Durability of chemically stabilized aggregate bases
Khoury, Naji N.
Zaman, Musharraf M.
Miller, Gerald A.
MetadataShow full item record
A comprehensive laboratory study was undertaken to investigate the effect of durability, namely, freeze-thaw (F-T) and wet-dry (W-D) cycles on raw and stabilized aggregate base. Four commonly used aggregates in Oklahoma, namely, Meridian (M), Richard Spur (RS), Sawyer (S), and Hanson (H), were used in this study. Aggregates were stabilized with different stabilizing agents. Resilient modulus and unconfined compressive strength (UCS) were the only measurements used to evaluate the effect of these actions. Additionally, laboratory tests such as Los Angeles abrasion and moisturedensity were conducted to characterize the aggregates. The study was divided into two phases. Phase I consists of evaluating the effect of F-T cycles and W-D cycles on Class C Fly Ash stabilized Meridian aggregate. One F-T cycle consisted of placing a specimen in a rapid F-T cabinet, then freezing it at -25°C (13°F) for 24 hours and thawing at 2 l.6°C (71°F) for another 24 hours with a relative humidity ranging between 90% and 95%. One wet/ dry cycle consisted of placing a sample 1 in an oven at 71°C (l 60°F) for 24 hours, then placing it in a water bath for 24 hours at room temperature. The effect of F-T/W-D on stabilized samples was examined on 3-day and 28-day cured samples stabilized with 10% CFA. Two sets of samples were prepared. The first set, called Same Specimen (SS) set, was subjected to selected sequences of freeze-thaw or wet-dry cycles. The second set, called Different Specimens (DS) set, was subjected to a given sequence of F-T or WD action and tested for Mr, followed by UCS test. The primary goal of this effort was to optimize/reduce the total number of samples needed for the testing program. The number of F-T/W-D cycles for the first set of specimens was 0, 4, 12, 30, and 60, while for the second set they were 0, 4, 12, and 30. Based on the results, it was observed that the resilient modulus of CF A-stabilized samples increased as the number of F-T /W -D cycles increased, up to a certain number, beyond which it started to decrease. Also, it was seen that the same samples could be used to evaluate the effects of F-T/W-D cycles on resilient modulus of CF A-stabilized specimens as long as the number of Mr tests is low. In addition, the deleterious effect of W-D cycles on Mr values was higher than the effect of F-T cycles. And the effect of these actions had more deleterious effects on 28-day cured samples than on 3-day cured samples. The UCS tests were also used to identify the effect ofF-T/W-D cycles. Tests were conducted on two sets of samples. For the first set (called Mr samples), tests were conducted on samples after subjecting them to a desired number of F-T or W-D cycles, followed by Mr testing. For the second set (called virgin samples), tests were directly conducted on samples after being subjected to 0, 4, 12, and 3 0 F-T or W -D cycles. It was seen that the unconfined compressive strength and modulus of elasticity values increased as the number of F-T/W-D cycles increased. It was also observed that samples subjected to resilient modulus tests had higher UCS and modulus of elasticity values than samples tested for only UCS. The effect of F-T and W-D cycles was observed on raw specimens. Specimens were subjected to 4, 12, and 30 F-T cycles. It was observed that Mr values decreased as the number of F-T cycles increased. The maximum reduction in resilient modulus values was approximately 20%. On the other hand, raw samples could not withstand even one cycle of wetting/drying and Mr testing. From the observations in Phase I, the same specimens scheme was used in Phase II, in which, specimens were compacted at OAC, and cured for only 28 days. Phase II consists of evaluating the effect of F-T cycles and W-D cycles on Meridian aggregate stabilized with CKD, FBA, and PC.; Richard Spur and Sawyer aggregates stabilized with CKD, CF A, and FBA; and Hanson aggregate stabilized with CKD, and FBA. In this phase, one F-T cycle consisted of placing a 28-day cured sample in a rapid F-T cabinet, then freezing it at -25°C (l 3°F) for 24 hours and thawing at 2 l .6°C (71°F) for another 24 hours with a relative humidity approximately 98%. During this phase, the membranes around the specimens were removed, so that moisture migration to specimens occurs more readily. One wet/dry cycle consisted of placing a 28-day cured specimen in an oven at 71°C ( l 60°F) for 24 hours, then placing it in a water bath for 24 hours at room temperature. It was observed that Mr values decreased as the number of F-T cycles increased. The percentage decrease varied with the stabilized agents and aggregate type. CKD-stabilized specimens subjected to F-T/W-D cycles had the lowest Mr values, followed by CF A, FBA, and then PC. In addition, the performance of stabilized aggregate base under F-T cycles is a function of aggregate mineralogy. For example, Meridian, a limestone aggregate, had lower Mr values than Sawyer, a sandstone aggregate. A commercially available software, Kenlayer, was used to evaluate the structural response of pavement as the resilient modulus of base changes due to stabilization and F-T/W-D cycles. Tensile strain at the bottom of the asphalt concrete layer and the compressive strain at the top of the subgrade were used to calculate corresponding allowable load repetitions (i.e., equivalent single axle load (ESAL)) and to evaluate these effects. Results showed that ESAL increased due to stabilization. It was also observed that a negative effect of F-T and W-D cycles on Mr produces a negative effect on ESAL and vice versus. In other words, an increase in Mr due to F-T/W-D cycles produces an increase in ESAL, while a decrease in Mr due to these cycles decreases the number of ESAL. In addition, the layer coefficients were determined using the traditional equation recommend by AASHTO, 1986. The effect of stabilization, aggregate mineralogy, F-T and W-D cycles, were observed on ESAL and the layer coefficient. Regression equations in tabular and graphical form are presented for predicting ESAL and layer coefficient of stabilized aggregate base for practical applications in pavement design. Such applications illustrated with design examples. The reference intensity ratio (RIR) method was employed to identify and quantify the mass percent of minerals and cementing compounds in the stabilized specimens. Results show that the cementing compounds such as ettringite, gismondine, straetlingite, and tobermorite, among others, responsible for modulus increase, were formed and their amount varied from one stabilized specimens to another. The Mr values correlate fairly well with the sum of the cementing compounds. Finally, the SEM micro graphs show the same trend as the XRD where the intensity of crystals formation is lower in CKD specimens than CF A, followed by FBA specimens. The cost for constructing a section ( 1.83 m by 30.48 m; 6 ft by 100 ft), in the area of Oklahoma City, having an ESAL value of approximately 2,000,000, was determined. The costs for the materials, hauling, and compaction were provided by the companies. Results showed that the cost for constructing the section with raw aggregate is more expensive than a section stabilized with CKD, CFA, FBA, or PC. This due to the fact the stabilization reduced the thickness of the base layer, and thus, the bulk materials. It was also found that constructing a section with FBA is cheaper than a section with CF A, followed by CKD, and then PC.