STEELWORK / CON4334
Learn

Foundations: understand the question before using a formula

Chinese–English terminology

1. Units, powers and the calculator Always examined

Simple explanation: Keep the units on the same scale

Changing units is like changing the ruler, not the object.

Keep the units on the same scale — Changing units is like changing the ruler, not the object.
Sketch, not to scale; use the original dimensions.
  1. 1kN=1,000N; 1m=1,000mm.
  2. Area needs two length conversions; inertia needs four.
  3. Convert before substituting, then check the final unit.

Remember: A kN·m moment becomes 1,000,000 N·mm, not 1,000.

Related concept and full method

Force is a push or pull, measured in newtons (N); 1 kN=1000 N1\text{ kN}=1000\text{ N}. Moment = force × perpendicular distance, in N·mm or kN·m. Stress = force ÷ area, in N/mm² (= MPa). Section properties (area, modulus, second moment) are geometry, not forces. The key to the conversions below: convert the length first, then square or cube it.

QuantityConversionWhy the power matters
Length1m=1,000mm;1cm=10mmConvert length before squaring or cubing.
Area1cm2=100mm2(10mm)2=100mm2。
Section modulus1cm3=1,000mm3Used in moment capacity pyS or pyZ.
Second moment of area1cm4=10,000mm4Used in EI and deflection.
Moment1kN·m=106N·mm1,000N×1,000mm。
Line load1kNm=1NmmBoth numerator and denominator divide by 1,000.
Area loadkNm2×tributary widthm=kNmWidth converts a floor load into a beam load.

Use kN and m for reactions and moments; switch to N and mm for stresses, strengths and deflections, because EE and pyp_y are in N/mm². Write the conversion on the same line as the substitution.

Example: 355Nmm2×1,096cm3×1,000mm3cm3106=389.08kN·m

Calculator tips: wrap the whole fraction in brackets under a square root, e.g. enter (275/355)\sqrt{(275/355)}, not 275/355\sqrt{275}/355. A fourth root is the power 0.25. For v=1/[1+0.05(λ/x)2]0.25v=1/[1+0.05(\lambda/x)^2]^{0.25} go: (λ/x)2(\lambda/x)^2 → ×0.05 and +1 → fourth root → reciprocal. Keep at least four significant figures in between.

Try it yourself. Convert 21,370cm4 and 350kN·m into mm units.

Reveal answer and reasoning

I=21,370×10,000=213,700,000mm4。M=350×1,000,000=350,000,000N·mm。

Animation labUnits and powers1 concept · 2 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. Force, length and stress must use compatible units before substitution.
  2. . An area has two length factors, so .
  3. A section modulus has ; second moment of area has . Use and for cm to mm.
  4. . Divide by to express that moment in .
Animation labUnits and powers1 concept · 11 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. Force, length and stress must use compatible units before substitution.
  2. . An area has two length factors, so .
  3. A section modulus has ; second moment of area has . Use and for cm to mm.
  4. . Divide by to express that moment in .

2. Rearranging and interpolating Always examined

Simple explanation: A value between two table rows

Move the same fraction of the way across both number ranges.

A value between two table rows — Move the same fraction of the way across both number ranges.
Sketch, not to scale; use the original dimensions.
  1. Choose the correct table, curve and strength column first.
  2. Find how far your input lies between the two rows.
  3. Apply that fraction to the change between their answers.

Remember: Teaching example: halfway between outputs 100 and 80 gives 90.

Related concept and full method

An equation’s two sides are equal; whatever you do to one side, do to the other. For example, to find the area from P=pAP=pA, divide both sides by pp: A=P/pA=P/p. Compare capacity and demand in the same units.

Given P=196kN and weld resistance per unit length q=1.05kNmm: L=Pq=196kN1.05kNmm=186.667mm

To choose the number of bolts: n≥P/capacity of one boltn\ge P/\text{capacity of one bolt}, then round up. When checking a design, 1.0061.006 is over 1 — never round it to 1.00 and call it a pass.

Interpolation assumes a straight line between two table values and takes the value in proportion. First choose the right column (correct steel strength, correct curve), then interpolate between rows. If xx lies between x0x_0 and x1x_1 with table values y0y_0 and y1y_1, the fraction is f=(x−x0)/(x1−x0)f=(x-x_0)/(x_1-x_0):

y=y0+f(y1−y0)

Made-up example (for practice only, not real table values): λ=70→200\lambda=70\to200 and λ=80→180\lambda=80\to180; find λ=74\lambda=74: f=4/10=0.4f=4/10=0.4, strength =200+0.4×(180−200)=192 N/mm2=200+0.4\times(180-200)=192\text{ N/mm}^2. Check: 192 lies between 180 and 200 ✓.

Try it yourself. Using that invented table, find the value at 77.

Reveal answer and reasoning

f=710=0.7; strength =200−0.7×20=186Nmm2. It lies between 180 and 200.

Animation labInterpolate between two table rows2 concepts · 2 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. Select the correct curve, grade and strength column before choosing the two bracketing rows.
  2. . The slider shows the fraction between the two endpoints.
  3. . A decreasing table needs a negative change in y.
  4. The example uses illustrative outputs 100 and 80. At the result is 90; do not extrapolate missing rows.
Animation labRearrange a capacity equation2 concepts · 7 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. For , P is force, p is stress and A is area. Here we want A.
  2. Divide both sides by the same non-zero stress p.
  3. The p on the right cancels, leaving .
  4. gives . This is required area, not a selected section or a complete design check.

3. Load path and reactions Always examined

Simple explanation: Why reactions balance the loads

Think of a seesaw that must neither fall nor turn.

Why reactions balance the loads — Think of a seesaw that must neither fall nor turn.
Sketch, not to scale; use the original dimensions.
  1. Upward and downward forces must balance.
  2. Take moments about a support to eliminate its reaction.
  3. Use perpendicular distance from the point to the force line.

Remember: A lateral restraint is not automatically a vertical support.

Related concept and full method

A slab carries an area load and passes reactions to secondary beams; they pass reactions to main beams; main beams to columns; columns push down on the foundations. A beam-end reaction pushes up on that beam and pushes down, by the same amount, on the member supporting it. Do not draw both arrows on the same beam, or you count it twice.

A free-body diagram draws one member on its own and replaces its surroundings with support reactions. A pin stops movement but allows rotation; a roller gives one reaction at right angles to its surface; a fixed support also stops rotation and can provide a moment. A lateral restraint only stops the flange moving sideways; it is not a vertical support.

The course’s simply supported beam has a pin at one end and a roller at the other, carrying vertical loads. If the loading is symmetric, the two reactions are equal. If not, use the two equilibrium equations below (anticlockwise positive here; any consistent sign convention gives the same answer).

Σvertical forces=0: RA+RB=total downward loadΣmoments about A=0: RBL=Σ(Piai)+wL(L2)

LL is the support spacing (m), PiP_i a point load (kN), aia_i its distance from A (m), and ww a full-span uniform load (kN/m). For overall equilibrium a UDL can be replaced by a single force wLwL at its middle; but for the shear and moment at a point, keep it spread out.

The tributary width is the strip of slab whose load reaches a beam. For a one-way slab with equal beam spacing ss, an interior beam takes s/2s/2 from each side, ss in total; an edge beam takes s/2s/2 from one side. Check the slab span direction first; do not take a dimension as the tributary width just because it is the nearest number.

Try it yourself. Invented beam: L=8m; a 60kN load at 5m from A; w=4kNm over the full span. Find reactions.

Hint

Take moments about A so its unknown reaction has zero lever arm.

Reveal answer and reasoning

RB=60×5+(4×8)×48=53.5kN. RA=60+32−53.5=38.5kN. The two reactions sum to 92kN, equal to total downward load.

