2. Seat structure geometry preparation

Impact simulation of an automotive seat begins long before the first LS-DYNA keyword is written. The quality of the finite element model is bounded by the quality of the geometry that feeds it. A seat frame is a welded assembly of thin-gauge stamped steel, tubular members, recliner mechanisms, and track sub-assemblies, and the CAD delivered by the design group is built for manufacturing tolerance and packaging studies — not for explicit crash analysis. This section covers the practical pipeline that converts that CAD into a solver-ready mesh: cleanup, midsurface extraction, connection representation, and mesh strategy for frames and recliners.

Why the CAD Cannot Be Used Directly

Production CAD for a seat typically contains solid bodies with fillets, chamfers, draft angles, embossed stiffening ribs, and small features such as bolt holes, locating tabs, and trim attachment slots. For an explicit shell model, these features are either irrelevant or actively harmful. Small fillets force tiny element sizes, which collapse the stable time step; embossed ribs may be geometrically important for stiffness but are often simplified; and overlapping or non-manifold surfaces prevent clean midsurface generation. The goal of cleanup is therefore not cosmetic — it is to remove features that do not influence the impact response while preserving the load paths that do.

A useful rule of thumb for seat structures: features smaller than roughly 3–5 mm in a 5–8 mm target element size rarely contribute to global crash behavior and should be removed or idealized. Bolt holes that carry no load can be filled; holes that define a joint must be retained and represented by a fastener model.

CAD Cleanup Workflow

Cleanup is normally performed in a dedicated pre-processor (ANSA, HyperMesh, or SpaceClaim) rather than in the native CAD system. A reproducible sequence is:

1. Import and audit. Import the assembly, check for duplicate surfaces, free edges, gaps, and sliver faces. Run a geometry check to quantify the number of free edges before and after each operation.
2. Remove non-structural detail. Delete fillets below the threshold, chamfers, logos, and cosmetic embossing. Suppress or fill small holes and slots.
3. Repair topology. Stitch gaps, remove slivers, and merge coincident edges so that each structural panel forms a closed, manifold shell region.
4. Simplify stiffening features. Where a stamped rib is retained, keep its footprint but remove the draft; where it is not, flatten it and compensate stiffness if necessary.
5. Partition for meshing. Split large panels along natural boundaries — flanges, beads, and weld lines — so that mesh transitions and connection points fall on clean edges.

Midsurface Extraction

Because seat frames are thin-walled (typically 0.8–2.5 mm), shell elements are the correct idealization. Midsurfacing extracts a surface at the mid-plane of each solid, and the original wall thickness is assigned as a section property. Extraction quality depends on the solid being a clean, constant-thickness wall. Where thickness varies — for example, a tailored blank or a hydroformed tube — the midsurface must be offset or the thickness field mapped onto the shell.

Practical checks after extraction:

• Verify that the midsurface lies within the original solid envelope; a surface that drifts outside indicates a bad offset.
• Confirm that adjacent panels meet at a common edge or a defined gap, not with an arbitrary overlap.
• Assign thickness per part and cross-check against the drawing or a measured section.
• For tubular members (seat back frame, recliner shaft), extract a cylindrical midsurface and assign the wall thickness as a shell property.

Where a part is genuinely thick or solid — recliner pawl, pivot boss, track slider — solid elements are appropriate and should not be forced into shells.

Weld and Fastener Representation

Seat frames are predominantly spot-welded, with some MIG/MAG and laser welds on the recliner and track. In LS-DYNA, spot welds are commonly represented by one of the following:

Two rules matter more than the choice of card. First, the weld must connect the correct nodes: a spot weld placed on a node that is not shared by both panels will not transfer load. Second, the weld spacing must reflect the physical pitch — typically 25–50 mm on seat frames — because weld spacing controls the local stiffness and the failure sequence. A model with welds every 10 mm will be artificially stiff and will not predict the correct collapse mode.

Meshing Strategy for Frames and Recliners

Seat frames are meshed with a target element size of 5–8 mm for the global model, refined to 2–3 mm in the recliner and pivot regions where stress gradients and failure are expected. The mesh should be predominantly quad-dominant, with triangular elements kept below about 5% of the total and avoided in load-path regions. Aspect ratios should stay below 5:1, and warpage below 15°.

For the recliner mechanism, the mesh must resolve the tooth engagement and the pawl-to-sector contact. This usually means a local refinement to 1–2 mm on the sector teeth and the pawl tip, with a contact definition that captures the tooth flank geometry. Because the recliner is a load-bearing joint, its mesh quality directly determines whether the model predicts the correct failure mode — tooth shear, pawl bending, or bracket tearing.

Track assemblies are meshed as shells for the rails and solids for the sliders and locking pawls. The ball or roller elements in a rolling track can be idealized as rigid cylinders with contact, since their deformation is negligible compared to the rail.

Verification Before Running the Model

Before submitting the model, run a short set of checks that catch the majority of geometry-related errors. First, a free-edge check on the assembled mesh: every free edge should correspond to a physical boundary, not an unintended gap. Second, a thickness audit: list all parts with assigned thickness and compare against the drawing. Third, a connection audit: count spot welds and verify that each connects two distinct parts. Fourth, a mass check: compare the model mass against the CAD mass; a deviation beyond a few percent usually indicates missing parts or wrong thickness.

A practical acceptance criterion for a seat frame model is that the first natural frequency and the static stiffness under a representative load fall within about 10% of a validated reference or test. If they do not, the error is almost always in the geometry preparation — a missing weld, a wrong thickness, or an over-simplified rib — rather than in the material card.