Engineer Suggests Controlling Cracks with Jointed Asphalt Pavements
BY Sandy Lender
While pavement engineers understand the weakest portion of a pavement structure is the joint, one gentleman wants to make use of that area. Tom Amon of Elkhorn, Wisconsin, wants to pioneer a movement to design joints in the surface course of asphalt mats for the purpose of directing and controlling future low-temperature cracking.
Amon currently works for Bob Kordus at Asphalt Contractors Inc., Union Grove, but he’s been involved in the asphalt industry for more than 50 years. He started out working with his grandfather at B.R. Amon & Sons Inc., of Elkhorn, and served as the chairman of the Wisconsin Department of Transportation (WisDOT) asphalt spec writing committee for 26 years.
During his tenure with WisDOT, “I began to see that thinking the liquid could be made to prevent long-term low temperature cracking was not going to work as it aged past five or 10 years,” Amon said. “That led me to look at how joints could be placed in the pavement. After a series of events, I discovered a method to do it behind the paver without any sealer…The results could be compared to saw-and-seal, but the spacing is 3 to 4 feet without crack filler. I used this on portions of several DOT roads I paved.”
Amon’s explanation hits a number of concepts we can explore when discussing the theory of placing joints in an asphalt mat for the purposes of:
• controlling future thermal cracking in pavement systems, which can then reduce pavement maintenance costs;
• simplifying liquid asphalt binder selection; and
• reducing requirement of some modifiers, thus higher mixing temperatures.
Saw-and-Seal of the Past
While reducing cracks and costs are laudable goals, Amon has an uphill battle when suggesting we put joints in asphalt pavements. Amon mentioned comparing the jointed flexible pavement results to those of saw-and-seal, which is a process whereby joints are built by sawing into the pavement and then sealing the resulting gap with an adhesive product. The technical report “Sawing and Sealing Joints in Bituminous Pavements to Control Cracking” by David W. Janisch and Curtis M. Turgeon, published March 1996, provides a review of more than 50 saw-and-seal test sections in Minnesota. Those test sections were constructed from 1969 to 1990 and included hot-mix asphalt (HMA) overlays of jointed concrete pavement, HMA overlays of HMA pavements, and newly constructed HMA pavements.
“The results show that in over 76 percent of the test sections, the formation of cracking was controlled by the sawing of joints,” the authors summarized. “The unsuccessful sections were those where a deep saw cut was not made, those where the existing joints were badly deteriorated and those where the underlying joints were poorly re-located. All of these factors can be minimized through proper project selection and good design.”
Depth of the saw cut is a factor in past success. For the slices Amon proposes making behind the screed today, depths of ½ to 1 inch are recommended. He has designed a prototype machine to create the creases rather than running the pizza cutter/wheel across the lane behind the paver. Let’s look at how else this concept has matured over the years.
Bringing Design to the Present
About 26 years ago, WisDOT paved a 10-mile section of Highway 12 near Lake Geneva, including a 1,000-foot test section of surface course with joints. Amon has learned that the road is scheduled for mill-and-fill in 2020, and is, according to Amon, the only surviving example of jointed asphalt’s success.
“In 1993, we put a 3 ½-inch asphalt overlay on the concrete,” Amon said. “Now there’s cracks all over the place outside the 1,000-foot test section, but very little cracking in between the pre-determined asphalt joints. The test section shows that it can control low-temperature cracking, and with only a 3 ½-inch overlay on a concrete that was placed in the 1960s, it shows that it may help in reflective cracking.”
Pinpoint the Pressure
Amon pointed to the research of Cliff Richardson, who wrote in his 1910 book “Modern Asphalt Pavements” that when asphalt contracts under cold temperatures, it does so horizontally. When it expands under heat, it does so vertically. As Richardson observed, “[t]he familiar transverse cracks in northern climates generally begin with low temperatures but do not completely close again when temperatures increase.”
This is reiterated by Janisch and Turgeon in their technical report:
“When a new bituminous pavement is subjected to cold temperatures, it contracts according to its coefficient of thermal contraction. This contraction causes tensile stresses to develop in the pavement. As the temperature decreases, these tensile stresses increase. If the thermal tensile stress exceeds the tensile strength of the bituminous material, a crack will form.”
We turn the methodology over to Dr. Hussain Bahia, professor and director of the Modified Asphalt Research Center (MARC) at the University of Wisconsin, and to Dr. Ray Brown, director emeritus of the National Center for Asphalt Technology (NCAT) at Auburn University. Both shared the mechanics of cold-weather cracking, confirming pavement will shrink in all directions, but has a tendency to contract along the horizontal plane.
“The lower the temperature, the greater the tensile stress within the pavement and the greater the chance for cracking and the greater the stress on the pavement,” Brown said. “The cracking relieves the tensile stress caused by reduced temperature.”
“Come summer, the pavement is trying to expand back,” Bahia said. “The molecules don’t understand direction, so they just start vibrating.” They take the path of least resistance, which is upward.
“Another theory is that when the cracks open in the winter, they get filled with sand and dirt and salts,” Bahia said. “Over the years, there is an accumulation that gets locked in under seals and maintenance.”
Mechanics vs Binder Selection
For the 2001 Transportation Research Board (TRB) annual meeting, Amon submitted an abstract titled “A Method to Construct Thermal Stress Relief Joints in Asphalt Pavement.” Included in Report 01-0266 were conclusions from comparisons of thermal contraction in HMA and PCC pavements on the useful properties of pavement joints. Specifically, he used points from the 1995 report “Temperature and Thermal Contraction Measurements as Related to the Development of Temperature Cracking on the Lamont Test Road” to show how average values for 12 studies proved movements in pavement tend to follow joints.