Animation labBalance reactions and moments1 concept · 2 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. This demonstrator is a simply supported 6 m beam with a 60 kN point load; it is not the page’s original loading diagram.
  2. . Moving the load towards B increases .
  3. . The two upward reactions must sum to P.
  4. With the load at midspan, . End couples, UDLs and overhangs require their own equilibrium terms.
Animation labFollow the floor load in 3D2 concepts · 3 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. The floor carries pressure in . The highlighted strip belongs to one secondary beam.
  2. Multiply pressure by tributary width: . The illustration uses .
  3. A primary beam receives the secondary beam reaction at their connection, not a new full-span UDL.
  4. Trace reactions down to columns and foundations. Count each loaded area once.

4. Shear force, bending moment and deflection Always examined

Simple explanation: What shear is trying to do

Imagine cutting the beam: shear stops the two cut faces sliding past each other.

What shear is trying to do — Imagine cutting the beam: shear stops the two cut faces sliding past each other.
Sketch, not to scale; use the original dimensions.
  1. Find the design shear from the load analysis.
  2. Use the section’s applicable shear area and strength.
  3. Compare demand with resistance in the same force units.

Remember: The web usually carries much of the vertical shear in an I-section.

Related concept and full method

Shear VV is the internal sideways force needed to keep either part of a cut beam in balance. Moment MM is the internal turning effect at the cut. Deflection δ\delta is how far the beam has moved from where it started. At a cut xx from A (count only the loads left of the cut):

Cut at distance x from A: V(x)=RA−wx−ΣPiInclude only point loads to the left of the cut.M(x)=RAx−wx22−ΣPi(x−ai)Again include only loads to the left of the cut.

Under a UDL the shear diagram is a sloping straight line and the moment diagram a parabola. At a point load the shear jumps, while the moment is continuous with a sharp corner. An applied couple makes the moment diagram jump. Between point loads the maximum moment is where V=0V=0; also check supports, load points and the roots of overhangs. The maximum moment is not always under a point load.

Full-span UDL alone: Vmax=wL2;Mmax=wL28Midspan point load alone: Vmax=P2;Mmax=PL4

These shortcuts assume a simply supported beam and the stated load position; for an off-centre load or an overhang, use equilibrium. In the elastic range, results from different loads at the same position can be added (superposition); adding maxima from different positions does not generally give the true maximum.

For a simply supported beam of constant section (constant EE and II):

δUDL,midspan=5wL4384EIδpoint,midspan=PL348EIThe second expression applies only to a point load at midspan.

E=205 000 N/mm2E=205\,000\text{ N/mm}^2 is the course Young’s modulus (how stiff the material is); II is the second moment of area about the bending axis (how stiff the shape is), in mm⁴. The course deflection check normally uses the unfactored imposed load: load factors enlarge the demand for strength checks, not the beam’s stiffness.

Try it yourself. A simply supported beam has w=10kNm and L=6m. What are its maximum shear and moment?

Reveal answer and reasoning

V=10×62=30kN. M=10×628=45kN·m. The moment has one extra length factor, so it cannot have units kN.

Animation labBalance reactions and moments3 concepts · 2 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. This demonstrator is a simply supported 6 m beam with a 60 kN point load; it is not the page’s original loading diagram.
  2. . Moving the load towards B increases .
  3. . The two upward reactions must sum to P.
  4. With the load at midspan, . End couples, UDLs and overhangs require their own equilibrium terms.
Concept map: Load path: from the weight of the slab to the shear and moment diagrams

Concept map

Load path: from the weight of the slab to the shear and moment diagrams

Every calculation question starts on this route: area load → load per metre → reactions → shear → moment → compare with capacity. Where the maximum moment is, and how a reaction passes to the next member, are both on this map.

Text version (from cause to result)

  1. Cause

    Area load (kN/m²)

    The slab’s dead and imposed loads per square metre. For design they are factored: 1.4gk+1.6qk1.4g_k+1.6q_k.

  2. What changes itWhere marks are lost

    bb Tributary width

    The strip of slab whose load reaches a beam. With equal spacing ss: an interior beam takes s/2s/2 from each side, ss in total; an edge beam takes s/2s/2. Check the slab span direction first.

  3. Key idea

    ww Load per metre of beam

    Area load × tributary width gives the load along each metre of beam (kN/m), plus the beam’s self-weight.

    Area load × width → line load
  4. Key idea

    Free-body diagram: draw one beam on its own

    Replace the surroundings with support reactions. A pin stops movement but allows rotation; a roller gives one reaction at right angles to its surface; a lateral restraint is not a vertical support.

  5. Key ideaWhere marks are lost

    RAR_A, RBR_B Reactions: take moments

    Take moments about one support for the other reaction, then use vertical equilibrium. For overall equilibrium a UDL can be replaced by one force wLwL at its middle.

    Formula

    ∑MA=0∑Fy=0\textstyle\sum M_A=0\qquad\sum F_y=0
  6. Key idea

    VV Shear-force diagram

    Cut the beam anywhere: the sum of all forces left of the cut is the shear. It jumps at a point load and slopes under a UDL.

    Shear stops the two sides of a cut sliding
  7. Key ideaWhere marks are lost

    MM Bending-moment diagram

    Moments about the cut of all forces to its left. The change in moment = the area under the shear diagram: a parabola under a UDL, a sharp corner at a point load.

  8. ResultWhere marks are lost

    Shear changes sign → maximum moment

    Because the moment is the running area under the shear, it stops growing where the shear is zero (changes sign) — that is the maximum. Also check supports, load points and the root of any overhang.

  9. How design deals with itWhere marks are lost

    A reaction becomes the next member’s load

    A beam-end reaction pushes up on that beam and down, by the same amount, on the member below: slab → secondary beam → main beam → column → foundation.

  10. How design deals with it

    Demand ≤ capacity

    The largest shear and moment are the “demand”, compared with VcV_c and McM_c (or MbM_b for LTB). Worked numbers: floor-beam map.

  11. Exam answer

    Exam: label every value

    Label the shear on both sides of every load point and the moment at every load point and the maximum; show the RA+RBR_A+R_B check. The first 7 marks of the LTB question are here.

Tour steps

  1. 1. From area load to load per metre

    Area load × tributary width → load per metre (plus self-weight).

  2. 2. Draw one beam, find the reactions

    The free-body diagram shows only support reactions; take moments, then check the sum.

  3. 3. Shear and moment

    Shear = forces left of the cut; moment = area under the shear diagram.

  4. 4. The load keeps travelling down

    This beam’s reaction is the load the next member receives.

  5. 5. Check it and write it up

    Compare the largest VV and MM with the capacities; label every point on the diagrams.

Animation labBuild the shear and moment diagrams3 concepts · 7 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. For the illustrative 60 kN point load on a 6 m simply supported beam, balance the reactions before making a cut.
  2. and before the point load. Moment grows linearly.
  3. Shear jumps down by P. To its right, and .
  4. For this case, moment is continuous and returns to zero at B. A concentrated applied couple instead creates a moment jump.

5. Steel: strength, stiffness and ductility Background

Source: Chapter 1 pp. 2–3; L1 slides 3 and 10.Steel is mostly iron (about 98%) with small additions: carbon up to 0.25%, manganese up to 1.6%, silicon 0.1–0.5%, sulphur and phosphorus up to 0.05% each. More carbon raises strength but lowers ductility and weldability, which is why structural steel limits its carbon content.

Yield strength: the stress at which noticeable permanent deformation starts. Stiffness: resistance to elastic deformation — a higher steel grade does not give a higher EE. Ductility: the ability to deform a lot before breaking. Weldability: whether good welded joints can be made with suitable procedures. A beam can be strong enough and still sag too much.

