FEA Crane Hook Design & Validation

The Question

Can an FEA-optimised crane hook — validated against hand calculations — be manufactured and physically proven to safely carry a 30kN load without yielding, at minimum mass?

Why It Matters

Lifting equipment is safety-critical: a crane hook must reliably carry its rated load without yielding, while minimising mass and material cost. This project follows the same design-verification loop used in real structural certification work — simulate, validate the simulation independently, manufacture, then physically test to confirm the prediction — rather than trusting a single FEA result in isolation.

What I Did

Validated the ANSYS Workbench FEA methodology against hand calculations on a known test specimen before trusting it on the real design
Designed the hook geometry in Autodesk Inventor, to a fixed spec: withstand a 30kN static load without yielding, fit specified bolt/pin mounts, minimum mass
Iterated through 7 design variants in ANSYS Workbench, tracking yield-load-to-mass ratio at each step to converge on an optimised geometry
Cross-checked the final design's predicted stress against independent hand calculations using stress-concentration factors
Had the final design laser-cut from 8mm Strenx 700MC high-strength steel
Physically tested the manufactured hook to destruction and reconciled the real result against the FEA prediction

The first four design iterations, each refining geometry based on the previous round's stress distribution and yield-load-to-mass ratio:

ANSYS equivalent stress contour for Hook design iteration v1 — v1 — Yield load/mass ratio: 28.06

ANSYS equivalent stress contour for Hook design iteration v2 — v2 — Yield load/mass ratio: 28.00

ANSYS equivalent stress contour for Hook design iteration v3 — v3 — Yield load/mass ratio: 39.50

ANSYS equivalent stress contour for Hook design iteration v4 — v4 — Yield load/mass ratio: 38.17

ANSYS Workbench equivalent stress contour and boundary condition setup for the final chosen hook design, Hook v5 — The final selected design (Hook v5) after 5 rounds of iteration — chosen for the best balance of yield load to mass across all 7 variants tested.

Key Result

Load vs extension graph from the physical tensile test, showing yield load of 42kN and failure load of 52.62kN — Physical load-extension test result on the manufactured hook — failure occurred at 52.62 kN, comfortably clear of the 30kN design load.

Photograph of the manufactured steel hook specimen at the moment just before complete fracture during the destructive tensile test — The manufactured specimen moments before complete fracture at the designed fail-safe point in the hook's neck.

FEA predicted a yield load of 34.6kN at 700g; the physically tested hook yielded at ~32.9–42kN (depending on method used) and failed at 52.62kN — within single-digit percent of the FEA prediction, and comfortably exceeding the 30kN design requirement.

Impact / What's Next

Beyond meeting the spec, the project's real value was in the reconciliation step: rather than treating the ~5% FEA-vs-test discrepancy as noise, I traced it to a specific, quantified cause — the yield stress assumed in ANSYS (700MPa) versus the steel's actual observed yield stress (660MPa) — and distinguished it from the far larger error introduced by simplified hand-calculation assumptions (over 200% discrepancy, due to modelling the hook as a two-notch flat plate rather than its true single-notch geometry). That habit of quantifying and explaining every source of error, rather than just reporting a result, is the same discipline this DPhil's CFD-vs-experiment validation work relies on.