A hole in a vehicle axle or a groove in an aircraft or machine part might seem insignificant. But when subjected to repeated forces, these small features can become high stress points where damage develops and cracks form.
A Concordia researcher has developed a faster way to predict what happens beneath the surface of metal parts where cracks are most likely to begin, addressing a longstanding challenge in fatigue engineering.
Ayhan Ince, a professor in the Department of Mechanical, Industrial and Aerospace Engineering at Concordia's Gina Cody School of Engineering and Computer Science, developed a new three-dimensional (3D) modelling framework to better understand what happens inside the metal surrounding these vulnerable areas.
In the study's computer-based comparison using a relatively simple notched steel bar, the new method completed its calculations in 4.5 minutes, compared with 28 minutes for a more detailed nonlinear simulation that accounts for permanent deformation, making it approximately six times faster. Since full nonlinear simulations become more demanding as engineering models grow in size and loading complexity, Ince expects the framework to provide even greater computational efficiency for large engineering components.
The advance could help researchers develop more accurate models of how damage accumulates in mechanical parts more quickly. Engineers could eventually use those models to accelerate durability assessment and improve component designs in industries ranging from automotive manufacturing to aerospace, the researchers say.
Seeing beyond the surface
Mechanical components are constantly subjected to forces that can gradually weaken the most highly stressed location at a hole, groove or geometric transition. Yet damage and crack growth depend on how material deformation evolves throughout a surrounding volume of material near those critical locations.
A vehicle axle, for example, may be pulled, compressed, bent and twisted, sometimes simultaneously. Over thousands or millions of repetitions, these forces can cause fatigue damage and eventually cracks.
For decades, researchers have developed efficient ways to calculate how materials respond at vulnerable points, such as holes and grooves. But those simplified methods could only predict what happens at the surface.
Detailed nonlinear computer simulations can show what happens inside the surrounding material, but they require considerable computing power, particularly when modelling many load histories, material choices or design variations.
Ince addressed this limitation by integrating several mechanics models into a new framework that efficiently calculates how forces are distributed and how the metal deforms throughout the three-dimensional region near the critical areas even when several types of loading act together and change over time.
The study represents an "important contribution to advanced notch analysis," says Daniel Kujawski, a mechanical engineering researcher at Western Michigan University who studies how materials become damaged and crack under repeated forces. He did not participate in the study.
Kujawski says the method "provides a potential bridge" between calculations focused on individual points and assessments that consider damage throughout the surrounding material. That could eventually help engineers assess how long components will last and identify where cracks could begin or spread.
Ayhan Ince: 'The work represents a landmark advance in computational mechanics.'Further validation ahead
"The work represents a landmark advance in computational mechanics, addressing a 3D field prediction challenge that researchers have pursued for decades," Ince says. "But more work remains."
So far, the researchers compared their framework with detailed nonlinear computer simulations for one steel material and one notched bar under three loading patterns. Next, they plan to examine other materials and geometries and validate their predictions through physical experiments.
Those tests will help determine how widely the method can be applied to real engineering components, Ince says.
Nearly two decades of groundwork
Ince's interest in the problem began while working in the vehicle industry, including at General Dynamics Land Systems-Canada.
He assessed the durability of components such as axles, suspension systems and structural supports, where complicated loading conditions made predicting damage particularly challenging.
While working full time, he pursued a PhD at the University of Waterloo under study coauthor Grzegorz Glinka's supervision. Ince's doctoral research, completed in 2012, laid the foundation for the methods he has continued developing throughout his academic career.
Read the cited paper: "Three-dimensional modeling framework for elastic-plastic stress-strain prediction ahead of notches under general multiaxial loading".
Discover Concordia's Department of Mechanical, Industrial and Aerospace Engineering.