The course uses S275, S355 and S460 (European grades); Chinese grades are Q235, Q345, Q390 and Q420. The grade name is not automatically the design strength for every thickness; read the table by thickness (section 18).

Animation labElasticity, yielding and ductility1 concept

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. Stress is force divided by area. Strain measures change in length relative to original length.
  2. Before yielding, stress is approximately . E controls the elastic slope.
  3. Further strain includes permanent deformation. Higher yield strength does not by itself increase E.
  4. Strength, stiffness and ductility answer different questions. This schematic is not a measured stress–strain curve.

6. Fatigue: repeated loading slowly grows cracks Background

Source: Chapter 1 pp. 2 and 14.Fatigue is failure by cracks growing step by step under repeated changes of stress. Typical examples are crane girders and bridges. It is different from one overload: loads that each look harmless can still grow a crack when repeated many times. Avoid sudden changes of section and stress concentrations. Ordinary wind fluctuations usually do not need a fatigue design unless there are very many stress cycles.

Model answer: “Fatigue is failure caused by progressive crack growth under fluctuating stress. Avoid abrupt section changes and stress concentrations, and detail welded connections appropriately.”

Animation labRepeated loading grows a crack1 concept

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. A notch or abrupt detail can concentrate stress near the connection.
  2. Many load cycles can initiate a crack even when one application does not cause static failure.
  3. The illustrative crack length increases with the animation stages. It is not a fatigue-life calculation.
  4. Use the specified fatigue method and detail category. Avoid abrupt changes and identify whether repeated loading is relevant.

7. Corrosion protection (theory question) Often examined

Source: Chapter 1 p.2.Corrosion removes steel and makes the effective section thinner. Paint and metallic coatings separate the steel from its environment; zinc or aluminium coatings improve resistance to corrosion and abrasion. Structures permanently in water can use cathodic protection. Weathering steel forms a protective oxide layer because of its alloy content — it is a choice of steel, not a paint.

Model answer (two methods): “Apply a protective paint or zinc coating to isolate the steel from corrosive exposure. For continuously immersed structures, use cathodic protection to reduce electrochemical corrosion.” Give two methods that work differently, each with a short reason.

Concept map: Corrosion protection: why steel rusts, and the two ways to stop it

Concept map

Corrosion protection: why steel rusts, and the two ways to stop it

The question asks for two methods. The key is that they must work differently: one keeps water and air away, the other lets another metal corrode instead of the steel.

Text version (from cause to result)

  1. Cause

    Moisture + oxygen

    Steel is exposed to damp air or water.

  2. Key idea

    Electrochemical corrosion: rust

    The iron in the steel reacts electrochemically with water and oxygen and slowly turns into rust.

  3. Result

    Thinner section, less strength

    Rust eats the steel away; the effective section gets thinner and weaker, and durability suffers (a serviceability limit state, SLS).

  4. Key idea

    Two ideas: barrier / sacrifice

    (1) Barrier: keep water and air away from the steel; (2) sacrifice: let another metal (e.g. zinc) corrode instead of the steel.

  5. How design deals with it

    Paint coating (barrier)

    A paint system forms a barrier on the surface that keeps out moisture and oxygen.

  6. How design deals with it

    Hot-dip galvanizing (barrier + sacrifice)

    The zinc layer both isolates the steel and corrodes first, sacrificing itself to protect the steel.

  7. How design deals with it

    Cathodic protection (immersed)

    For structures permanently in water; it reduces electrochemical corrosion.

  8. How design deals with it

    Weathering steel (material choice)

    Its alloy content forms a stable protective oxide layer. It is a choice of steel, not a coating.

  9. Exam answerWhere marks are lost

    Exam: two methods that work differently

    One line each on how it protects. The safest pair: paint (barrier) + galvanizing (sacrifice), or add cathodic protection.

Tour steps

  1. 1. Why does steel rust?

    Moisture + oxygen → electrochemical reaction → rust → a thinner section.

  2. 2. Two ideas

    Cut off the cause (keep water and air out), or find a stand-in (a sacrificial metal).

  3. 3. Four methods

    Paint, galvanizing, cathodic protection, weathering steel — which idea does each belong to?

  4. 4. The exam answer

    Pick two methods that work differently and give one line each on how they protect.

Animation labHow corrosion protection works1 concept

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. Corrosion can consume steel and reduce its effective thickness.
  2. Paint and metallic coatings protect the exposed surface; coating integrity matters.
  3. Edges, damaged coatings and water-trapping details deserve attention. The model shows layers schematically.
  4. Weathering steel and cathodic protection have specific environmental applications; they are not universal substitutes for paint.

8. Fire protection: learn this sketch (theory question) Always examined

Chapter 1 p.3, Figure 1: solid casing, hollow casing and profile casing around an I section.
Chapter 1 p.3, Figure 1: solid casing, hollow casing and profile casing all surround the I-section.Open full-size image

Source: Chapter 1 p.3.Steel loses strength as it heats up; fire protection delays the heating of the load-bearing steel. In the figure above: solid casing — the rectangle around the I-section is completely filled; hollow casing — a box around it with air inside; profile casing — a layer following the steel outline. In your answer, draw the I-section clearly inside each outline and label it. The lecture gives no thickness or fire rating, so do not add numbers.

Try it yourself. Does a hollow fire-protection casing mean the structural steel member itself must be a hollow section?

Reveal answer and reasoning

No. The source figure shows an I section inside a hollow protective enclosure. Distinguish the steel shape from the protection shape.

Concept map: Fire protection: why wrap the steel, and how to sketch the three methods

Concept map

Fire protection: why wrap the steel, and how to sketch the three methods

The most frequent theory question (4 times in 16 papers). Understand the cause — steel softens when hot — and the three methods become three ways of doing the same thing: insulation.

Text version (from cause to result)

  1. Cause

    A fire heats the steel

    Steel conducts heat well, so it heats up quickly in a fire.

  2. Key idea

    Steel softens: strength and stiffness drop

    The hotter the steel, the lower its strength and stiffness. At a critical temperature the member can no longer carry its load.

  3. Result

    Members fail → collapse, less time to escape

    Beams and columns fail early; the structure may collapse before people can get out.

  4. How design deals with it

    The principle: insulate to delay heating

    Fire protection does not stop the steel heating; it gives it an insulating coat that delays it reaching the critical temperature, buying time.

  5. How design deals with it

    Solid casing

    The whole I-section is encased in solid material, e.g. concrete. Sketch: the I-section inside a filled (hatched) rectangle.

  6. How design deals with itWhere marks are lost

    Hollow casing

    Fire boards form a box around the I-section with an air space inside. Sketch: the I-section inside a rectangular outline.

  7. How design deals with it

    Profile casing

    A sprayed coating or intumescent paint forms a thin layer following the I shape. Sketch: a thin layer drawn around the I outline.

  8. Exam answerWhere marks are lost

    Exam: one-line principle + sketch two

    Write the principle in one line, then sketch two methods. Draw the I-section in every sketch and label it. The lecture gives no thickness or fire rating, so do not invent numbers.

    Lecture figure: solid, hollow and profile casing

Tour steps

  1. 1. Why protect against fire?

    Fire → steel heats up → strength and stiffness drop → members fail.

  2. 2. One principle: insulation

    The aim is to delay the steel heating, buying time for escape and fire-fighting.

  3. 3. Three ways to do it

    Solid (filled), hollow (an outline), profile (a thin layer following the shape). The difference is all in the outline you draw.

  4. 4. The exam answer

    One line of principle + two labelled sketches is the skeleton of a full-mark answer.

