Computational Engineering / Validation, verification and model assurance
Predict before you analyse
Why the estimate has to come first, and how to make one for a structure that has no formula.
Lesson review: expert verification remains open
This lesson has received a full-context editorial review and a check of its technical explanations. Selected arithmetic has been checked separately. This is not independent engineering sign-off.
Checks needed for project use
- A computational engineering reviewer must check algorithms, convergence claims, statistical assumptions and numerical examples beyond the cases explicitly recorded.
- Confirm truss web layout and diagonal angle in the 24 m worked example; 63.4° is not established by the stated eight panels and 3 m depth.
- Formal checking categories and project acceptance criteria require the responsible engineer.
These are educational models and exercises. Check their assumptions before using the results. The full technical content and every diagram have not yet been independently verified.
Concrete and steel calculations retain their stated first-generation Eurocode teaching basis. Use the standards and amendments specified for your project; do not mix generations without an appropriate design basis.
Read the lesson register and outstanding review notesWhy this matters
Record an independent estimate before viewing the analysis when practical. It gives a clear comparison and reduces the temptation to adjust expectations to the output. Checks made afterwards can still be valuable; retain their assumptions and method so that another engineer can assess their independence.
By the end of this lesson you should be able to
- Explain why prediction must precede analysis rather than follow it
- Reduce a complex structure to one that has a formula
- Predict a deflected shape and a bending moment diagram qualitatively
- Judge how close an estimate needs to be to count as agreement
Simplify for an independent estimate
Use a model whose assumptions and limitations you can explain. A truss can often be screened with chord force approximately M/d and diagonal force approximately V/sin θ. Exact member forces require an appropriate section or joint equilibrium calculation for the actual web arrangement.
A building may be represented as a cantilever for a first estimate of overturning and drift, with an effective stiffness appropriate to its load path. Strip models for slabs, subframes for frames and beam or arch models for shells can provide useful comparisons, but are not automatically upper or lower bounds.
Sketch likely deflections, moment signs and force paths as well as estimating magnitudes. Choose comparison tolerances according to the quantity, simplification and consequence of error. A universal 10% or factor-of-two allowance is not appropriate.
Investigate disagreements in both the estimate and the model. Possible causes include different assumptions, implementation errors and software limitations. Record the explanation and any resulting correction.
Worked example
Predicting a truss before modelling it
Given
- A parallel-chord truss, 24 m span, 3 m deep, 8 panels
- Point loads of 40 kN at each of the 7 interior bottom-chord nodes
- Diagonals at 63.4° to the horizontal
Find
Predictions for the peak chord force and the end diagonal force, before building anything
Practice
A 30 m truss, 3.5 m deep, carries a midspan moment of 1 400 kN·m. Predict the peak chord force in kN.
Practice
A 12-storey building carries a wind load of 180 kN at each floor, with floors at 3.6 m centres starting at 3.6 m. Predict the base overturning moment in kN·m. (The sum of 1 to 12 is 78.)
Check yourself
Why must a prediction be recorded before the analysis runs, rather than simply thought about?
Check yourself
A prediction and a model disagree by a factor of three. What is the right first response?
Summary
- Record an early prediction when practical and document the independence of later checks
- Reduce a dimension until the structure has a formula
- Predict the shape and the signs, not only the magnitude
- Decide what counts as agreement before you compare
- An unexplained disagreement stays open until it is closed in writing