 3292. 6.00 31.10 .00 1.212 879. (.47 .00 2.000 m m 747.0 SAND 13.80 3459. 6.50 30.40 .60 1.232 935. 5.41 .00 2,000 m m 709.9 SAND 14.00 3615. 8.00 33.50 .00 1.251 893. 4.40 .00 2.000 I I I I I I 759.4 SAND 14.20 3581. 7.60 25.50 .00 1.271 616. 3.06 .00 2.000 m m 524.0 SILTY SAND 14.40 2312. 4.20 20.20 .00 1.290 547. (.34 .00 1.900 m m 465.2 SAND 14.60 2468. S.30 22.00 .00 1.310 573. 4.(7 .00 2.000 m m 486.3 SAND 14.30 2590. (.40 25.60 .60 1.320 (64. 4.26 .00 2,000 m m 544.3 SAND 15.00 2230. 5.60 23.00 .00 1.349 598. 4.59 .00 2.000 m m 508.5 SAND - 107-4127. 8.10 32.80 .00 1.506 964. 4.34 .00 2.000 ttttti 734.6 SAXD 16.80 3949. 8.40 29.10 .80 1.526 718. 3.33 .00 2.000 m m 610.7 SAND 17.00 3514. 7.30 32.50 .00 1.546 682. 5.22 .00- 2.000 t u t u 750.1 SAND 17.20 3626. 6.10 32.10 .00 1.565 912. 7.27 .00 2.000 « m i 774.9 SAND 40 4349. 9.10 39.70 .00 1.585 1079. 4.89 .00 2.000 \" t t t i 917.2 SANO .60 4572. 8.30 39.50 .00 1.605 1101. 5.75 .00 2.000 ttttti 935.9 SANJ, 17.80 4733. 14.00 33.10 .60 1.624 879. 2.20 .01 2.150 H H I I 747. C SILTY SAND 19.00 3893. 8.50 29.40 .00 1.644 726. 3.38 .00 2.000 •***•! 616.9 SAND 18.20 5051. 13.20 40.00 .00 1.663 941. 2.56 .01 2.150 ttttti 739.6 SILTY SAND 18.40 4928. 10.60 40.00 .00 1.633 1035. 3.81 .00 2.150 880.2 SAND 18.60 4516. 15.10 40.00 .00 1.703 872. 2.00 .01 2.150 \" t i n 740.8 SILTY SAND 18.80 4461. a.30 38.40 1.30 1.722 1043. 5.03 .00 2.000 ttttti 886.4 SAND 19.00 5574. 14.70 40.00 .00 1.742 886. 2.12 .01 2.150 ttttti 753.2 SILTY SAND 19.20 5574. 11.10 40.00 .00 1.762 1017. 3.54 .01 2.150 ttttti B64.7 SANS 19.40 5296. 8.40 10.40 .00 1.781 37. .16 .00 1.800 ttttti .11 .00 -.54 .163 31.6 CUY 19.(0 4182. 14.40 39.60 .00 1.801 882. 2.17 .01 2.150 ttttt 750.1 SILTY SAND 19.30 3715. 6.30 11.20 1.03 1.820 143. .89 .00 1.800 H H I I .06 .00 -.55 .093 121.4 CLAYEY SILT 20.00 2302. 6.60 21.10 .00 1.840 493. 3.22 .00 2.000 ttttti 418.7 SILTY SAND 20.20 2691. 5.40 14.20 .00 1.860 285. 2.36 .00 1.900 ttttti 242.2 SILTY SAND 20.40 2023. 5.00 8.10 .00 1.879 77. .67 .00 1.700 ttttti .04 .00 -.55 .066 65.7 CLAYEY SILT 20.60 1344. 5.90 7.30 .00 1.899 15. .10 .00 1.700 ttttti .05 .00 -.55 .091 13.0 CLAY 20.80 720. 6.50 B.20 1.35 1.919 26. .16 .00 1.700 ttttt .06 .00 -.55 .106 22.3 CUY 21.00 6.50 7.60 .00 1.938 4. .03 .00 1.500 ttttti .06 .00 -.55 .106 3.7 HUD - 108-l THRUST A 8 C UO ED ID KD 6AHKA 5V PC OCR CO CU PHI II SOIL TYPE IN) (KG) (BAR) (BAR) (BAR) (BAR) (BAR) (1/113) (BAR) (BAR) (BAR) (DE6) (BAR) H I M m m m i l u r n n u t i i m i i n n u r n m m m m m m i n n u r n I I I I I u r n u r n m m i m i m m i 21.00 - ->o .0 21.60 21.80 22.00 22.20 22.40 -22.60 -22.30 23.00 23.20 23.40 23.60 23.80 24.00 24.20 24.40 24.60 24.80 25.00 25.20 25.40 25.60 25.80 26.00 25.20 ) 26.60 26.60 27.00 27.20 \"\" 40 oO 27.80 28.00 28.20 2B.40 28.60 28.80 29.00 29.20 29.40 29.60 29.80 30.00 30.20 30.40 30.60 30.80 31.00 31.20 31.40 31.60 31.80 32.00 32.20 32.40 1 ji.dO 33.00 33.20 33.40 33.60 33.80 '34.00 34.20 34.40 34.60 34.80 35.00 35.20 35.40 35.60 6.30 9.80 8.80 10.90 10.80 10.00 9.80 11.80 13.70 12.00 10.20 9.40 12.70 14.80 13.80 12.50 12.90 12.10 12.00 13.00 14.00 14.00 12.00 16.00 12.40 13.20 15.50 16.00 14.80 15.10 13.50 12.40 11.60 13.30 16.20 12.30 20.00 12.00 12.00 17.00 12.80 18.50 11.20 16.00 17.00 14.20 14.30 14.00 15.70 15.40 18.20 15.50 16.80 14.60 19.00 13.60 14.00 14.00 14.00 15.00 17.70 13.90 13.50 13.30 13.20 12.90 11.80 13.00 13.10 14.90 13.60 IB.00 16.60 16.00 29.70 18.90 23.60 24.10 16.70 25.60 20.50 17.80 21.60 23.50 23.50 30.20 28.40 20.80 26.50 23.30 29.60 29.50 37.70 31.60 36.70 28.50 25.00 26.20 32.80 30.00 33.70 40.00 40.00 37.80 36.20 40.00 31.00 40.00 40.00 40.00 40.00 40.00 39.20 34.30 37.80 32.70 40.00 40.00 22.20 40.00 40.00 36.60 37.00 40.00 40.00 40.00 33.50 37.30 40.00 39.50 40.00 39.60 40.00 38.20 40.00 40.00 38.00 40.00 40.00 37.50 32.50 40.00 40.00 40.00 40.00 40.00 39.10 40.00 .00 4.30 3.90 4.30 3.90 4.00 4.50 5.00 4.50 4.70 4.60 4.50 4.50 5.20 5.10 5.20 4.B0 5.20 5.40 4.80 5.00 4.90 4.90 4.70 4.20 4.30 5.20 4.20 4.60 5.20 5.20 6.00 6.40 6.30 5.00 6.30 6.50 4.80 5.60 5.60 6.00 4.30 6.00 6.70 3.80 4.70 5.80 8.00 7.20 7.10 6.70 7.20 7.10 7.60 7.20 7.60 6.40 7.00 5.70 6.60 7.50 7.40 7.10 6.50 7.60 6.20 6.00 7.50 B. 30 1.93B 1.958 1.977 1.997 2.017 2.036 2.056 2.076 2.095 2.115 2.134 2.154 2.174 2.193 2.213 2.233 2.252 2.272 2.291 2.311 2.331 2.350 2.370 2.390 2.409 2.429 2.448 2.46B 2.488 2.507 2.527 2.547 2.566 2.5B6 2.605 2.625 2.645 2.664 2.684 2.704 2.723 2.743 2.763 1009 2.782 634 2.802 2.821 2.841 2.661 2.880 2.900 2.920 2.939 2.959 2.978 2.998 3.018 3.037 3.057 3.077 3.096 3.116 3.135 3.155 3.175 3.194 3.214 3.234 3.253 3.273 3.292 3.312 3.332 3.351 3.371 612. 291. 499. 440. 175. 528. 349. 178. 247. 379. 444. 717. 532. 178. 422. 353. 568. 594. 696. 637. 787. 488. 433. 335. 723. 572. 623. 634. 878. 767. 787. 965. 666. 932. 827. 969. 688. 9B0. 951. 590. 870. 477. 149. 900. 896. 7B3. 736. 856. 754. 852. 568. 787. 725. 903. 907. 892. 907. 605. 772. 911. 852. 932. 936. 656. 714. 943. 940. 878. 921. 761. 779. 834. 