The new concrete pavement for an airport facility in Memphis, TN had an abnormal and alarming amount of cracking. Also, there were significant difficulties during construction in preparing the subgrade. As a result of these construction and performance problems, MEA, a UES Company, was requested to conduct a forensic geotechnical investigation.
Loess (a wind-blown soil deposit) was the immediate subgrade support for the about 16 acres of new pavement. Upon visual examination, this loess subgrade appeared to be moist and a typical loess soil with significant amount of silt. In the lab, however, when the above loessal soils were tested across the typically recommended and specified compaction ranges, this soil was found to have extremely detrimental characteristics for pavement support and construction. In addition to having low bearing strength at both modified and standard Proctor compaction limits, there was also significant swell. Also, the soaked bearing strengths for this soil were always lower than the recommended range for pavement design.
Without testing at the compaction limits of +95%, the results can be deceiving and produce exceedingly high soaked CBR values with no evidence of the soil’s swell potential. This was especially the case with regard to this loess. From soaked bearing strength testing it was found that this soil had an egregious sensitivity to the compacted moisture-density where the peak strength was up to 10.9-fold greater than at the 95% standard compaction limits (9.4 ave). For example, see Figure 1 which depicts the drop in strength when the soils became soaked.
The very bad engineering properties of these subgrade soils can be attributed to the significant amount of smectite present in the loess. From X-ray diffraction tests the <1 μm fraction of the loess was found to have clay consisting of an average 40% smectite. Clay minerals are known for their dramatic expansion with water absorption.
The use of a subgrade soil with this amount of smectite resulted in a myriad of construction difficulties. Shortly after the pavement was placed, significant damage was evident. The pavement cracking became significantly worse in less than a year’s time. Differential surface elevations measured across the concrete pavement demonstrates the effects of the swelling/shrinking and softening process of the pavement subgrade which is accompanied by soil moisture change (see Figure 2). Also, the soft to hard subgrade stiffness determined from falling weight deflectometer tests in correspondingly wet and slightly moist areas validated this soil behavior.
Classification of the swell potential of soils based merely on plasticity index values could be misleading. In fact, pavement construction soils classified by plasticity with even a low to medium swell potential could have a significant impact on the final product. The effect of swell on damage is a function of structure type and the construction activities. Soil swell also causes changes in grades during construction—and if unsuspected—results in thinner pavement courses than specified, not to mention the future performance problems of the pavement.