2. LS-DYNA spot weld modeling landscape

Before any spot weld can be calibrated or validated, the analyst must decide how the joint will be represented in the model. LS-DYNA offers several distinct joint representations, and each one embeds a different set of kinematic and constitutive assumptions. Choosing the wrong representation is one of the most common root causes of poor correlation in both crash and durability simulations — not because the material data was wrong, but because the joint's load-transfer mechanics were never captured by the chosen idealization.

The Four Principal Joint Representations

In automotive body-in-white (BIW) practice, four representations dominate: nodal rigid body (rigid spider / constrained spot weld), beam or solid element weld, spot weld with a dedicated failure criterion (MAT_100 / MAT_SPOTWELD family), and cohesive-zone or tied-contact interface. Each maps to a different physical scale and a different failure phenomenology.

The rigid spider ties a set of shell nodes on one sheet to a single node (or node set) on the other using *CONSTRAINED_NODAL_RIGID_BODY or *CONSTRAINED_INTERPOLATION. It transfers force and moment but has no compliance and no failure — it is a "perfect" weld. This is appropriate only when the weld is far from the failure zone and the analyst wants to suppress local deformation. Using it in a crash model where the weld is expected to pull out will produce non-physical load paths and artificially stiff joints.

The beam weld (typically *MAT_100 or *MAT_SPOTWELD on a short beam, or a solid hex weld) introduces axial, shear, bending, and torsional stiffness. It is the workhorse for crash models because it can carry the combined tension–shear–peel loading that real spot welds experience during impact. The beam's cross-section, length, and material card determine its stiffness; the failure criterion determines when it releases.

The cohesive-zone / tied-contact approach (*MAT_COHESIVE_MIXED_MODE, *CONTACT_TIED_*) is preferred when the weld nugget itself is being studied — for example, in durability or weld-quality investigations where crack initiation and propagation through the nugget or heat-affected zone matter. It resolves the traction–separation law directly and can capture mixed-mode fracture, but it demands finer mesh and more calibration data.

Assumptions That Drive Selection

Every representation carries assumptions that must be checked against the simulation's intent:

  • Scale assumption: rigid and beam welds assume the nugget is small relative to the structural wavelength; cohesive zones assume the process zone is resolved by the mesh.
  • Failure assumption: rigid welds never fail; beam welds fail by a resultant-force or equivalent-plastic-strain criterion; cohesive zones fail by energy release rate.
  • Loading assumption: beam welds assume the dominant load path is through the beam axis; if peel dominates, the beam may over-stiffen the joint unless calibrated.
  • Mesh assumption: beam welds tolerate coarse meshes (5–10 mm shells); cohesive zones typically require 1–3 mm or finer near the nugget.
Selection rule of thumb: If the weld is not expected to fail and is not in the crash energy-management path, use a rigid spider. If the weld may fail and the mesh is coarse, use a beam weld with a calibrated failure criterion. If the weld's internal fracture process is the subject of study, use a cohesive zone.

Selection Criteria for Crash vs. Durability

Crash and durability simulations impose different requirements, and the same physical weld may need different representations in each.

Criterion Crash Simulation Durability Simulation
Primary concern Energy absorption, joint rupture sequence Crack initiation, fatigue life
Typical representation Beam weld (MAT_100) or solid weld Cohesive zone or detailed solid nugget
Mesh size near weld 5–10 mm shells 1–3 mm or finer
Failure criterion Resultant force / plastic strain Traction–separation, S-N or strain-life
Loading rate High (10–100 /s) Low, cyclic
Calibration data Component-level impact tests Coupon fatigue tests, weld quality data

In practice, a crash model may use beam welds throughout the BIW, while a durability model of the same joint uses a submodel with cohesive elements. The two models are not interchangeable, and the analyst must document which representation was used and why.

Practical Selection Workflow

A defensible selection process follows four steps. First, classify the joint by its role: load path (carries primary structural load), attachment (locates a panel), or sacrificial (designed to fail). Second, check the mesh resolution available at the joint. Third, check whether failure is expected and, if so, what mode (pull-out, shear, peel, nugget fracture). Fourth, match the representation to the required output: force–displacement, rupture time, or fatigue life.

For example, a B-pillar-to-rocker joint in a side-impact model is a load-path joint expected to fail in combined tension–shear. With a 6 mm shell mesh, a beam weld using *MAT_100 with a force-based failure criterion is appropriate. The same joint in a durability load case, where the concern is fatigue cracking at the nugget edge, would be modeled with a cohesive zone and a finer mesh, because the beam weld cannot resolve the local stress gradient that drives crack initiation.

Common pitfall: Using a rigid spider in a crash model because it is fast and stable. This suppresses joint failure entirely and can shift the predicted failure location to the parent metal, producing a non-conservative or non-physical result. Always verify that the chosen representation can fail if the physical joint can fail.

Representation Map

The following diagram summarizes how the four representations relate to mesh scale, failure capability, and simulation type. It is intended as a selection aid, not a substitute for engineering judgment.

Spot Weld Representation Selection Map Mesh scale coarse → fine Rigid Spider No compliance No failure Coarse mesh OK Beam Weld Axial/shear/bending Force-based failure 5–10 mm shells Solid Weld 3D stress state Strain-based failure 2–5 mm solids Cohesive Zone Traction–separation Mixed-mode fracture 1–3 mm or finer Attachment only No failure expected Crash / impact Joint rupture sequence Crash / detailed Nugget stress state Durability / fatigue Crack initiation Selection guidance: • Load-path joint + coarse mesh + failure expected → Beam weld (MAT_100) • Nugget fracture study + fine mesh + mixed-mode → Cohesive zone • Non-critical attachment + no failure → Rigid spider (fastest, most stable)

Documenting the Choice

Because the representation determines what the model can and cannot predict, the selection must be recorded in the simulation report. A minimal record includes: joint ID, representation type, element formulation, material card, failure criterion and its parameters, mesh size at the joint, and the calibration source. Without this record, a later reviewer cannot tell whether a discrepancy is a modeling error or a genuine physical effect.

The next chapter moves from representation to the constitutive side: how the beam weld's force–displacement response is calibrated against coupon and component tests, and how the failure criterion parameters are extracted from those tests.

← Previous 📖 Contents Next →