6.54 1.08 2.22 1.47 .57 2.01 1.33 .52 .61 1.13 1.64 3.14 1.51 .40 1.07 1.01 1.61 1.B3 2.93 1.81 2.08 1.24 1.33 .72 2.29 1.60 1.43 1.89 2.21 1.91 2.22 3.14 2.27 2.75 1.86 3.22 1.18 3.39 3.28 1.23 2.72 .89 3.94 1.94 .30 2.48 2.45 2.17 1.75 2.12 1.49 2.10 1.22 2.09 1.36 2.69 2.60 2.56 2.61 2.08 1.61 2.67 2.58 2.93 2.98 2.79 2.60 3.10 3.05 2.38 2.84 1.57 1.80 2.03 .34 .74 .61 .82 .84 .71 .72 .92 1.09 .91 .73 .62 .94 1.18 1.05 .94 .94 .fit .81 .93 1.00 1.04 .86 1.23 .55 .94 1.14 1.15 1.03 1.07 .92 .60 .76 .87 1.14 .77 1.49 .74 .74 1.22 .81 1.36 .65 1.09 1.25 .92 .92 .90 05 ,01 26 01 1.15 .94 1.22 .83 .86 .86 .85 .95 1.18 .63 .60 .77 .76 .74 .66 .74 .74 .89 .78 1.16 1.04 .19 1.500 1.950 2.000 1.950 1.900 2.000 1.950 1.900 1.950 1.950 1.950 2.000 2.100 1.900 2.100 1.950 2.100 2.150 2.150 2.150 2.150 2.100 1.950 2.100 2.150 2.100 2.100 2.150 2.150 2.150 2.150 2.150 2.150 2.150 2.150 2.150 2.100 2.150 2.150 2.100 2.150 2.100 2.150 2.150 1.900 2.150 2.150 2.150 2.100 2.150 2.100 2.150 2.100 2.150 2.100 2.15U 2.150 2.150 2.150 2.150 2.100 2.150 2.150 2.150 2.150 2.150 2.150 2.150 2.150 2.150 2.150 2.100 2.100 5.150 10.500 10.516 10.538 10.557 10.574 10.594 10.613 10.630 10.649 10.668 10.686 10.706 10.727 10.745 10.767 10.785 10.807 10.830 10.852 10.875 10.897 10.919 10.938 10.959 10.982 11.003 11.025 11.047 11.070 11.093 11.115 11.138 11.160 11.183 11.205 11.226 11.250 11.272 11.295 11.316 11.339 11.360 11.363 11.406 11.423 11.446 11.468 11.491 11.513 11.535 11.557' 11.579 11.601 11.623 11.645 11.668 11.690 11.713' 11.735 11.758 11.780 11.802 11.625 11.847 11.870 11.892 11.915 11.938 11.960 11.983 12.005 12.027 12.048 15.071 7.35 2.23 10.25 11.79 2.73 11.10 11.15 3.16 4.12 3.12 11.32 9.81 12.90 4.73 3.96 3.31 13.00 I2.3B 11.09 12.78 12.91 13.84 12.50 5.13 11.87 13.19 14.79 14.29 13.20 13.75 12.51 11.16 11.54 11.90 14.46 11.07 7.09 10.63 10.92 15.91 11.72 6.21 10.18 14.25 5.51 12.65 12.73 12.88 14.42 13.68 16.61 13.77 15.72 13.31 17.25 12.19 12.47 12.51 12.46 13.54 15.91 12.38 12.27 11.88 11.80 11.83 11.52 11.64 11.72 13.13 12.13 16.35 14.69 i4. ia .70 .21 .97 1.12 .26 1.05 1.05 .30 .39 .29 1.06 .92 1.20 .44 .37 .31 1.20 1.14 1.02 1.18 1.18 1.27 1.14 .47 1.08 1.20 1.34 1.29 1.13 1.24 1.13 1.00 1.03 1.06 1.29 .99 .63 .96 .97 1.41 1.03 .55 .89 1.25 .48 1.10 1.11 1.12 1.25 1.19 1.44 1.19 1.36 1.15 1.48 1.04 1.07 1.07 1.06 1.15 1.35 1.05 1.04 1.00 .99 1.00 .97 .97 .98 1.10 1.01 1.36 1.24 1_!7 .45 .12 .57 .60 .16 .59 .59 .20 .786 28.6 25.4 25.0 25.1 25.0 .889 .26 1.096 .19 .59 .53 .61 .29 1.224 .25 .20 .61 .60 .55 .60 .59 .63 .60 .31 1.313 .57 .61 .64 .61 .59 .60 .58 .54 .57 .56 .61 .54 .40 .53 .53 .65 .55 .35 1.541 .51 .60 .32 1.402 .57 .57 .56 .61 .59 .65 .59 .64 .59 .66 .56 .56 .56 .56 .59 .63 .56 .56 .54 .54 .55 .55 .54 .54 .57 .55 .64 .61 .59 25.0 27.0 25.0 25.0 25.1 27.3 25.5 26.3 25. C 25.0. 