Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Determination of drained and undrained soil parameters using the screw plate test Berzins, William Edward 1983

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1983_A7 B47.pdf [ 5.15MB ]
JSON: 831-1.0062977.json
JSON-LD: 831-1.0062977-ld.json
RDF/XML (Pretty): 831-1.0062977-rdf.xml
RDF/JSON: 831-1.0062977-rdf.json
Turtle: 831-1.0062977-turtle.txt
N-Triples: 831-1.0062977-rdf-ntriples.txt
Original Record: 831-1.0062977-source.json
Full Text

Full Text

DETERMINATION OF DRAINED AND UNDRAINED SOIL PARAMETERS USING THE SCREW PLATE TEST by WILLIAM EDWARD BERZINS B.A.Sc. The University of British Columbia 1979 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE IN THE FACULTY OF APPLIED SCIENCE DEPARTMENT OF CIVIL ENGINEERING We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October 1983 © WILLIAM EDWARD BERZINS, 1983 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 i t 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 i s understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of C./V/L. w<E^/^f^V(^ The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 D a t e O O T D O g ^ / 7 j DE-6 (3/81) - i i -ABSTRACT An evaluation of the screw plate test for use in the determination of s o i l deformation behavior is presented. Current methods of analysis are used to interpret the screw plate data, and the limitations of each method are discussed. As a result of the extensive test programme, the screw plate is found to provide reproducible estimates of drained moduli in sands, and undrained shear strengths in clays. Screw plate tests were performed at three research sites where clay, s i l t , and sand lithologies are encountered. Comparisons are made between deformation parameters obtained from the screw plate test, and those obtained from the pressuremeter, cone penetrometer, dilatometer, vane shear, and t r i a x i a l tests. Published correlations are used where applicable, and confirm the s u i t a b i l i t y of the screw plate test for f i e l d investigations in a variety of s o i l conditions. The development of a unique test apparatus is also detailed. Special features of this system include the automation of plate insertion, the controlled application of rapid cyclic load histories, and the retrieval of the plate upon completion of the sounding. The adaptation of this system to a conventional hydraulic jacking unit is also described. Based upon this systematic evaluation of the screw plate test, recommended test procedures and methods of analysis are also summarized. Suggested topics for future research are presented which would further enhance the applicability of the test to foundation design problems. - i i i -TABLE OF CONTENTS ABSTRACT i i LIST OF FIGURES v i LIST OF TABLES l x ACKNOWLEDGEMENTS x 1. INTRODUCTION 1 1.1. Test Principles 1 1.2. Historical Review 1 1.3. Purpose and Scope 2 2. THEORETICAL ANALYSIS OF SCREW PLATE DATA 4 2.1. Drained Parameters . . . 4 2.1.1. Drained Modulus; Janbu Analysis 4 2.1.2. Drained Young's Modulus; Schmertmann Analysis 9 2.1.3. Coefficient of Consolidation 13 2.2. Undrained Parameters 14 2.2.1. Undrained Elastic Modulus I 7 2.2.2. Undrained Shear Strength I 9 3. DEVELOPMENT OF TEST APPARATUS AND PROCEDURE 2 2 3.1. Development of Test Apparatus 2 2 3.1.1. Screw Plate Configuration 2 2 3.1.2. Effect of Plate Stiffness 2 7 3.2. Installation System 2 9 3.3. Data Acquisition System 3 0 3.4. Development of Test Procedures 3 8 3.4.1. Drained Tests in Sand 3 8 3.4.2. Undrained Tests in Clay 4 0 3.4.3. Factors Affecting Test Procedure and Results . 4 0 - i v -TABLE OF CONTENTS (cont'd) Page 4. DESCRIPTION OF FIELD PROGRAMMES 43 4.1. McDonald - Sand and S l l t y Clay Site 43 4.1.1. General Geology and Site Description . . . . 43 4.1.2. Description of Test Programme 43 4.1.3. Test Results - Drained Behavior 46 Constrained Moduli in Sands . . . . 46 Comparison with Laboratory and Pressuremeter Moduli 51 Modulus Determination from Cyclic Loads 56 Young's Moduli in Sands 62 Cyclic Loading for Liquefaction Assessment 64 4.1.4. Test Results - Undrained Behavior 66 Undrained Shear Strength 6 6 Repeated Undrained Loading in S i l t . 68 4.1.5. Torsional Resistance during Plate Installation 70 4.2. Langley Sensitive Clay Site 7 0 4.2.1. General Geology and Site Description . . . . 7 0 4.2.2. Description of Test Programme 7 2 4.2.3. Test Results 7 2 4.2.4. Undrained Modulus 7 6 4.2.5. Undrained Shear Strengths 7 9 - v-TABLE OF CONTENTS (cont'd) Page 4.3. Cloverdale Sensitive Clay Site 79 4.3.1. General Geology and Site Description . . . 79 4.3.2. Description of Field Programme 81 4.3.3. Test Results 81 Undrained Elastic Modulus . . . . 81 Undrained Shear Strength . . . . 85 5. SUMMARY AND CONCLUSIONS 88 5.1. Measurements of Drained Behavior 88 5.1.1. Constrained Modulus 88 5.1.2. Young's Modulus 88 5.2. Determination of Drained Modulus from Unload, Reload Curves 88 5.3. Measurement of Undrained Shear Strength - c . . . 89 5.4. Measurement of Undrained Modulus . . . . 89 6. SUGGESTED TEST PROCEDURE AND ANALYSIS 91 6.1. Test Equipment 9 1 6.1.1. Screw Plate 91: 6.1.2. Installation System 9 1 6.1.3. Loading System 9 2 6.1.4. Measurement System 9 2 6.2. Suggested Procedure for Plate Installation . . . . 9 3 6.3. Suggested Test Procedure in Sand 9 3 6.4. Suggested Test Procedure in Clay 9 7 7. SUGGESTIONS FOR FUTURE RESEARCH 1 0 0 BIBLIOGRAPHY 1 0 1 - v i -LIST OF FIGURES Page 2.1. Determination of Screw Plate Modulus Number k . . . 6 m 2.2. Values of Janbu's Settlement Number S 8 2.3. Simplified Distribution of Vertical Strain Beneath Loaded Circular Area 11 2.4. Determination of Coefficient of Consolidation - Janbu Method 15 2.5. Determination of Coefficient of Consolidation - Kay & Avalle Method 16 3.1. Schematic Representation of UBC Screw Plate System . . 23 3.2. Example of Double Pitched Screw Plates 26 3.3. Position of Torque Motor 31 3.4. Schematics of Hydraulic Control System 32 3.5. Torque Load Cell 34 3.6. Axial Load Cell 35 3.7. Reference Beam in Position 36 3.8. Screw Plate Recording System 37 4.1. General Location of Research Sites 44 4.2. Typical Cone Profile at McDonald Site 45 4.3. Location Plan McDonald Farm Sand and S i l t y Clay Site . 47 4.4. Typical Load-Displacement Curve in Sand 48 4.5. Range of Janbu Modulus Numbers at McDonald Site . . . 49 4.6. Range of Cone Resistance at McDonald Site 50 4.7. Comparison of Screw Plate, Pressuremeter and Laboratory Modulus Numbers 52 4.8. Corrected k verses (k_).. , and (k_) a m cl lab E pressuremeter ->H LIST OF FIGURES (cont'd) Page 4.9. Variation in Rebound Modulus E^ with Stress Amplitude at a Constant Maximum Stress Level 57 4.10. Variation in Rebound Modulus E^ with Stress Level at a Constant Stress Amplitude 59 4.11. Relationship between Rebound Modulus and I n i t i a l Modulus at Various Levels of Plate Load . . . 60 4.12. Influence of Stress Range and Load on Laboratory Modulus Determination 61 4.13. Observed Relationship Between Cone Bearing and Vertical Young's Modulus 63 4.14. Incremental Platic Strain per Cycle i n Sand 65 4.15. Comparison of Undrained Shear Strengths 67 4.16. Reduction of Strength during Load Test i n S i l t 69 4.17. Relationship between Installation Torque and Cone Resistance 71 4.18. Typical Cone Profile at Langley Site 73 4.19. Location Plan at Langley Site 74 4.20. Load Test in Langley Sensitive Clay 75 4.21. Effect of Plate Rotation during Tests i n Langley Sensitive Clay 77 4.22. Comparison of Undrained Moduli at Langley Clay Site . . 78 4.23. Comparison of Undrained Shear Strengths at Langley . . . 80 4.24. Typical Cone Profile in Cloverdale Clay Site 8 2 4.25. Location Plan at Cloverdale Clay Site 8 3 4.26. Load Test in Cloverdale Clay 8 Z t - v i i i -LIST OF FIGURES (cont'd) Page 4.27. Comparison of Undrained Moduli at Cloverdale Clay Site . 86 4.28. Comparison of Undrained Shear Strengths at Cloverdale Clay Site 87 - ix _ LIST OF TABLES Page 2.1. Summary of Modulus Factors A 18 3.1. Observed Plate Stresses at Failure 39 4.1. Summary of Test Programme - MacDonald Site 46 4.2. Summary of Test Programme - Langley Site 72 4.3. Summary of Field Programme - Cloverdale Site 81 - x-ACKNOWLEDGEMENTS The author i s grateful for the assistance and guidance provided by his Research supervisor , Dr. R.G. Campanella. I would like to thank colleagues Don Gillespie, Peter Robertson and Steve Brown for their assistance during the f i e l d programme. Further thanks must go to Art Brookes and Dick Postgate, whose expertise proved invaluable in the development of the test equipment. The author i s also grateful to Merrill Blackmore, who patiently and s k i l l f u l l y prepared the figures for the text. Test data from the clay sites was obtained by the Graduate students in Soil Mechanics, whose contribution i s acknowledged here. Financial support was provided by N.S.E.R.C. and was greatly appreciated. Heartfelt thanks i s also extended to my parents, to whom this thesis is dedicated. 1 1. INTRODUCTION 1.1. Test Principles The screw plate usually consists of a single f l i g h t of a helical auger, with a cross-sectional area ranging from 250 cm2 to 2000 cm2. This plate Is screwed down to the test depth and a load applied at the surface. The load-deformation behavior of the s o i l i s then recorded, and interpreted to yield drained or undrained moduli, as well as consolidation characteristics and undrained shear strengths. 1.2. Historical Review Use of the screw plate test appears to have originated approximately 25 years ago. Kummeneje (1956) used the screw plate test to predict settlements of petroleum tanks on sand. Kummeneje and Eide (1961) used the device to assess changes in soil porosity and settlements associated with s o i l densification through blasting. Gould (1967) presented favourable comparisons between screw plate tests and large plate bearing tests i n granular deposits. Webb (1969) conducted screw plate tests in fine to medium sands well below the water table, and produced compressibility correlations for the cone penetration test using compressibility measurements obtained from the screw plate test. Schmertmann (1970) proposed a method of predicting settlements using the Cone Penetration Test, employing the screw plate test as a means of assessing the in-situ deformability of cohesionless soils. Janbu and Senneset (1973) also produced a significant contribution by presenting a detailed method 2 of analysis of the screw plate test for determining deformability and consolidation characteristics of both cohesive and non-cohesive so i l s . More recent studies have concentrated on the sophistication of interpretive methods, both in the correlation of test results with other ih-situ tests, and their application to more specific foundation problems. Dahlberg (1975) used the test to determine the preconsolidation pressure i n sands. Marsland and Randolph (1975) compared the results of screw plate tests with pressuremeter tests. Schwab (1976) and Schwab and Broms (1977) examined the time-dependent and time-independent behavior of s i l t y clays. Selvadurai and Nicholas (1979) provide a comprehensive review of the theoretical assessment of the screw plate test in cohesive so i l s , and provide a framework within which the screw plate test can be compared with other in-situ tests in cohesive soils. Bodare and Massarch (1982) use a screw plate modified for impulse loading to study in-situ shear moduli. Kay and Parry (1982) evaluated the use of the screw plate test for the determination of moduli, shear strength and coefficient of consolidation in s t i f f clays. Recent developments include the measurement of Installation torque using a torque load c e l l , and the application of rapid cyclic loads (Berzins and Campanella, 1981). 