The Effect of Base Course Gradation on the Results of Laboratory Pumping Tests

Abstract

This paper reports the results of a laboratory study initiated to investigate the performance of a variety of base course samples with different gradations, when those samples are placed over a standard subgrade soil and subjected to repetitions of load in such a way as to induce the pumping of fine soil to the base course surface and the intrusion of subgrade soil into the interstices between base course particles. To accomplish the purpose of this investigation, eight gravel base course samples varying from extremes of open-graded to dense-graded and seven coarse sand base course samples, similarly graded, were placed, each at a relative density of approximately 90 per cent, over a silty-clay subgrade soil compacted to 90 per cent of maximum modified AASHO dry unit weight, and subjected to 40,000 repetitions of a 25 psi load applied at the base course surface. In each case, the load was applied through a loading head which at all times remained in contact with the base course surface. Measurements of the total deflection of the subgrade- base course system were taken intermittently during each test. At the end of each test, the weight of soil pumped to the base course surface was measured and a grain size analysis was performed on the base course sample to determine the increase in soil finer than the No. 200 mesh sieve above the subgrade. A description of the repetitive load equipment is included and the results of density tests performed on each sample are presented. A good correlation was found to exist between large deflections of a subgrade-base course system and either base course pumping or subgrade intrusion. Test specimens with very open-graded gravel bases were subject to intrusion of subgrade soil and specimens with dense-graded gravel cases with an excess of 3 per cent by weight finer than the No. 200 mesh seive demonstrated pumping of fine soil sizes to the base course surface. There appeared to exist an optimum gradational range which exhibited neither base course pumping nor significant subgrade intrusion. Test specimens with course sand base course samples were found to perform satisfactorily over a wider range of gradation than those with the larger sized gravel bases. Test results were compared with existing filter criteria for thin base courses.