26.4 25.0 25.0 26.2 26.6 26.1 26.4 27.5 26. ! 27.2 26.2 27.5 27.7 27\". 5 25.0 27.0 28.0 26.2 26.8 26.8 26.3 25.6 26.4 25.0 26.4 25.0 26.2 25.0 27.0 25.9 26.9 26.9 26.2 25.4 27.0 26.8 27.2 27.2 26.9 26.4 27.3 27.2 26.6 27.1 25.0 25.8 •Jk.7 690.3 247.4 424.0 374.4 148.3 448.8 297.0 151.4 210.3 321.8 377.5 609.8 451.8 151.4 353.9 300.1 482.8 504.5 761.5 541.7 669.6 414.7 368.2 284.6 629.4 485.9 529.3 708.9 746.1 663.6 663.6 620.4 566.4 792.5 702.7 823.5 585.0 832.3 806.0 501.4 739.3 405.4 857.6 706.9 126.7 764.6 761.5 665.5 625.3 727.5 640.8 724.4 482.8 668.6 616.0 767.7 770.8 756.4 770.8 684.1 656.2 773.9 724.4 792.5 795.6 727.5 606.7 801.8 798.7 746.1 763.2 647.0 662.4 70B.1 SAND SILT SILTY SAND SANDY SILT SILTY CLAY SILTY SAND SANDY SILT SILTY CLAY CLAYEY SILT SILT SANDY SILT SILTY SAND SANDY SILT SILTY CLAY SILT SILT SANDY SILT SILTY SAND SILTY SAND SILTY SAND SILTY SAND SANDY SILT SANDY SILT CLAYEY SILT SILTY SAND SANDY SILT SANDY SILT SILTY SAND SILTY SAND SILTY SAND SILTY SAND SILTY SAND SILTY SAND SILTY SAND SILTY SAND SILTY SAND SILT SAND SILTY SAND SANDY SILT SILTY SAND CLAYEY SILT SAND SILTY SAND CLAY SILTY SAND SILTY SAND SILTY SAND SANDY SILT SILTY SAND SANDY SILT SILTY SAND SANDY SILT SILTY SAND SANDY SILT SILTY SAND SILTY SAND SILTY SAND SILTY SAND SILTY SAND SANDY SILT SILTY SAND SILTY SAND SILTY SAND SILTY SAND SILTY SANO SILTY SAND SILTY SAND SILTY SAND SILTY SAND SILTY SAND SANDY SILT SANDY SILT 5!l TV SAND -109-Z THPJST * 5 C UO ED ID KD GAMRA SV PC OCR KO CV PHI H SOIL TYPE m) (KG) (MR) (FAR) (MR) (BAR) (BAR) (T/H3! (PAR) (FAR) (BAR) (DEG) (MR) «««* IMtft Ml»# ««tl» HIM (Itlll IttIM (Hll IIHW f»f»*« (tfttl tUtt »*•+! *»•» tttft «*IH HUH HIDHIIUI 15.00 2900. 8.60 2B.00 .40 1.374 6B1. 3.05 4.71 2.000 1.367 4.93 3.64 .76 37.3 1257.3 SILTY SAND 16.00 4450. 13.60 42.50 .60 1.472 1020. 2.65 7.52 2.150 1.473 11.34 7.70 1.07 33.4 2290.8 SILTY SAND 17.00 4600. 14.40 42.00 .30 1.570 980. 2.43 7.34 2.150 1.5B6 11.7? 7.43 1.06 38.3 2174.6 SILTY SANO 1B.00 4650. 13.60 40.50 1.40 1.668 954. 2.55 6.34 2.150 1.6?8 9.72 5.73 .93 38.4 1995.5 SILTY SAND 19.00 4075. 7.80 25.00 2.00 1.766 601. 3.23 2.97 20.00 2850. 7.00 10.00 3.00 1.845 83. .47 2.73 20.20 1300. 5.00 11.40 1.40 1.884 215. 2.08 1.56 20.40 1560. 8.30 18.40 1.60 1.904 342. 1.62 3.15 20.60 1640. 6.40 8.60 1.70 1.923 54. .34 2.34 20.80 1140. 7.20 8.80 3.40 1.943 32. .17 2.74 21.00 640. 5.90 7.30 3.80 1.963 25. .IB 2.06 21.20 480. 6.30 7.30 4.30 1.982 11. .07 2.25 21.40 350. 5.40 6.90 2.80 2.002 29. .24 1.76 21.60 750. 6.20 11.90 1.80 2.022 1B2. 1.28 2.03 21.80 700. 4.00 7.70 2.70 2.041 36. .26 2.00 22.00 700. 