COMPARATIVE STUDY OF VTM AND AASHTO TEST ..., Study notes of Design

Download COMPARATIVE STUDY OF VTM AND AASHTO TEST ... and more Study notes Design in PDF only on Docsity! Page 1 of 18 COMPARATIVE STUDY OF VTM AND AASHTO TEST METHOD FOR CBR Shabbir Hossain, Gale M. Dickerson and Chaz B. Weaver, CO Materials Division, VDOT 1.0 INTRODUCTION 1.1 Background The California Bearing ratio (CBR) test has been widely used in pavement design since mid 1940’s. It is a relative measure of subgrade soil or base/subbase aggregate strength. The Virginia Department of Transportation (VDOT) has been using its own Virginia Test Method-8 (VTM-8) to perform CBR testing. On the other hand, private industry and other DOT’s are using AASHTO (American Association of State Highway and Transportation Officials) test method (T 193) to measure CBR. Although there are some differences between these two test procedures, both of them measure the same soil property. According to Thomas Stanton (Materials and Research Engineer, CA Division of Highways), the Bearing Ratio Test was adopted in California in 1930. VDOT’s test method for CBR was first introduced in the Soils Laboratory’s (Division of Tests, Virginia Department of Highways) Virginia’s Test Method of Conducting C.B.R. Tests in January 1952. The current standard VTM-8 follows the original 1952 method exactly. AASHO introduced T 193 in 1963; which was revised and updated by AASHTO in 1972, 1981 and 1999 to meet industry requirements and advances in theory and technology. The test procedures in AASHTO is internationally recognized and widely used in textbooks, scientific studies and industry literatures. Moreover, they are constantly updated to reflect the knowledge gained from research and field experience. The AASHTO procedure for CBR test is more flexible than Virginia method and it allows the designer to request other information such as CBRs for range of moisture content and density. VDOT is currently equipped to conduct AASHTO version of CBR test in all of its seven soils labs and at least four of these labs are accredited by AASHTO Materials Reference Laboratory (AMRL). Therefore it would be beneficial for VDOT to investigate the possibility of using AASHTO method in order to maintain comparable industry standard. 1.2 Objective The main objective of this study is to compare the CBR values measured according to VTM and AASHTO procedures, respectively. The significance of any difference, if any existed, also needs to be investigated. The other objectives are as follows: • Identify the factors that may affect the CBR results as measured by VTM and AASHTO procedures • Develop recommendations so as to get a comparable CBR value from VTM and AASHTO methods. 1.3 Scope The major factors influencing the CBR values measured by VTM and AASHTO methods were evaluated at Central Office (CO) Soils Laboratory (at Elko) with only one source of soil from Fredericksburg District. In addition to these samples, three other district laboratories participated in the direct comparison of CBR values measured with VTM and AASHTO procedures. 2.0 COMPARISON OF TEST METHODS Although the CBR test is one of the most widely used tests for evaluating subgrade support value, there are variations in the procedure followed by different agencies. In addition to agency specific standards, ASTM and AASHTO have developed standards over the years. The Virginia Department of Transportation mostly follows the AASHTO standard with some modification as stated in the standard Virginia Test Method-8 (VTM-8). Both AASHTO and VTM are primarily intended to determine the CBR value of soil and soil- aggregate having a maximum particle size of less than ¾”. VTM does not have any separate procedure for soils having particles larger than ¾”; it follows AASHTO recommendation. Some of the major modifications of the AASHTO method incorporated into VTM are discussed below: Page 2 of 18 2.1 Equipment and Accessories The size of the mold used in VTM and AASHTO procedures is different. Although the diameters are same, there are differences in heights as summarized in Table 1. The height of the spacer disk in AASHTO method is 2.416-inches which produces a compacted sample with a net height of 4.584 inches. It is important to note that the height of the compacted sample from moisture-density relationship test procedure (Proctor: T 99 or Modified proctor: T 180) is also same as AASHTO CBR test (4.584 inches). On the other hand, VTM uses 6” and 7” molds with 1” and 2” spacer disks, respectively. Therefore, the net sample height