1.3. Purpose and Scope The primary objective of this dissertation i s to compare deformation parameters obtained from the screw plate test, with those derived from other in-situ and laboratory tests. Review of . 3 the screw plate data w i l l be completed using existing analytical techniques, with an emphasis on determining moduli and shear strengths of cohesive and non-cohesive s o i l s . In order to provide a more comprehensive evaluation of factors affecting the test results and interpretation, an improved installation system was developed, and i s discussed herein. Use of this installation system allowed the author to evaluate deformation behavior at a variety of research sites. These test results are reviewed, and suggested guidelines for test procedures and further research are identified. 2. THEORETICAL ANALYSIS OF SCREW PLATE DATA 4 Screw plate tests have b een analyzed to provide data on the deformation behavior of s o i l . Analytical techniques are summarized in the following sections as they pertain to drained and undrained behavior. 2.1. Drained Parameters The drained analysis of the test assumes that a l l consolidation-induced strains occur during the test. This i s normally accomplished using incremental loading, with sufficient time allowed between load increments for pore pressure dissipation. The parameters which can be derived from drained analyses are the constrained modulus, M, Young's Modulus, E, and coefficient of consolidation for radial drainage, c f . Attempts have also been made to measure an in-situ bearing capacity i n sands, (Berzins and Campanella, 1981). It was generally observed that high plate loads in sands resulted in considerable deformation of the plate i t s e l f , and that failure loads could not be achieved. Consequently, peak strengths could only be inferred from extrapolation of the test data using a hyperbolic relationship. 2.1.1. Drained Modulus; Janbu Analysis A method of determining a drained constrained modulus i s presented by Janbu and Senneset (1973). The constrained modulus i s defined by the expression: 5 p a where M = constrained tangent modulus (Janbu, 1963) k = modulus number m p = reference stress, (normally 1 bar) p' = vertical effective stress a = stress exponent = 1 for O.C. clays = .5 for sand and s i l t = 0 for N.C. clays It should be noted here that intuitively the stress exponent 'a' should vary with the O.C.R. in overconsolidated clays; however this range is not documented. Janbu and Senneset (1973) then use a construction as illustrated in Figure 2.1 to determine the modulus number k m, using the following equation: 6 = 1- _2_ k p m a in which: 6 = plate deflection S = dimensionless settlement number k = modulus number m p = reference stress (normally 1 bar) Si P n = net stress on plate = p - p^ B = plate diameter. p = applied stress on plate p' = i n i t i a l vertical effective overburden o pressure. (2.2) J, D p-1 B P L A T E D I S P L A C E M E N T <J FIGURE 2 1 - DEFINITION OF JANBU MODULUS NUMBER km 7 By measuring the net plate stress 'p ' and deflection '6' n in the f i e l d , and using the solutions for 'S' given in Figure 2.2, a modulus number 'k ' can be determined. From this, a constrained m modulus can be calculated at each depth for a uniform s o i l . Janbu's analysis assumes that the s o i l is homogeneous, elastic, isotropic, and that a l l strains are vertical. In addition, their theory does not compensate for variations in plate stiffness; consequently in dense sands the constrained modulus w i l l be underestimated. An advantage of their method is that i t recognizes that the modulus i s very much dependent upon stress level; consequently, i t provides a method whereby the variation of modulus with depth is accounted for. In addition, the use of the i n i t i a l tangent portion of the load curve, rather than subsequent unload-reload curves, reduces the compounding effects of continued disturbance. Since the modulus of the s o i l is very much dependent upon the stress level and history, the use of repeated loading curves can only introduce further uncertainty in the determination of a representative modulus. Janbu's method, on the other hand, u t i l i z e s the i n i t i a l portion of the loading curve, which represents the soil's "least disturbed" state. The primary uncertainty i n defining the i n i t i a l portion of the curve is related to the selection of an applied stress level which results in the in-situ "undisturbed" state. Arching of stresses immediately behind the plate, which generally occurs in sands, may effectively reduce the residual load on the back of the plate, hence the applied stress may not necessarily be superimposed upon the i n -situ stress level. 1.0 -t s OS-oc - CLAY: M« 0-0 10 20 30 0-1 1 1 1 1 0 -, tO 20 30 4 0 O 10 2 0 30 40 EFFECTIVE OVERBURDEN STRESS (Po' IN l/m2) FIGURE 2 2 - VALUES OF JANBU'S SETTLEMENT NUMBER "S " (AFTER JANBU AND SENNESET, 1973) 9 For the purpose of this study, i t was assumed that arching did occur. Consequently, constrained moduli were obtained by drawing tangents to the test curve when the applied stress equalled the overburden stress. 2.1.2. Drained Young's Modulus; Schmertmann Analysis Schmertmann (1970) presents a method whereby the elastic modulus "E " can be determined using a simplified strain influence s factor. His analysis assumes a homogeneous, isotropic, elastic half space. The Schmertmann (1970) analysis i s derived from Ahlvin and Uhlery (1962) who define the vertical strain at depth beneath a uniform circular load by: e = £ ( l + v) [(l-2v)A + FJ (2.3) Z E where: = vertical strain at depth p = vertical stress on circular plate of radius 'r' E = elastic modulus v = Poisson*s ratio A,F = dimensionless factors which are a function of the point location. By assuming "p" and E are constant, the vertical strain i s shown to depend on a vertical strain influence factor, I , where: I = (1+v) [(l-2v)A + FJ (2.4) The distribution of the influence factor i s shown in Figure 2.3. This simplified distribution i s based upon Schmertmann's review of theoretical and experimental analysis. Schmertmann then proposes corrections to the strain influence factor which consider the effects of strain r e l i e f due to embedment and settlement due to creep. The embedment correction applied to compensate for an assumption inherent in the elastic model; in which the overlying material can produce tensile stresses which reduce the strain f e l t at the bearing level. A linear correction factor is applied, whereby: p C = 1 - 0.5(—) (2.5) Ap where = embedment correction factor p = in-situ vertical pressure *o r Ap = P~PQ = net foundation pressure increase. The correction* for creep can be f i r s t attributed to Nonveiler (1963), who suggested the following logarithmic relationship: >t = P 0 [ l + 0 log(t/t ) j (2.6) where p = settlement at time t = reference settlement at t o o 3 = constant = 2 to 3. Based on case histories, Schmertmann suggests that t = 3.2 x i o 6 o 11 FIGURE 2-3 - SIMPLIFIED DISTRIBUTION OF VERTICAL STRAIN BENEATH LOADED CIRCULAR AREA AFTER SCHMERTMANN (1970) 12 seconds, hence for a rapid plate load test in sand, the correction factor: C 2 = 1 + B log(t/t o) (2.7) does not converge. Consequently, creep cannot be considered a factor i n rapid tests i n sand, and = 1. Application of the embedment correction to the general equation describing vertical strain, and integration over the depth of influence yields: 2B I P = / e z dz « Ap j dz o o E s 2B I = C Ap I (-*) Az (2.8) 0 E s The determination of E can be made by back calculation from the s J results of the screw plate test, and the above equation. An assumption of a constant modulus within the strain area beneath the plate yields the following expression from which E g can be determined: E s = C x ^ 1.2 B (2.9) P where: E g = equivalent Young's modulus = embedment factor = 1 - (,5)(—) Ap Ap = applied plate stress p = measured plate deflection B = plate diameter. The modulus can then be calculated by assuming a tangent or secant to the stress-deformation curve, depending upon the stress level of interest. It should also be noted that Schmertmann (1970) recommends modulus be determined over a stress range 1 to 3 tsf (- 1 to 3 kg/cm2), which i s a typical design value for shallow foundations. The method cannot be confidently applied at low i n i t i a l plate stresses, particularly at depth since the method has been developed for a typical stress range. 2.1.3. Coefficient of Radial Consolidation The screw plate test can also be used to determine the coefficient of consolidation (Janbu and Senneset, 1973, Kay and Avalle, 1982). Janbu and Senneset (1973) present a method whereby axisymmetric or one-dimensional consolidation theories can be applied to incremental load tests in cohesionless s o i l . By using the basic relationship: Td 2 (2.10) c = t where: c = coefficient of consolidation d = drainage path t = time after load increase T = dimensionless time factor, 14 they analyze the consolidation beneath the loaded plate, which i s assumed to be governed by essentially radial drainage. By using the construction shown in Figure 2.5, a coefficient of radial drainage can be determined by using the following equation: D 2 T J 2 c r = T90 * 3 3 5 ( 2 > 1 1 ) C90 fc90 coefficient of radial consolidation plate radius = length of drainage path "d" time for 90% consolidation time factor for 90% consolidation = 0.335. Kay and Avalle (1982) propose a revised method for determining c , whereby more r e a l i s t i c drainage conditions are assumed. Their method i s summarized in Figure 2.4. The field measurement of the coefficient of consolidation requires incremental load tests of lengthy duration. As a result, verification of the aforementioned theories was not undertaken in this study. The analyses have been presented, however, for completeness. 