4.90 6.60 2.40 2.061 36. .35 1.44 22.20 430. 4.60 4.10 2.80 2.080 29. .32 1.28 22.40 425. 5.10 6.90 2.70 2.100 40. .37 1.50 22.60 360. 4.50 7.80 3.90 2.120 21. .14 2.16 22.80 790. 5.80 9.20 2.10 2.139 98. .77 1.75 23.00 670. 4.30 7.40 4.10 2.159 14. .10 2.03 23.20 490. 4.50 7.80 4.20 2.17? 21. .14 2.09 23.40 360. 4.70 '7.90 4.90 2.199 18. .11 2.17 23.60 3C0. 7.3) 8.50 5.40 2.218 21. .12 2.37 23.80 300. 7.20 8.30 5.40 2.237 14. .OB 2.35 24.00 320. 7.20 8.20 5.50 2.257 11. .06 2.33 24.20 380. 7.20 8.30 5.30 2.277 14. .08 2.31 24.40 370. 4.90 7.80 5.10 2.2?6 7. .04 2.16 24.60 340. • 7.20 8.50 5.00 2.316 21. .12 2.26 24.80 590. 4.50 10.60 2.20 2.336 124. .86 1.86 25.00 650. 7.20 8.30 5.0Q 2.355 14. .08 2.22 25.20 400. 4.70 9.00 3.10 2.375 58. .38 1.94 25.40 930. 8.30 15.00 2.20 2.395 218.. 1.0? 2.53 25.60 920. 8.30 12.40 2.70 2.414 124. .61 2.56 25.80 1700. 9.30 13.50 2.40 2.434 127. .54 2.96 26.00 1240. 7.50 8.60 5.00 2.453 14. .08 2.23 26.20 700. 7.30 8.80 5.00 2.473 29. .17 2.11 26.40 550. 7.50 9.00 5.20 2.493 29. . .16 2.19 26.60 790. 7.20 11.00 2.90 2.512 109. .66 2.02 26.80 960. 8.10 12.80 2.80 2.532 145. .76 2.32 27.20 420. 8.50 10.00 5.00 2.571 29. .14 2.50 27.60 500. 8.30 io.ro 4.00 2.610 36. .18 2.37 27.80 730. 9.50 11.50 6.00 2.630 47. .1? 2.83 28.20 520. 8.90 10.20 4.30 2.649 21. .10 2.55 2.000 1.804 2.84 1.59 .50 39.7 868.6 SILTY SAND 1.800 1.892 3.08 1.63 .73 .615 97.4 SILTY CLAY 1.9O0 1.910 2.42 1.27 .53 31.5 182.4 SILTY SAND 1.950 1.929 5.32 2.76 .74. 30.7 476.7 SANDY SILT 1.800 1.944 2.49 1.28 .63 .521 . 54.9 CLAY 1.700 1.958 3.20 1.63 .73 .638 38.0 CLAY 1.700 1.972 2.06 1.04 .56 .449 22.1 CLAY 1.500 1.982 2.38 1.20 .61 .504 10.2 WD 1.700 1.995 1.63 .82 .48 .374 24.5 CLAY I.BOO 2.011 4.27 2.12 .74 25.1 170.2 SANDY SILT 1.700 2.025 2.D3 1.00 .55 .447 30.8 CLAY 1.700 2.033 1.22 .60 .38 .298 30.7 SILTY CLAY 1.700 2.052 1.02 .50 .33 .259 24.5 CLAY 1.70O 2.066 1.32 .64 .40 .317 33.8 SILTY CLAY 1.700 2.080 2.35 1.13 .59 .505 20.0 CLAY 1.800 2.095 1.71 .82 .48 .391 83.3 CLAYEY SILT 1.700- 2.109 2.15 1.02 .55 .471 12.3 OFF CHART 1.700 2.123 2.28 1.07 .57 .494 19.3 CLAY 1.700 2.137 2.42 1.13 .59 .519 16.6 CLAY 1.700 2.150 2.81 1.31 .64 .586 22.0 CLAY 1.700 2.164 2.79 1.29 .64 .583 14.5 OFF CHART 1.700 2.178 2.77 1.27 .63 .580 10.6 iwo 1.700 2.192 2.74 1.25 .62 .574 14.2 OFF CHART 1.500 2.201 2.47 1.12 .59 .532 6.4 HUD 1.700 2.215 2.48 1.21 .61 .568 21.0 CLAY 1.800 2.231 1.99 .89 .51 .448 105.0 CLAYEY SILT 1.700 2.245 2.44 1.17 .60 .562 13.6 OFF CHART 1.800 2.260 2.16 .96 .53 .480 49.2 SILTY CLAY 1.950 2.279 3.28 1.44 .68 247.4 SILT 1.800 2.295 3.37 1.47 .68 .686 . 