is 6 inches for VTM method compared to 4.584 inches in AASHTO. This difference in height of the sample may affect measured CBR values. There are minor differences in the height and diameter of swell plate and the size of plate spindles. This may not produce any significant difference in measured soaked CBR values. Although the default surcharge weights are the same in both methods for normal condition, AASHTO allows use of higher loads as desired by the client. Therefore the difference between the two methods for swell potential should be minimal for normal surcharge weights. Table 1: Differences in Equipment and Accessories Between VTM and AASHTO Methods. CBR Test Standard Equipment and Accessories Virginia Method (VTM-8) AASHTO Method (T 193) Remarks Mold Size Diameter = 6 inches Height = 7 or 8 inches Diameter = 6 inches Height = 7 inches Spacer Disk Height 1 inch Tall for 6”x7” Mold 2 inches Tall for 6”x8” Mold (results in same sample height for both mold) 2.416 inches Swell Plate Diameter = 5.75 inches Thickness = 0.125 inches Perforated but diameter or number of holes are not specified. Diameter = 5.875 inches Thickness = 0.25 inches Perforated with 42 Holes Hole diameter = 0.0625 inches If not specified, VTM usually follows AASHTO. Swell Plate Spindle Length = 3.75 inches Diameter = 0.375 inches Length = Adjustable (≈ 4” to 7”) Diameter = 0.375 inches Surcharge Weight Diameter = 5.75 inches Weight = 5 lb Hole at the center; Always uses two surcharges. Diameter = 5.875 inches Weight = 5 lb May be slotted or split; Usually uses two surcharges. More surcharges could be used if specified by the client 2.2 Procedure Although the basic mechanism of measuring CBR value is the same in both VTM and AASHTO methods, there are some differences in the sample preparation techniques as presented in Table 2. Virginia Test Method (VTM) uses more specific values for moisture content and density for the samples whereas AASHTO allows the user agency to specify these values. Moreover, AASHTO has a provision to conduct the test for a range of moisture contents and densities as requested by the client. The specified value in VTM method is always at optimum moisture content and maximum dry density as determined by VTM-1 (Proctor) method. AASHTO also has the option of running the test at optimum moisture content and maximum dry density as determined by either AASHTO T 99 or AASHTO T 180 method. Even in this situation the samples in AASHTO method are prepared at the optimum moisture content but the CBR value at maximum dry density is interpolated from a range of CBR values measured at different densities. There Page 5 of 18 3.0 VARIABILITY OF TEST METHOD Every laboratory test method has inherent variability that can’t be avoided in a standard condition. The precision and bias statement in any standard test method usually addresses such issues. Unfortunately such a statement is not available in either VTM or AASHTO method. But some of the variability for both VTM and AASHTO methods could be estimated from other available sources. There are two types of variability considered for any test method: within-laboratory (single operator) and multi-laboratory. The within-lab variability measures the repeatability characteristics and multi-lab variability defines the reproducibility of a test method in different laboratories. In this study, replicate samples are tested in the same laboratory to compare VTM and AASHTO methods. Therefore, only the within-lab component of variability may influence this study results. The variability of AASHTO method is estimated from AMRL proficiency data available at ARML’s website (www.amrl.net). The single operator (within-lab) coefficient of variation for samples from the last several years is presented in Figure 1. The coefficient varies from 13% to 22% with the exception of a few values. For the last several years, the Virginia Department of Transportation also conducted a round robin study similar to AMRL proficiency tests for VTM method of CBR testing. Unfortunately it was not designed to measure within lab variability. Different VDOT districts and several private laboratories have participated in this study. As required by VTM method, both soaked and unsoaked CBR were measured. The results are presented in Figures 2 and 3, respectively. From these data multi-lab variability was estimated as shown in Figure 4. In general it varies from 17% to 38% over several years. Single operator/within-lab variability will be much smaller than these values and may be comparable to AMRL data. The ASTM