2.2. Undrained Parameters Undrained parameters are obtained from rapid load tests in fine s i l t s and clays in which the rate of load application precludes significant pore pressure dissipation. Undrained analyses of screw plate data yields estimates of an undrained modulus and undrained shear strength. where: '90 r90 i IMPERVIOUS PERMEABLE C r = Q335 .f£ *90 IMPERVIOUS 03 0£ z 5! On y IX) TIME, mm. - » (v^--scole) 1 2 4 6 15 , _ . '90 = 2.7 min. VELDS C r * 420 m2/yeor (RNE SAND) 131 x FIGURE 2-4- DETERMINATION OF COEFFICIENT OF CONSOLIDATION JANBU METHOD (AFTER JANBU AND SENNESET, 1973) FIGURE 2-5 - DETERMINATION OF COEFFICIENT CONSOLIDATION K A Y A N D A V A L L E M E T H O D , 1 9 8 2 2.2.1. Undrained Elastic Modulus Selvadurai and Nicholas (1979) provide a theoretical assessment of the screw plate tests in a homogeneous, isotropic, elastic medium. They present a number of closed-form solutions which consider the effects of plate r i g i d i t y and the plate-soil interface. The closed form solutions follow the basic relationship: P^E" = A <2'12> u in which: 6 = plate displacement p = average stress on plate a = plate radius E u = undrained, elastic Young's modulus A = modulus factor. The modulus factor 'X' is a function of Poisson's ratio and the degree of bonding with the plate. The results of a review by Selvadurai et a l . (1979) are presented in Table 1.1. They conclude that the undrained modulus can be approximated by the expression: — — = 0.60 to 0.75 (2.13) Pa/E u The upper limit applies when the plate is partially bonded to the s o i l , which might be the case in a sensitive clay. Kay and Parry (1982) suggest that a value of .66 be adopted as a reasonable TABLE 1.1. SUMMARY OF MODULUS FACTORS 'A'. Solution * - A pa/E u Reference (a) 0.630 Kelvin (1890) (b) 0.589 Collins (1962, Kanwal and Sharma (1976), Selvadurai (1976) (c) 0.750 Hunter and Gamblen (1974) (d) 0.750 Keer (1975) (e) 0.648 Selvadurai (1979 a) (f) 0.585 Selvadurai (1976) (8) 0.730 Christian and Carrier III (1978) Pells and Turner (1978) (h) 0.525 Christian and Carrier III (1978) Pells and Turner (1978) a = radius of screw plate P = Eu = average stress on undrained modulus screw plate = —S-. IT a 6 = plate displacement Solution Remarks (a) average displacement of uniform load (b) displacement of fu l l y bonded rigid disc (c) displacement of smoothly embedded rigid disc (d) displacement of partially bonded rigi d disc (e) central displacement of flexible disc (f) displacement of ri g i d spheroidal region (g) average displacement of deep borehole subjected to uniform load (h) displacement of rigid plate at base of deep borehole (After Selvadurai et a l . 1979) approximation for partial bonding in most clays. Selvadurai et a l . (1979) studied the disturbance associated with plate installation by performing f u l l scale model tests. They concluded that the disturbance induced by rotation of the plate through a clay was minimal. This finding i s reflected in the analysis they propose, whereby the effect of stress r e l i e f is not specifically addressed. Selvadurai et a l . (1979) attempt to eliminate this uncertainty by analyzing the unload-reload portion of the test curve for the modulus determination, rather than the i n i t i a l tangent portion. The valid i t y of this method cannot be verified theoretically, however Selvadurai as well as Kay and Parry (1982) found reasonable agreement between the i n i t i a l tangent moduli and subsequent reload moduli. The uncertainty in the determination of an undrained elastic modulus in an ideal media w i l l be further complicated when any of Selvadurai et al.'s (1979) assumptions are violated. This is particularly true in many s o i l deposits which exhibit heterogeneity and strength anisotropy. The combination of stress r e l i e f , s o i l v a r i a b i l i t y and preshearing during plate installation make an accurate determination of undrained modulus tenuous at best. 2.2.2. Undrained Shear Strength The undrained shear strength can be determined from the screw plate test by using the expression of the bearing capacity of a deep circular footing, whereby: undrained shear strength ultimate average plate stress total overburden stress bearing capacity factor. Again the undrained shear strength depends upon boundary conditions including the soil-plate interface and plate stiffness. Selvadurai et a l . (1979) reviewed classic theoretical and empirical solutions, and concluded that: p u l t = ^ - 9.0 for partial bonding (2.15) c u = 11.35 for f u l l bonding It should be noted here that factors including strain rate effects, strength anisotropy and progressive failure are not specifically addressed in current methods of screw plate analysis. Failure loads were achieved in the clays tested during this study. However, i t i s recognized that load limitations with a more conventional test apparatus may preclude the development of sufficient plate stress to cause failure, particularly In s t i f f clays. Kay and Parry (1982) propose a method whereby the load-displacement curve can be extrapolated to obtain the ultimate plate capacity, and hence undrained shear strength. By measuring the plate deflection at two specific points on the stress-displacement in which: c u P u l t a vo \ = curve, they estimate the ultimate plate load using the following hyperbolic relationship; Pult = 2 ' 5 4 Py " X ' 5 4 Px <2'16> where: P u^ t = the ultimate plate stress, p = the plate stress at a strain equal to 1.5% of the plate diameter (B) Py = the plate stress at a strain equal to 2% of B. 22 3. DEVELOPMENT OF TEST APPARATUS AND PROCEDURE 3.1. Development of Test Apparatus Upon i n i t i a t i o n of the research project, the available literature was collated in an effort to define the design c r i t e r i a for the plate configuration and installation system. Development of an automated test apparatus concentrated upon the adaptation of the test to a hydraulically operated CPT r i g developed at U.B.C. A schematic representation of the screw plate system i s presented i n Figure 3.1, and i s elaborated upon i n subsequent sections. The system developed incorporated several features which extended i t s testing capability. The servo-controller system was used to apply cyclic loads to the plate, and could also be used to apply strain-controlled loading i f a displacement transducer was used. Other special features included the measurement of torque during plate installation, and the capability to vary rates of rotation and thrust in a controlled fashion. 3.1.1. Screw Plate Configuration Previous researchers studying the screw plate have utilized a single flight of helical auger, ranging in area from 250 to 2000 sq. cm. The aspect ratio, ( i . e . , the ratio of the half-pitch to diameter), is reported to be between 0.1 and 0.2. Thickness of the plate, hence i t s relative r i g i d i t y , has been discussed by Selvadurai and Nicholas (1979), but has not been extensively treated in published literature. In order to provide a pivotal point around which the helical auger can rotate, the screw plate has a conical TORQUE MOTOR - 2000 Nm (17,000 In.lbs.) TORQUE OUTPUT ot 140 bar (2000 p«l) INPUT PRESSURE - 0-100 RPM HYDRAULIC PISTONS; TOTAL THRUST AREA 122 (18 88 »q In) MAXIMUM SAFE THRUST 75 KN (17,000 lb«.) WITHOUT ANCHORS SCREW PLATE FLOW FORWARD-SYSTEM RATE — REVERSE CUTOFF CONTROL CONTROL MAIN TRUCK HYDRAULICS MANUAL PISTON CONTROLl SERVO-LOOP CONTROLLER SCREW PLATE RODS 4-45 cm. (1-75 In.) O.D. - 1015 GRADE MECHANICAL STEEL TUBING 127 cm (-5 In) ID. 21 SPLINE COUPLINGS MAXIMUM TENSILE LOAD ACROSS PINNED COUPLINGS 22 kN (5000 Ibl.) TORQUE PLATE LOAD CELL LOAO CELL -•OUTER RING CONNECTED TO PISTONS 'INNER RMG ATTACHEO TO BASE OF TORQUE MOTOR -GAUGES ATTACHED TO SPOKES CELL OUTPUT 10 mV ot 2000 Nm THRUST BEARING HOUSING FEMALE SPLINE TO RECEIVE MOTOR HUB ' ROLLER BEARINGS MALE SPLINE ± 2cm TRAVEL DISPLACEMENT TRANSDUCER - TRANSTEK LVDT ± 3" ( 76 cm) - MAGNET ATTACHED TO RODS FIGURE 3*1-SCHEMATIC REPRESENTATION 0 F SCREW P L A T E SYSTEM Is} U> point i n advance of the plate i t s e l f . In order to assess the v i a b i l i t y of a particular plate design, one must consider the various components of plate resistance. Preliminary calculations based upon the mechanisms of resistance, and available CPT data on the tip and sleeve resistance of various soil deposits revealed that the plate should satisfy the following c r i t e r i a : 1) reduced f r i c t i o n between the plate surface and the s o i l , 2) reduced surface area on the leading (cutting) edge of the plate, 3) reduced edge area, 4) reduced diameter of the conical t i p . Preliminary calculations indicated that a single flighted auger would require approximately 50% less torque to i n s t a l l than a double helical plate. During the course of the f i e l d testing, however, i t was found that plate size (for a single flighted auger), had a minimal effect on the total torque which was required during installation i n most s o i l s . The total installation torque became an important factor in dense sands, where the capacity of the system may be reached, and cyclic torsional loads required. The plate stiffness is a key parameter in s t i f f soils, as increased f l e x i b i l i t y results In higher measured deformations; hence an underestimation of the elastic modulus. The effect of plate r i g i d i t y i s further discussed when the test data are presented. It was generally observed that the double flighted plate was 25 self-centering, and tended to "wander" less during installation. As a result, i t is believed that the use of a double flighted plate leads to reduced s o i l disturbance, particularly in dense sands, and also permits symmetrical loading on the plate. The "wandering" phenomena observed during the installation of the single-flighted plate was most l i k e l y the result of a tendency for the plate to rotate about a point located on the cutting edge, rather than the central conial t i p . The eccentricity of the center of rotation increased the total torsional force required during installation, since the applied torque was no longer being used solely to overcome the bearing resistance of the cutting edge. The conical tip has to be quite large to overcome the tendency to wander; hence the zone of disturbance below the plate i s increased considerably. The use of a self-centering, double-flighted plate allows the size of the conical tip to be reduced; consequently s o i l disturbance w i l l be less. Another important consideration i n the selection of the plate is that i t should be recovered at the end of the sounding. This was found to be quite important, as with the Swedish cast-iron plate, (see Fig. 3.2) which was found to have broken during a load test in dense sands. By recovering the plate, one can examine i t for evidence of buckling or breakage, which would significantly alter the interpretation and hence r e l i a b i l i t y of the data. In order to accommodate this improvement during this study, splined rods with pinned connections were used to permit removal of the plate. Figure 3.2 Illustrates the double helical plates employed In this study. F i g u r e 3.2 - Example of Double P i t c h e d Screw P l a t e s n 3.1.2. E f f e c t of Plate S t i f f n e s s Selvadurai and Nicholas (1979) studied the r e l a t i v e e f f e c t of plate r i g i d i t y i n determining an undrained modulus. By examining t h e i r a n a l y s i s , we can gain some i n s i g h t i n t o the e f f e c t of varying plate s t i f f n e s s . Selvadurai (1979) developed an expression for the behavior of a f l e x i b l e plate i n contact with an e l a s t i c medium: w(3-4v)(l+v)( 144-60 5 + 90R) = X (3.1) pa/E 16(l-v)(64+90R) where: 6 = plate d e f l e c t i o n p = average plate load a = plate radius E = s o i l modulus v = Poisson's r a t i o for s o i l X = modulus f a c t o r £ = radius of rods behind plate a R i s defined as the r e l a t i v e r i g i d i t y of the plate, where: R = ' (3-4v)(l-fv) 5> (Hj3 ( 3 . 