Module 12 · Lesson 12.1
The checks that come first
Four things to do before reading any result, none of which takes longer than a minute.
Why this matters
Most modelling faults are found in the first minute or not at all. Not because the later checks are weak, but because once you start reading results you start explaining them, and a plausible explanation for a wrong number is easier to construct than to resist.
The four checks below cost almost nothing and are done before the results are opened. They are ordered so that each one, if it fails, tells you where to look next.
By the end of this lesson you should be able to
- Run the equilibrium check and interpret the residual correctly
- Use the deflected shape as a qualitative test rather than as reassurance
- Compare quantities against the plausible bands
- Read a solver's warnings rather than dismissing them
What you should already know
- The stiffness method and where reactions come from (Module 9)
- Deflected shapes and load paths — Structural Behaviour teaches the qualitative reasoning this lesson uses
Check 1: total loads against total reactions
Sum every applied load. Sum every reaction. They must cancel, direction by direction.
In exact arithmetic the residual is zero. In floating point it is a rounding-level number — parts per billion of the applied load in a healthy model. A residual you can see is not rounding, and there are only two ways it arises: something is not connected, or the conditioning is so poor that the digits carrying the answer have been thrown away.
This check takes one subtraction. It catches lost members, orphaned nodes, coincident nodes, loads applied outside the model, and severe ill-conditioning. Nothing else in structural computing has that ratio of cost to coverage.
Run it on every model. Run it again after every change. A change that breaks the balance is a change that broke something.
Check 2: does it deflect the right way
Sketch the deflected shape you expect before looking at the model's. Then compare.
The value is in the order. Looking first and then judging is not a check — it is a rationalisation, and a model that sags where it should hog will be explained away by anyone who has already seen it sag.
What to look for:
- Direction. Does it go the way the load pushes?
- Symmetry. A symmetric structure under symmetric load must deflect symmetrically. Anything else is a modelling asymmetry you did not intend.
- Continuity. Look for discontinuities — a jump where the structure is continuous means a connectivity fault.
- Magnitude. Note the automatic magnification factor the software chose. A small factor means the real deflections are large.
Check 3: is it the right order of magnitude
Some quantities have bands that real structures occupy. Knowing them is faster than re-deriving anything:
- Span over deflection at serviceability: 200 to 1 000.
- Building floor natural frequency: 4 to 10 Hz.
- Steel unit weight: 77 kN/m³ — if the model's self-weight implies anything else, the density field is in the wrong unit.
- Concrete short-term modulus: 27 to 39 GPa, and the long-term value is roughly a third of it. Two models of the same structure differing by that factor have almost certainly disagreed about which to use.
- Structural steel modulus: 210 GPa, one of the few genuinely fixed numbers in the subject.
A quantity outside its band is an input error until proved otherwise.
Check 4: read the warnings
Solvers warn about things their authors anticipated. Those are not the dangerous faults — the dangerous ones run clean — but they are free information and they are routinely dismissed.
Three levels, and they mean different things:
- A warning says you have modelled something unusual. Sometimes you meant to. Decide, and record the decision.
- A severe warning says something is probably wrong. Act on it or write down why not.
- An error stops the analysis. You have no choice.
The habit worth building is to read them before looking at results, for the same reason as the deflected shape: afterwards, they get explained away.
Worked example
Four checks on a model that looks fine
Given
- A six-storey frame, 32 m × 18 m on plan
- Design permanent plus imposed load at serviceability: 9.5 kN/m²
- Model reports total vertical reaction 32 780 kN
- Maximum floor deflection reported: 11 mm on a 9 m span
- First natural frequency reported: 6.8 Hz
- One severe warning: 'element 4471 length below tolerance'
Find
Whether the results can be used
Practice
A model of a 24 m × 15 m floor plate applies 7.5 kN/m² and reports a total vertical reaction of 2 610 kN. What percentage of the applied load is unaccounted for?
Check yourself
A model's total loads and reactions balance to within 10⁻¹¹ of the applied load, and the deflected shape looks entirely sensible. What has been established?
Worked example
Three minutes on a model you have never seen
Given
- A colleague sends a model and asks whether the results look reasonable
Find
What to do in the first three minutes
Check yourself
Why is looking at the deflected shape a genuine engineering check rather than a formality?
Summary
- Total loads against total reactions, every model, every time, before reading anything
- Predict the deflected shape before looking, or it is not a check
- A quantity outside its plausible band is an input error until proved otherwise
- Read the warnings before the results, for the same reason
- Passing three of the four checks means one problem is still hidden
This is educational material. It uses simplified examples to teach principles, and must not be relied on for real design or safety-critical decisions. Module overview and checkpoint