Animation labSee the fire-protection enclosure1 concept

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. Steel loses strength as it heats. Protection aims to delay heat reaching the section.
  2. Solid casing fills the enclosure around the I-section.
  3. A hollow enclosure surrounds the section while leaving an air space.
  4. Profile casing follows the section outline. The supplied sketch does not determine protection thickness or fire duration.

9. Section shapes, axes and names (theory question) Often examined

Simple explanation: Why I, Z and S are different

The same section has different numbers for different jobs.

Why I, Z and S are different — The same section has different numbers for different jobs.
Sketch, not to scale; use the original dimensions.
  1. I describes how area is spread: it affects elastic stiffness.
  2. Z is the elastic section modulus; S is the plastic section modulus.
  3. Choose the property, axis and units required by the formula.

Remember: Do not substitute S for Z merely because it is larger.

Related concept and full method

Chapter 1 p.4, Figure 2: UB, UC, channel, equal/unequal angles, tee and hollow sections.
Chapter 1 p.4, Figure 2. Labels identify UB (universal beam), UC (universal column), channel, equal/unequal angles, tee cut from UB and hollow sections. Selectable equivalents of the dimension labels follow. These retain the source ranges, not design inputs for a particular calculation: UB D×B is 203×133 to 914×419; UC D×B is 152×152 to 356×406; channel D×B is 76×38 to 432×102; equal angle A×A is 25×25 to 250×250; unequal angle A×B is 40×25 to 200×150; tee B×A is 133×102 to 305×457; circular hollow section D is 21.3 to 457; square hollow section D×D is 20×20 to 400×400; rectangular hollow section D×B is 50×30 to 450×250. The individual arrows identify overall depth, overall width or angle-leg length. Do not measure additional dimensions from drawing scale.Open full-size image

A UB (universal beam) is deep and efficient in bending about its strong axis; a UC (universal column) has wider flanges, suited to compression and bending both ways. Channels and angles are used in bracing, trusses and compound members; a tee can be cut from a UB or UC. CHS, SHS and RHS are circular, square and rectangular hollow sections. In this course x–xx\text{–}x is the strong axis and y–yy\text{–}y the weak axis.

A name such as 356×127×39 UB means roughly 356 mm deep, 127 mm wide and 39 kg per metre. These are nominal; read the real dimensions from the same table row. Self-weight: 39×10/1000=0.39 kN/m39\times10/1000=0.39\text{ kN/m} (with g=10g=10).

Chapter 1 p.5, Figure 3: welded plates forming I/H and box members.
Chapter 1 p.5, Figure 3: separate plates are welded into I-, H- and box sections. The web connects upper and lower flanges; a box is assembled from its surrounding plates.Open full-size image

Built-up sections are plates welded into an I, H or box. In a sketch, show the separate web and flange plates and mark the welds. Compound sections combine rolled sections or add cover plates — e.g. strengthened beams, crane girders, battened and laced columns.

Chapter 1 p.5, Figure 4: compound beam, crane girder, battened and laced members.
Chapter 1 p.5, Figure 4: the labels identify a compound beam, crane girder, battened members and laced members.Open full-size image
Chapter 1 p.6, Figure 5: zed, sigma and lipped channel sections.
Chapter 1 p.6, Figure 5: the labels identify zed (Z), sigma (Σ) and lipped channel sections.Open full-size image

Cold-formed sections are bent from thin sheet and are often used as purlins and sheeting rails; do not apply UB formulas to these thin-walled shapes. If the exam asks for four hot-rolled sections, UB, UC, channel and angle are the safest choice.

Concept map: Hot-rolled sections: the shape decides the job

Concept map

Hot-rolled sections: the shape decides the job

The theory question “sketch and name four hot-rolled sections” has been asked twice. Once you see where the steel sits, it is clear why a UB is a beam and a UC a column — and how ryr_y links to LTB and column buckling.

Text version (from cause to result)

  1. Cause

    Steel far from the axis works hardest

    In bending, steel far from the neutral axis carries the most stress with the longest lever arm. That is why an I shape puts most of its steel in the flanges.

  2. Key idea

    II, ZZ, SS, rr Section properties come from the shape

    II governs stiffness (deflection), ZZ the elastic moment, SS the plastic moment, and r=I/Ar=\sqrt{I/A} buckling. They are not interchangeable.

    Why I, Z and S are not interchangeable
  3. Key idea

    Strong axis x–x, weak axis y–y

    For the same section II and rr about xx are far larger than about yy. In this course x–xx\text{–}x is always the strong axis.

  4. What changes it

    How to read a section name

    356×127×39 UB: about 356 mm deep, 127 mm wide, 39 kg per metre (self-weight 39×10/1000=0.39 kN/m39\times10/1000=0.39\text{ kN/m}). These are nominal; read the real dimensions from the table row.

  5. Result

    UB: deep and narrow → beams

    Very efficient in strong-axis bending; but the narrow flanges give a small ryr_y, so it is prone to LTB and needs lateral restraint (see the LTB concept map).

  6. Result

    UC: wide flanges, width ≈ depth → columns

    Stocky in both directions, so ryr_y is not too small — suited to compression and bending both ways (see the column-buckling map).

  7. Result

    Channel

    “[” shape: one web with both flanges on the same side. Used for bracing, secondary members and compound sections.

  8. Result

    Angle

    “L” shape with equal or unequal legs. Used in bracing, trusses and connection pieces. Connected through one leg it needs a deduction (see the net-area map).

  9. Exam answerWhere marks are lost

    Exam: sketch and name four

    The safest four: UB, UC, channel and angle. Sketch each cross-section, label it and give one line on its use.

    Lecture figure: common rolled sections

Tour steps

  1. 1. Where the steel sits

    The further from the axis the more useful, so the shape sets I,Z,S,rI,Z,S,r.

  2. 2. Strong and weak axes

    Strong about xx, weak about yy — that is what separates beams from columns.

  3. 3. UB for beams, UC for columns

    A UB is deep and narrow — efficient in bending but with a small ryr_y; a UC is stocky both ways.

  4. 4. Channels and angles

    Mostly for bracing, trusses and connections.

  5. 5. The exam answer

    Four labelled sketches, one line of use each.

Animation labExplore section geometry and axes1 concept · 2 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. The flanges are the wide plates; the web connects them. Rotate the I-section to see both.
  2. The same section has different stiffness and resistance about its two principal axes.
  3. I controls elastic curvature; is elastic section modulus. Plastic modulus S comes from plastic stress blocks.
  4. Nominal section labels are not every actual dimension. Keep the row, axis and units together.

10. Limit states: ULS and SLS Often examined

Simple explanation: Strong enough and stiff enough are two questions

A shelf can avoid breaking yet still sag too much.

Strong enough and stiff enough are two questions — A shelf can avoid breaking yet still sag too much.
Sketch, not to scale; use the original dimensions.
  1. ULS checks safety against the relevant failure modes.
  2. SLS checks the specified everyday-use limit.
  3. Use the load case required for each check.

Remember: Passing bending resistance does not prove deflection passes.

Related concept and full method

Source: Chapter 1 p.7, Table 2.1 extract.A limit state is a condition beyond which the structure no longer meets a requirement. ULS is about collapse or damage that cannot be repaired: yielding, rupture, buckling, mechanism, overturning, sliding, uplift, fire, brittle fracture, fatigue. SLS is about use: excessive deflection, vibration, wind-induced oscillation and durability.

The design process: characteristic loads × load factors → work out their effects → compare with design resistance. Imposed load gets a bigger factor (1.6) because it is less predictable. The material safety margin is already inside the course design strengths, so do not apply an extra Eurocode material factor.

For every applicable failure mode, satisfy load demand≤resistance.

Try it yourself. A beam has adequate strength but cracks its plaster finish through excessive movement. Which check has failed?

