A feel for bumps

May 4, 2015

Research attempts to determine cause of problem

Crack sealing has been used as a very cost-effective method of pavement preservation for decades.

Crack sealants reduce moisture and debris infiltration into the pavement structure, thereby, improving pavement performance.  During the life of most asphalt concrete pavements, overlays are placed to rehabilitate and further extend pavement life. During breakdown rolling of the overlay, transverse bumps and cracks have been observed in the new overlay in the vicinity of where the crack sealant was placed in the underlying pavement. It is believed that multiple reasons may cause this phenomenon, including mixture design, climatic conditions, paving and compaction equipment, timing of the overlay with respect to sealant placement, sealant type and pavement grade. Heat from the overlay is transferred down into the substrate pavement and crack sealant. The heated substrate pavement expands and transverse cracks shrink, exuding the crack sealant upward toward the overlay. The adhesive nature of the crack sealant produces a resistant force to the forward movement of the new asphalt concrete in front of the breakdown roller. When this forward movement is restrained due to the adhesive sealant underneath, the roller passes over the excess material and creates a bump. The bump is then located slightly in front of the transverse crack containing the sealant. With more focus on ride quality and pavement smoothness, paving contractors, asphalt concrete providers, paver and roller manufacturers, engineering firms and agencies have all investigated ways to prevent bumps.

This article summarizes a recent study done to evaluate several factors thought to contribute to transverse-bump formation.

The bow wave

Although bumps and transverse cracks have appeared in new asphalt overlays on top of crack sealant for some time, little objective research has been done to determine the cause and prevention. Suggestions at solutions by interested parties in the asphalt industry based on observations and anecdotal evidence include overlay mixtures with high frictional properties such as open-graded mixtures, stone mastic asphalt or dense-graded mixtures with highly angular and fractured aggregate, which tend to experience less shoving than mixes containing low-angularity aggregate. The use of compaction equipment with nondriven front rollers tends to push the mixture, creating a larger “bow wave” resulting in transverse bumps. Use of stiffer tack coats has resulted in less overlay shoving and less bump formation. Hard, stiff sealants may not adhere to the overlay, while soft, low-melt-temperature sealants may soften enough when heated by the overlay to not restrain the mix if it displaces during compaction. However, medium-stiffness sealants with elastic properties may have a tendency to soften, adhere and restrain the overlay bow wave.  

A recent study indicated that the speed of the vibrating steel roller during breakdown influenced bump formation as well as the number of roller passes. A study conducted for the Colorado DOT found that bumps accompanied by transverse cracking occurred after the crack sealants had been in service for two years in one test pavement. The number of passes of the vibrating steel rollers further exacerbated the presence of the bumps and cracks. The same rollers used in static mode reduced the effect, and pneumatic rollers used for breakdown eliminated the effect. The ambient temperature and temperature of the substrate pavement during construction was reported to have little effect. Transverse bumps over crack sealant on a flat gradient pavement have been reported. However, a relatively large bow wave also was reported during breakdown rolling during this earlier research. A diagram of what is meant by a bow wave is shown in Figure 1.  

The size of this bow wave may be a key to understanding the cause or causes of the transverse bumps.

Golden opportunity

The following five factors were evaluated on two pavements in Golden, Colo., in the summer of 2011. Although only one crack sealant (ASTM D 6690 Type II) was used in this first experiment, the results may offer some clues regarding the cause or causes of transverse bumps, which could deal with the following:

Sealant application (recessed,
flush, overbanded, overbanded
with release)

Breakdown roller (vibrating steel,
static steel, pneumatic)

Breakdown roller speed (200 fpm,
300 fpm)

Overlay type (hot mix, warm mix)

Pavement grade (0-1%, 3-4%)

Asphalt concrete used in the overlay was both hot-mix asphalt (HMA) and warm-mix asphalt (WMA). The same mixture was used for both processes, that is, a foaming process was utilized with the HMA mixture to allow lower mixing and compaction temperatures with the same mixture. Properties of this mixture are shown in Table 1.

Dependent variable

The dependent variable in this experiment is the appearance of transverse bumps and cracks on top of the sealants in the substrate pavement. Bumps and cracks were evaluated quantitatively depending on when the bump or cracks appeared after breakdown rolling as shown in Table 2. These bumps were visually identified by the author and verified by the paving crew.

Two pavements in Golden, Colo., were selected for evaluation in this experiment. Location 1 on Yank Street was selected because of the 0 to 1% grade. Location 2 on 55th Place was selected for the 3 to 4% grade. Both pavements had transverse cracks of approximately the same severity of 1⁄4-in. wide traversing the entire pavement width. 

Each crack to be sealed was identified prior to installation and numbered on the edge of the pavement. Installation was done by the Jefferson County Colorado Road and Bridge Division at both pavement locations on March 17, 2011. The sealant was installed in accordance with recommendations supplied by Deery American Corp. for the crack sealant.  

