I'm running a drilling simulation in Abaqus/Explicit on an Al2024-T3 plate to study the residual stress near the hole and compare it with results from my real experiment. I’ve already implemented all the key material behaviors, including Johnson–Cook plasticity and damage with element deletion enabled, plus proper contact and boundary conditions. The drill bit spins and feeds correctly through the plate, and I'm capturing stress at a point about 2 mm from the hole edge — exactly where I placed strain gauges in the actual experiment. When I look at the results, the range of residual stress is quite close between simulation and experiment: in the real test, the stress (converted from strain) ranges from about −14 MPa to +25 MPa, and in the simulation, it fluctuates between −10 MPa and +16 MPa. So in terms of magnitude, it's a good match.

The problem is in the shape of the curve. In the experiment, the stress builds up smoothly and stabilizes over time — but in the simulation, the curve is full of sharp fluctuations that go up and down like noise. It doesn’t reflect the physical behavior I observed. I’m wondering if this is due to mesh quality, local element behavior, or maybe a lack of damping in the model. I’ve already tried refining the mesh and checking the contact settings, but the stress at the measurement point still doesn’t settle into a realistic pattern. If anyone has experienced similar issues or has suggestions to smooth out the stress response in an explicit drilling model, I’d really appreciate the input.