Inertia Relief in Nonlinear Analysis

Engineering Tools — Inertia Relief in Nonlinear Analysis

Why this page exists

A common request is "I have a free-free nonlinear model — how do I turn on inertia relief in SOL 106?" The short answer is that you cannot, at least not with PARAM,INREL. That parameter is a linear-static feature, and SOL 106 (nonlinear static) is not on its supported-solution list. Nastran does not issue an obvious error when you try — it simply does not do what you expect, which is exactly why this trips people up.

PARAM,INREL is linear-static only

Both Nastran lineages are explicit about this.

• Simcenter / NX Nastran (QRG 2401): INREL = -1 is "valid for SOLs 101, 105, and 200"; INREL = -2 is "valid for SOL 101 only". SOL 106 is not listed.
• MSC Nastran (QRG 2024.2): "INREL controls the calculation of inertia relief or enforced acceleration in linear static analysis and buckling."

For reference, the linear PARAM,INREL options are:

  PARAM,INREL,-1   needs SUPORT / SUPORT1, total DOF <= 6   (SOL 101/105/200)
  PARAM,INREL,-2   automatic, free-free model, no SUPORT     (SOL 101 only)
  PARAM,INREL,-4   MSC only - general, any DOF, handles mechanisms

None of these do anything inside SOL 106.

Doing inertia relief WITH nonlinearity

The documented path: MSC SOL 400, IRLOAD = QLINEAR

MSC Nastran's supported route to inertia relief in a nonlinear run is SOL 400 with the IRLOAD Case Control command:

  SOL 400
  ...
  IRLOAD = QLINEAR        $ global - applies to all SUBCASES
  ...
  ANALYSIS = NLSTATIC

Key points from the MSC QRG (2024.2, IRLOAD p.439):

• Active only in SOL 400 with ANALYSIS = NLSTATIC.
• Requires a set of STATIC supports that constrain all six rigid-body motions.
• IRLOAD = QLINEAR is a quasi-linear (small-displacement) inertia-load calculation. Combining it with large displacement (PARAM,LGDISP,1) is a fatal error, and superelements with IRLOAD also fatal.
• IRLOAD = NONE (default) deactivates the inertia load calculation.

So even the "nonlinear" inertia relief is small-displacement in its load balancing — the nonlinearity is in the material and boundary behavior, not in the inertia-relief load computation itself.

The pragmatic workaround if you must stay in SOL 106

If moving the model to SOL 400 is not practical, the field workaround is to do the inertia-relief balancing by hand: compute the rigid-body acceleration that balances your applied load, turn that into an equivalent inertia load set (mass times acceleration over the model), and apply it as an explicit load alongside the real load so the model is self-equilibrated. This is exactly what PARAM,INREL does internally for you in linear statics — here you are just doing the bookkeeping yourself. It is approximate (the mass distribution and the balancing acceleration are evaluated up front, not updated as the model deforms), but for many free-free nonlinear cases it is good enough.

Which to use

• Linear free-free model — PARAM,INREL in SOL 101. Done.
• Nonlinear, and you have SOL 400 — SOL 400 + IRLOAD = QLINEAR + a 6-DOF STATIC support.
• Nonlinear, stuck in SOL 106 — build the inertia-balancing load set by hand and apply it explicitly.

The thing to remember: there is no PARAM,INREL that works in SOL 106. If you put it in the deck, it is quietly ignored.