2. Technical foundations

Before any LS-DYNA model can be trusted, the engineer must share a precise vocabulary with the simulation. Battery pack crash protection sits at the intersection of three disciplines — structural mechanics, electrochemistry, and explicit finite element analysis — and each brings its own terms. This chapter establishes the working definitions, the coordinate and sign conventions, the material and element concepts, and the regulatory framework that every later chapter depends on. The goal is not encyclopedic coverage but operational fluency: knowing what a term means, where it appears in a keyword deck, and how it influences the predicted result.

2.1 The physical object: cells, modules, and pack

A traction battery pack is a hierarchical assembly. At the lowest level sits the cell — a cylindrical, prismatic, or pouch unit with a nominal voltage typically between 3.2 V (LFP) and 3.7 V (NMC). Cells are grouped into modules, which provide mechanical retention, electrical interconnection via busbars, and often cooling channels. Modules are mounted inside a pack enclosure, usually an aluminium or steel tray with a bolted or welded lid, which is itself attached to the vehicle body through mounting points.

For crash simulation, the relevant mechanical idealisation differs by level. A single 21700 cylindrical cell behaves as a thin-walled shell with an internal jelly-roll; under axial compression it exhibits a characteristic force–displacement curve with an initial elastic rise, a peak, and a long plateau from progressive folding. A module is often represented as a homogenised block with calibrated crush stress, because modelling every cell explicitly is computationally prohibitive for full-vehicle runs. The pack enclosure, by contrast, is modelled in detail — it is the primary load path and the component most likely to be redesigned.

2.2 Crash loading regimes and terminology

Automotive crash events are classified by direction and severity. The terms below recur throughout this course and in regulatory documents:

• Frontal impact — full-width rigid barrier (FWDB), offset deformable barrier (ODB, 40 % overlap), and small overlap (25 %).
• Side impact — moving deformable barrier (MDB) at 50–60 km/h, and pole side impact.
• Rear impact — typically lower severity but relevant for pack retention.
• Underbody / ground impact — kerb strike, speed bump, or rock impact, which drives bottom-plate intrusion.
• Thermal runaway propagation — a post-crash electrochemical event, not a structural one, but triggered by mechanical deformation.

Two quantitative measures dominate acceptance criteria. Intrusion is the maximum displacement of the enclosure or a module boundary into the cell region, measured along a defined axis. Cell crush is the local deformation of an individual cell, often expressed as a percentage of its original diameter or height. A third measure, peak acceleration at the pack mounting points, governs fastener and bracket design.

2.3 Explicit FEA concepts specific to LS-DYNA

LS-DYNA solves the equations of motion using an explicit central-difference time integration scheme. The stable time step is governed by the Courant condition, Δt ≤ L_c/c, where L_c is the characteristic element length and c is the wave speed in the material. For steel, c ≈ 5000 m/s; for a 5 mm element, Δt ≈ 1 µs. This single relationship explains why mesh refinement is expensive and why mass scaling is sometimes applied — and why mass scaling must be used with caution, since it alters inertia and can mask real behaviour.

Several LS-DYNA-specific concepts appear repeatedly:

• Hourglassing — spurious zero-energy deformation modes in reduced-integration solid and shell elements. Controlled via *CONTROL_HOURGLASS and monitored through the hourglass energy ratio in glstat.
• Contact — defined with *CONTACT_AUTOMATIC_SINGLE_SURFACE or *CONTACT_AUTOMATIC_SURFACE_TO_SURFACE. Penetration and contact energy are checked in glstat and sleout.
• Erosion — element deletion when a failure criterion is met, controlled by *MAT_ADD_EROSION. Erosion is a numerical convenience, not a physical prediction, and must be justified.
• Spotweld and tiebreak — *CONSTRAINED_SPOTWELD or *CONTACT_SPOTWELD for joining, with failure parameters calibrated against coupon tests.

2.4 Material models and their applicability

Three material families cover most battery pack crash work. Metals (enclosure, busbars, brackets) use *MAT_PIECEWISE_LINEAR_PLASTICITY (MAT_024) or *MAT_PLASTIC_KINEMATIC (MAT_003), with strain-rate enhancement via the Cowper–Symonds coefficients C and p. Polymers and composites (module housings, thermal pads) use *MAT_ELASTIC_PLASTIC or *MAT_ORTHOTROPIC. Cell homogenised material is often represented with *MAT_CRUSHABLE_FOAM (MAT_063) or a user-defined curve, calibrated so that the plateau stress matches cell-level test data.

The table below summarises the mapping between component, typical material model, and the key parameter that must be calibrated.

2.5 Regulatory and testing framework

Two standards dominate. UN ECE R100.3 specifies mechanical integrity tests including vibration, mechanical shock, and crush, with the requirement that no fire or explosion occurs. GB 38031-2020 (China) adds a dedicated battery pack crush test and a thermal propagation requirement. In the United States, FMVSS 305a governs electrolyte spillage and electrical isolation after impact. These standards define the load cases that a simulation must reproduce, and their pass/fail criteria become the simulation's acceptance thresholds.

The figure below maps the relationships between the physical hierarchy, the crash load cases, and the simulation outputs that are checked against regulatory criteria.

2.6 From terminology to application: a worked example

Consider a side pole impact on a pack with a 2 mm aluminium tray and a 30 mm gap between the tray inner wall and the nearest module. The simulation must answer one question: does the pole intrude far enough to crush a cell?

The workflow proceeds as follows. First, the pole is modelled as a rigid cylinder with a 150 mm radius, positioned at the vehicle's mid-height and given an initial velocity of 32 km/h toward the pack. Second, the tray is meshed with 5 mm shell elements using MAT_024, with the flow curve taken from tensile tests at 0.001, 0.1, and 10 s⁻¹. Third, the module is represented as a homogenised block with MAT_063, its plateau stress calibrated to 12 MPa from a cell-stack compression test. Fourth, contact is defined between the pole and the tray, and between the tray and the module, using *CONTACT_AUTOMATIC_SURFACE_TO_SURFACE with a friction coefficient of 0.15.

After the run, the engineer inspects three outputs. The glstat file reports the hourglass energy ratio; if it exceeds 5 % of total energy, the result is suspect and the mesh or hourglass control must be revisited. The nodout file gives the displacement of the tray inner wall at the pole's centreline; if this exceeds 30 mm, the module is contacted. The secforc file gives the force at the mounting points, which is compared against the fastener's rated capacity. Only when all three checks pass — energy balance, intrusion below the gap, and mounting force below capacity — can the design be considered acceptable for this load case.

2.7 Summary of key terms

The vocabulary established here — cell, module, pack, intrusion, cell crush, hourglassing, erosion, spotweld, plateau stress, and the standards ECE R100.3, GB 38031, and FMVSS 305a — forms the shared language for the remainder of this course. Each term maps to a specific keyword, output file, or acceptance criterion. When a later chapter refers to "calibrating the plateau stress" or "checking the hourglass ratio," it assumes the definitions given here. The next chapter builds on this foundation by examining the geometry and mesh preparation required before any of these concepts can be applied in a working LS-DYNA deck.