Module 19 · Lesson 19.1
What governs where
Eighteen modules have each taught a check. This lesson puts four members side by side and asks which check actually decides each one — and the answers are all different.
Why this matters
A course is made of modules, and a module teaches one thing at a time. That is the only way to learn the material, and it has one serious side effect: it makes every check look equally likely to matter.
They are not. In a real building, each member type has a limit state that decides it, and those limit states are different for different members. A designer who has absorbed the modules but not the pattern will check everything for bending, size everything for strength, and produce a building that is wrong in several different directions at the same time.
This lesson puts four members from one building next to each other and asks a single question of each: what actually decides this?
By the end of this lesson you should be able to
- Name the governing check for each member type and say why
- Read the gap between the top two checks, and act on it
- Recognise a member carrying steel that is not working
- Say why each governing limit state has a different remedy
What you should already know
- Deflection and serviceability (Module 5)
- Bending, shear and lateral-torsional buckling (Modules 6 and 8)
- Flexural buckling of columns (Module 7)
- Fatigue (Module 17)
Four members, one building
The building is a three-storey industrial and office structure: 9 m floor beams on a braced frame, with a crane bay along one side. Nothing exotic; the sort of thing that gets built constantly.
Take one member of each kind and assemble every check that applies.
The floor beam. Bending at 39%, shear at 14%, lateral-torsional buckling at 40%, deflection at 56%. Deflection governs, and by a clear margin. Nothing here is close to failing — but if the span grows, or the loading rises, it is deflection that arrives first.
The column. Flexural buckling at 88%, bending at 13%. The column is a stability problem, and it is the tightest member in the building.
The brace. Flexural buckling at 33%. It is nowhere near working, which is normal: braces are usually sized by minimum slenderness, by the connection, or simply by what is available, rather than by the force in them.
The crane girder. Bending at 54%, fatigue at 335%. It fails, and it fails on a check that has nothing to do with strength.
Three different limit states govern across four members. Each one has a different remedy, and applying the wrong remedy is worse than useless — it costs money and moves nothing.
Notice also that two of the four members are running below 60%: the brace at 33% and the floor beam at 56%. A low utilisation is information rather than a verdict. Neither of those members is sized by the force in it — one by slenderness and availability, the other by a deflection limit — and both are perfectly normal. What a low utilisation asks for is a reason, not a smaller section.
Worked example
The same beam, decided by three different checks
Given
- A 610×229×113 UB floor beam carrying 30 kN/m
- Deflection limited to span/360
- Restraint spacing and span both variable
Find
Which check governs, and what changes it
Try it
What governs where
Four members in one building. Each bar is a member's governing check, and the label on it is which limit state that is. They are not the same limit state — which is the whole argument of this module.
Inspect a member
- Distinct governing limit states
- 3
- Which ones
- Deflection, Fatigue, Flexural buckling
- All members pass
- NO
- Failing
- crane-girder
- Carrying steel that is not working
- floor-beam, brace
- — inspecting —
- Floor beam
- Deflection
- 56 %
- Lateral-torsional buckling
- 40 %
- Bending
- 39 %
- Shear
- 14 %
- Gap to the next check
- 16 points
- Top two are close
- no
Deflection governs clearly. Deflection is a stiffness problem, and stiffness comes from depth. A stronger steel does nothing: E is the same for every grade. Go deeper, or accept a longer span with a different floor system.
Things worth trying
- Start at the defaults — 9 m span, 3 m restraint spacing, category 71 weld. Read the top chart first: three different limit states are governing across four members.
- That is the argument of the module in one picture. A designer who checks everything for bending has made four different mistakes, not one.
- Inspect the floor beam. At 9 m it is not a bending problem at all — deflection governs, and bending is well behind. Read the advice: stiffness comes from depth, and a stronger steel does nothing because E is the same for every grade.
- Now walk the span out to 15 m. Deflection grows with the fourth power of span while bending grows with the square, so the gap widens rather than closing.
- Bring the span back to 9 m and take the restraint spacing out to 9 m. LTB takes over from deflection — a completely different remedy, reached without touching the section.
- Find a restraint spacing where LTB and deflection are within ten points of each other. The verdict changes: this is now a section problem, because relieving one hands the governing role straight to the other.
- Inspect the brace. It runs at about a third, and the audit flags it as carrying steel that is not working — braces are usually sized by minimum slenderness or by what is available, not by force.
- Note that the FLOOR BEAM is flagged too, at 56%. A low utilisation is information, not a verdict: a beam governed by deflection is often under 60% on every strength check, and that is not waste — it is what a serviceability-governed member looks like.
- Inspect the crane girder at category 71. Fatigue is at 335% while bending sits at 54% — it fails by a factor of three on a check that is not bending at all.
- Now raise the weld detail category to 125. Fatigue falls to 61% and the girder passes. The section never changed; only the quality of one weld detail did.
Practice
A floor beam has bending at 82%, shear at 31%, LTB at 79% and deflection at 44%. What is the gap between the governing check and the next tightest, in percentage points?
Check yourself
A brace in the stability system is running at 33% of its buckling resistance. What does this most likely indicate?
Summary
- Four members in one building were governed by three different limit states
- The 9 m floor beam is a deflection problem at 56%, with bending at 39%
- Taking the restraints to 9 m hands the governing role to LTB at 89% — the beam never changed
- Near 6 m spacing the two are within a point: a section problem, not a single-check problem
- Read the GAP as well as the governor — it decides whether there is one thing to fix
- Four of the five common remedies do not match the instinct to use a bigger, stronger section
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