Reveal answer and reasoning

Serviceability, specifically deflection. Passing ULS does not establish SLS adequacy.

Concept map: Limit states: “will it collapse?” and “does it work?” are two different routes

Concept map

Limit states: “will it collapse?” and “does it work?” are two different routes

ULS and SLS are not only a theory question (asked twice); they decide which load each calculation uses: strength with factored loads, deflection with unfactored loads. This map links the two.

Text version (from cause to result)

  1. Cause

    Limit state

    A condition beyond which the structure no longer meets a design requirement. Design makes sure none of them is reached.

  2. Key idea

    ULS: will it collapse?

    Ultimate limit state: safety — collapse or damage that cannot be repaired.

  3. Key idea

    SLS: does it work in everyday use?

    Serviceability limit state: it does not collapse, but it does not work properly in normal use.

  4. Result

    ULS examples

    Bending or shear failure of a beam (yielding), column buckling, overturning or sliding, brittle fracture, fatigue, fire.

  5. Result

    SLS examples

    Excessive deflection cracking finishes, vibration causing discomfort, wind-induced oscillation, durability (e.g. corrosion).

  6. How design deals with itWhere marks are lost

    1.4Gk+1.6Qk1.4G_k+1.6Q_k ULS uses factored loads

    Characteristic loads × load factors, compared with the design resistance. Imposed load is less predictable, so its factor is larger (1.6). The material safety margin is already inside the course design strength pyp_y — do not apply another.

  7. How design deals with itWhere marks are lost

    QkQ_k SLS uses unfactored loads

    Deflection looks at the real everyday load: imposed load only, unfactored, compared with a limit (L/360L/360 with brittle finishes).

  8. Exam answer

    The floor-beam question tests both

    Shear and bending checks = ULS (factored); the deflection check = SLS (unfactored). Both routes run side by side in one question — see the floor-beam map.

  9. Exam answer

    Theory question: definition + two examples each

    One definition, then two examples each of ULS and SLS.

Tour steps

  1. 1. Two questions

    ULS asks “will it collapse?”; SLS asks “does it work normally?”.

  2. 2. Examples of each

    Collapse, fracture and buckling are ULS; too much deflection, vibration or rust are SLS.

  3. 3. The two routes use different loads

    ULS: 1.4Gk+1.6Qk1.4G_k+1.6Q_k; SLS: the real imposed load QkQ_k, unfactored.

  4. 4. Calculation and theory

    The floor-beam question uses both routes; the theory question needs a definition + two examples each.

Animation labTwo different design questions1 concept · 1 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. ULS compares factored actions with the applicable resistance to yielding, rupture or instability.
  2. SLS checks movement, vibration or another specified use requirement.
  3. A beam can be strong enough but deflect too much. The two checks need their own loads and denominators.
  4. A resistance pass cannot stand in for a serviceability pass. Complete all requested checks.

11. Loads and load factors Always examined

Simple explanation: Why dead and imposed loads stay separate

Keep two shopping baskets until their different multipliers are applied.

Why dead and imposed loads stay separate — Keep two shopping baskets until their different multipliers are applied.
Sketch, not to scale; use the original dimensions.
  1. Put self-weight and permanent finishes in the dead-load basket.
  2. Put the specified use load in the imposed-load basket.
  3. For this course’s stated gravity combination: 1.4G+1.6Q.

Remember: That ULS combination is not the imposed-load deflection load.

Related concept and full method

Source: Chapter 1 pp.8–9 and 13–14; Data File p.1 onward.GkG_k = characteristic dead load (permanent: self-weight, finishes); QkQ_k = characteristic imposed load (people, furniture); WkW_k = wind load. The main adverse combinations are:

Dead + imposed load: 1.4Gk+1.6QkDead + wind load: 1.4Gk+1.4WkDead + imposed + wind load: 1.2Gk+1.2Qk+1.2Wk

These assume every load is unfavourable. If dead load helps resist uplift or overturning (favourable), its factor is 1.0, and a favourable imposed load may be left out. The full table also covers cranes, earth pressure, temperature and accidents — use those only when asked; do not force 1.4 / 1.6 onto everything.

If the question says ultimate / factored / design load, it is already factored — do not factor again. If it gives dead and imposed loads separately, factor each once. If self-weight is already included, do not add it; otherwise self-weight = mass per metre × gg ÷ 1000, as dead load.

Slab dead-load intensity: thickness(m)×unit weight(kNm3)+finishes load(kNm2)Beam ULS line load: (1.4gk+1.6qk)×tributary width+1.4×beam self-weight

Try it yourself. Invented slab: 150mm concrete at 24.5kNm3, finishes 0.5kNm2 and imposed load 3kNm2. Find factored floor intensity.

Reveal answer and reasoning

Concrete: 0.150×24.5=3.675kNm2. Dead total =3.675+0.5=4.175. ULS=1.4×4.175+1.6×3=5.845+4.8=10.645kNm2.

Animation labFrom characteristic to design load2 concepts · 3 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. G is permanent load; Q is imposed load. A surface load and a line load also have different units.
  2. This illustration uses the course gravity case . Other combinations in the original text retain their own factors.
  3. For illustrative , change Q and watch each separate contribution.
  4. Do not carry this ULS total automatically into deflection. Follow the stated SLS load case.
Animation labFrom characteristic to design load2 concepts · 2 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. G is permanent load; Q is imposed load. A surface load and a line load also have different units.
  2. This illustration uses the course gravity case . Other combinations in the original text retain their own factors.
  3. For illustrative , change Q and watch each separate contribution.
  4. Do not carry this ULS total automatically into deflection. Follow the stated SLS load case.

12. Simple, continuous and semi-continuous design Background

Source: Chapter 1 p.10.Simple design treats beam ends as pinned, so there are no continuity moments at the beam ends; sideways loads go to a bracing system or a rigid core. But note: a beam reaction that is off the column centre still bends the column — “pinned beam” does not mean “no moment in the column” (see columns section 8).

Continuous design uses joints stiff and strong enough to pass moments through. Semi-continuous design explicitly models the real stiffness and strength of the joints (from tests or calibrated analysis); you cannot simply assume how “fixed” a joint is.

Animation labRestraints, sway and imperfections2 concepts

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. A frame needs a defined path for horizontal force as well as gravity.
  2. Pinned joints alone do not provide frame moment resistance; sway restraint needs a real structural system.
  3. Diagonal bracing carries horizontal action through axial forces. This sketch does not assign a numerical frame classification.
  4. Use the specified imperfection/notional-force model and critical-load criteria. Avoid counting alternative imperfection models twice.

13. P–Δ and P–δ: why deformation adds moment (theory question) Always examined

Simple explanation: Moving sideways gives the load a new lever arm

Once a compressed member moves sideways, the same compression creates extra moment.

Moving sideways gives the load a new lever arm — Once a compressed member moves sideways, the same compression creates extra moment.
Sketch, not to scale; use the original dimensions.
  1. P–Δ concerns overall frame or storey movement.
  2. P–δ concerns bowing relative to the member’s end line.
  3. Amplify only the first-order moments specified by the method.

Remember: Do not amplify a moment that the question already gives as amplified.

Related concept and full method

Chapter 1 p.11, Figure 6. Δ is global sideways movement; δ is member bow relative to its end chord.
Chapter 1 p.11, Figure 6. Legend: Δ = global sideways movement; δ = member bow relative to the line joining its ends. Other source labels: P is downward axial compression; h is the frame height shown; Le is column effective length. The source lower-case subscript e is retained. The diagram title concerns column effective length and P–Δ and P–δ moments (Column effective length, P–Δ and P–δ moments) . Open full-size image

First-order analysis works out equilibrium on the original, undeformed shape. If an axial compression PP acts through a displaced point, it creates an extra moment P×displacementP\times\text{displacement}. P–Δ goes with frame sway; P–δ with the bowing of a single member. Look at which deformation the sketch shows before naming the effect.

