Surface Tech Technical Blog — September 2026
For years, the asphalt industry has had a quiet problem. We know fiber reinforcement improves cracking resistance — the field performance is there, the mechanism is well understood, and engineers who've watched a reinforced overlay outlast the section next to it don't need convincing. But when a DOT lab runs a standard cracking test on a fiber mix and the number barely moves, the conversation stalls. The material didn't fail. The test did.
A new peer-reviewed study from the University of Canterbury, published in Transportation Research Procedia and presented at the 12th International Conference on Bearing Capacity of Roads, Railways and Airfields (BCRRA 26), puts hard data behind that suspicion — and offers a better way forward.
Professor Mofreh Saleh and his team built dense-graded AC14 mixes with two binder grades (PG64-16 and PG58-22), at 5.1% binder content, compacted to air-void contents ranging from roughly 2% to 9%. Half the specimens were control; half were reinforced with aramid ACE fibers at 66 and 132 g/tonne. Then they ran every mix through three cracking protocols:
The question wasn't just "does fiber help?" It was "can our tests even tell?"
Here's the finding that should stop every spec writer in their tracks: in this study, the Ideal-CT index went up as air voids increased — for both binder grades. That's backwards. Every pavement engineer knows more air voids means worse cracking performance. A test that rewards a poorly compacted mix isn't measuring cracking resistance; it's measuring something else and calling it cracking resistance.
The author's language is direct: this is "an unexpected and counterintuitive trend, which represents a major limitation of this test method."
The consequence for fiber evaluation is predictable. On PG64-16, the fiber specimens averaged a CT index of 191 against 92 for control — more than double — but the fiber specimens also happened to have higher air voids, so the test's bias was pushing in the same direction as the fiber. On PG58-22, where the fiber specimens were compacted tighter, the numbers went the other way and showed no benefit at all. Same fiber, same dosage, opposite conclusions, and the difference was the test's sensitivity to compaction, not the reinforcement.
The monotonic SCB parameters didn't do better. Fracture toughness at least trended the right way with air voids, but it couldn't distinguish fiber from control, and at one void level it increased with notch depth — another physical impossibility. The author's summary: these indices "demonstrated limitations, including inability to reliably predict the influence of air voids, inconsistency with notch depth effects, and insensitivity to fibre reinforcement."
So the team built something closer to a road.
Instead of pulling a specimen apart once at 0.5 mm/min, the cyclic SCB test applies a haversine load at 10 Hz — traffic-frequency loading — at 50% of the monotonic peak, at 25 °C, and records cumulative deformation cycle by cycle. The result is a three-phase curve that looks like a dynamic creep test: an initial settling phase, a long linear phase, and a runaway phase where the crack takes over.
Two numbers fall out of that curve:
Both are intuitive, both are physically meaningful, and both behaved exactly as they should. CGR rose and CN fell as air voids increased, for every mix tested. The test saw compaction. Now the question was whether it would see fiber.
It did.
In a dense, well-compacted mix at roughly 2% air voids, the aramid fiber specimens reached a Cracking Number of about 10,100 cycles. Control reached about 1,400. That's roughly a 7× improvement in cycles before crack onset, in the same mix, under the same loading.
Across the full 2–9% air-void range, the fiber mix's Crack Growth Rate tracked at roughly half of control. At 9% voids — a badly compacted mix by any standard — control specimens hit crack growth rates of 2.0–2.3 × 10⁻³ mm/cycle while the fiber trend sat near 1.1 × 10⁻³. Even in the worst case, the fiber was slowing the crack.
The author's conclusion is worth quoting in full: "although higher air voids consistently accelerate crack growth in both mixtures, fibre reinforcement effectively slows crack propagation, lowers the CGR, and enhances cracking resistance, thereby extending pavement service life."
That last phrase — extending pavement service life — is the whole point. That's what aramid fiber is for, and this is a university test finally reporting it in a number a specification could use.
Three things, and they matter in different rooms.
In the lab: If your only cracking test is Ideal-CT, you are not equipped to evaluate fiber reinforcement. This isn't a Surface Tech opinion; it's now a peer-reviewed finding. A flat Ideal-CT result on a fiber mix is not evidence the fiber isn't working. It's evidence the test can't see it. Agencies that have run Ideal-CT on ACE XP mixes and shrugged should look again with a test that loads the specimen the way traffic does.
In the spec: Balanced Mix Design is built on the promise that performance tests discriminate between good and bad mixes. A cracking test that can be gamed by under-compaction undermines that promise for every material, not just fiber. The cyclic SCB protocol — simple specimen, standard equipment, 10 Hz loading, two clear outputs — is a serious candidate for the cracking side of BMD, and it deserves a look from any agency writing performance specs.
In the field: Nothing here changes what contractors already see. Fiber-reinforced overlays hold together longer. What changes is that the gap between field experience and laboratory numbers just got a published explanation, and a published fix.
We'd rather you hear these from us. The cyclic-test numbers above are read from the published figures, not tables, so treat the 7× and 50% as strong approximations rather than certified values. The Cracking Number advantage is largest in well-compacted mixes and narrows as voids climb toward 9%, where both mixes are failing fast. And the study didn't break out cyclic results by binder grade, so we can't claim the cyclic numbers for PG58-22 specifically — only that Ideal-CT couldn't evaluate it fairly.
None of that softens the central finding. It sharpens it: fiber works, compaction matters, and the test you use determines whether you can see either one.
The full paper is open access: Saleh, M. (2026), "Evaluating cracking resistance of fibre-reinforced asphalt mixes using a repeated cyclic SCB test," Transportation Research Procedia 99, 317–325, DOI 10.1016/j.trpro.2026.07.060.
If you're an agency engineer or a lab manager who has been evaluating fiber with Ideal-CT and wants to talk through the cyclic SCB protocol, we'd welcome the conversation. If you're a contractor who has been waiting for the lab data to catch up to what you see on the road — it just took a big step.