2 ) 12(l-v ) ( l - v ) E a i n which: h = plate thickness {vp} and {Ep} = e l a s t i c constants for plate. The s o l u t i o n of t h i s equation f o r varying plate s t i f f n e s s e s y i e l d s 28 .589 for log R = 2 £ = .25 (3.3) .883 for log R = 0 (After Selvadurai (1979)) This formula provides a basis whereby a stiffness correction factor can be applied to the test data. By re-arranging 3.3, the relationship becomes: E = X SI (3.4) 6 From this, a relationship between moduli determined from s t i f f and flexible plates i s determined, where X = pa/E where: E g = modulus determined by a s t i f f plate test, Ej = flexible plate modulus, 8 = stiffness correction factor = 1.5 for a flexible plate = 1.0 for a s t i f f plate. Admittedly, Selvadurai's solution i s for undrained behaviour, consequently volume change due to shear as observed i n sands i s not accounted for. Nevertheless, the stiffness factor represents a method by which the screw plate moduli can be normalized for plate stiffness. An increase in plate r i g i d i t y , R, or the radius of the push rods, £ , w i l l reduce the correction factor and produce more 3. consistent E determinations. 3.2. Installation System  Screw Plate Rods In evaluating various alternatives for the screw plate rods, consideration was given to satisfying the forces required during installation of the plate and during load tests. The cross-sectional area of the loading rods chosen ensured a minimal amount of elastic compression during loading, (0.25 centimeters at 90 kN load for 20 metres of rod). In addition, in anticipation of conducting repeated torsional shear tests in-situ, the torsional twist was minimized. The connections between rods are splined i n order to allow clockwise and counter-clockwise rotation during installation and removal of the plate. Set screws are employed in order to prevent accidental disengagement of the rods, and to permit the rods to be pulled up where the s o i l provides insufficient reaction during withdrawal of the plate. The rods are hollow in order to allow for future instrumentation of the plate i t s e l f , as well as permitting the use of internal rods in the future in order to apply the load directly at the plate, and thus eliminate the effect of rod f r i c t i o n . Torque Motor Preliminary calculations indicated that installation of the screw plate would require approximately 2000 Nm of torque. Selection of a suitable torque motor was governed not only by required power, but also by space limitations within the pre-existing hydraulic pistons, as well as the input pressures provided by the truck hydraulics. The motor selected is capable of operating continuously at low rates (1 to 100 RPM) and high loads (up to 20 kN). Figure 3.3 illustrates the positioning of the torque motor. Hydraulic Loading Pistons The screw plate system was essentially designed 'around' the hydraulic pistons i n i t i a l l y developed for use with the CPT test. F u l l details of the hydraulic system are given elsewhere (Campanella and Robertson, 1981). The system is limited to producing 75 kN of thrust, which i s the reaction force provided by the truck. To be compatible with the screw plate system, the hydraulic pistons had to provide a variable rate of advance, with an additional control of pressures applied to the rods during installation. The advance of the pistons can be controlled either manually, or through a servo-loop system. Figure 3.4 illustrates the schematics of the hydraulic control system. In order to enable the torque motor to apply torque to the screw plate rods, and at the same time apply an axial load, a thrust bearing was designed. This bearing permits the application of either a compressive or tensile force on the rods during screw plate advancement or withdrawal. 3.3. Data Acquisition System A particular refinement of the measuring system over conventional screw plate systems was the design of a torque load 32 L E V E L I N G J A C K S (ONE OF FOUR) PENETRATION HEAD ROTARY HEAD LOCKING VALVE 9-FLOW DIVIDERS AD R E L '.EF ! j L — l _ l I l _ j O II CONTROL ADJ. i U-J RELIEF L_ HIGH - LOW PRESSURE CONTROLS ( lOOOpsi and 2000ps i ) liL (6900kPo and IJBOOkPo) VARIABLE VOLUME PRESSURE COMPENSATED PUMP INTAKE STRAINER FIGURE 3-4-SCHEMATIC OF HYDRAULIC CONTROL SYSTEM after Campanella 8. Robertson ( 1 9 8 1 ) 33 c e l l which enable the operator to measure Installation torque during the advancement of the plate. Figure 3.5 shows the principle of the torque load c e l l . Axial loads, hence the load on the plate, were measured using an axial load c e l l shown in Figure 3.6. Axial loads were also estimated using a pressure transducer which recorded the pressure in the hydraulic system. This pressure was multiplied by the cross-sectional area of the pistons to obtain an estimate of the total applied load. Axial displacements were recorded using a direct current displacement transducer (DCDT) mounted on a reference beam (Figure 3.7). The reference beam was employed to ensure that changes in surface load beneath the truck pads did not influence the deformation measured at the rods. The parameters measured during the test consist of: 1) installation torque 2) axial load 3) axial displacement X-Y-Y' chart recorders were used to record the installation torque and the load-displacement curves. Loads and displacements were also routinely recorded on time plots as well. Figure 3.8 shows a typical layout of recording equipment. The transducer outputs were also used as input signals to an MTS servo-controller. In this manner, strain or stress controlled tests could be performed, with variable amplitudes and magnitude. The hydraulic system could routinely provide 1 hz load cycles in dense sands. Lower frequency loading of 0.1 hz was used in s i l t s 34 F I G U R E 3.5 T O R Q U E L O A D C E L L 35 FIGURE 3.6 A X I A L LOAD CELL FIGURE 3 . 8 SCREW PLATE RECORDING SYSTEM where large amplitude strain resulted i n considerable f l u i d displacement in the pistons. Square, triangular or sinusoidal waveforms could also be applied. 3.4. Development of Test Procedures 3.4.1. Drained Tests in Sand Gould (1967) identifies a suggested load procedure whereby loads are incrementally applied to the plate, with consolidation/compression permitted before each additional load increment is applied. Dahlberg (1975) specifies that a t ^ be reached for each load increment, as the coefficient of consolidation can be determined during the load test using Eqn. 2.1.3. Generally, the screw plate test in sands can be considered a f u l l y drained test, hence the tests conducted during this study were done at a rate of 0.1 Hz. Cyclic stress amplitudes as well as maximum stress levels were varied periodically. The effect of these variations is discussed later. The general test procedure adopted during this study is summarized below. (1) Plate Installation - the plate was rotated into the s o i l under it s own impetus, in other words i t was allowed to "pull i t s e l f " downward. Efforts to advance i t at a rate equal to the pitch times the rate of revolution resulted only in preloads on the plate which were detected during installation and the f i r s t load increment. The rate of revolution was approximately 10 revolutions per minute. The torque required during installation was generally not affected by the rate of 39 revolution. (2) Load App l i c a t i o n During Tests - loads were applied using a load c e l l which provided an analogue s i g n a l to a servo c o n t r o l l e r . C y c l i c loads were applied at 0.1 Hz, generally to the maximum capacity of the t e s t i n g v e h i c l e . Higher frequency loading was found d i f f i c u l t to achieve given the l i m i t a t i o n of the hydraulic system. Deformations were recorded using a displacement transducer. Loads were generally applied u n t i l f a i l u r e of the s o i l was observed. The f a i l u r e stresses observed i n t h i s study during tests i n cohesive and noncohesive s o i l s are summarized i n Table 3.1 below. TABLE 3.1. Observed Plate Stresses at F a i l u r e . STRESS AT FAILURE (bars) S i l t y Clay Sensitive Clay Fine Sand, Loose Medium Sand, Dense 4-6 1-2.5 8-14 greater than 14 As mentioned previously, f a i l u r e loads were not attained during plate load tests i n dense sands. (3) Repetition of Load Sequence - upon completion of the load t e s t , the plate would be advanced one to two metres, and the test repeated. It was considered that a one metre (or 4 plate diameters) difference between test depths was s u f f i c i e n t to eliminate s i g n i f i c a n t superposition of s t r a i n on successive t e s t s . (4) Upon completion of the profile, the plate rotation was reversed and the entire down hole apparatus retrieved. At this time, i t was observed that invariably, the disturbed zone through which the plate passed was unable to support the weight of the plate plus rods. This seemed to confirm the assumption that rod friction was insignificant since i t was less than the total weight of the rods. This observation was also made by Schmertmann (1970). (5) The development of the aforementioned test equipment and procedures resulted in a system which proved to be a rapid investigative tool. A 15 metre profile, including torque measurements and cycled tests at 1 metre intervals could be completed in 8-10 hours; thereby providing valuable deformation parameters over the depth of influence beneath most conventional shallow foundations. 3.4.2. Undrained Tests i n Clay A similar test procedure was employed in clays to determine undrained shear strengths and moduli. Drained tests were not attempted because of the length of time i t normally took to achieve 90% consolidation. Researchers studying consolidation parameters in clays generally conduct slow incremental load tests (Janbu and Senneset (1973), Kay and Parry (1982)). 3.4.3. Factors Affecting Test Procedure and Results The contribution of rod weight to the i n i t i a l plate load depends upon the amount of fr i c t i o n along the rods- When i t was discovered that the rods had to be clamped at the surface during withdrawal of the plate, i t was conservatively assumed that the f r i c t i o n on the rods was negligible, hence a l l load calculations included rod weight. The installation system reported by Janbu and Senneset (1973) u t i l i z e s a down-hole hydraulic piston to apply the load to the plate, thereby eliminating any possible effects on f r i c t i o n on the rods. A restriction of their system is that i t precludes possible instrumentation of the plate through inner cables, and further complicates the installation procedure. Consequently, a simpler system using only exterior rods was adopted in this study. Further refinement of the test might involve the use of inner rods to apply the load to the plate, in a fashion similar to the mechanical f r i c t i o n cone. This should be considered If rod friction becomes excessive. The effect of rod compression on measured plate deformation was also evaluated. Calculations showed that the screw plate rods would be compressed by .1 mm per metre at the f u l l capacity of the loading system. At a depth of 10 metres, rod compression would typically contribute only 5% of the observed plate deformation at the maximum load, hence rod compression effects were neglected during modulus determinations. The effect of rod buckling could not be quantified. The use of inner rods on down-hole hydraulic pistons could reduce the possible influence on measured displacements. The stiffness of the plate can also has a significant effect on deformations measured in s t i f f s o i l s , as discussed in 3.1.2. Several times during the installation of the plate in dense sands, cyclic torsional loads were required, in which the plate r o t a t i o n repeatedly was reversed during i n s t a l l a t i o n . This was found to be the only way i n which the plate could be advanced through the denser sands. This c y c l i c t o r s i o n a l loading undoubtedly pre-shears the s o i l immediately below the plate, and hence would tend to reduce the s t i f f n e s s measured during the i n i t i a l load c y c l e . 