Curl and Warp Analysis of The LTPP SPS-2 Site in Arizona

SUMMARY

This study examined the roughness and roughness progression of 21 test sections on the LTPP SPS-2 site in Arizona. The site included 12 test sections from the standard experiment and 9 supplemental test sections selected by ADOT. The standard test sections included a matrix of variations in surface layer thickness, surface layer strength, and base type. The supplemental sections included four test sections that were similar to their counterparts from the standard experiment but without dowel bars. The supplemental test sections also included three other jointed plain concrete pavement designs of interest to ADOT and two AC test sections.

Traditional profile analyses revealed roughness caused by transverse and longitudinal cracking on some test sections and some localized roughness caused by built-in defects. However, spectral analysis and filtered profile plots showed that curl and warp contributed to, and in some cases dominated, the roughness on many of the test sections. In addition, the progression of roughness throughout the experiment often followed a disorderly trend because of diurnal and seasonal changes in slab curl and warp. This study applied objective profile analyses to quantify the level of curl and warp on each section. These automated algorithms estimated the gross strain gradient needed to deform each slab into the shape present in the measured profile and produced a PSG value. The levels of curl and warp within each profile are summarized by the average PSG value.

For the 19 jointed concrete test sections, variations in average PSG over time and slab-by-slab PSG along the pavement explained many of the changes in IRI over time. This included diurnal variations in slab curl, which often caused the overall progression in IRI to appear disorderly throughout the experiment. PSG analysis also revealed that the level of curl and warp increased overall throughout the life of the experiment, with commensurate increases in the IRI.

LOW-STRENGTH SECTIONS, STANDARD EXPERIMENT

Table 6 summarizes the observations for the low-strength sections in the base experiment. The table lists key structural factors of each section and provides the net change in IRI values from the initial visit (0.32 years) to the final visit (16.32 years) for the left and right side. The table also lists the net change in PSG values from the initial to the final visit and the prevailing direction of curl in the final visit. For these test sections, a negative value indicates an increase in upward curl and a positive value indicates a decrease in upward curl.

Table 6 shows that a strong relationship exists between the change in PSG and the change in IRI, even though the specific relationship depends on the pavement structural properties (see table 4). For example, the highest change in PSG of the 11-inch-thick sections occurred on section 0215, which showed the highest change in IRI. Further, section 0217, where the level of curl decreased, showed a net decrease in IRI.

On the six low-strength test sections in the standard experiment, the correlation between IRI and PSG was very high. This permitted statistical analysis that isolated the portion of the overall roughness caused by curl and warp from the remainder of the roughness. With the influence of curl and warp removed, the balance of roughness often followed a more orderly trend. This allowed investigation of sources of roughness, such as distress, that were otherwise obscured by the influence of curl and warp. It also identified instances in which the roughness progressed due solely to changes in curl and warp, where the test section had not deteriorated despite the increase in overall IRI.

Table 6 also lists the change in IRI with the influence of curl removed. In this case, the balance of the roughness held steady throughout the monitoring period on sections 0215, 0217, and 0221. This was also the case on sections 0219 and 0223, with the exception of two left IRI values from years 9 and 10 that stood out as higher than the rest. Profile analysis showed that the additional roughness was isolated to the 8–11-ft wavelength range. Otherwise, the curl-removed IRI on these sections held steady at a relatively low value or increased slightly from an initially low value.

With the influence of curl and warp removed, the balance of the roughness on section 0213 held steady throughout the experiment on the left side but increased after 11 years on the right side. Patches of narrow dips caused by longitudinal cracking appeared in the left-side profiles in the second half of the monitoring period, and they caused localized roughness starting in year 12 of the experiment.

Other observations from the low-strength sections include the following:

• Section 0213: Section 0213 included a long bump followed by a long dip in all profiles. This feature of the profile included a 0.9-inch transition from the peak of the bump to the lowest part of the dip over 40 ft of pavement.
  
  
• Section 0217: Section 0217 included transverse and longitudinal cracking starting in year 3 of the experiment. The transverse cracks caused narrow dips in the profiles at some mid-slab positions that increased in severity throughout the rest of the monitoring period.
  
  
• Section 0221: One of the slabs within section 0221 experienced severe aggregate loss early in the experiment. The distress surveys listed scaling along the right side of the slab. This area and a few others were sealed with fiberglass by year 14. Nevertheless, the slab with the most severe aggregate loss caused localized roughness in the right-side profiles by the end of the monitoring period.

HIGH-STRENGTH SECTIONS, STANDARD EXPERIMENT

Table 7 lists summary results for the high-strength sections. As a group, the high strength sections exhibited smaller increases in roughness. As in the low-strength group, the sections with lean concrete base (LCB) either increased in roughness less than counterparts with other base types or decreased in roughness.

Table 7 lists the net change in PSG for each section. On sections 0218, 0220, and 0222, a positive number indicates a reduction in upward curl. The IRI either reduced or held steady on these sections. On sections 0214 and 0216, a positive value indicates a change from upward curl at the start of the experiment to downward curl at the end. On section 0214, the transition toward downward curl was more aggressive on the first half of the section, and a group of slabs in the second half of the section maintained a reduced level of upward curl. In this case, the effect on IRI depended on the initial level of downward curl. On section 0224, upward curl increased overall.

SUPPLEMENTAL SECTIONS

Table 8 lists summary results for the supplemental sections. Sections 0262–0265 match the designs of sections 0213, 0221, 0223, and 0215, respectively, except that they are undoweled. All of the supplemental sections were curled upward (on average) in the initial visit (0.32 years) and the final visit (16.32 years), but some exhibited downward curl in the second visit (1.42 years). Sections 0262 and 0265 faulted, and the faulting progressed from virtually none to an average value of about 0.15 inches on section 0262 and 0.10 inches on section 0265.

AC SECTIONS

The IRI increased on section 0260 by 64 inches/mi on the left side and 9 inches/mi on the right side. Two areas of localized roughness on the left side of the section account for the increase in IRI. These areas included small potholes, patching, longitudinal cracks, and pools of sealant.

The IRI increased on section 0261 by 17 inches/mi on the left side and 58 inches/mi on the right side. Dips appeared at several transverse cracks starting 10 years into the experiment that exacerbated the roughness of the section in both wheel paths. In addition, a distressed area on the right side developed into a pothole. This caused localized roughness in the final profiling visit.