Bahia seemed to concur, explaining, “We build asphalt continuous. The length builds up friction with the base.” He listed the spacing of joints at 12 to 15 feet apart as offering a lessening of the friction with the subbase and lower layers, “theoretically.”
By providing a release for the friction, joints offer a place from which cracks can form their track. Different spacing between joints has been tried. The technical report from Janisch and Turgeon showed joint spacings of 40, 60 and 100 feet, but ultimately recommended “[s]aw cuts should be spaced 12 m (40 ft) apart. Pavements on granular soils may benefit from having the joints spaced at 9 m (30 ft) apart. Further research is needed in this area to determine what joint spacing will work best for a given pavement.”
Amon’s 1,000-foot test section in Wisconsin has joints spaced at 3 to 4 feet apart.
“You have to put the cuts close enough together to relieve the stress for it to prevent cracks from occurring in between the joints,” Brown said.
Janisch and Turgeon’s 1996 technical report stated:
“The frequency of these thermal cracks primarily depends on the asphalt stiffness…the frequency of thermal cracking increases each year until some optimum spacing is achieved. This spacing is the distance at which tensile stresses that develop between successive cracks is below the tensile strength of the bituminous.
“A saw cut in a bituminous pavement produces a weakened plane, due to the pavement’s reduced cross section. When thermal stresses develop, the pavement will crack at the sawed joint because this weakened plane cannot withstand the same stresses as the unsawed portion of the pavement without cracking.”
The orientation of the transverse joint may influence the thermal cracking as well. Amon said he cut joints at both right angles to the centerline and at 30-degree angles to the centerline. He states that both have performed similarly—in other words, both have fewer cracks than the control section of pavement without transverse joints. “We designed the pavement to crack where it was directed,” he said.
Amon believes directing and controlling potential cracks may be more reliable than designing mixes that are crack-proof. In Report 01-0266, he stated: “By providing a mechanical method to control transverse cracking, the selection process of the asphalt binder can be changed. The criteria would be simplified to allow other performance properties to be chosen when disregarding the thermal cracking characteristics. The effects of climate on oxidation or aging of the asphalt binder may not be so critical to the pavement performance. By not relying on the chemical properties of the binder to resist thermal cracking, higher viscosity asphalt binders could improve performance.”
He also sees a reduction in maintenance work for crack filling if the pavement is designed with optimum cracking spacing. One cannot deny that reducing cracks is a good goal. “We need to work more on not letting the crack even start,” Bahia said.
Control the Pressure
“If the asphalt wants to crack, let’s control where it cracks,” Amon said. “Let’s control those pressures. The mixes don’t have to change.”
For engineers considering the danger of cutting into pavement joints that will allow water to infiltrate the system, Amon has an answer. He refers to the cutting as a slice-and-seal without the sealer.
“It’s similar to a saw-and-seal, but the hot mix seals itself,” Amon explained. “We have a pavement that heals itself because of the liquid asphalt in the mixture. The pavement structure remains undisturbed.”
Amon’s Report 01-0266 explained, “[t]he depth of the cut impressions placed in the hot mat is generally ½-inch to 1-inch deep. The hot roller compresses the joint closed with little evidence of the installation left behind. After several months, the evidence of a joint becomes difficult to see by a person walking along the pavement.”
Bahia doesn’t consider the slice action as benign as that, and offers an aspect of the process to be studied and perfected.
“They [the cuts] are joints,” Bahia said. “They are certainly joints. These joints have not been successful in even concrete pavements because they are weakness points and offer very little load transfer, and could allow water to go inside the pavement. Many engineers look at it as a plan for disaster. We can’t even build a good long-lasting longitudinal joint, so it’s very unlikely that we can verify that the joint Mr. Amon is building is actually closing.”
Amon did, however, have a core of asphalt with a joint in it sent to the FHWA lab at Turner Fairbanks in 2000. “There was a detailed scan on various layers in the mix and found no visible disturbance of the aggregate or binder matrix,” Amon explained. “This helps to explain that the load is carried across the joint and it could remain waterproof if it remained closed.”
The technical report from Janisch and Turgeon brings up the question of whether or not the healing action is desirable. In the section “Timing of Saw Cut,” the authors describe problems with adhesion failure on two saw-and-seal projects that “were sawed less than 24 hours after the mix was placed.”
Considering the key to closing the transverse joint behind the screed is to do so immediately, timing is critical to the success of the slice-and-seal concept of jointed asphalt pavement. Janisch and Turgeon found: “It is believed that the saw blade did not ‘cut’ through the aggregate particles in the fresh mix but ‘plowed’ through them instead.”
Again, plowing—or moving—the aggregate particles appears to be what is desired in the concept Amon has proposed. Janisch and Turgeon found the opposite to be preferred. “This resulted in spalling along the edge of the joint and led to adhesion failure. The special provisions were changed to require 72 hours between the time the HMA is placed and when it can be sawed.”
Keep in mind, the timing discussed there is for joints that are being sealed with an adhesive product and not merely rolling closed to re-seal and heal the joint.
At this time, of course, agencies and contractors work together to decide the best way to build longitudinal joints for various projects. The transverse joint may require more engineering or a better public relations campaign. Either way, adding a cutter behind the screed, but before the breakdown roller, will ask for careful planning and timing on the part of the contractor who wants to try out joints in flexible pavements. In the meantime, Hwy 12 near Lake Geneva in Wisconsin has a 26-year-old test section to be studied before the 2020 paving season sees it milled away.