Fatigue behavior and modulus growth of cementitiously stabilized pavement layers
MetadataShow full item record
The objective of National Cooperative Highway Research Program (NCHRP) Project 04-36 was to recommend laboratory procedures to measure performance-related characteristics of pavement layers stabilized with cement, lime, and fly ash and to provide validated distress models to be incorporated into the Mechanistic-Empirical Pavement Design Guide (MEPDG). The objective of this specific study was to identify a procedure for determining the fatigue behavior of cementitiously stabilized materials (CSM) and modulus growth with time. A laboratory based, third-point flexural beam test and a non-destructive ultrasonic pulse velocity test were developed and applied to cementitiously stabilized materials. Ultrasonic pulse velocity measurements were conducted to assess non-destructively the flexural strength and flexural modulus of base and subgrade soil stabilized with cement, fly ash, and lime. The host materials selected for this study are classified as gravel (GM), sand (SP), silt (ML), and clay (CL) based on the Unified Soil Classification System (USCS). The host materials were stabilized with four binders: cement, Class C fly ash, lime-Class F fly ash, and lime. Laboratory tests to monitor the fatigue behavior of CSMs involved flexural strength, flexural modulus, and fatigue cracking tests. These tests were performed using a third-point flexural beam test for all CSMs. Prismatic molds of dimensions 102 mm � 102 mm � 400 mm were used to fabricate the beam specimens. Different curing procedures were applied to different mixtures depending on the binder. Cement-stabilized mixtures (gravel, sand, silt, and clay) were cured in the moist room (100% relative humidity, 23 �C) for 28 d (ASTM D558). Fly ash-stabilized mixtures (sand, silt, and gravel), clay-lime and silt-lime-Class F fly ash were sealed with plastic wrap and cured in an oven set to 40 �C (ASTM C593) for 7 d. Nine different mixtures were studied. The effect of density, binder content, and curing time was also studied. Additionally, modulus growth tests were conducted on these specimens. Resilient modulus tests were performed in general accordance with NCHRP 1-28A on lightly stabilized materials and flexural modulus tests were performed on heavily stabilized soils. Cylindrical specimens were prepared for the resilient modulus tests. An ultrasonic velocity test system ?PUNDIT (Portable Ultrasonic Nondestructive Digital Indicating Tester)-Plus? manufactured by CNSFARNELL was used to measure the propagation speed of a pulse of ultrasonic longitudinal stress waves. The beam and cylindrical specimens were tested by the direct transmission of the pulse of ultrasonic longitudinal stress waves. P-wave velocity (or constrained modulus) of the CSMs was studied using the ultrasonic pulse velocity test. The ultrasound velocity tests were used to assess nondestructively the flexural strength and modulus as well as modulus growth. Results from the flexural strength tests showed clay-cement and silt-fly ash to have the highest and the lowest flexural strength among these nine mixtures, respectively. For the three class C fly ash stabilized soils (sand-fly ash, gravel-fly ash and silt-fly ash), the sand-fly ash specimen had the highest flexural strength, while the silt-fly ash specimen had the lowest flexural strength. The flexural strength was found to decrease when the specimens were under-compacted and to increase when the binder content and curing period increased. The flexural modulus test results indicate that determination of flexural modulus should be done at a stress level of 30% to obtain consistent results. Among the nine mixtures, the highest and the lowest flexural modulus for the 30% stress level (i.e., applied stress divided by flexural strength) was observed for the clay-cement and the silt-fly ash specimens, respectively. The non-plastic silt-fly ash specimens were the weakest mixture. For the fly ash-stabilized soils, the gravel-fly ash had the highest flexural modulus. The flexural modulus was found to increase with binder content for all stress levels. A linear relationship was developed between the flexural strength and flexural modulus. Also, a relationship between the unconfined compressive strength and flexural strength was found. In general, the fatigue life ranged between 1 and 75,537 cycles in this study. The modulus kept degrading as the fatigue test continued, whereas the displacement increased. The fatigue life for different mixtures ranged from 44% to 64% of the initial modulus. The specimens compacted to a reduced dry density showed a lesser fatigue life than the specimens compacted to the target dry density. The sand-cement specimens with higher binder content did not perform well as compared to the sand-cement specimens with lower binder content. The fatigue life of sand-cement (6% binder) was greater than the fatigue life of sand-cement (8% binder) at the same stress level. For the gravel-cement specimens, there was not much improvement in the fatigue life with increase in binder content, but the silt-fly ash specimens performed better with increased binder. A stress-based fatigue model was developed using two parameters ? stress ratio (SR) and fatigue life (N) that resulted in a good fit for estimating the fatigue life. The R2 values for the different specimens ranged between 0.70 and 0.95. The fatigue data from this study fitted very well to the MEPDG fatigue model, and also to most of the concrete fatigue models with varying degrees of success (R2 ranged between 0.50 and 1.00). An attempt to develop a strain-based model using the initial strain (which is calculated from the fatigue test data) was also made. The model was a good fit (R2 = 0.79) for only the sand-fly ash mixture. The strain-based fatigue data was not found to be a suitable fit in this study. Results from the ultrasonic pulse velocity tests showed that with decrease in density of the specimens, constrained modulus and P-wave velocity decreases, whereas, with increase in binder content and curing time of the specimens, the constrained modulus and P-wave velocity increases. A relationship was found between the flexural strength and P-wave velocity (or constrained modulus) as well as between the flexural modulus and P-wave velocity (or constrained modulus). Flexural modulus as measured in beam tests was found to increase with time. Resilient modulus did not follow the same trend of increase probably due to the selected method that proved to be destructive. This might be due to the testing of the same specimen at different curing periods, which may have damaged the specimens. The ultrasonic pulse velocity tests showed a clear trend of increasing stiffness (constrained modulus) with time for all mixtures. This study showed that the developed third-point flexural beam test is appropriate for all combinations of binder and soil mixtures; i.e., lightly and heavily cementitiously stabilized materials. A stress-based fatigue model using two parameters, stress ratio (SR) and fatigue life, represent the fatigue behavior satisfactorily. A strain-based model was not suitable for all mixtures. The ultrasonic pulse velocity test proved to be a convenient non-destructive method for estimation of the flexural strength and flexural modulus. Furthermore, the ultrasonic velocity test allows for a convenient procedure to study modulus growth with time.