2. Crash energy management fundamentals

Every crashworthy body structure performs one primary job: absorb the kinetic energy of a moving vehicle in a controlled, predictable way, while keeping the passenger compartment intact and limiting the forces and accelerations transferred to occupants. Generative design for energy-absorbing structures only becomes meaningful once these physical constraints are explicit, because the optimizer must be told what "good" looks like in terms of deformation mode, force level, and occupant load.

Kinetic energy and the stopping distance budget

The energy that must be dissipated in a barrier impact is the vehicle's kinetic energy at the moment of contact:

E_k = 1/2 * m * v^2

Example (compact BEV, 1,800 kg, 50 km/h = 13.89 m/s):
E_k = 0.5 * 1800 * 13.89^2 ≈ 173.6 kJ

That 173.6 kJ is not destroyed — it is converted into plastic work through folding, bending, tearing, and friction. The average crush force required over a given deformation stroke follows directly:

F_avg = E_k / s

If the front structure is allowed 600 mm of crush stroke:
F_avg = 173,600 J / 0.6 m ≈ 289 kN

This single relationship drives most front-end design decisions. A longer stroke lowers the average force, which lowers occupant deceleration. A shorter stroke forces higher forces into the same mass. Generative design typically treats crush stroke as a hard geometric constraint and searches for a wall thickness, rib topology, and material grade that produce a stable force plateau near the target F_avg rather than a single peak.

Deformation modes: what the structure must actually do

Energy absorption is only useful if the collapse is progressive. Three modes dominate automotive energy absorbers:

  • Progressive folding (axial crushing). Thin-walled columns and crash boxes fold in a repeating pattern of lobes. Force oscillates around a mean value with a peak-to-mean ratio ideally below about 1.3. This is the preferred mode for front rails and rear rails.
  • Global bending. The member buckles sideways as a beam. Energy is absorbed, but the stroke is short and the force is unpredictable. In a front rail this usually means the rail folds instead of crushing, and the remaining stroke is lost.
  • Fracture and tearing. Material cracks before full folding. Energy per unit mass is high but the mode is unstable and can produce sharp edges and sudden force drops.

The classic design lever for forcing progressive folding is the ratio of wall thickness t to the square root of the section width, combined with a trigger. A common practice is to add a shallow dent, notch, or chamfer at the front of the crash box so the first fold initiates at a controlled location and at a lower peak force. Generative design can place these triggers automatically, but the objective function must reward a low peak-to-mean force ratio, not just total absorbed energy.

A structure that absorbs 170 kJ but peaks at 600 kN may fail the occupant criterion, while a structure that absorbs 150 kJ with a flat 300 kN plateau may pass. Energy alone is never the objective.

Occupant protection targets by impact direction

Front, side, and rear impacts load the body in fundamentally different ways, and the targets differ accordingly.

Impact direction Primary structure Dominant injury mechanism Typical target
Frontal (full / offset) Front rails, crash boxes, bumper beam, firewall Restraint-coupled chest and head deceleration Limit occupant chest deceleration; keep intrusion below a defined threshold
Side (pole / MDB) B-pillar, rocker, door beams, floor cross members Direct chest and pelvis loading; small crush distance Limit B-pillar intrusion velocity and residual intrusion
Rear Rear rails, rear crash boxes, back panel Whiplash and seat/fuel-system integrity Protect fuel system and seat anchorage; limit rear intrusion

In frontal impact, the available stroke is large (often 500–700 mm), so the design problem is one of tuning a force plateau. In side impact, the available stroke between the door outer panel and the occupant is often only 200–300 mm, so the structure must reach a high force level quickly and hold it. This is why side structures use high-strength steel or extruded aluminum with multi-cell sections rather than long folding columns. Rear impact is usually a lower-severity event for the occupant but a high-severity event for the fuel system and rear structure, so the target shifts from occupant deceleration to containment and anchorage integrity.

From targets to design variables

For a generative design study, the occupant targets must be translated into measurable structural responses. A practical mapping is:

  • Frontal: minimize peak force and maximize plateau stability subject to a minimum absorbed energy and a maximum intrusion at the firewall.
  • Side: minimize B-pillar intrusion at chest and pelvis height subject to a maximum mass and a minimum stiffness at the door latch.
  • Rear: maximize energy absorbed before rear rail fracture subject to a maximum intrusion into the fuel tank zone.

The SVG below shows how a frontal impact energy budget is distributed across the front structure and how the force–displacement curve relates to the deformation modes.

Frontal Impact Energy Budget and Force–Displacement Behavior Energy path through front structure Bumper Crash box folding Front rail axial crush Firewall Occupant cell must remain intact Impact ~25% E_k ~55% E_k ~20% E_k Force–displacement response Displacement (mm) Force (kN) target plateau Progressive folding (stable) Global bending (unstable) peak force Progressive folding: low peak-to-mean ratio Global bending: high peak, short stroke

Practical verification and common failure symptoms

When a generative design result is transferred to a crash simulation, the first check is the force–displacement curve. A healthy front structure shows a short initial rise to a peak, then a plateau that stays within roughly ±15% of the mean until the crush stroke is exhausted. If the curve shows a sharp peak followed by a deep drop, the structure is bending globally rather than folding. If the curve rises continuously without a plateau, the structure is too stiff and the occupant deceleration will exceed the target.

Typical symptoms and their causes:

  • Peak force too high, plateau too low. The trigger is too weak or the wall thickness is too large at the front. Add or deepen the trigger, or reduce the front section thickness.
  • Force drops after the first fold. The second fold is not initiating. Check the fold wavelength; it should be roughly 1.5–2 times the section width.
  • Rail folds sideways. The rail is too long unsupported or the section is too narrow. Add a mid-rail support or increase the section aspect ratio.
  • Intrusion at the firewall exceeds target. The front structure has absorbed insufficient energy before bottoming out. Increase the crush stroke or raise the plateau force within occupant limits.
When setting up a generative design study, define the occupant target as a constraint (maximum intrusion, maximum deceleration) and the energy absorption as an objective. This keeps the optimizer from trading occupant safety for mass reduction.

The key takeaway is that crash energy management is a force-and-stroke problem, not just an energy problem. The structure must absorb the required kinetic energy over the available stroke, in a stable deformation mode, while keeping the force plateau within the occupant's tolerance. Front, side, and rear impacts each impose different stroke budgets and different injury mechanisms, so the generative design objectives and constraints must be defined per impact direction rather than as a single global target.