Module 13
Validation, verification and model assurance
Two different questions that get the same word. Predicting before analysing, the three-stage check, and a record somebody else can review.
What this module covers
- State the difference between validation and verification, and give an example of failing each
- Predict a result before running the analysis, and say why the order matters
- Apply the three-stage check: look and think, check another way, check independently
- Choose a simplification that gives an independent estimate
- Complete a model record that a reviewer can actually use
- Recognise when the software itself is invalid for the problem
Lessons
Validation and verification, separated properly — and why a model that passes one can fail the other completely.
Start lesson →Why the estimate has to come first, and how to make one for a structure that has no formula.
Start lesson →Look and think; check another way; check independently. In that order, because they cost different amounts and catch different things.
Start lesson →- 13.4
The model record
What has to be written down for a model to be reviewable, and why the exclusions matter as much as the inclusions.
Start lesson →
Module checkpoint
Check what you have taken in
3 questions
Question 1
A model of a floor is analysed, and the engineer then hand-calculates the midspan moment and finds it agrees within 3 %. What has this established?
Question 2
A 10 m beam carries 30 kN/m. A model reports end moments of 250 kN·m and a midspan moment of 125 kN·m. Compute the free moment wL²/8 in kN·m, and say whether the model is self-consistent.
Question 3
A frame's beams are 8 m and carry 24 kN/m. If the columns were infinitely stiff, what end moment would the beams carry, in kN·m? (Fully fixed ends give wL²/12.)