standard for CBR testing also provides some measure of variability. The within-lab variability is estimated to be 8% but it is based on measurements of seven replicate samples from one source in one lab. The variability of VTM and AASHTO methods are comparable as discussed above and summarized in Table 4. Based on the data discussed above, an estimation of single operator/within lab coefficient of variability of 8% may be considered conservative. In that case, CBR measured on two replicate samples by the same operator in the same laboratory within a reasonable time period may statistically (at 5% significance level) vary a maximum of 22% (8* 2√2) of their mean. Such variability of the test method needs to be considered in comparing VTM and AASHTO methods. Table 4: Variability of Test Methods. Estimated Variability (Coefficient of Variation) Test Method Within-laboratory (single operator) Multi-laboratory VTM-8 Not Available Approximate Range: 17 to 38% AASHTO T 193 Approximate Range: 13 to 22% Approximate Range: 35 to 91% ASTM D 1883 8.2% (based on very limited data) Not Available Page 6 of 18 AMRL Single-operator Precision (CV) 0.1 inch CBR 0 10 20 30 40 50 60 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Year Single Operator Precision (sample #1): Avg. 19.3 (Range 13-34) Single Operator Precision (sample #2): Avg. 21.6 (Range 13-49) Figure 1: Variability of AASHTO method from AMRL database. Page 7 of 18 VDOT Round Robin Study (VTM-8) 0 5 10 15 20 25 V 1 V 2 V 3 V 4 V 5 V 6 V 7 PL1 PL2 PL3 PL4 PL5 Participating Laboratory Identifier 1998 1999 2000 2002 Figure 2: Soaked CBR values from VDOT round robin study. Page 11 of 18 4.0 FACTORS INFLUENCING CBR RESULT As discussed above, there are several key factors such as mold/sample height, compaction effort, etc. influencing the CBR results. These factors may produce a significant difference between the final CBR values from VTM and AASHTO methods. An experimental program as shown in Table 5 was designed to evaluate these factors. A red sandy lean clay soil sample from Stafford County of Fredericksburg District was used for this experimental evaluation. The variables considered in the experimental design were mold/sample height, number of layers and number of blows per layer. The optimum moisture content and maximum dry density were determined using VTM-1 method which is essentially same as AASHTO T 99 for this particular soil because of 100% passing #4 sieve. The soil properties are presented in Table 6. As suggested in VTM-8 method, all samples were compacted at approximately 1.6% below the optimum moisture content and only unsoaked condition was used to measure CBR values. Table 5: Experimental Program to Evaluate Factors Influencing CBR Results. Variables: Mold Type: VTM and AASHTO Compaction Effort (Total blow count): 120, 225 and 270 Number of Layers: 3 and 5 (2x3x2 = 12 samples, no replicates) Mold Type Total Number of Blows Number of Layers Blows per Layer VTM AASHTO 3 40 Sample 1 Sample 2 120 5 24 Sample 3 Sample 4 3 75 Sample 5 Sample 6 225 5 45 Sample 7 Sample 8 3 90 Sample 9 Sample 10 270 5 54 Sample 11 Sample 12 Table 6: Soil Properties for Experimental Program Soil Properties District Fredericksburg, VA County Stafford Color Red Visual Classification Red Clay w/ fine sand & trace silt Percent passing no. 4 100 Percent passing no. 200 55 Specific Gravity 2.765 Liquid Limit 35 Plasticity Index 16 Unified Soil Classification CL (Sandy Lean Clay) Page 12 of 18 To incorporate the variation in sample height, both VTM and AASHTO molds were used to prepare the samples. Use of VTM and AASHTO molds produced sample heights of 6” and 4.584”, respectively. Although VTM method allows both 7” and 8” mold, a 7” mold was used for this comparison. The number of layers to fill the mold was another variable used in this experiment. The samples were compacted at both 3-layers and 5-layers as done in AASHTO and VTM methods, respectively. As required in AASHTO method, three levels of compaction effort were used to prepare CBR samples with total blow counts: one of them being same as VTM method; this scheme covers both VTM and AASHTO methods for a relative comparison. The total number of blows for each compaction effort was distributed evenly on 3 layers and 5 layers as needed. Samples were prepared and tested in the CO Soils Lab at Elko. Although the use of replicate samples was planned, only one sample for each combination was tested due to shortage of soil. The results of CBR tests are summarized in Table 7. It is important to note that none of the