4. DESCRIPTION OF FIELD PROGRAMMES The f i e l d testing programmes were conducted on three research sites shown on Figure 4.1. The generalized site geology i s given in Table 4.1 below. SITE SOIL DESCRIPTION McDonald Site - Sea Island Langley Clay Site Cloverdale Clay Site Sand and clayey Si l t Sensitive Clay Sensitive Clay TABLE 4.1. General Lithology at Research Sites. 4.1. McDonald - Sand and Si l t y Clay Site 4.1.1. General Geology and Site Description The McDonald site is situated on sea island, and i s approximately 2 metres above sea level. The deposits represent various stages of development of the Fraser Delta. The general s o i l profile consists of: 0-2 m - clay; s i l t y , soft, organic 2-13 m - sand; medium to coarse, variable density, concretionary layers at depth 13-15 m - fine sand; transition zone 15—1-300 - s i l t ; clayey, normally consolidated, soft A typical cone profile Is shown in Figure 4.2. 4.1.2. Description of Test Programme A detailed site investigation has been carried out at the site as part of an on-going research effort. Details on equipment and FIGURE 4 2 - TYPICAL CONE PROFILE AT McDONALD (AFTER CAMPANELLA, ROBERTSON AND GILLESPIE, 1983) SITE procedures are given in Robertson (1982). The tests under consideration in this study are summarized in Table 4.2 below. TEST TYPE NO. Cone penetration profiles 6 Standard penetration tests 13 FIELD Self boring pressuremeter profile 3 Dilatometer profiles 2 push-in cone pressuremeter profile 2 Screw plate profiles 12 LAB cyclic t r i a x i a l tests t r i a x i a l compression tests 7 5 TABLE 4.2. Summary of Test Programme - McDonald Site. A location plan is given in Figure 4.3. 4.1.3. Test Results - Drained Behaviour Constrained Moduli in Sands A typical screw plate load displacement curve is presented i n Figure 4.4. These curves were analyzed using Janbu and Senneset's (1973) method of analysis, using the i n i t i a l tangent modulus. The modulus numbers k within the McDonald sands are presented in m Figure 4.5 and range from k = 120 to 550 . m The variation i n k^ corresponds quite closely with the variation i n cone bearing values at the site (Figure 4.6). Both profiles indicate peaks at approximately 9 to 10 metres, where the sand density appears greatest. The presence of the dense layer at 12 to 13 metres i s not reflected i n the screw plate tests because of the presence of the softer layer at depth. 1-4 P C 2 P C S A S B P M T - 3 O S B P M T - 2 O P C 6 ^ O S B P M T -P P M T - I GRAVEL ROAD STEEL REFERENCE POST P P M T - 2 BCHI A P C 4 H D M T Z A P C > a 3 or tf) "o 4 A PIEZOMETER CONE TEST • FLAT PLATE OILATOMETER TEST O SELF BORING PRESSUREMETER T E S T # PUSH IN CONE PRESSUREMETER TEST 53 BOREHOLE, STANDARD PENE TRATION TEST, SAMPLES + SCREW PLATE TEST DMT-1 SCALE 4 m FIGURE 4-3 - LOCATION PLAN AT MACDONALD FARM SAND AND SILTY CLAY SITE after Robertson (1982) 1.0 2'0 PLATE DISPLACEMENT (cm) FIGURE 4.4-TYPICAL LOAD DISPLACEMENT CURVE IN SAND J A N B U MODULUS NUMBER km 100 2 0 0 3 0 0 4 0 0 5 0 0 - f 1 •— 1 1— FIGURE 4 5 - RANGE OF JANBU MODULUS NUMBER km AT MACDONALD SITE 50 B E A R I N G R E S I S T A N C E ( q , . ) B A R S R A N G E O F C O N E R E S I S T A N C E FIGURE 4-6 - RANGE OF CONE RESISTANCE AT MCDONALD'S FARM a f t e r R o b e r t s o n ( 1 9 8 3 ) This variation can be attributed to variations i n lithology, occasional cementing within the beach-deposited sands (concretions), and localized high K q values arising out of the deposition of the sand in a high energy environment. Comparison with Laboratory and Pressuremeter Moduli The screw plate kffl values are compared to laboratory (kg^iab and pressuremeter (k_) values as reported by Robertson (1982) in E pmt Figure 4.7. The use of the nondimensional modulus number "k" eliminates the effect of stress va r i a b i l i t y . The laboratory k £ values were obtained from the i n i t i a l tangent portion of the tr i a x i a l tests performed on 'undisturbed' samples. It i s recognized that the measurement of deformation behaviour at low levels of stress and strain i s d i f f i c u l t unless specialized procedures are used. The pressuremeter k^ was obtained from unload-reload tests performed at the site. When screw plate k values are corrected r r m for plate r i g i d i t y , using equation 3.5, they compare favourably with laboratory and pressurementer values as shown In Fig. 4.8. It i s recognized that the various moduli numbers represent different loading mechanisms and stress paths. However, for the purposes of discussion, the k m values wi l l i n i t i a l l y be compared to the average kg value. A differentiation between the various moduli numbers w i l l be made after further discussion. The apparent relationship between the various modulus numbers can be summarized as: (.9 to 1.2) (k E) in fine sand 52 J A N B U M O D U L U S N U M B E R k m P M T A N D L A B O R A T O R Y K c 100 — I — 2 0 0 — I — 3 0 0 — I — 4 0 0 — I — 5 0 0 — I — 6 0 0 5 + o o o o A o o A A o O' a. u o 10 A (kE) PRESSUREMETER • (kg) LABORATORY O (kj JANBU'S METHOD O O O O O O o 15 o o 100 Z O O 3 0 0 o — I — 4 0 0 o o o o 5 0 0 6 0 0 FIGURE 47- COMPARISON OF SCREW PLATE, PRESSUREMETER AND LABORATORY MODULUS NUMBERS k m (.75 to 1.0) (kg) in medium dense sand. (4.1) Intuitively, one would expect that (kg) t would be lower than k m, unless the in-situ K Q value is high, because the pressuremeter measures a horizontal modulus, whereas the screw plate measures a vertical constrained modulus. Ladd et a l . (1977) observed that the vertical modulus can be approximately twice the horizontal. Similarly, i f we examine the relationship between the constrained modulus and vertical elastic modulus, using: M (l+v)(l-2v) <4'2> in which: M = constrained modulus E = elastic Young's modulus v = Poisson's ratio, we would expect that kffi = (1.5 to 2.1) ( k E ) ^ a ^ , provided that the laboratory k was obtained from a truly undisturbed sample. It should be recognized that the effects of sample disturbance generally result in an underestimation of the modulus determined from laboratory tests, consequently i n practise one would expect that the in-situ k m values would be even greater than twice the laboratory k . The discrepancy in the modulus numbers observed ln Figure 4.8 may arise out of a violation of the assumptions in Janbu's analysis. Factors which would reduce the measured stiffness of the s o i l , as discussed previously, Include s o i l disturbance during installation, J A N B U M O D U L U S N U M B E R k m P R E S S U R E M E T E R A N D L A B O R A T O R Y k E and lateral strains during loading which result i n greater measured vertical displacements. These factors cannot be quantified without model studies or parametric f i n i t e element studies which would indicate the variation in modulus which is expected with various degrees of s o i l disturbance. Robertson (1982) observed that the pressuremeter unload-reload modulus is approximately equal to the in-situ horizontal modulus. If we extend this observation to the screw plate data and assume that ( E ) v e r t i c a l = 2 ( E ) h o r i z o n t a l > then we would expect that: k = 1.5 to 2.1 (k_) , . (4.3) m E vertical v ' and k m - 3.0 to 4.2 ( k ^ . (4.4) In order to obtain a more r e a l i s t i c vertical constrained modulus from Janbu's analysis, i t i s necessary to adjust the modulus number accordingly: (k ) m r U — = (2.7 to 3.6) in fine sand m^^  Janbu = (2.3 to 3.0) In medium dense sand. (4.5) The relative rigidity of the UBC plate was evaluated using equation 3.2, and i t was concluded that the U.B.C. plate was very flexible. Consequently, the use of this plate leads to an underestimation of modulus in dense sands. To obtain a corrected modulus, E should be multiplied by B = 1.5 when a flexible plate i s used. Based on this analysis, i t can also be shown that an optimal ( s t i f f e r ) plate design can be achieved by: (1) increasing the plate thickness, (2) decreasing the plate diameter, and (3) increasing the diameter of the rods in contact with the plate. These would be primary considerations in the optimization of plate design. Modulus Determination from Cyclic Loads Several researchers have indicated that elastic moduli can be determined from the unload-reload portion of in-situ tests. Hughes (1982), for example, presents a theoretical basis for the determination of shear moduli from reload tests using the self boring pressuremeter. A number of cyclic screw plate load tests were performed in the sands at the site to obtain a preliminary assessment of the s u i t a b i l i t y of the screwplate test to determine a drained elastic modulus in a similar fashion. Figure 4.9 presents an example of the variation in a rebound modulus, E^ > with stress amplitude at a constant upper limit. E p ranges from: E i E R = {.73 to 1.1} E i f = I n i t i a l elastic modulus. (4.6) PLATE DISPLACEMENT (cm) FIGURE 4 9 - VARIATION IN ER WITH STRESS AMPLITUDE AT A CONSTANT MAXIMUM STRESS LEVEL In this instance, and F^ represent the slope of the stress displacement curve. Figure A.10, on the other hand, shows how the modulus also varies with maximum attained load for constant stress amplitudes. This phenomena is summarized in Figure 4.11, where a considerable variation i n rebound modulus i s shown. An optimum level of plate load was found to be approximately 60% of the ultimate plate load. In practise, this plate load can be estimated from the cone resist-ance profile. Considerable evidence exists of variations i n screw plate rebound modulus, much more so that that observed in pressuremeter and laboratory tests. Perhaps the most important consideration i s the boundary stress condition, including the load acting on the back of the plate during unloading. The resistance of the s o i l column above the plate would lead to less elastic rebound; hence an increase i n the apparent r i g i d i t y of the s o i l . This increase would be dependent upon installation procedure as well as s o i l parameters, and has not been accounted for through closed-form solutions. Similarly, E R varies considerably with stress level at constant stress amplitudes, as was shown in Figure 4.10, and can be up to twice the i n i t i a l value. This behavior Is consistant with the observation of Makhlouf and Stewart (1965), which is shown in Figure 4.12. They found that sands were s t i f f e r with increasing stress level, and with decreasing strain amplitude. Consequently, the determination of an "elastic" modulus is very much dependent upon the stress level and amplitude during the unload-reload portion of FIGURE 4-10 EXAMPLE OF THE VARIATION IN E„ WITH STRESS LEVEL AT A CONSTANT STRESS AMPLITUDE P L A T E D I S P L A C E M E N T (cm) 60 FIGURE 411 - RELATIONSHIP BETWEEN REBOUND MODULUS AND INITIAL MODULUS AT VARIOUS LEVELS OF PLATE LOAD (data from a l l tests) INFLUENCE UPON E OF THE LOWER LIMIT OF A CONSTANT RANGE OF DEVIATOR S T R E S S INFLUENCE UPON E OF THE RANGE OF DEVIATOR S T R E S S WITH A CONSTANT UPPER LIMIT FIGURE 4 12 INFLUENCE OF STRESS RANGE AND LEVEL ON LABORATORY MODULUS DETERMINATION ( A F T E R M A K H L O U F a S T E W A R T , 1 9 6 5 ) test. Consistent modulus determinations cannot be obtained by selecting arbitrary stress levels or amplitudes, hence a standard procedure should be adopted. The Janbu analysis takes into consideration the effect of stress level with the modulus factor, k . As a result, i t * m represents one of the better methods of determining an in-situ modulus, particularly at low stress levels. The use of an i n i t i a l tangent to the loading curve reduces the possible errors inherent in methods which use the unload-reload curves. 