|Institution:||University of Texas – Austin|
|Keywords:||Expansive soils; Pavements; Central Texas geology; Swell potential|
|Full text PDF:||http://hdl.handle.net/2152/31799|
Expansive soils are a significant issue in Central Texas due to a high potential to shrink and swell which leads to cracking of roadways. A significant amount of research has been conducted on expansive soils, which has led to the development of direct and indirect methods to determine a soil’s swelling potential. The methods for direct measurement of the swell potential are typically both time consuming and expensive, which has led to the underutilization these methods. Indirect methods, which use index geotechnical properties to predict the swelling behavior of a soil, are empirically based correlations that are only approximations that don’t take into account variables such as the mineralogical composition of the soil and include the Texas Department of Transportation's (TxDOT) approach, Tex-124-E, which is based solely on Atterberg Limits and grain size distributions to determine the potential vertical rise of an expansive deposit beneath a pavement system. The purpose of this study is to develop an approach that both directly measures an expansive soil’s swelling potential using centrifuge technology (DMS-C) and determines a potential vertical rise (PVR) for use in site characterization. This study consists of eleven soils sampled from ten sites in Bexar, Atascosa, and Guadalupe Counties of the San Antonio TxDOT district to determine the PVR using the DMS-C and Tex-124-E approaches. Soil characterization tests were conducted including Atterberg Limits and compaction tests, as well as, over 300 specimens tested in the centrifuge testing program. The centrifuge testing program consisted of compacting samples into the double infiltration setup at initial conditions of 3% dry of optimum moisture content and 100% relative compaction and testing the samples at three separate artificial g-levels that correlate to three effective stresses to generate a swell-stress curve that was defined over a range of stresses typically found in the active zone. The results from the centrifuge tests for samples from each site are verified with the traditional free swell tests (ASTM D4546.) At each site, the swell-stress curve and stresses for the soil profile were used to determine the PVR for the DMS-C approach. From the results, seven of the sites received a high or severe degree of concern for potential damage to the pavement. Of these seven sites, six of the sites correlated to soils derived from the Navarro/Marlbrook Formation, which is a major geologic formation in both the San Antonio region as well as the rest of Central Texas east of the Balcones Fault zone. The same stresses, as well as, the liquid limit, plastic limit, and moisture content are used to predict the PVR with the traditional Tex-124-E approach. These results were analyzed and compared to the values to the PVR from the direct measurements taken in the DMS-C approach for each site. From the comparisons, the approximate prediction of PVR for Tex-124-E does not correlate to the direct measurements of swelling results to determine the DMS-C approach. … Advisors/Committee Members: Zornberg, Jorge G. (advisor), Cox, Brady R (committee member).