FatDyn-X
Time-domain response Solver routes Frequency response Random vibration Verification

Time domain

Response histories from the solve you already have

An inertia-relief / static run is enough for dynamics: the unit load cases and the load histories give every response below. The same responses come from inertia-relief via normal modes and from the modal-transient solve, for load histories of any length.

Node time histories

Displacement, velocity and acceleration at any node. Total acceleration includes the rigid-body part.

Relative motion and contact

Relative displacement and strain between node pairs: contact and interference checks, twist and bending tracking, over more than a million time steps.

Element forces and stresses

Spring, rod, beam and bush force histories (CELAS, CBUSH, CBAR, CBEAM). Shell and solid stress histories at top and bottom surfaces.

Mass-force time histories

Force and moment on every concentrated mass (CONM2). Mounts of battery packs, motors, payloads and avionics can be sized from calculated forces.

Inertia-relief acceleration contours

Inertia-relief accelerations combined with a trusted mass, inertia and CG of the vehicle, written as contours and as CONM2 forces and moments.

Modal coordinates and amplification

Modal coordinates, the loads acting on each mode and a mode-amplification summary from the modal-transient solve.

Solvers inside FatDyn-X

Three routes to the dynamic response

Start from the inertia-relief / static run, or from the normal modes. With the modes, loads are projected to modal space and solved inside FatDyn-X.

Inertia relief / static

Unit load cases from the solver (SOL 101), combined with the load histories. Gives node, element-force, stress and mass-force histories, and acceleration contours. The route for stiff structures and preliminary loads. More

Inertia relief via normal modes

A single normal-modes run (SOL 103) gives the same quasi-static answer as the unit-load inertia-relief cases (SOL 101), which remain available for stiff structures and preliminary loads. Rigid-body modes drive the accelerations and mass forces; residual-static modes supply the non-resonant shapes.

Modal transient via normal modes

The physics of a solver's modal-transient analysis (SOL 112), without the FORCE, TLOAD and TABLED decks. Damping per mode. Accurate from t = 0 with no start-up oscillation, and non-zero initial loads are handled correctly. Vehicle loads start from 1 g, so this matters.

One set of modal stresses serves the quasi-static and the dynamic events of a duty cycle. See the solver choices.

Frequency domain

Frequency response analysis

Modal frequency response for force excitation, computed inside FatDyn-X from the normal modes. NVH-type responses need only the modes and a short keyword block, in place of a SOL 111 deck with its RLOAD and TABLED cards.

What is computed

  • Complex transfer functions H(f): magnitude and phase
  • Modal damping applied per mode
  • Node displacement, velocity and acceleration against frequency
  • Spring, rod, beam and bush element forces against frequency
  • Shell and solid element stresses against frequency

Output

  • One CSV per node and per element
  • Consolidated stress-response scan for large element sets
  • The same transfer functions feed random response and PSD fatigue

Frequency domain

Random vibration response

Response to PSD / ASD input at nodes, in element forces and in stresses, by two input paths.

Load (force) PSD, via normal modes

  • The frequency response is built internally from the modes
  • The frequency grid is derived from the band of the input PSD
  • PSD fatigue starts automatically when the response is complete

Acceleration PSD (base excitation)

  • Step 1: FatDyn-X writes a runnable Nastran / OptiStruct deck
  • Step 2: with the solver's punch file, it computes the PSD response and fatigue

Response statistics

Spectral moments m0, m1, m2 and m4; RMS, irregularity factor and peak rate.

Any signal to a PSD

A stand-alone tool converts acceleration or load time histories to PSDs and reports the number of averaged segments.

No accelerometer data?

Acceleration and mass-force histories can be computed from the normal modes and the load histories, then converted to PSDs.

Verification

Checked against commercial FEA