Modal + harmonic vibration setup for a 6S4P battery pack structure — how to model cells and apply base excitation

Background:

This continues the same custom 6S4P 21700 Li-ion pack (~22 V nominal, ~1000 Wh) I’ve been simulating for a portable industrial device. I’ve looked at the thermal side already; now I want to evaluate the structure under operational and transport vibration — specifically the cell carrier, the nickel-strip interconnects, and the mounting points to the enclosure.

Goal:

  1. A modal analysis to find the natural frequencies of the pack assembly.

  2. A frequency-response/harmonic step to estimate stresses at the mounts and interconnects under a base excitation profile [e.g. sine sweep 10–500 Hz, or a PSD per the relevant transport standard].

Setup so far:

  • Geometry simplified — cell carrier and enclosure mounts modeled as solids; I’m unsure how to best represent the 24 cells.

  • Bonded contacts between carrier and cells; fixed support at the [4] enclosure mounting bolt locations.

  • Second-order tets, mesh refined at fillets and mount features.

My specific questions:

  1. For modal accuracy without an unworkable mesh size, is it better to model each cell as a solid cylinder with effective (smeared) density/modulus, or as a point/remote mass coupled to the carrier? What do people find reliable here?

  2. For base excitation (the whole pack shaken at the mounts), what’s the correct workflow — a harmonic analysis with prescribed base acceleration? And is random vibration (PSD input) supported, or should I approximate with a sine sweep?

  3. What’s the recommended way to apply the constraint and excitation at the mounting bolts so the boundary condition is realistic rather than over-stiff?

What I’m not asking:

I’m not asking whether vibration matters — I know it does for a portable device. I want the setup approach that gives trustworthy natural frequencies and mount/interconnect stresses.