What is a performance-based property? It is an engineering property that can be directly used to predict performance. Historically we have used empirical properties and have correlated them to performance.
For example, crushed faces are usually specified in hot mix and we know that all things being equal, the more crushed faces in a mix, the more rut resistant the mix will be. But we can not predict rut depth directly from crush count. On the other hand, Superpave was designed to have properties which can be used directly to predict rut depth.
The performance based asphalt binder specification, "PG asphalts", is built around performance based properties in the asphalt binder. Three separate properties are used to specify how much contribution to rut resistance the asphalt binder will supply, how much resistance to fatigue cracking the mix will have and how much resistance to low temperature cracking the mixture will have.
On the mix design side there are three levels of mix design. Measuring performance based properties of the mixture is more involved than would be justified in all circumstances. Therefore, the mix design is split into three levels, Level 1, Level 2 and Level 3.
Level 1 mix design does not contain performance based properties. Volumetric properties (air voids, asphalt content, etc.) and aggregate properties (crushed faces, fine aggregate sharpness, etc.) are the basis of Level 1. In some ways, Level 1 mix designs are similar to Marshall mix designs but they give a much better indication of rut resistance. In Marshall mix design we see how much the specimens compact but in Superpave Level 1 we watch how the mix compacts and the way it compacts indicates how strong the aggregate skeleton is inside the mix.
Both Level 2 and Level 3 mix design measure performance based properties. Level 3 is a more detailed analysis than Level 2 but both methods use the performance based properties to predict performance. A Superpave shear tester is used to run the tests which measure the engineering properties. The tests are non-destructive, that is the specimen can be tested again and again.
What's the main limitation? To predict permanent deformation and fatigue cracking requires a shear tester. The shear tester costs $300,000 to buy. To predict low temperature cracking requires an indirect tensile creep tester. It costs $150,000. So there is a price to pay for the ability to predict performance. On the other hand, if a project goes "belly up" hundreds of thousands of dollars, sometimes millions of dollars are wasted. The Superpave system can not make sure the pavement is built correctly, but it can make sure the pavements are designed correctly. If the shear tester can prevent 25% of the failures we see, the benefits will far outweigh the costs.
For more information regarding performance based testing, contact Gerry Huber or Bob McGennis.
Return to Articles about Superpave
Return to Superpave Home Page