Problem definition
minimize f(x), subject to g_i(x) ≤ 0, h_j(x) = 0 and x_min ≤ x ≤ x_max.
x can contain fiber fraction, ply angles, ply counts or process variables supported by the selected study. State the objective and units. A mass objective, a stiffness target and a failure constraint answer different questions.
Feasible comparisons
Compare candidates with the same loading, material basis and model assumptions.
Record which variables changed, the feasible bounds, calculated outputs and rejected constraints. Discrete ply counts and manufacturability limits require explicit treatment. Do not compare a failed or incomplete numerical run with valid candidates as though it were a feasible design.
Interpretation and validation
Best sampled candidate ≠ proof of global optimum.
Grid and bounded searches only establish the best evaluated candidate. Repeat at finer resolution and test sensitivity around the candidate. Calibration uncertainty and unmodeled failure mechanisms can dominate a small objective improvement. Preserve each run and its inputs for an auditable study.
These explanations describe general mechanics and transport principles. Follow the selected Workbench model’s supported mechanisms, calibration and limitations.