Introduction to practical FEA.

A: Option A fits the duty and the audit context. You're checking stresses against allowables, not post-yield behavior, and linear elastic keeps the model transparent under HSE scrutiny. B sounds attractive because welds do yield locally, but that drags you into material data justification and plastic strain acceptance criteria you don't need for an occasional lift. C solves a problem you don't have; nothing is time-dependent here, and explicit methods raise questions during review. D has value for vibration problems, yet it doesn't answer whether the lug meets stress limits under load.

A: A explains why stress kept rising with finer mesh and why test failure didn't match the predicted location. B would shift global stiffness and deflection, not create infinite stress at a corner. C can inflate stresses, but you'd still see convergence with mesh refinement. D is a common mistake and does elevate local stress, yet it doesn't produce the classic non-converging pattern tied to a mathematical singularity.

A: A comes straight from first principles: 10 kN over 500 mm² gives 20 MPa, a quick filter for whether the FEA output is even plausible. B mixes nominal stress with local peak stress without geometry to justify it. C quietly changes the load path and invents an effective area. D confuses modulus with stress, a trap when reading colorful contour plots under pressure.

A: A protects you when every decision is traceable. The model must reflect the controlled document. B feels safe but silently departs from the design basis, which auditors dislike. C has no engineering or procedural basis. D guarantees pain later when the mismatch is found without a formal trail.