3. Resolve Contact Energy Error from Overlapping Parts

Why Overlapping Parts Inflate Contact Energy

In LS-DYNA, contact energy (glstat / matsum "internal energy" of contact, reported as sliding interface energy and total contact energy) should stay small relative to internal energy. When two parts are modeled with initial geometric overlap, the penalty-based contact algorithm sees a nonzero penetration at t = 0. The penalty force is proportional to penetration depth, so the solver immediately applies a large restoring force, injecting spurious energy into the system. That energy shows up as abnormal contact energy growth in the first few cycles, often accompanied by a spike in the global energy error and a nonzero initial penetration warning in the messag file.

The distinguishing signature is timing: genuine contact energy from impact grows gradually as parts approach and deform, whereas overlap-induced energy appears almost instantly at the start of the run and is not balanced by corresponding kinetic or internal energy changes. This case traces that signature back to the geometry and removes it.

Step 1 — Confirm the Energy Signature

Run a short diagnostic (a few hundred cycles is enough) and inspect the energy summary. A healthy model shows contact energy near zero at early time. An overlapping model shows a step increase. Use the following checks:

• In glstat, compare contact energy against internal energy. A ratio climbing above a few percent within the first milliseconds is suspicious.
• In messag, search for initial penetration and note the reported penetration depth and the part/segment IDs involved.
• In sleout (if *CONTACT_..._FORCES output is requested), look for large contact forces at t = 0.

Step 2 — Locate the Overlap

Overlap usually comes from one of three sources: parts modeled to share a face but positioned with a small positive gap that was later closed by a translation, CAD assembly tolerances that were not respected, or a thickness offset applied in the wrong direction. To find it, request a contact-force probe or use the initial penetration report to identify the offending segment pair, then measure the actual geometric gap between the two surfaces in the preprocessor.

For a shell-to-shell pair, the effective gap is the surface-to-surface distance minus the sum of the contact thicknesses. If that value is negative, the parts overlap in the contact sense even if the mid-surfaces do not intersect.

Step 3 — Apply the Correction

Choose the correction that matches the physical intent of the model. Three options are common:

1. Translate the part. If the parts are meant to be separate, move one part so the effective gap is positive. In a keyword deck this is done by adjusting the node coordinates or by applying a small *INITIAL_VELOCITY that separates them before contact is active. In a preprocessor, use the part translate/offset tool and re-export.
2. Exclude the contact pair. If the overlap is intentional (e.g., a bonded interface modeled with coincident surfaces) and contact is not physically meaningful there, remove that pair from the contact definition using *CONTACT_..._EXCLUDE or by excluding the part set from the contact's part set.
3. Adjust the thickness offset. For shell contacts, the contact thickness (SST, MST, SFST) controls the effective surface. If the offset sign is wrong, the effective surfaces overlap even when mid-surfaces are separated. Correct the sign and magnitude so the effective gap is non-negative.

Step 4 — Verify the Fix

Re-run the short diagnostic and confirm the following:

If the contact energy still steps up, the overlap was not fully removed — re-check the effective gap including thickness offsets, and confirm the exclusion set actually covers the offending segments. Once the early-time contact energy is flat, proceed to the full simulation.