The Geometry Didn't Move, but the Physics Did: Auditing Dynamics Models Across a Supplier Transfer

A supplier transfer is usually managed as a purchasing event: a new purchase order, a new quality clause, a new set of first-article inspections. For the dynamics model, however, it is a model-change event. The part number may not change. The drawing revision may not change. The mesh may still import cleanly. But the process that produced the part has changed, and the parameters that once described that process may now be describing a factory that no longer makes the component.

Reshoring makes this problem especially visible. A part that returns from an overseas supplier to a domestic shop may keep the same nominal geometry while the casting method, heat-treat cycle, machining sequence, and assembly fixturing all change. The model does not know this. It continues to run with the assumptions entered years ago, and those assumptions can be wrong in ways that are difficult to see in a contour plot.

Treat the Supplier Change as a Model Change Request

The first cultural step is procedural. A supplier transfer should trigger a model-change request, not merely a purchasing checklist. The request should identify every simulation model that references the transferred part, including system-level models where the part is a minor component. A bracket that appears in a full-vehicle dynamics model may carry joint stiffness assumptions that were calibrated to a previous supplier's weld process. If those assumptions remain untouched, the model has effectively been re-baselined without review.

The model-change request should ask a simple question: which parameters in this model were derived from the manufacturing process rather than from the design intent? Design intent travels with the drawing. Process-derived parameters do not.

Five Parameters That Rarely Survive the Move

Joint stiffness is the most common casualty. A bolted joint's stiffness depends on clamp load, surface finish, fastener coating, and the compliance of the joined members. A new supplier may use a different torque strategy, a different thread-locking compound, or a different surface treatment. The CAD assembly looks identical, but the joint's dynamic behavior may not be. If the model uses a linear stiffness value calibrated from test or from a previous supplier's process, that value deserves a fresh look.

Damping is the second. Material damping and interface damping are often entered as engineering judgments rather than measured values. A change in casting porosity, coating, or joint preload can shift the effective damping in a assembled structure. The shift may be small at the component level and significant at the system level, particularly near resonance.

Material anisotropy is the third. A forged part and a machined-from-solid part can share the same geometry and alloy designation while exhibiting different grain flow and directional properties. If the dynamics model assumes isotropic behavior, the transfer may not invalidate the assumption. If the model uses directional properties calibrated to a prior forging line, the new process may not reproduce them.

Residual stress is the fourth. Heat treatment, welding, and machining sequences leave residual stress fields that affect stiffness, distortion, and fatigue. A supplier change can alter those fields even when the material certificate matches. The model may not include residual stress directly, but its boundary conditions and preload assumptions may have been tuned around a particular stress state.

Tolerance assumptions are the fifth. A new shop may hold the same drawing tolerances but distribute variation differently. The mean clearance in a bushing may shift, or the worst-case stack may move from one side of nominal to the other. Dynamics models that use nominal clearances or rigid contacts can miss the resulting change in impact behavior and rattle.

A Pre-Run Transfer Checklist

Before the next solver run, the model owner should document the following:

• Which process-derived parameters are present in the model, and who owns each one.
• Whether joint stiffness values were calibrated to a specific supplier process, and whether that calibration still applies.
• Whether damping values were measured, estimated, or inherited from a legacy model.
• Whether material directionality assumptions match the new supplier's forming process.
• Whether residual stress or preload assumptions were tied to a prior heat-treat or machining sequence.
• Whether tolerance and clearance assumptions were based on the previous supplier's capability data.
• Whether any test-correlation cases used hardware from the old supply chain.

The checklist is not a demand to revalidate everything. It is a filter. Some parameters will survive the transfer intact. Others will need a targeted test, a supplier conversation, or a sensitivity study. The goal is to make the decision deliberately rather than inherit it by default.

The Cost of Assuming Geometry Is the Model

The reshoring shock is not that parts come home. It is that models often stay exactly where they were. A dynamics model is a chain of assumptions, and manufacturing process is one of the largest assumption sources in that chain. When the process moves, the model should move with it.

A transfer audit does not need to be a months-long program. It can be a focused review of the parameters that are most sensitive to process change, completed before the next design review or solver run. The alternative is a model that looks current because its geometry is current, while its physics quietly describes a supplier that is no longer in the chain.