A P–Δ-only second-order analysis includes the frame’s sway but leaves the member bow to be allowed for separately. A P–Δ–δ second-order analysis includes frame deformation, member bow, stiffness change and imperfections. Advanced analysis also includes yielding. The lecture allows first-order results to be used with effective lengths and amplification factors; do not amplify moments that are already amplified.

Model answer: “The analysis establishes equilibrium in the deformed position, including additional moments from axial forces acting through frame sway Δ and member bow δ. Frame/member imperfections and stiffness changes are included in the second-order P–Δ–δ model.”

Concept map: P–Δ and P–δ: why deformation makes the moment bigger

Concept map

P–Δ and P–δ: why deformation makes the moment bigger

One of the most frequent theory questions (3 times in 16 papers). Read from top to bottom: compression plus an offset → the two effects → the snowball → how analysis and design deal with it → the exam answer.

Text version (from cause to result)

  1. Cause

    PP Axial compression

    A column carries a large compression PP. As soon as it does not push straight down the centre line, it adds a moment.

  2. Cause

    Δ\Delta The frame sways

    Horizontal loads such as wind, or a frame that is not perfectly plumb (design assumes a lean of h/200h/200), move the top of a storey sideways by Δ\Delta relative to its base.

    Source: Lecture: global imperfection Δ=h/200\Delta=h/200, or a notional horizontal force of 0.5% of the vertical load.

  3. Cause

    δ\delta The member bows

    No member is perfectly straight (initial bow about L/550L/550 to L/200L/200, depending on the buckling curve), and it bows out by δ\delta between its ends under moment.

  4. Key idea

    P–ΔP\text{–}\Delta P–Δ effect (global)

    The compression acts through the storey sway Δ\Delta and creates an extra moment P×ΔP\times\Delta. It is an effect of the whole frame.

    Left: frame sway Δ; right: member bow δ
  5. Key ideaWhere marks are lost

    P–δP\text{–}\delta P–δ effect (member)

    The compression acts through the member’s own bow δ\delta and creates an extra moment P×δP\times\delta. It is an effect of one member.

  6. Result

    A snowball: more bending, more moment

    The extra moment bends the member more, the offset grows, and so does the extra moment. Deflections increase, stiffness drops, and in the worst case the frame becomes unstable.

  7. How design deals with it

    Second-order analysis: equilibrium on the deformed shape

    First-order analysis writes equilibrium on the original, undeformed shape and misses the extra moments; second-order analysis writes it on the deformed shape and includes them. A P–Δ–δ analysis also includes member bow and imperfections.

  8. What changes itWhere marks are lost

    λcr\lambda_{cr} How far from instability

    The elastic critical load factor: how many times the loads must be multiplied before the frame becomes unstable. λcr≥10\lambda_{cr}\ge10 is non-sway and P–Δ may be simplified; λcr<5\lambda_{cr}<5 is ultra-sensitive and needs second-order or advanced analysis.

    Source: Columns lecture section 5

  9. Exam answer

    Four load–displacement curves

    First-order elastic: a straight line; second-order elastic: stiffness gradually drops; first-order plastic: reaches a plateau (mechanism); second-order elastic-plastic: a lower peak, then the load falls.

    Load–displacement curves of four analysis types

    Source: Short theory: four analysis types (asked once)

  10. How design deals with itWhere marks are lost

    Mamp=k MM_{\text{amp}}=k\,M Amplification factor: add it in one go

    Without a second-order analysis, multiply the first-order moment by an amplification factor to include the extra moment. Exam questions usually give the factor (e.g. 1.1).

    Source: How it is used: columns crash course.

  11. Exam answer

    Exam answer: definition + both effects + sketch

    Write one definition (equilibrium on the deformed shape), one line each for P–Δ and P–δ, and a sketch: a frame swaying by Δ\Delta on the left and a column bowing by δ\delta on the right.

    Lecture figure: P–Δ and P–δ

Tour steps

  1. 1. The cause: compression plus an offset

    Compression alone adds no moment; it must act through an offset. There are two kinds of offset: storey sway Δ\Delta or member bow δ\delta.

  2. 2. The two effects

    Compression × storey sway = P–Δ (global); compression × member bow = P–δ (member).

  3. 3. The snowball

    Extra moment → more deformation → more extra moment. That is why it cannot be ignored.

  4. 4. Analysis

    First-order analysis cannot see it; second-order analysis works on the deformed shape. The four analysis curves may also be asked on their own.

  5. 5. How design deals with it

    Use λcr\lambda_{cr} to judge how sensitive the frame is; in calculations multiply the first-order moment by the amplification factor, once only.

  6. 6. The exam answer

    Definition + P–Δ + P–δ + sketch, about 4 marks. No essay needed.

Animation labSeparate P–Δ from P–δ1 concept

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. The dashed line marks the original frame position; axial compression acts downwards.
  2. Global displacement Δ creates the additional moment PΔ.
  3. Local displacement δ is measured from the line joining the displaced ends, and adds Pδ.
  4. Follow the original analysis route. Do not amplify an already amplified moment or confuse critical-load factor with member slenderness.

14. Imperfections and the notional horizontal force Background

Source: Chapter 1 pp.12–13 and L1 slide 8.Global imperfection: the course uses an equivalent lean of Δ=h/200\Delta=h/200. Alternatively, apply a notional horizontal force equal to 0.5% of the relevant design vertical load. This models the imperfection; it is not a new gravity load.

0.5%=0.5100=0.005Invented demonstration: if the vertical load is 2,000kN, the notional horizontal force is 0.005×2,000=10kN. h=4,000mm→h200=20mmThe final value is equivalent out-of-plumb displacement.

A member’s initial bow depends on its buckling curve: a0,a,b,c,da_0,a,b,c,d give L/550, L/500, L/400, L/300, L/200L/550,\,L/500,\,L/400,\,L/300,\,L/200. When you design with effective lengths and amplification factors, these imperfections are already included; do not add an extra imaginary moment.

Animation labRestraints, sway and imperfections2 concepts · 2 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. A frame needs a defined path for horizontal force as well as gravity.
  2. Pinned joints alone do not provide frame moment resistance; sway restraint needs a real structural system.
  3. Diagonal bracing carries horizontal action through axial forces. This sketch does not assign a numerical frame classification.
  4. Use the specified imperfection/notional-force model and critical-load criteria. Avoid counting alternative imperfection models twice.

15. Strength, overall stability and robustness Background

Source: Chapter 1 pp.14–15.Strength checks compare the demand on members and connections with their resistance. Overall stability prevents overturning, sliding, uplift or uncontrolled sway; bracing, rigid joints, shear walls and cores can provide sideways resistance. Robustness stops accidental local damage spreading into a disproportionate collapse.

Ways to improve robustness: vertical and horizontal ties, resistance to a minimum horizontal load, alternative load paths if a vertical member is removed, and stronger key elements. In an answer, link each measure to what it prevents: e.g. a tie keeps the structure continuous if a support is lost.

Animation labUnderstand an alternative load path1 concept

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. Gravity load passes through beams to their columns.
  2. Local accidental damage interrupts one path. The highlighted column fades to show the interrupted route.
  3. Ties and continuity can provide another load path, subject to design and deformation capacity.
  4. This schematic explains the principle; it does not prove that a particular frame survives column removal.

