Run and plot a sweep
Start with bounded study inputs, then create a named sweep over that study:
sweep load_envelope over study.bracket_response {
axis service_load {
half = 0.5 kN
nominal = 1.0 kN
proof = 1.5 kN
}
combine = product
output peak_by_case = collect(case.output.peak_stress.value)
output governing = max(case.output.peak_stress.value, over = cases)
}Check the sweep root and inspect its case count:
runmat check bracket.fea --sweep load_envelope
runmat plan bracket.fea --sweep load_envelope
runmat run bracket.fea --sweep load_envelopeCases may finish in any order; publication restores canonical case order and retains stable case coordinates. Plot the collected compact values for trends. Retain whole fields only when comparison needs them:
output stress_fields = retain_each(case.output.stress)retain_each can multiply storage by the case count. Prefer attributed extrema or bounded sampled
outputs when the engineering question does not require every field page.
Local execution runs independent cases concurrently within admitted worker and memory bounds. Cluster placement can fan them out without changing study or case identity.