Modeling Transverse Squeeze Flow in Thermoplastic AFP: A Realistic Approach to Interfacial Slip and In-Situ Consolidation

Thermoplastic composites are increasingly viewed not just as alternatives to traditional materials, but as a critical enabler of faster, cleaner, and more adaptable manufacturing workflows. Their potential is especially visible in automated fiber placement (AFP), a technique that’s been gaining momentum for its ability to lay down thermoplastic prepreg tapes directly onto curved or contoured surfaces. The key advantage is consolidation on the fly—no autoclave, no long post-processing cycles which is ideal for industries chasing efficiency without compromising structural integrity. But in practice, it’s rarely that simple. AFP is still an evolving process, and with thermoplastics in particular, it comes with a unique set of limitations. Among the most difficult to manage are gaps and overlaps between adjacent tapes—flaws that often go unnoticed until a part fails inspection. While machine misalignment and surface curvature play a role, the issue runs deeper. It turns out, much of the variability stems from how the tape itself responds as it’s pressed into place. The thermoplastic matrix, softened by heat, doesn’t just sit where it’s placed and moves and spreads. And this lateral deformation under compaction is difficult to anticipate using current models because most of those models simplify away too much. They often assume the material flows like a Newtonian fluid—its viscosity constant and predictable—or that the interface between tape and substrate behaves in an ideal way: either perfectly stuck or completely free to slide. But real materials don’t behave like that. Not under the high pressures and tight time scales of AFP. The viscosity of thermoplastics shifts dramatically with strain rate, and the friction at the interface is neither zero nor infinite. It’s evolving, contingent, and messy.

To this account, new research paper published in Composites Part A: Applied Science and Manufacturing and conducted by Mahmoud Fereidouni, and Professor Suong Van Hoa from the Concordia Center for Composites, Department of Mechanical, Industrial and Aerospace Engineering at Concordia University in Canada, designed a new model that accounts for how shear-thinning fluids like molten composites actually behave. More importantly, they introduced a novel way to capture slip at the interface—not as an all-or-nothing condition, but as a continuum shaped by the development of intimate contact.  First the researchers used finite element analysis in COMSOL as a verification tool for the theoretical framework, setting up two-dimensional models that tried to capture the complex, high-speed behavior of molten carbon-fiber-reinforced PEEK as it was squeezed beneath a compaction roller. The authors found that when they assumed a no-slip interface—that is, no relative motion between tape and roller—the internal pressure predictions ballooned to levels no AFP machine could possibly generate. We’re talking about values upwards of 900 MPa under conditions that should have been entirely routine which didn’t add up. On the flip side, when they modeled perfect slip, the pressure all but vanished. The width changes still looked plausible, but the underlying mechanics didn’t track with observed behavior from prior experimental work.

This mismatch pointed to a deeper problem: the conventional assumptions of either complete slip or no slip at all were too simplistic to capture the true complexity of the system’s behavior. So, instead of forcing the data to fit these extremes, the authors developed a middle-ground solution. They introduced an imperfect slip model, where the degree of friction at the interface wasn’t fixed but evolved based on how much intimate contact had formed during consolidation. Rather than using a generic curve, they based this on real surface roughness. Specifically, they imagined surface asperities shaped like triangles gradually collapsing into trapezoids under load. It was a physically intuitive way to link surface geometry with evolving slip conditions. And when they plugged that into the simulation? The agreement was excellent—not just with their own final experimental measurements, but with experimental trends reported in previous work. The model indeed hit the right pressure values and also tracked how the tape’s width changed under different forces and speeds.

In conclusion, for years, engineers working with AFP of thermoplastic composites have had to make do with rough approximations when trying to predict how molten tapes would spread under heat and pressure and what makes the new study of Professor Suong Van Hoa and Mahmoud Fereidouni stand out is they successfully developed a new physics-based squeeze flow model that incorporates the non-Newtonian behavior of thermoplastic composites during automated fiber placement and also introduced a novel imperfect slip framework linked to intimate contact evolution, using a trapezoidal asperity model to capture interfacial friction dynamics with high fidelity. The new model now can enable manufacturers to accurately predict tape deformation and minimize defects like gaps and overlaps during in-situ consolidation, which are critical for structural integrity. It allows for smarter process control and more consistent, high-quality composite production, by reflecting real material behavior and interfacial conditions. The proposed trapezoidal asperity model deserves a special mention—not because it’s novel for novelty’s sake, but because it brings tangibility to something that’s often treated abstractly. Intimate contact is hard to measure and harder to predict, but here it’s broken down into something you can quantify and track. That matters, especially in aerospace industry, where even minor variations in material behavior can ripple into performance or safety concerns.