4.1.3.A Young's Moduli i n Sands The load-displacement curves were also analyzed using Schmertmann's (1970) method, and equation 2.8. Young's moduli were obtained by assuming a secant modulus at 2 bars (-2 tsf) as suggested by Schmertmann, and neglecting the effects of compression due to creep in sands during a test of short duration. The results are presented in Figure A.13. The values obtained during this study exhibit a scatter similar to that found in the literature (Dahlberg (1975) and Schmertmann (1970)). Young's modulus, E , i s found to vary by a factor of 2-3 at the site, again possibly due to the variation in the lithology at the site. Perhaps the most significant observation here i s the tendency for the McDonald site data to increasingly underestimate E g in high q c ( s t i f f ) sands. Returning again to the effect of plate stiffness discussed in Section 3.1, the appropriate correction factor has again been applied, whereby 63 1000 10 J 1 >- ' 1 1 10 20 40 100 200 400 CONE BEARING CAPACITY <L. (bar) FIGURE 4-13 - OBSERVED RELATIONSHIP BETWEEN CONE BEARING q c AND VERTICAL YOUNG'S MODULUS, E s (SCHMERTMANN^ METHOD) MCDONALD SITE 64 OS.) s'corr E„ x B (4.7) where: s corr Young's Modulus corrected f o r plate s t i f f n e s s , and E s uncorrected modulus, and corr e c t i o n factor for plate s t i f f n e s s , 1.5 f o r f l e x i b l e plates 1 for r i g i d plates. A p p l i c a t i o n of t h i s c o r r e c t i o n f a c t o r to the trend obtained i n t h i s study brings i t well within the range reported i n the l i t e r a t u r e , and h i g h l i g h t s the importance of plate r i g i d i t y during the evalua-ti o n of test data. C y c l i c Loading for L i q u e f a c t i o n Assessment Hughes et a l . (1980) propose that repeated c y c l i c load t e s t s can be performed with the pressuremeter to assess the l i q u e f a c t i o n resistance. This assessment can be made through an estimate of the c y c l i c stress r a t i o to cause l i q u e f a c t i o n (Robertson 1982). The constant stress amplitude po r t i o n of the load curve of Figure 4.10 has been reduced, and the r e s u l t s shown i n Figure 4.14. Here the po t e n t i a l f o r using the screw plate test i n a s i m i l a r manner can be seen. Plots of cumulative or incremental s t r a i n can be used i n a manner s i m i l a r to that suggested f o r the pressuremeter. The maximum plate load reached during each cycle obviously a f f e c t s the cumulative s t r a i n to a large extent, however standardization of te s t 2 4 6 8 10 NUMBER OF CYCLES " N " FIGURE 4-14 - INCREMENTAL PLASTIC DEFLECTION PER CYCLE procedures could permit similar comparison to be made between the response in loose and dense sands. It should be noted here that several cyclic tests were also done to compare results from various waveforms. No appreciable change was observed with either triangular, square or sinusoidal loading. Sinusoidal loading was generally adopted for the cyclic tests because i t placed less strain on the hydraulic system. 4.1.4. Test Results - Undrained Behavior Undrained Shear Strength Undrained shear strengths were obtained using a variety of correlations with the screw plate, cone, pressuremeter and dilatometer data. Some disagreement between the various shear-strength determinations i s to be expected, due to inherent differences in the orientation of failure planes, induced disturbance and strain rate. Nonetheless, comparisons are presented herein. A summary of the undrained shear strength profile i s presented in Figure 4.15. The strength values agree reasonably well, within a range of 10 kPa of the median value. An N^ value of 9 for the screw plate data appears to give good agreement with other in-situ strength determination. The cone factor N£ = 17 was selected using a r i g i d i t y index of G/cu = 300, and the correlation developed by Baligh (1975). There is some evidence of scatter i n the data, which i s bri e f l y discussed in the next section. It is also important to note that the tests done in the s i l t y clay at depth were done while the author U N D R A I N E D S H E A R S T R E N G T H C u k P o • S C R E W P L A T E N K = 9 C O N E N c = 17 A P R E S S U R E M E T E R D I L A T O M E T E R FIGURE 4-15 - COMPARISON OF UNDRAINED SHEAR STRENGTHS MCDONALD S ILTY-CLAY SITE was s t i l l developing a suitable test procedure; consequently the possible effect of plate rotation during the application of load, which i s discussed in a later section, was not specifically addressed. In addition, at this particular site the effect of rod friction within the overlying dense sands could be significant, and result i n lower calculated values of c^. A proper assessment would require field vane tests at depth, as well as modifications to the installation system, including inner rods which would eliminate the effect of rod fr i c t i o n . Repeated Undrained Loading i n S i l t During undrained cyclic load tests i n s i l t , a significant reduction in ultimate plate capacity was noted with successive load increments. This behavior i s typified in Figure 4.16, where strain softening appears to be occuring. Continued strain also occurs after the load i s released, which would indicate that some consoli-dation may be taking place after the drop in the plate load. There i s l i t t l e evidence in the other in-situ tests to suggest that this behavior should be expected. The effect of strain-softening on a determination of c u would vary according to the degree of disturbance associated with each installation. Conse-quently, unless this phenomena i s accounted for, the selection of a suitable N„ or N factor is somewhat arbitrary. This factor K. C highlights the importance of carefully controlled installation of the plate , particularly in strain-softening materials. 69 FIGURE 4-16 - REDUCTION OF STRENGTH DURING LOAD TESTS IN SILT AT MACDONALD SITE 4.1.5. Torsional Resistance during Plate Installation Measurements of installation torque were made during several of the soundings at the site. Torque soundings for a 500 cm2 plate and a 250 cm2 plate are presented in Figure 4.17, along with the appro-priate cone bearing profile. The Installation torque "T" appears to correlate quite closely with the cone bearing q £, and varies according to the following approximation: T/q c = 17.6 in s i l t y clay = 6.6 In sand (4.8) where: T = torque in Nm q^ = cone bearing in bars. This correlation can be used to predict the site-specific installa-tion torque required given the cone bearing profile. In addition, the torque record can be used to indicate when pre-determined test depths are reached, and provide a rough log of stratigraphy at the site. 4.2. Langley Sensitive Clay Site 4.2.1. General Geology and Site Description This site Is adjacent to the #1 Highway, and situated at the base of a 5 m cut for the road right of way. The water table i s at 2.3 i depth. The general geology consists of Capilano sediments, raised marine and f l u v i a l deposits. The site stratigraphy i s : (T/qt) SILTY CLAY • 17.5 500 cm2 plate 2 50 cm' (T/q,J SAND « «.6 NOTt -. TOOObt IN Nm g c IN IMS FIGURE 4-17 - RELATIONSHIP BETWEEN INSTALLATION TORQUE AND CONE RESISTANCE AT SAND AND SILTY CLAY SITE 0-2 m - gravel and clay f i l l 2-10 m - clay, over-consolidated with interbedded s i l t y sand; sensitivity = 10-15 based on S. = (R.. = f r i c t i o n t R^ r ratio) 10-30 m - clay, slightly over-consolidated to normally consolidated. A typical cone profile i s presented in Figure 4.18. 4.2.2. Description of Test Programme A detailed site investigation was carried out, and i s summarized below: TEST TYPES NUMBER Cone penetrometer 1 Dilatometer 1 Vane 1 Piezometer Cone 1 Cone Pressuremeter 1 Screw Plate 1 TABLE 4.2. Summary of Test Programme - Langley Site. Location of the probings are shown in Figure 4.19. 4.2.3. Test Results A typical screw plate load test in sensitive clay i s presented in Figure 4.20. The achievement of a peak and residual load is more typical of the response which might be observed through laboratory testing, and has not previously been reported in literature covering the screw plate. FIGURE 4-18 - CONE PROFfLE AT LANGLEY CLAY SITE 74 L E G E N D P I C P T I O C T . 14 , 1 9 8 2 P 2 D I L A T O M E T E R N O V . 4 , 1 9 8 2 P 3 V A N E T E S T N O V . 18, 1 9 8 2 P 4 P O R E P R E S S U R E D I S S I P A T I O N S N O V . 18, 1 9 8 2 P 5 P U S H IN C O N E P R E S S U R E M E T E R N O V . 2 5 , 1982 P 6 S C R E W P L A T E D E C . 2 , 1 9 8 2 FIGURE 4-19 - LOCATION PLAN AT LANGLEY CLAY SITE DEPTH 7 m FIGURE 4-20- LOAD TEST IN LANGLEY SENSITIVE CLAY During the tests at this particular site, i t was observed that the torque rods were rotating slightly during the application of load. When the rods were clamped as shown in Figure 4 . 2 1 , the peak load was again achieved. A possible explanation of this phenomena i s that the f r i c t i o n a l resistance of the thoroughly remolded s o i l in contact with the plate i s negligible, hence the plate has a tendency to rotate during loading because of i t s screw pitch. This will tend to reduce the measured strength and modulus. The irregularities i n the loading curve beyond the peak are possibly related to failure planes which develop beneath the plate and result in a discontinuous load-deformation curve. Again, this phenomena has not been extensively treated in the literature. Note also that the apparent undrained modulus does not appear to have been significantly affected by loading the plate past the peak or "failure" load. This i s an important consideration when examining the relationship between i n i t i a l and reload modulus values from undrained tests i n clay. 4 . 2 . 4 . Undrained Modulus A comparison of undrained modulus values i s shown In Figure 4 . 2 2 . The screw plate moduli reported are E^Q, taken as a secant modulus to a load of 50% of P and E, , which is taken from ult' loop' the f i r s t reload cycle. The screw plate moduli give modulus values approximately twice that estimated from the cone. This seems quite reasonable, since the cone modulus is based on laboratory correlations which generally underestimate the in-situ modulus. DEPTH = 8 m .2 JL_ , 1 1 -0 5 10 P L A T E D I S P L A C E M E N T ( c m ) FIGURE 4-21 - EFFECT OF PLATE ROTATION DURING TESTING IN LANGLEY SENSITIVE CLAY FIGURE 4-22 - COMPARISON OF UNDRAINED MODULI AT LANGLEY CLAY SITE More test data i s required before a conclusion can be drawn. 