Mastering Contact Stress: The Non-Linear Reality of Pins and Bores

 

In the traditional world of machine design, we are often comforted by the simplicity of linear equations. We apply a force, divide it by an area, and obtain a stress value that we compare against a material’s yield strength. However, when we move away from simple beams and into the intricate world of mechanical assemblies, where parts "push" into one another, these linear assumptions begin to crumble. This is the world of contact stress, a complex field where the relationship between load and deformation is rarely a straight line.

Why Linear Thinking Fails in Contact?

In many mechanical connections, such as a pin supported in a bore, we are dealing with parts that are not bonded together but are instead in a state of constant, variable interaction. Standard linear equations are designed for static, unified structures, but they fail to capture the reality of surfaces that may only touch at a single point or across a narrow arc.

When we simulate these interactions using Finite Element Analysis (FEA), we must shift from "Linear Static" to "Non-Linear Contact". This transition is necessary because the contact area itself changes as the load increases, requiring the computer to execute a series of iterations to find where the parts actually touch and how much they deform at that specific interface.

The Myth of "Average Pressure" in Bushing Design

For decades, many designers have relied on a "rule of thumb" equation to calculate the pressure in a bushing or bore: σ=F/(w×d), where the force is divided by the width and diameter of the bore. This is widely referred to as the average pressure equation.

The source material reveals a critical limitation of this approach: this equation is mathematically derived from an integral that assumes a uniform pressure distribution across a full 180 degrees of the bore. In the real world, this is seldom the case. Physical evidence, such as wear patterns on used bushings, suggests that contact often stops at about 90 degrees.

Because the actual contact area is much smaller than the theoretical 180-degree arc, the true peak pressure can be as much as three times higher than the calculated average. For this reason, engineers typically limit the "average" pressure to less than one-third of the material’s yield strength to account for these hidden peaks. Using non-linear FEA allows us to stop guessing and actually map the true pressure distribution at the interface.

The Impact of Running Clearance and Hertzian Stress

A second layer of complexity arises from the running clearance, the tiny gap between the pin and the bore required for assembly or motion. In a linear simulation, parts are often assumed to be perfectly matched, but in a non-linear contact simulation, the pin must first deflect to "take up" that clearance before it ever begins to push against the bore surface.

This interaction creates what is known as Hertzian compressive stress, named after Heinrich Hertz, who pioneered the study of contact mechanics in 1882. As the load increases, the pin is forced deeper into the bore surface, and the contact area spreads.

The sources provide a startling demonstration of this non-linearity:

• In a low-load scenario, the contact pressure might be spread over only 20-30 degrees.

• When the load is increased by ten times, the contact area spreads toward 90 degrees.

• However, because the contact area is growing, the peak pressure does not simply increase tenfold; the relationship is non-linear.

Furthermore, while the surface experiences peak compressive stress, the most critical area for failure is often below the surface, where high levels of shear stress develop. This is why simply checking the surface pressure is often insufficient to prevent long-term fatigue or "scoring".

Strategies for Achieving Convergence in Complex Simulations

Because non-linear contact simulations require the solver to match pressure distributions with interface deflections through multiple steps, they are notoriously "heavy" on CPU capacity. A simulation that previously took seconds in a linear mode can tie up a computer for significantly longer in a non-linear mode.

The goal of the solver is to reach convergence, the point where the calculated deflection matches the pressure distribution at the interface. To help the solver reach this point without the model going "unstable," several strategies are recommended:

1. Use Guided Planes: Adding a guided plane to a pin, for example, can keep it from moving left to right (x-direction) while allowing it to move freely up and down (y-direction) to settle into the bore.

2. Symmetry Planes: Whenever possible, use half or quarter symmetry. By cutting the model in half and adding a symmetry boundary constraint, you not only save a massive amount of CPU power but also add structural stability to the model, preventing parts from "flying away" during early iterations.

3. Load Balancing: Ensure that applied loads and reactions are perfectly balanced. Unrepresentative boundary constraints (like fixing a bore rigidly) can create artificial stress "hotspots" that prevent the solver from converging.

The Responsibility of Verification

Ultimately, the power of modern FEA tools allows us to visualize complex contact states that were invisible to previous generations of engineers. However, as author Anthony Rante emphasizes, "the FEA method is powerful, but it is easy to misuse".

In real-world joints, local yielding often occurs at the contact surface, which helps redistribute the load and reduce peak stresses. This is a function of material toughness and elongation. While the software can give us a "pretty picture" of these stresses, the engineer must still verify the results using load balances and a foundational understanding of mechanics.