Interactions of expansive clay soils with transportation facilities have caused billions of dollars of damage. In the process of determining agents and methodologies to overcome expansive clay behavior, multiple injections of chemical agents into the subgrade to modify clay behavior are most promising. A study was conducted to optimize a potassium-rich solution of four parts. The modifying solution has been injected into expansive clay subgrades with success since the early 1970s. A laboratory methodology was developed to inject specimens multiple times and to conduct swell tests. Screening tests were conducted to evaluate linear effects that each of the four constituents had on swell response through interpretation of a factorial design. Analysis of replicated tests, randomly arranged in blocks, revealed that one constituent had no effect on swell behavior. A central composite experimental design then was used to optimize the remaining constituents. The resulting response surface equation led to an optimal combination for swell abatement, confirmed during short- and long-term verification testing. The optimal blend of constituents was determined to reduce swell from 7.2 percent to 0.84 percent under injection situations simulating the field, and it was found to improve the reduction in swell by 2.6 percent compared with the original blend. In addition, XRD confirmed that modification of clay occurred. This study was unique in its use of laboratory simulation of multiple field injections and in its use of statistical analysis to minimize the number of tests needed to ensure confidence in the optimization sought.


Civil, Architectural and Environmental Engineering

Publication Status

Available Access

International Standard Serial Number (ISSN)


Document Type

Article - Journal

Document Version


File Type





© 2024 SAGE Publications, All rights reserved.

Publication Date

01 Jan 1998