Editor’s Note: One of the best places to test and prove the benefits of warm-mix asphalt (WMA) should be a few miles south of the Arctic Circle. Although WMA has proved its merits time and again for most pavement engineers in the United States, the researchers at NCC Roads AB and NCC Roads Sweden North have been working with an ongoing, 15-mile new construction project for the past two years and have results to share. Nils Ulmgren, the development manager for contractor NCC Roads AB presented findings from the project at the 2nd International WMA Conference in St. Louis in October 2011. Here are some lab practices all engineers can examine from his company’s experiences.
Prior to the beginning of the project that we’ll examine in this article, officials in Sweden experimented for about a decade with minor low-temperature asphalt (WMA) trials and different techniques. In 2010, they chose to pave a new interstate highway that would have a speed limit of 110 kilometres (km) per hour (68 mph) and a traffic load of 7,500 YDT—23 percent of that is heavy traffic of more than 3.5 tons.
The completed project will have two lanes in each direction with a total length of 24.5 km, or 15.2 miles (mi), in the northern part of Sweden at Hudiksvall. That is about 350 km (217 mi) north of Stockholm, which is similar to 100 km (62 mi) north of Anchorage, Alaska.
The contractor for the ongoing asphalt works is NCC Roads AB. The team is using a WMA based on foam-technique with only amine added as an adhesion agent.
Before production, our team performed a thorough lab study to establish the optimal composition of the asphalt. This included all the typical testing such as determination of parameters such as resistance to deformation, stiffness, water sensitivity, workability and, very important in this area, resistance to abrasion against studded tires.
On site, the team performed a test trial of 1,000 meters (m) or 1,094 yards of the WMA alongside a reference section of 3,000 m (3,281 yards). We performed the same lab tests to these test- and reference-sections as we did in the pre-study. The results from the analyses from the pre-trials show that the quality was as good for the WMA as it was for the HMA.
We could proceed.
During 2010 and 2011, we produced about 120,000 tonnes of WMA in a refurbished Astec Double Barrel asphalt plant and paved 19 km in three layers of base, binder and wearing course evenly distributed between the two years. We use the same regimen of testing during paving as we used in pre-production. The rest of this article will present experiences from production, but will focus on the quality aspects as shown from our testing of samples from mix design and paving.
To date, the test results show that the WMA is as good as the HMA, but we should pay special attention to the moisture content in the WMA mix as it is somewhat higher than that of the HMA (0.15 percent compared to 0.05 percent). It has an influence on water sensitivity. That makes it vital for us to add an adhesion agent. The lab analyses results of pavement samples included herein are as of mid-August 2011.
We kept normal hot-mix production above 150 degrees Celcius (302 degrees Fahrenheit ) for two reasons. First, we found it necessary in order to have a viscosity of the binder that made it soft enough to cover the aggregate easily. Second, that’s what we had to do to dry the aggregate so no moisture was left. We found that if the aggregate was heated to no more than 120 to 130 degrees Celcius (248 to 266 degrees Fahrenheit) there was no real guarantee that all moisture had been removed.
Here’s how we handled adding the RAP.
We dried the aggregate at a temperature more than 150 degrees Celcius (302 degrees Fahrenheit) and then added the unheated RAP so that the mixed components ended up with a temperature of 120 to 130 degrees Celcius (248 to 266 degrees Fahrenheit). Adding RAP made it easier to produce WMA when the aggregate had been fully dried. The RAP itself contains some moisture so it’s important to keep this as low as possible—less than 2 percent. Even then, we found some moisture left—about 0.1 percent. We handled that by adding a little more of the anti-stripping agent.
The number of cores we took to test water sensitivity ITSR were few—a total of four over all HMA courses and 12 over all WMA courses—showing only the WMA binder course failing to meet the owner requirements. Again when testing the change in softening point from the tank to the pavement, we had few samples—three total for the HMA courses and 13 for the WMA courses. What we learned from the change in softening point numbers is that the RAP content counts for an increase of about 6.6 degrees Celcius (11.9 degress Fahrenheit) in the AC32 base, which has 30 percent RAP, and for about 2.8 degrees Celcius (3.6 degrees Fahrenheit) in the AC16 binder, which has 20 percent RAP. This means that the increase due to the production is limited to about 2 to 3 degrees Celcius (3.6 to 5.4 degrees Fahrenheit), and it is the same for both WMA and HMA. The difference is normally some degrees higher in batch plants.
The one main risk parameter for WMA seems to be the water sensitivity. Be cautious of higher moisture and higher void content. We have found that RAP is important for the process, but we stress that you must keep it dry.
The pavement structure being built starts with an unbound subbase of 600 mm (23.6 inches) and an unbound base of 80 mm (3.15 inches). For the slow lane, the base course is 76,000 tons of 86 mm (3.39 inches) of AC 32 with 30 percent RAP. The binder course is 29,000 tons of 50 mm (1.97 inches) of AC 16 with 20 percent RAP. The wearing course is 36,000 tons of 35 mm (1.38 inches) of SMA 11 with 5 percent RAP.