Technical and Engineering Analysis
The Technical and Engineering Analysis domain tests the technical core of consulting practice: defining a design basis, selecting and applying codes, performing and checking calculations, validating models, and communicating assumptions and uncertainty.
Core rules and concepts
- Record the design basis before calculating: functional requirements, loads, environmental conditions, interfaces, applicable codes, tolerances, and exclusions.
- Apply load and resistance factors exactly as the governing code requires. Strength design and allowable-stress design use different factors and must not be mixed.
- Factor of safety = capacity / demand. Show the governing load combination, not only the service-level case, and check the controlling limit state.
- Maintain dimensional consistency and significant figures. Convert units explicitly and label them in every step; a unit error can survive a plausible-looking number.
- Verify and validate software: verification checks the model against theory or hand calculation; validation checks that the model represents the real system. Independent checking and peer review are standard quality controls.
- Document assumptions, bounding cases, sensitivity, and any deviation from code. Vendor data and proprietary software output require engineering review.
- Consider constructability, operability, maintainability, and lifecycle cost, not only analytical elegance.
Worked example
A member has a factored demand of 168 kN and a factored resistance of 180 kN. Utilization = 168 / 180 = 0.93, so the member passes with 7% margin. If the demand were unfactored at 120 kN and the nominal capacity were 200 kN, the service-level factor of safety would be 1.67, but that does not replace the code check. The governing check is the factored combination; report both the controlling ratio and the assumptions behind the loads.
Traps that cost marks
Mixing allowable and ultimate values; dropping units mid-calculation; assuming software output is correct without verification; using unverified vendor data; ignoring second-order or dynamic effects; reporting excessive precision; leaving assumptions undocumented.