Crack preparation method included blowing out the cracks using 100-psi compressed air. Sealant was applied to the cracks by hot pouring using a pressure wand and either sealing to level with the surrounding pavement or sealing to slightly over full and then spreading the excess off the surface with a V-shaped squeegee creating the “over-band” application. Two-ply toilet paper was used as a release agent on top of specific overbanded cracks prior to overlay construction on Aug. 31, 2011.

The weather conditions and pavement temperature during installation were clear and dry with no moisture present in the cracks. Pavement temperatures ranged from 94º to 102ºF during construction at both sites. A 94% relative to maximum theoretical density was achieved on all sections. It was achieved with fewer passes of the vibratory breakdown roller, but still achieved with static breakdown.

The HMA and WMA was produced by Asphalt Paving Co. of Golden, Colo. The materials were delivered to the jobsites in covered tandem 12-ton dump trucks operated by Jefferson County. All paving was accomplished by Jefferson County using a Caterpillar AP1055D paving machine, a Caterpillar CB534D vibratory steel wheel roller with drum amplitude set at the No. 1 position and a Caterpillar PS150C pneumatic-tire roller adjusted to 75-psi tire pressure. Temperatures of the HMA and WMA ranged from 255ºF to 280ºF and from 235ºF to 255ºF, respectively, for the Yank Street and 55th Place pavements. Paving operations occurred in the downhill direction for 55th Place.

Transverse bumps appeared after breakdown rolling. Sometimes these bumps appeared after the first pass; sometimes multiple passes were required to manifest the bumps. Bumps tended to be more severe, i.e., larger, when they were manifested after a single pass of the breakdown roller. Therefore, the number of passes of the breakdown roller required to cause the bumps was observed and noted during construction. The results of these individual observations for each crack are shown in Figure 2. If no bump occurred regardless of breakdown roller activity, the numeral 0 is shown in the figure. If four passes were required to manifest a bump (the least severe), the numeral 1 appears, if three passes, the numeral 2 appears, and so on. Figure 3 is a graphical representation of the average values of bump generation for the breakdown roller operating at 300 ft per minute on the 3-4% grade pavement.  

Breaking down the issue

There was a significant difference whether bumps were created during breakdown rolling between the two sites. No bumps were generated at the 0-1% grade site on Yank Street. This was true regardless of crack-seal preparation method, asphalt mixture type or the type or speed of breakdown roller used. However, at the 3-4% site on 55th Place bumps and transverse cracks were created. These bumps and transverse cracks were dependent on roller type, mixture type and crack-seal preparation method. The most significant reduction in bump appearance occurred when the static steel wheel roller was used for breakdown rolling over the recessed and flush-sealed crack sealants. However, only very minor bumps and transverse cracking occurred with static rolling over the overbanded crack sealants. Vibratory breakdown rolling produced the most significant bumps and cracks over the overbanded and overbanded with release agent crack sealant for the HMA overlay. However, bumps and cracks also appeared over the recessed and flush-sealed cracks after two or three passes of the roller. Bumps and cracks also occurred in the WMA overlay over all four types of crack preparation, but generally required one additional pass of the breakdown roller to occur.

Observations on Yank Street (0 to 1% grade) indicate the size of the bow wave in front of the breakdown roller was very small or nonexistent, but on 55th Place (3 to 4% grade) the bow wave was larger. This could mean the bow wave or pushing of the asphalt mixture is directly related to the propensity of the mixture to form a bump over crack sealant. The relatively stiff asphalt mixture used in this research, as indicated by the properties shown in Table 2, may provide evidence for the lack of bumps on Yank Street where a small bow wave was observed and the occurrence of bumps on 55th Place, where a larger bow wave was generated due to the downhill paving operation. 

Pavement grade had a significant effect on the appearance of transverse bumps and cracks appearing in both HMA and WMA overlays placed over crack sealants. Regardless of the crack-sealant preparation method, breakdown roller type or speed and overlay mixture type, no bumps were created on the test pavement with 0 to 1% grade. However, when the same asphalt mixtures were placed on a pavement with 3 to 4% grade bumps could be generated at will when the steel breakdown roller was used in vibrating mode at 300 ft per minute. WMA was slightly less susceptible to bump generation than HMA with both static and vibrating rollers. Recessed and flush-sealed crack preparation was slightly effective in reducing bumps when vibratory breakdown was used.

The bow wave generated in front of the breakdown roller appears to contribute to bump formation in overlays placed over transverse crack sealant. This bow wave was related to the pavement grade in this study but could be caused by asphalt-mixture properties. Therefore, further study of bump creation based on bow wave size is recommended. Numerous factors could be studied which contribute to the size of the bow wave.AT

Sponsored Recommendations