Heterodoxy
History is often unkind to conventional wisdom.
For centuries, educated people, people who should have known better, believed the sun revolved around the Earth. Surgeons once scoffed at the idea of washing their hands before operating. Medical authorities prescribed cigarettes to soothe irritated throats. And as recently as the 1980s, the medical establishment was so certain that stomach ulcers were caused by stress and diet that Barry Marshall had to drink a beaker of H. pylori bacteria and give himself an ulcer to get anyone to take his research seriously. He later won the Nobel Prize.
In every case, the radical idea eventually won.
So, I read with great interest the growing chorus of voices in our industry acknowledging that lab-compacted air voids, the cornerstone of asphalt mix design and quality assurance that we have relied upon for about five decades, is not actually a reliable predictor of pavement performance. We should look elsewhere, they say: new tests, new parameters, a new paradigm. Blasphemy. Or is it?
The simplicity of air void specifications is appealing. Mixtures with air voids that are too high are likely to be under asphalted, leading to durability problems. When air voids are too low, you can get flushing, shoving, and rutting. That logic is baked into virtually every QA program in every state DOT in the country.
But when we look for relationships between air voids and other volumetric properties and actual field performance, more and more data show that they are not good indicators of performance. Numerous examples from the NCAT Test Track and other field studies have shown that mixtures with lab-compacted air voids below 2% have stood up to extreme heavy traffic; and other cases where two mixes with identical in-place air voids perform dramatically differently in the field.
One of the most compelling examples comes from a study NCAT completed in 2024 for a state DOT that asked us to review its policies and procedures for handling asphalt acceptance sublots that failed to meet the agency’s air void specification. We examined field performance data from 16 projects that were at least eight years old and had one or more sublots with failing lab-compacted air voids, ranging from 1.67% to 2.23%. We then compared the field performance of those failing sections to the sections on the same projects that had passed the air void specification. The result? No practical difference in rutting. Not a single section with failing air voids had accumulated more than 0.2 inches of rutting after eight years of service. The pavements did not know it had failed the test.
It is worth remembering the world in which volumetric criteria were developed more than 50 years ago. Asphalt binders were unmodified. Fine aggregates were predominantly natural sands. RAP was rarely used in surface mixtures. The criteria developed in that era were suitable for the materials of that time. But today's mixtures are different and often more complex. Applying criteria conceived for one generation of materials as the primary performance screen for a very different generation is, at minimum, worth questioning.
Building specifications around practical performance-related tests is no longer just a concept. A handful of states have already successfully implemented them; others are doing pilot projects and BMD validation projects. The data are accumulating. Although the new tests may not be perfect, confidence is growing in our ability to eliminate mixtures susceptible to rutting and cracking.
If air voids have any remaining value in modern QA, it is as an indicator of production consistency, a role better served by asphalt content, gradation, and maximum specific gravity. I am ready to say, without apology, that it is time to move beyond air voids as the primary indicator of mix quality.
Healthy skepticism of conventional wisdom, even our own conventional wisdom, is not blasphemy. It is how science works. It is how our pavements get better.
Randy C. West, Ph.D., P.E. | Director and Peter and Lisa Wilson Distinguished Professor