16. Brittle fracture Background

Source: Chapter 1 p.15.Brittle fracture is sudden cracking under tension with very little plastic deformation. Risk factors: welding, stress concentrations, rapid loading, high stress, thick material and low temperature. Prevention: choose a suitable steel quality and thickness and use good welding practice. In an answer, describe the mechanism and list several factors; do not mix it up with fatigue (repeated loading, slow crack growth).

Animation labBrittle fracture versus yielding1 concept

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. A notch or weld detail can concentrate tensile stress.
  2. Steel quality, thickness, low temperature and rapid loading affect brittle-fracture susceptibility.
  3. Brittle fracture can occur with little plastic warning. It is distinct from progressive fatigue crack growth.
  4. Follow the required steel quality and welding requirements. The animation is qualitative, without an invented fracture threshold.

17. Deflection limits: pick the right row Always examined

Source: Chapter 1 pp.16–17, Table 5.1; Data File p.2.The course normally checks the extra deflection caused by the unfactored imposed load. Simply supported beams with plaster or brittle finishes: limit span/360\text{span}/360; other beams (not purlins or sheeting rails): span/200\text{span}/200; cantilevers: length/180\text{length}/180. Purlins and sheeting rails must suit their cladding.

The full table also has limits for construction stage, composite slabs, frame sway and crane rails. The slide’s H/500H/500 is the row for sway at the top of a frame; storey drift is a different row (storey height/400\text{storey height}/400). In an answer, say which displacement and which load you are checking. Exam questions state the finish condition (usually brittle finishes → L/360L/360).

Limit calculation only: L=7,000mm→L360=19.444mm.

Try it yourself. A beam has factored imposed load 48kN but the original characteristic imposed load is 30kN. Which enters the usual course deflection formula?

Reveal answer and reasoning

30kN. The factor 1.6 belongs to ULS demand; the serviceability check uses the unfactored imposed load.

Animation labSee stiffness and deflection1 concept · 1 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. Use the specified SLS load, span and support arrangement. The demonstrator has a full-span UDL.
  2. The loaded beam bends; the deformation is exaggerated so its shape can be seen.
  3. For a simply supported full-span UDL, . Double L with w, E and I unchanged: δ becomes 16 times as large.
  4. The readout uses , and . Select the finish/support-specific limit from the original table.

18. Design strength pyp_y depends on thickness Always examined

Simple explanation: S355 does not always mean 355

The grade name is a label; the table supplies the design strength.

S355 does not always mean 355 — The grade name is a label; the table supplies the design strength.
Sketch, not to scale; use the original dimensions.
  1. Identify the thickness that governs this check.
  2. Read the matching thickness band within the correct grade.
  3. Use that strength consistently in the calculation.

Remember: For the supplied S355 table, thickness above 16 mm can reduce py.

Related concept and full method

Source: Chapter 1 Tables 3.2/3.3 on pp.18–19; Data File p.1.In the supplied table, the S355 design strength depends on thickness:

Thickness intervalpy(Nmm2)
t≤16mm355
16<t≤40mm345
40<t≤63mm335
63<t≤80mm325
80<t≤100mm315

For beam bending, choose pyp_y from the flange thickness TT, as the lecture examples do; for a plate use that plate’s thickness; for web bearing use the web’s pywp_{yw}. Keep the different “strengths” apart: pyp_y steel design strength, psp_s bolt shear, pwp_w weld, pcp_c column compression, pbp_b LTB bending — same units, different questions.

Try it yourself. Which strength applies to S355 cover plates 12mm thick and a main plate 20mm thick?

Reveal answer and reasoning

Cover: 355Nmm2 because 12≤16. Main: 345Nmm2 because 16<20≤40. Do not use one value just because both say S355.

Animation labRead a table without losing the keys2 concepts · 2 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. Name the required property: material strength, section property, buckling strength or a moment factor.
  2. Keep section size, steel grade, thickness band, curve and axis as separate lookup keys.
  3. , and require different conversion powers. Do not use adjacent columns interchangeably.
  4. Use bracketing rows within the same valid column. The original page remains the source of all table values.
Animation labRead a table without losing the keys2 concepts · 5 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. Name the required property: material strength, section property, buckling strength or a moment factor.
  2. Keep section size, steel grade, thickness band, curve and axis as separate lookup keys.
  3. , and require different conversion powers. Do not use adjacent columns interchangeably.
  4. Use bracketing rows within the same valid column. The original page remains the source of all table values.

19. Local buckling and the four classes Always examined

Simple explanation: A thin part can wrinkle first

A thin plate may wrinkle before the whole steel member reaches its intended resistance.

A thin part can wrinkle first — A thin plate may wrinkle before the whole steel member reaches its intended resistance.
Sketch, not to scale; use the original dimensions.
  1. Check flange and web slenderness using their own definitions.
  2. Compare each ratio with the correct class limits.
  3. The less favourable element determines the section class.

Remember: Bending limits and uniform-compression limits are different.

Related concept and full method

Source: Chapter 1 pp.19–22; Data File p.1.A thin plate can wrinkle locally before the whole member buckles. Classification compares each plate’s width (or depth) ÷ thickness. The worst plate decides the whole section: a Class 2 flange with a Class 3 web → Class 3 overall. When the lecture says “lowest class” it means the worst behaviour, i.e. the highest class number.

ClassWhat can be developed?Exam implication
1: plasticFull plastic moment and enough rotation for a plastic hingePlastic analysis permitted where otherwise applicable.
2: compactFull plastic moment but limited rotationMay use plastic section resistance, not assume redistribution capability.
3: semi-compactExtreme fibre can reach py; local buckling limits plastic spreadingUse elastic resistance or explicitly permitted effective properties.
4: slenderLocal buckling can occur before extreme fibre reaches pyEffective/reduced section treatment required; gross Class 1 formulas do not apply.
ε=275pyWhen py=355Nmm2: ε=275355=0.880141

For a rolled I/H beam bending about its strong axis, use the Table 7.1 rows “outstand flange, compression due to bending, rolled” and “web, neutral axis at mid-depth”. Flange b/Tb/T limits: 9ε, 10ε, 15ε9\varepsilon,\,10\varepsilon,\,15\varepsilon (Classes 1, 2, 3); web d/td/t: 80ε, 100ε, 120ε80\varepsilon,\,100\varepsilon,\,120\varepsilon. The section table already lists b/Tb/T and d/td/t — use them directly; do not use the overall depth DD in place of the clear web depth dd.

For a web in compression plus bending (e.g. a column), use the general-web row with r1=Fc/(d t pyw)r_1=F_c/(d\,t\,p_{yw}), limited to between −1 and +1. The Class 1 limit is 80ε/(1+r1)80\varepsilon/(1+r_1) but not below 40ε40\varepsilon. Easy rule: if d/t≤40εd/t\le40\varepsilon it is always Class 1. For pure compression the lecture checks flange 13ε13\varepsilon and web 40ε40\varepsilon (this only proves it is not slender).

The beam examples also check d/t≤70εd/t\le70\varepsilon: it means the web will not buckle in shear, so the simple shear formula applies. It is not the 80ε80\varepsilon classification limit, and it does not replace the web bearing / buckling checks at a support.

Try it yourself. For S355 with py=355, a rolled beam has bT=8.5 and dt=60. Find its bending class.

Reveal answer and reasoning

9ε=7.921 and 10ε=8.801: flange is Class 2. 80ε=70.411: web is Class 1. Overall Class 2. The flange controls.