4.2.5. Undrained Shear Strengths The general agreement between undrained strengths from a variety of tests (excluding the dilatometer) is evidenced in Figure 4.23. In view of the previously stated difference in the orientation of the failure planes associated with each test, this agreement i s not necessarily expected. Naturally, the selection of bearing factors is somewhat arbitrary, nonetheless the general agreement beween various i n - s i t u determination of c u i s encouraging. The value of N £ = 17 chosen for the cone was based on a r i g i d i t y index of 1^ = G/cu - 200, and Baligh's (1975) solution. Adoption of a single N^ or N^  for a particular site would require a more comprehensive laboratory and field programme. The effects of strain rate and strength anisotropy were not evaluated at this site; and would, result in variations in undrained shear strengths. A s t a t i s t i c a l l y valid number of tests would be required in order to make a confident assessment of these factors. 4.3. Cloverdale Sensitive Clay Site 4.3.1. General Geology and Site Description The site is located on level ground approximately 2-3 metres above sea level. The site i s underlain by s i l t y clays and clays of the Cloverdale sediments situated on t i l l . The specific lithology at the site i s : 0-1.5 m - s i l t , sand and gravel f i l l 1.5-4.0 m - clay, slightly O.C., soft 80 U N D R A N E D S H E A R 10 2 0 3 0 4 0 i 1— -t-STRENGTH Cu kPa 5 0 6 0 1 1 — LEGEND 5 + i a. 10-VANE SHEAR . DILATOMETER — • — U.B.C. CONE NC=I7 0 SCREW PLATE ( N K = 9) — + — PUSH-IN CONE PRESSUREMETER ASSUMING P L - IQ=5CU —• — — — — PUSH-IN CONE PRESSUREMETER (GIBSON a ANDERSON) ( P L - PQ)= 52 Cu / FIGURE 4 23 - COMPARISON OF UNDRAINED SHEAR STRENGTHS AT LANGLEY SENSITIVE CLAY SITE 81 4-14 m - clay, slightly O.C., very sensitive, Sfc = 10 to 27 14-25 m - clay, N.C, soft s i l t y 25-25.3 m - dense sand 25.3-27.7 m - clay, N.C. 27.7 •»• - glacial t i l l A typical cone profile i s presented in Figure 4.24. 4.3.2. Description of Field Programme A series of in-situ tests were completed at the site. The test programme is summarized in Table 4.3 below. TEST DESCRIPTION NUMBER Cone penetration test 1 Dilatometer 1 Vane shear 1 Piezometer cone profile 1 Screw plate test profile 1 TABLE 4.3. Summary of Field Programme - Cloverdale Site. Test locations are given in Figure 4.25. 4.3.3. Test Results Undrained Elastic Modulus A typical load displacement curve i s presented in Figure 4.26. At this site the torque rods were not clamped, hence the true peak loads were probably not attained. Based upon the tests at Langley, this might lead to a 10-15% underestimation of the ultimate load. It i s also evident here that the selection of a suitable modulus PORE £ PRESSURE U (BAR) FC 0 24 0 FRICTJ^JJpiST 05 0 BEARING , RESISTANCE QC (BAR) 20 0 OIFF. RR. RAT" TIO AllAX 0 JBO • * SOIL PROFILE u r g a n i c , s i l t . s a n a & asstd. gravel f i l l S o f t . s e n s i t i v e , s l i g h t l y over-consolidated clay Soft,sensitive s l i g h t l y over-consolidated s i l t y clay. S u increases l i n e a r l y with depth from about 14 kPa-26 kPa Su/p* r a t i o - 0.54 at surface to 0.28 at 14 m Cv = 0.01 ctn2/min k - 4.9 X 10-8cm/sec Soft, s e n s i t i v e normally c o n s o l i -dated s i l t y - c l a y S u increases l i n e a r l y with depth about 26 kPa-47 kPa S u/P' r a t i o i s constant with depth at 0.28. Entire stratum has tandom & occa-s i o n a l lenses of s i l t SAND Normally consolidated clay Dense g l a c i a l t i l l End of Hole ~ CLOVERDALE UBC CONE NOV. 5,1981 UBC PIEZOMETER CONE 3. NO PORE PRESSURE CORRECTION REQ0. FIGURE 4-24- TYPICAL CONE PROFILES AT CLOVERDALE SITE 63 83 t LEGEND TEST HOLE NO i . 2. 3. 4. FUGRO CONE ^FSCK 225, OCT. 22/81 FLAT DILATOMETER, OCT. 29/81 SWEDISH VANE, OCT. 29/81 U.B.C. PIEZOMETER CONE *C6 FPST 3 UBC, NOV. 5/81 SCREW PLATE TEST, NOV. 19/81 GRAVEL ROAD •-BRIDGE PAVED ROAD NOT TO SCALE FIGURE 4-25 - LOCATION PLAN CLOVERDALE CLAY SITE E R - REBOUND MODULUS E| = INITIAL MODULUS DEPTH = 6m 50 4 PLATE DISPLACEMENT (cm) FIGURE 4-26 - LOAD TEST IN CLOVERDALE CLAY from a reload curve i s very much dependent upon the amplitude of the plate loadings reached during each cycle. Values of E U are compared in Figure 4.27. It appears that except for E ^ Q , the screw plate test produces higher values of the in-situ undrained modulus, as was the case at Langley. Again, this reflects the fact that the screw plate modulus may more closely correlate with the true in-situ modulus. An important point here is that the tests in the Langley clay showed no change in modulus for a f u l l unload-reload past failure. In contrast, the Cloverdale test shows that the modulus varies i f f u l l unload is not achieved. Undrained Shear Strength Undrained shear strengths are compared in Figure 4.28. A cone N = 12 was selected based on a best f i t with the corrected vane c strengths. Remarkably good correlation i s seen between the various in-situ tests, with the dilatometer again giving slightly higher values of c u in the sensitive clay. It should be noted, however, that at this site, plate rotation was not prevented, hence the ultimate load reached during the test may have been underestimated. 86 FIGURE 4-27 - COMPARISON OF UNDRAINED MODULI AT CLOVERDALE CLAY SITE UNDRAINED SHEAR STRENGTH Cu kPa FIGURE 4-28 - COMPARISON OF UNDRAINED SHEAR STRENGTHS AT CLOVERDALE, CLAY SITE 5. SUMMARY AND CONCLUSIONS 5.1. Measurements of Drained Behavior 5.1.1. Constrained Modulus The screw plate test appears to produce an underestimation of the in-situ constrained modulus, M. This underestimation arises because of several factors, most notably: 1) plate flexure which results ln greater deflections, 2) Janbu's assumption of vertical compression, 3) disturbance associated with plate installation. This combination yields values of k m which are 25 to 30 percent of expected k values. The k determinations are consistent, m m however, and indicate that reproducible results can be obtained. 5.1.2. Young's Modulus The determination of a Young's modulus, E g, using Schmertmann's method i s also sensitive to the effect of plate stiffness. The results shown herein indicate that the application of a modulus factor, A, according to the relative stiffness of the plate, w i l l result in E g values similar to those published elsewhere. This highlights the applicability of the test in sands of varying densities. 5.2. Determination of Drained Modulus from Unload Reload Curves Review of the cyclic loading curves indicates that the screw plate reload modulus Is very much dependent upon stress level and stress amplitude. Determination of an absolute reload modulus i s further complicated by the uncertain load condition acting on the back of the plate. It does appear, however, that careful selection of a maximum plate load and stress amplitude could yield a reasonably consistent modulus values. 5.3. Measurement of Undrained Shear Strength - c u The screw plate c y determinations agreed very well with other in-situ tests. An value of 9.0 was found to provide a reasonable estimate of c y in the clays tested, however further site specific correlations are required. It was also discovered that plate rotation during load testing in sensitive clays can lead to an underestimation of c u. In general, in-situ determinations of c u must address the issues of strain rate, strength-anisotropy, and progressive failure before confident correlation can be made. 5.4. Measurement of Undrained Modulus The most apparent uncertainty of the screw plate test i s i t s estimate of undrained modulus in sensitive clay. As with the determination of sand moduli, the plate boundary conditions including plate back load and torque-induced shear strain contribute to rather uncertain E u determinations. Part of the inaccuracy of existing interpretive techniques also l i e s in their assumption of homogeneous, elastic, isotropic media. The selection of a suitable undrained modulus from the screw-plate test has historically involved an arbitrary selection of the stress level at which an unload-reload modulus i s determined. The data presented during t h i s study h i g h l i g h t s the necessity f o r further work Into the r e s o l u t i o n of whether i n i t i a l tangent or reload moduli should be used. 91 6. SUGGESTED TEST PROCEDURE AND ANALYSIS As with any test, the standardization of test procedures and analyses w i l l improve the applicability of the test in a variety of s o i l conditions. Based on the results of this study, the following guidelines are proposed. 6.1. Test Equipment 6.1.1. Screw Plate The screw plate should satisfy the following c r i t e r i a : (1) double-flighted (2) cross-sectional area of 250 cm2 for use in sand; 500 cm2 i n clay (3) have a high relative stiffness, as defined in Section 3.1.2. 6.1.2. Installation System The torque required during installation of the plate depends primarily upon the s o i l type. During this study, i t was observed that the installation torque varied accordingly: 100 to 250 Nm - i n clay and s i l t 250 to 2000 Nm - in sand. The selection of a suitable torque motor w i l l be governed by the observed cone resistance. Where a higher torque is required during installation, the plate may be advanced by cy c l i c a l l y reversing the direction of rotation; alternatively a smaller plate should be used. The rate of advance of the plate can be up to a rate equal to the pitch times the rate of rotation. In general, a low f r i c t i o n bearing at the top of the rods should permit the unhindered advancement of the plate into the s o i l . The screw plate rods should be pinned so that the plate may be removed upon completion of the profile. A larger rod cross-sectional area i s required i n order to improve the relative stiffness of the plate, perhaps at the expense of Increased rod f r i c t i o n during loading. 6.1.3. Loading System The loads required for failure in cohesive materials ranged from 1.0 to 2.5 bars. In sands, failure loads may not necessarily be achieved, as the resistance of the sand w i l l increase during testing. Consequently, the loading frame should have a capacity of approximately: 12 kN i n s i l t and clays 50 to 75 kN in sand. 6.1.4. Measurement System The parameters which must be measured during the test are: (1) plate load (2) plate displacement. The load may be measured using a load c e l l , or by measuring the piston pressure with a pressure gauge. The capacity of the load c e l l should be greater than the overall capacity of the loading frame. The displacement may be measured using a displacement transducer or dial gauge. During this study i t was observed that displacements were generally: Total = 0 to 2.0 cm in sand = 2.0 to 5.0 cm i n clays. The range and accuracy of the displacement gauges should be selected accordingly. The use of automated recording equipment and electrical transducers allows the operator to apply rapid cyclic loads with a suitable loading frame. 