Run a drop test

A part is released from a height, lands on a rigid floor, and you want to know the peak stress and how much of the dent stays. That is explicit dynamics: an event measured in milliseconds, with plasticity throughout and deformation too large for an implicit solver to follow.

Solved by OpenRadioss. The free tier caps the event at 10 ms, which is enough for most single impacts and not for a sequence of them.

Pick the analysis

Press + New simulation.
Choose Explicit dynamics, then Drop test.

Impact is the same run stated differently: there you give the velocity the part already has, here you give the height it fell from and the deck carries sqrt(2gh) downwards. The fall itself is not simulated — only the landing — so the run spends its steps where the answer is.

Mesh the part

Explicit dynamics meshes as every other solid family does, with one restriction worth knowing before you spend time on it: first-order tetrahedra only. An explicit code carries no mid-side nodes, so the mesher is asked for order 1 and the reader refuses anything else rather than quietly dropping nodes.

Select Mesh in the tree.
Set Maximum size and press Generate mesh.

Order is fixed at linear here and the control says so. Every other solid family meshes quadratic by default; this one cannot, so bending accuracy is bought with more elements rather than better ones.

The element size sets the time step, not just the accuracy. An explicit solver takes a step no longer than the time a stress wave needs to cross the smallest element, so halving the element size both doubles the number of elements and doubles the number of steps. A mesh twice as fine costs four times as much.

Set the material

The material is elastoplastic — Johnson-Cook, on the Material step. Density, Young’s modulus, Poisson’s ratio and yield stress come from the library; the two hardening numbers and the two optional groups below them are the case’s own.

Yield stress is required. An impact that stays elastic bounces and tells you nothing about damage, and the validator says so.
Hardening modulus B and exponent n give the flow stress σy + B·εp ⁿ. Zero B is perfectly plastic.
Strain rate and Thermal softening are off unless you set their coefficients, and both are fitted to a specific test rather than being properties of the metal. Leave them at zero unless you have the constants from the same source as your yield stress.

State the drop

On the Models step:

Drop height — the fall the part is released from. The pane shows the impact speed it works out to, straight down.
Floor clearance — the gap between the lowest node and the floor when time starts. Small, so the run does not spend steps on free flight; not zero, so the first step is not already a penetration.
Event duration — how long to watch. Milliseconds, not seconds.
Animation interval — how often to write a state. It must fit inside the duration, or there is nothing to look at.

A drop always lands on something, so the rigid floor is always there: an infinite plane placed under the lowest node of the part. There are no boundaries to pick on the geometry.

Run it, and read it

Press RUN. Two programs run in order — the starter checks the deck and the engine advances it — and a third converts the last animation state for the viewport.

The result carries von Mises stress and plastic strain. The plastic strain is the one to look at for a dent: it is what did not spring back.

What this build does not do

Stated because a drop test is exactly the case where people assume more:

Not possible	Why
Two deformable bodies in contact	The floor is rigid and infinite; one part is what this build offers
Ballistic penetration	Needs failure and element erosion
Sheet-metal forming	Needs a die and a punch
Blast	A different analysis type in this family, not a drop — see Run a blast analysis

Analysis types and solvers — what each family solves and with what.
Explicit material and Explicit analysis — every control on the two panes.