samples needed any origin correction as explained earlier (in data interpretation section). Table 7: CBR Test Results for Laboratory Experiment at CO Soils Lab. Compaction Sample No. Test Procedure Number of Layers Blows per Layer Sample Density (% Optimum) Water Content (%) CBR (%) 1 VTM 3 40 86.8 10.2 14.2 2 AASHTO 3 40 87.6 10.2 13.3 3 VTM 5 24 86.3 10.5 13.7 4 AASHTO 5 24 94.8 10.5 16.6 5 VTM 3 75 91.1 10.2 23.7 6 AASHTO 3 75 94.8 10.2 25.0 7 VTM 5 45 92.6 10.5 22.3 8 AASHTO 5 45 94.7 10.5 25.8 9 VTM 3 90 91.3 10.2 22.0 10 AASHTO 3 90 95.9 10.2 25.8 11 VTM 5 54 94.6 10.5 28.3 12 AASHTO 5 54 96.3 10.2 32.0 • Moisture-Density relationship according to VTM-1: ! Maximum Density = 119.6 lb/ft3 and Optimum M.C. = 12% In general, higher compacted density should yield higher CBR value. The measured CBR values from this experiment are plotted against the achieved density of the sample in Figure 5. Data are grouped according to mold type (two levels: VTM and AASHTO) and number of layers (two levels: 3 and 5) to fill the mold. Lines for all four combinations in the Figure 5 are close to each other and have positive slopes indicating the expected trend. For a particular density the variation in CBR values for all four combinations is minimal and may well be within the normal variation (22% of the mean) of the test methods. A statistical trend line was drawn through the data. The coefficient of regression (R2) of 0.9 for the trend line indicates a fairly strong effect of density on CBR value. Two lines with a variation of 22% of mean CBR value based on trend line are drawn on the same plot. All the data points are within the range of these two lines indicating no significant difference between these two methods of measurement for CBR value as long as they are based on density. One important thing to note here is non-achievement of maximum dry density despite high compaction effort in both methods. Cutback on moisture content from optimum may account for this low density but such a small amount of reduction in moisture content should not have any significant effect in the relative comparison between these two methods. Page 15 of 18 Table 8: CBR Test Results for Direct Comparison. AASHTO T 99 Method A AASHTO T 99 Method D Density before Soaking Soaked CBR Value Sample No. District Description AASHTO Lab Classification OM MD OM MD VTM-8 AASHTO T 193 VTM-8 AASHTO T 193 1 Culpeper Reddish tan clayey SILT, trace sand and mica A-7-5(22) 24.20 97.10 22.40 100.00 96.90 98.40 9.60 9.00 2 Culpeper Olive brown to reddish brown micaceous SILT, trace sand A-4(0) 17.20 106.20 17.10 108.90 103.60 107.50 4.30 1.50 3 Lynchburg Red micaceous silt A-2-5(0) 20.40 99.10 99.20 97.80 4.20 4.3* 4 Lynchburg Tan micaceous silt A-2-5(0) 15.10 106.30 99.99 101.70 5.10 7.2* 5 Lynchburg Yellow tan silt with decomposed stone A-2-4(0) 14.40 114.20 101.60 102.40 2.70 5.6* 6 Lynchburg Red silty clay A-4(4) 24.10 96.80 101.90 99.00 7.40 8.3* 7 Staunton Red clay A-6(15) 19.60 105.60 100.68 n/a 5.00 2.27 8 Staunton Red silty clay A-6(5) 16.10 113.00 100.69 n/a 10.77 6.10 9 Staunton Brown silty clay A-3(0) 15.30 105.20 99.94 n/a 16.13 14.87 10 CO Soils Lab Red clay A-7-5(10) 22.00 100.10 98.84 97.17 11.33 6.67 11 CO Soils Lab Decomposed stone A-2-4(0) 12.30 117.00 99.45 100.85 22.50 15.83 12 CO Soils Lab Tan micaceous silt A-2-4(0) 13.10 103.00 97.55 97.55 7.17 4.17 AASHTO T-99 Method A is assumed unless otherwise noted *Corrected values Page 16 of 18 Table 9: Direct CBR Comparison Statistics for F-test and t-test. F-Test Two-Sample for Variances VTM-8 AASHTO T-193 Mean 8.85 7.150833333 Variance 33.0616 19.65122652 Observations 12 12 Df 11 11 F 1.68241916 P(F<=f) one-tail 0.200797698 F Critical one-tail 2.817927225 t-Test: Two-Sample Assuming Equal Variances VTM-8 AASHTO T-193 Mean 8.85 7.150833333 Variance 33.0616 19.65122652 Observations 12 12 Pooled Variance 26.35641326 Hypothesized Mean Difference 0 Df 22 t Stat 0.810715452 P(T<=t) one-tail 0.21310642 t Critical one-tail 1.717144187 P(T<=t) two-tail 0.426212839 t Critical two-tail 2.073875294 t-Test: Two-Sample Assuming Unequal Variances VTM-8 AASHTO T-193 Mean 8.85 7.150833333 Variance 33.0616 19.65122652 Observations 12 12 Hypothesized Mean Difference 0 Df 21 t Stat 0.810715452 P(T<=t) one-tail 0.213310992 t Critical one-tail 1.720743512 P(T<=t) two-tail 0.426621983 t Critical two-tail 2.079614205 Page 17 of 18 Comparison of VTM-8 vs AASHTO T-193 0 5 10 15 20 25 0 5 10 15 20 25 Soaked CBR (VTM-8) Line of Equality Figure 6: Comparison of CBR values measured according to AASHTO and VTM methods.