Press on it

Drag the plate down. The dent is not an animation — it is the finite-element solution for this foam, swept around the axis of the plate. Let go and it comes back.

Twin Mattress
0.95 × 1.90 × 0.20 m · 16.10 kg · 45 kg/m³ foam
0.0
newtons
Show mesh
Run the test
Indentation force is read at 25% — that is the condition the reference bands are quoted against.

Indentation force

0110155220 N
Force0.0 N
In pounds0.0 lbf
Plate travel0.0 mm
Plate area323.0 cm²
Plate radius101.4 mm

Why not just use a formula

Elastic half-space
This bonded layer
Stiffer by1.65×
The closed-form answer is 65% low. Foam in a real mattress sits on a solid base and cannot flow away under the plate. That gap is the whole reason this is simulated.

Mesh convergence

Where the shape comes from. The plate press is axisymmetric, so the problem reduces exactly to two dimensions — no approximation. We solve it with 4-node axisymmetric quadrilaterals and 2×2 Gauss quadrature, then sweep the surface displacement around the axis. Because the material model is linear, the field scales with travel, so every depth you drag through is the exact solution rather than a blend between keyframes.
We stop at 25% and will not quote 65%. At 65% compression foam cells buckle and densify. A linear model pushed that far would overstate the force badly, so no IFD65 and no support factor are published. It would need a hyperelastic model — Ogden or Hill. A missing number beats an unverified one.
What the shipped file does. MattressTwin_v1.usdz is a rigid body with one exact convex collider — right mass, right size, right shape. It does not squash at runtime. The figure here describes the foam the asset represents, and it is reproducible with one command: python3 harness/09_ifd_fem.py 0.20 7194 0.30