6.2. Suggested Procedure for Plate Installation After selection of a suitable plate and installation system, the installation of the plate i t s e l f is relatively straightforward. It can be rotated to the test depth at a variable rate, governed primarily by the a b i l i t y of the frame torque motor to follow the plate downward. The plate should be permitted to "pull i t s e l f " downward without the application of vertical loads. Load tests should be separated by at least four plate diameters in order to reduce the possibility of superposition of strains. Upon completion of the profile, the plate should be withdrawn and inspected. During withdrawal, i t may be necessary to clamp the rods to prevent the apparatus from slipping downward, though i t i s unlikely that the rods would drop completely down the hole. 6.3. Suggested Test Procedure in Sand Drained tests in sand should be carried on in the following manner. 1) An i n i t i a l load equal to approximately 60% of the capacity of the loading frame should be applied, and the load displacement curve recorded. 2) The constrained tangent modulus i s then obtained by: j / j l - a p a Janbu and Senneset (1974) M = k p (£-r~a ( 2 . D where: M = constrained tangent modulus k = modulus number m p = reference stress (1 bar) 3. p' = vertical effective stress a = stress exponent = .5. The modulus at the test depth can be estimated from a tangent drawn through the i n i t i a l portion of the test curve S P n B 6 = — — (2.2) k p m *a where: 6 = plate deformation, corrected for rod compression B = plate diameter p^ = net plate stress = applied stress + rod weight where necessary S = dimensionless settlement number, S, (Figure 2.2) . 3) The stiffness of the plate should be calculated using the stiffness factor R, whereby: 95 R = - * ( 3 - 4 v ) ( l - r V ) 5> ( h j 3 ( 3 > 2 ) 12 (1-v )(l-v) E 3 P Selvadurai et a l . (1979) in which: h = plate thickness a = plate radius {v} and {E} = s o i l elastic constants { V p } and {Ep} = plate elastic constants. The calculated modulus value should then be multiplied by the appropriate stiffness correction factor, B: 3 = 1 for log R = 2 B = 1.5 for log R = 0. 4) Based on the limited data obtained during this study, a r e a l i s t i c vertical constrained modulus, , can be estimated ' true' using: M t r u e = (2.3 to 3.6) in sand. (4.5) ^Janbu 5) Alternatively, Schmertmann's method of analysis can be applied to the data to obtain an equivalent Young's Modulus, E g, for vertical compression. E g is obtained by using a secant modulus over the range from 1 to 3 bars, and the relationship: 2B I = C Ap I — Az 1 o E s (2.8) where: p = plate settlement AP = P ~ P Q p = applied plate stress p Q = in-situ effective overburden pressure B = plate diameter I = strain influence factor (Fig. 2.3) z A = depth z = embedment correction Po { 1 - .5t-!4 } P-po Any consistent units may be used. This modulus has been verified for a number of case histories, and i s most applicable to shallow foundations which generally have design stresses of that order. 6) A complete unload-reload cycle should then be completed to the maximum capacity of the loading frame. The secant modulus during this load curve, should be approximately equal to the i n i t i a l tangent modulus. A deviation from this observation probably reflects the effects of s o i l disturbance. A significant difference would indicate that the s o i l has been disturbed during installation of the plate, and that only the rebound modulus w i l l be reliable. 97 6.4. Suggested Test Procedure ln Clay (1) Based on an understanding of the site stratigraphy, which can be obtained through cone penetration testing or more conventional investigative methods, the in-situ undrained shear strength should be estimated. The ultimate plate stress can be estimated using the following relationship: Pult * c u \ + 0 v o ( 2 ' 1 5 ) where: Pu]_t = u l t : * - m a t e plate capacity c u = undrained shear strength = bearing factor = 9 o = in-situ vertical effective stress, vo (2) A rapid undrained test should be performed, with the load applied to 60% of the estimated Puj_t» a n d dropped to zero. An estimate of undrained modulus, E , can then be made using Selvadurai et a l . (1979): pa/E u = X (2.13) Selvadurai and Nicholas (1979) where A = {0.60 to 0.75} = modulus factor (2.14) p = plate load a = plate radius E = undrained elastic modulus u 6 = plate displacement. This modulus should be calculated using the i n i t i a l portion of the load displacement curve. 98 (3) Care must be taken to prevent plate rotation during the test. (4) A final load cycle should be applied to the failure load. From this load, c can be determined from u p , - a ult V O / o i c \ cu = 57 ( 2 ' 1 5 ) k where: c = undrained shear strength u P ^ t = plate load plus rod weight a = in-situ vertical stress vo = bearing factor = 9. (5) Further load cycles should be applied to check for strain softening. (6) The plate should be advanced at least one metre to the next test depth, and the test repeated. (7) Upon completion of the profile, the plate should be withdrawn, and inspected for damage. (8) In s t i f f e r clays, where an ultimate load may not be achieved, the ultimate load may be estimated by: p , = 2.54 p - 1.54 p (2.17) *ult r y x Kay and Parry (1982) in which: p .^ = ultimate plate stress ult p = plate stress at a strain equal to X 1.5% of the plate diameter (B) 99 Py = the plate stress at a strain equal to 2% of B. (2) The constrained modulus number 'k ' can be estimated in a f u l l y m drained test from the i n i t i a l tangent portion of the curve and Janbu's formula: S Pn B 6 * I T — m a where: 6 = plate deformation k = constrained modulus number m S = dimensionless settlement number (see Figure 2.2) p = net plate stress n P a = reference stress (= 1 bar) B = plate diameter. (3) Corrections should be made for rod compression and effect of plate r i g i d i t y . These corrections are detailed in 3.1.2. 100 7. SUGGESTIONS FOR FUTURE RESEARCH Future research related specifically to the research sites discussed herein should be directed towards obtaining more f i e l d vane shear data to correlate with the screw plate estimates of undrained shear strength. This i s particularly true of the s i l t y clays at depth at the McDonald site, where no vane tests were completed. Through these correlations, the effects of variable strain rate and strength anisotropy can be examined. Furthermore, efforts should be made to eliminate rod f r i c t i o n and plate rotation in the additional screw plate tests at those depths. Further research into the s o i l behavior in drained tests in sand should be directed at achieving a more fundamental understanding of the relationship between the screw plate - derived modulus number k , and the true k_ value of the s o i l , m' E Parametric finite element studies, including a study of the effect of load on the back of the plate, would be most applicable here. Additional f i e l d testing could be used to verify the predicted response, and directed towards improvements in the test apparatus, including plate geometry and stiffness. As our fundamental understanding of s o i l behavior improves, more attention can then be directed towards interpretting the results of cyclic load tests performed using the plate. This could include relating the unload-reload curves to such parameters as i n -situ moduli and sand densities, and their application to machine foundations. BIBLIOGRAPHY Berzins, W.E. and Campanella, R.G., 1981, "Development of the Screw Plate Test for In-sltu Determination of Soil Properties", Dept. of C i v i l Engineering, University of British Columbia, Soil Mechanics Series No. 48. Bjerrum, L., 1973, "Problems of Soil Mechanics and Construction on Soft Clays", Proceedings 8th International Conference on Soil Mechanics and Foundation Engineering, Vol. 3, Moscow. Bodare, A. and Massarch, R., 1982, "Determination of Dynamic Soil Properties i n the f i e l d " , VBB Report, Sweden. Campanella, R.G. and Robertson, P.K., 1981, "Applied Cone Research", Sym. on Cone Penetration Testing and Experience, Geotechnical Engineering Div., ASCE, Oct. 1981, pp. 343-362. Campanella, R.G., Robertson, P.K. and Gillespie, D., 1983, "Cone Penetration Testing in Deltaic Soils", Canadian Geotechnical Journal, Vol. 20, February. Dahlberg, R. 1974. "Penetration, Pressuremeter and Screw Plate Tests in a Preloaded Natural Sand Deposit", Proceedings of the European Symposium on Penetration Testing, ESOPT I, Stockholm, Vol. 22. Gould, J.H., 1967, "A Comparative Study of the Screw Plate and Rigid Plate Bearing Tests", M.Sc. Thesis, Dept. of C i v i l Engineering, University of Florida. Hughes, J.M.O., 1982, "Interpretation of Pressuremeter Tests for the Determination of the Elastic Shear Modulus", Engineering Foundation Conference on Updating Subsurface Sampling of Soils and Rocks and Their In-situ Testing, Santa Barbara, California. Janbu, N., (1963), "Soil Compressibility as Determined by Oedometer and Triaxial Tests", Proceedings 4th European Conference on Soil Mechanics and Foundation Engineering. Janbu, N. and Senneset, K., 1973, "Field Compressometer - Principles and Applications, Proceedings 8th International Conference on Soil Mechanics and Foundation Engineering, Moscow, Vol. 1.1. Kay, J.N. and Avalle, D.L., 1982, "Application of Screw-Plate to Stiff Clays", Journal of the Geotechnical Engineering Division, ASCE, Vol. 108, No. GT1, January. Kay, J.N. and Parry, R.H.G., 1982, "Screw Plate Tests in a S t i f f Clay", Ground Engineering, September. Kummeneje, 0., 1956, "Foundation of an Oil Tank in Drammen", Norwegian Geotechnical Institute Publication No. 12. Kummeneje, 0. and Eide, 0., 1961, "Investigation of Loose Sand , Deposits by Blasting", Norwegian Geotechnical Institute Publication No. 45. Ladd, C.C., Foot, R., Ishihara, K., Schlosser, F., and Poulous, 1977, "Stress-Deformation and Strength Characteristics", Proceedings, Ninth International Conference on Soil Mechanics and Foundation Engineering, Tokyo, Japan, Vol. II. Makhlouf, H.M., and Stewart, J.J., "Factors Influencing the Modulus of Elasticity of Dry Sand", Proceedings Sixth International Conference on Soil Mechanics and Foundation Engineering, Montreal. Marsland, A. and Randolph, M.F., 1977, "Comparison of the Results from Pressuremeter Tests and Large In-situ Tests in London Clay, Geotechnique 27, NO. 1. Robertson, P.K., 1982, "In-situ Testing of Soil with Emphasis on its Application to Liquefaction Assessment", Ph.D. Thesis, Dept. of C i v i l Engineering, University of British Columbia, 1982. Schmertmann, J.H., 1970, "Static Cone to Compute Static Settlements Over Sand", Journal of the Soil Mechanics and Foundation Division, ASCE, Vol. 96, May. Schwab, E.F., 1976, "Bearing Capacity, Strength and Deformation Behavior of Soft Organic Soils", Dept. of Soil Mechanics and Rock Mechanics, Royal Institute of Technology, Stockholm. Schwab, E.F. and Broms, B.B., 1977, "Pressure-Settlement-Time Relationship by Screw Plate Tests In-situ", 10th International Conference on Soil Mechanics and Foundation Engienering, Stockholm. Selvadurai, A.P.S. and Nicholas, T.J., 1979, "A Theoretical Assessment of the Screw Plate Test", 3rd International Conference on Numerical Methods in 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, November. Webb, D.L., 1969, "Settlement of Structures on Deep Alluvial Sand Sediments in Durban, South Africa", British Geotechnical Society Conf. on In-situ Investigations in Soils and Rocks, Session III, Paper 16. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items