136.2 CLAYEY SILT 1.8O0 2.310 4.24 1.84 .78 .830 159.3 SILTY CLAY 1.700 2.324 2.75 1.18 .60 .585 13.7 OFF CHART 1.700 2.338 2.55 1.0? .57 .551 26.1 CLAY 1.700 2.352 2.48 1.14 .59 .575 27.0 CLAY 1.600 2.367 2.41 1.02 .55 .528 94.5 CLAYEY SILT I. BOO 2.383 3.00 1.26 .63 .630 147.7 CLAYEY SILT 1.700 2.412 3.42 1.42 .67 .703 31.1 CLAY 1.300 2.442 3.18 1.30 .64 .665 37.0 CLAY 1.800 2.458 4.22 1.72 .75 .834 56.7 CLAY 1.700 2.487 3.44 1.46 .68 .743 23.7 OFF CHART - 110-C a l c u l a t i o n of Young's Modulus* from Screw P la te Test Sea Island Embankment S i t e Pla te Diameter = Gamma Sand -Gamma Water = Water table at 18 cm 17 .5 kPa/m 9.81 kPa/m •1.25 m Depth Po' AP (kg) AP A E Pul t fm) (kPa) Measured Rods T o t a l (kPa) (cm) (MPa) (kPa) 3 35.33 1400 52.78 1452.78 570 0.10 123 4 43.02 1600 63 1663 652 0.11 128 t 5 50.71 750 73 823 323 0.10 70 t 6 58.40 2500 83 2583 1013 0.20 109 t 7 66.09 4250 92 4342 1704 0.40 92 t 8 73.78 3760 102 3862 1516 0.20 164 t 9 81.47 4240 112 4352 1708 0.30 123 t 10 89.16 4350 122 4472 1755 0.40 95 t 11 96.85 4250 132 4382 1720 0.30 124 t 12 104.54 4500 142 4642 1822 0.40 98 t 13 112.23 5400 152 5552 2179 0.40 118 t 14 119.92 5010 162 5172 2029 0.40 110 t 15 127.61 4550 172 4722 1853 0.40 100 t 16 135.30 3350 182 3532 1386 0.30 100 t 17 142.99 4900 192 5092 1998 0.50 86 t 18 150.68 4900 201 5101 2002 0.50 86 t 19 158.37 4150 211 4361 1711 0.40 92 t 20 166.06 1700 221 1921 754 0.20 25 933 21 173.75 2800 231 3031 1189 0.30 26 1295 22 181.44 2300 241 2541 997 0.20 33 1357 22.5 185.29 1940 246 2186 858 0.20 28 1301 23 189.13 2320 251 2571 1009 0.22 30 1224 23.5 192.98 2760 256 3016 1183 0.26 30 1120 24 196.82 2400 261 2661 1044 0.24 29 1233 24.5 200.67 2030 266 2296 901 0.29 21 1359 * An equivalent Young's Modulus i s c a l c u l a t e d for the sand l a y e r ; an undrained Young's Modulus i s c a l c u l a t e d for the c layey s i l t , t An ul t imate p l a t e load , Pu l t , was not reached i n the sand l a y e r Po' = e f f e c t i v e v e r t i c a l overburden pressure AP = measured screw p l a t e load [from slope (AP/A) of f i e l d SPLT l o a d - d e f l e c t i o n curve] A = measured screw p l a t e d e f l e c t i o n [from slope (AP/A) of f i e l d SPLT l o a d - d e f l e c t i o n curve] E = Young's Modulus (Es i n sand; Eu i n c layey s i l t ) "@en ; edm:hasType "Thesis/Dissertation"@en ; edm:isShownAt "10.14288/1.0062529"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Civil Engineering"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Prediction of embankment performance using in-situ tests"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/28495"@en .