Animation labWhy thin elements buckle locally1 concept · 3 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. The flange outstand and web have different widths, thicknesses and edge support conditions.
  2. A thinner plate can wrinkle locally before the complete member loses stability.
  3. Class 1 allows plastic rotation; Class 2 reaches plastic resistance; Class 3 reaches elastic resistance; Class 4 requires effective properties.
  4. Check every relevant compression element with the supplied limits and stress distribution. The deformation shown is qualitative.
Concept map: Local buckling and the four classes: thin plates wrinkle first, so sections are graded

Concept map

Local buckling and the four classes: thin plates wrinkle first, so sections are graded

Classification is not arithmetic for its own sake: it answers “are the plates thick enough to reach full strength before they wrinkle?”. The answer decides whether you may use SS or only ZZ.

Text version (from cause to result)

  1. Cause

    Plates in compression

    A section is made of plates. In bending the compression flange is squeezed; the web is bent (and in a column also compressed).

  2. Key idea

    Local buckling: thin plates wrinkle

    If a plate is too thin it wrinkles locally before it fully yields, so the section cannot reach its full strength. This is different from the whole member buckling.

    A thin plate in compression wrinkles locally
  3. What changes itWhere marks are lost

    b/Tb/T, d/td/t Width-to-thickness ratios

    The larger the ratio, the thinner the plate and the easier it wrinkles. The section table already lists them.

  4. What changes it

    ε\varepsilon Stronger steel, stricter limits

    ε=275/py\varepsilon=\sqrt{275/p_y}. Stronger steel must reach a higher stress before wrinkling, so it needs chunkier plates; the limits are multiplied by ε\varepsilon to tighten them.

  5. Result

    Class 1 plastic

    Reaches the full plastic moment with enough rotation to form a plastic hinge. I-section flange b/T≤9εb/T\le9\varepsilon, web d/t≤80εd/t\le80\varepsilon.

  6. Result

    Class 2 compact

    Reaches the full plastic moment but with limited rotation. Flange within 10ε10\varepsilon, web within 100ε100\varepsilon.

  7. Result

    Class 3 semi-compact

    Only reaches the elastic moment pyZp_yZ (first yield at the edge); it wrinkles before the plastic moment. Flange within 15ε15\varepsilon, web within 120ε120\varepsilon.

  8. Result

    Class 4 slender

    Wrinkles before yielding; effective (reduced) properties are needed. The usual exam UB/UC sections are never Class 4.

  9. How design deals with itWhere marks are lost

    The worst plate decides

    Classify the flange and the web separately and take the higher (worse) class. E.g. flange Class 2, web Class 3 → Class 3.

  10. How design deals with it

    Decides which moment resistance

    Class 1 or 2 → Mc=min⁡(pyS,1.2pyZ)M_c=\min(p_yS,1.2p_yZ); Class 3 → pyZp_yZ; Class 4 → effective section. The LTB βw\beta_w also depends on the class. Worked numbers: classification map.

  11. Exam answer

    Theory: purpose of classification

    One line of purpose (width-to-thickness limits decide whether local buckling stops the section reaching its yield or plastic resistance), then one line for each of the four classes.

Tour steps

  1. 1. Thin plates in compression wrinkle

    A section is made of plates; a thin plate in compression wrinkles locally (local buckling).

  2. 2. What makes it wrinkle more easily?

    A larger width-to-thickness ratio; and stronger steel tightens the limits (ε\varepsilon).

  3. 3. The four classes

    From “can form a plastic hinge” to “wrinkles before yield”, each class is worse than the one before.

  4. 4. The whole section takes the worst

    Compare flange and web separately and take the worse; that decides between SS and ZZ.

  5. 5. The theory answer

    One line of purpose + one line per class.

Animation labWhy thin elements buckle locally1 concept · 7 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. The flange outstand and web have different widths, thicknesses and edge support conditions.
  2. A thinner plate can wrinkle locally before the complete member loses stability.
  3. Class 1 allows plastic rotation; Class 2 reaches plastic resistance; Class 3 reaches elastic resistance; Class 4 requires effective properties.
  4. Check every relevant compression element with the supplied limits and stress distribution. The deformation shown is qualitative.

20. Which code the course uses: do not mix Background

The lecture uses the Hong Kong Code of Practice for the Structural Use of Steel 2011 for calculations; the syllabus lists the 2011 (2023 edition). The Data File tables come from the 2011 Code, and the section properties from an SCI / BS 5950 design guide. All exercises on this site follow those formulas, tables and symbols.

Lecture pp.23–26 introduce BS 5950, GB 50017, Eurocode 3, AS 4100 and AISC and compare their symbols. The Eurocode family: EN 1990 basis, 1991 actions, 1992 concrete, 1993 steel, 1994 composite, 1995 timber, 1996 masonry, 1997 geotechnical, 1998 earthquake, 1999 aluminium. Each country can add a National Annex to set some parameters.

MeaningCourse HK notationEurocode notation in the lecture
Major cross-section axisx–xy–y
Minor cross-section axisy–yz–z
Along memberNot labelled x in this course tablex
Elastic section modulusZWel
Plastic section modulusSWpl
Axial actionPN
Design yield stresspyfy
Column compressive strengthpcχfy
LTB bending strengthpbχLTfy

Eurocode calls dead load a permanent action and imposed and wind loads variable actions, with combinations such as 1.35G1.35G and 1.5Q1.5Q plus combination factors — an overview only; do not use them in HK-code calculations. (The p.26 label “Wrapping Index” is a typo for warping.)

Animation labRead a table without losing the keys1 concept

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. Name the required property: material strength, section property, buckling strength or a moment factor.
  2. Keep section size, steel grade, thickness band, curve and axis as separate lookup keys.
  3. , and require different conversion powers. Do not use adjacent columns interchangeably.
  4. Use bracketing rows within the same valid column. The original page remains the source of all table values.

Appendix: Eurocode combination-factor arithmetic Background

Source: Chapter 1 p.25, Table 11.This comparison assumes a combination factor ψ=0.5\psi=0.5 for wind and ψ=0.7\psi=0.7 for imposed load, each multiplied by the load factor 1.5 (both have no units).

Accompanying wind coefficient

1.5×0.5=0.75.
Animation labRearrange a capacity equation1 concept · 1 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. For , P is force, p is stress and A is area. Here we want A.
  2. Divide both sides by the same non-zero stress p.
  3. The p on the right cancels, leaving .
  4. gives . This is required area, not a selected section or a complete design check.

Accompanying imposed coefficient

1.5×0.7=1.05.
Animation labRearrange a capacity equation1 concept · 1 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. For , P is force, p is stress and A is area. Here we want A.
  2. Divide both sides by the same non-zero stress p.
  3. The p on the right cancels, leaving .
  4. gives . This is required area, not a selected section or a complete design check.

So the two illustrative combinations contain 0.75W0.75W and 1.05Q1.05Q. They are not unfactored loads and not HK-code factors. For HK floor questions go back to 1.4G+1.6Q1.4G+1.6Q.

Try it yourself. Under those illustrative assumptions only, what accompanies a 20kN wind action?

Reveal answer and reasoning

0.75×20=15kN. This answer belongs to the lecture’s Eurocode comparison, not the HK worked-example loading method.

Animation labRearrange a capacity equation1 concept · 1 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. For , P is force, p is stress and A is area. Here we want A.
  2. Divide both sides by the same non-zero stress p.
  3. The p on the right cancels, leaving .
  4. gives . This is required area, not a selected section or a complete design check.
Animation labRearrange a capacity equation1 concept · 2 source expressions

Supplement to the original lesson. Enable JavaScript to play, step through calculations and rotate 3D models. The following explanation remains readable offline.

  1. For , P is force, p is stress and A is area. Here we want A.
  2. Divide both sides by the same non-zero stress p.
  3. The p on the right cancels, leaving .
  4. gives . This is required area, not a selected section or a complete design check.

Practise this chapter: tutorials, assignments and past-paper answers →

Recent-paper practice for this chapter

Older-paper answers for this chapter (reference)