Structures and materials: stress, stiffness, strength and why things break: Stiffness, strength and toughness: three different questions about a material
All courses / Understanding Engineering / Module 4: Structures and materials: stress, stiffness, strength and why things break Part 2 of 5Stiffness, strength and toughness: three different questions about a material
Stiff, strong and tough are used loosely in daily talk, but they are three separate properties with different numbers. This part defines each from the stress-strain curve, gives real values for steel, concrete, timber, aluminium and plastics, and shows why a strong material is not always the right one.
2.1 Stiffness is how much stress it takes to stretch a material by a given fraction, and steel is stiff whatever its grade
Press a steel scale and a plastic scale with the same force. The plastic one bends much more. That difference is stiffness. Within small deformations most solids obey a simple rule, known as Hooke's law: the strain is proportional to the stress. The constant of proportionality is Young's modulus, written E: stress = E x strain. Because strain has no unit, E has the unit of stress. The values are large, so they are quoted in gigapascals, GPa: one GPa is 1,000 MPa.
Steel has a modulus of about 200 GPa, and the figure is nearly the same for a soft structural grade and for a high-strength rebar. Making steel stronger by alloying or by working it hardly changes its stiffness. The reason is that E comes from how strongly neighbouring atoms resist being pulled apart or pushed together, and alloying leaves that practically unchanged. Strength, in contrast, depends on defects, which engineers can control. So you can never make a floor stiffer by buying a stronger grade of the same metal.
The table puts numbers on it. A 3 m rod, 16 mm across, is pulled with 2 kN, about the weight of a 200 kg load. Steel stretches by 0.15 mm, about one and a half sheets of office paper. An aluminium rod stretches three times as much, oak about 17 times, nylon about 67 times, and a polyethylene rod 250 times as far, nearly 4 cm.
The lesson is that stiffness, not strength, often sets the size of a member. Floors that bounce, doors that jam as a beam sags, and glass facades that crack when the frame moves are stiffness problems. Notice also that stiffness of a part is E x area / length for a straight pull, so a short, fat part is stiffer than a long, thin one of the same material.
| Material | E (GPa) | Strain = stress / E | Stretch over 3 m |
|---|---|---|---|
| Steel | 200 | 9.95 MPa / 200,000 MPa = 0.00005 | 0.15 mm |
| Concrete (M25), squeezed, not pulled | 25 | 9.95 / 25,000 = 0.0004 | 1.2 mm |
| Aluminium alloy | 69 | 9.95 / 69,000 = 0.00014 | 0.43 mm |
| Oak, along the grain | 12 | 9.95 / 12,000 = 0.00083 | 2.5 mm |
| Nylon | 3 | 9.95 / 3,000 = 0.0033 | 10 mm |
| HDPE (milk-can plastic) | 0.8 | 9.95 / 800 = 0.0124 | 37 mm |
Stress first: 2,000 N / 201 mm2 = 9.95 MPa, the same for every row. All strains and stretches calculated. Concrete is shown squeezed, because a pull of 9.95 MPa is nearly three times its tensile strength and would crack it; its E uses the IS 456 rule E = 5,000 x square root of the cube strength in MPa. Check: a plastic washing-line rope does stretch by several centimetres under a wet shirt load, a steel wire does not.
Source: Kosky et al., Exploring Engineering (Academic Press, 2013), ch. 11, 'Materials Engineering'- Kosky et al., Exploring Engineering (Academic Press, 2013) · ch. 11, 'Materials Engineering'
- Bloomfield, How Things Work (Wiley, 2015) · ch. 3, 'Mechanical Objects, Part 1'
- Bureau of Indian Standards, IS 456:2000, Plain and reinforced concrete · cl. 5.6.3, 6.2.2, 6.2.5.1, 23.2, 26.4 (table 16), 26.5.1.1, 36.4.2, 38.1 and table 18; https://law.resource.org/pub/in/bis/S03/is.456.2000.pdf
2.2 Strength is where a material gives up, toughness is the energy it takes to get there, and ductility is the warning
Pull a steel bar slowly and record the stress against the strain. The result is the stress-strain curve, the main tool of materials engineers. For small strains it climbs along a straight line, the elastic range, where letting go returns the bar to its length. At the yield stress the line bends over. Beyond it, the bar stretches permanently, a change called plastic deformation. The stress still rises a little to a peak, the ultimate tensile strength, then the bar necks down at one spot and parts.
Strength usually means the yield stress, because a part that has yielded has changed shape and may no longer work. Ductility is how far the material can stretch before it breaks, and it is a warning: a ductile steel beam sags, stretches and cracks visibly before it fails, giving people time to leave. Brittle materials such as glass, cast iron, brick and unreinforced concrete show almost no plastic range. They snap near the end of the straight line.
Toughness is the energy absorbed per unit volume up to fracture, the area under the whole curve. A tough material can take a blow or a sudden overload by deforming and spending that energy as heat. The worked estimate compares a rebar's elastic area with its total. Most of what a Fe500 bar can absorb lies beyond yield, more than a hundred times the elastic part.
That explains rules in the Indian steel standard, IS 1786, which asks rebar to stretch at least a certain percentage before it breaks. In an earthquake, a frame that can bend and absorb energy survives, and a brittle frame breaks without notice. Strength and toughness are usually in tension: processes that raise strength, such as cold working or adding carbon, tend to reduce stretch.
| Step | Working | Result |
|---|---|---|
| Yield strain of the bar | 500 MPa / 200,000 MPa | 0.0025, or 0.25 per cent |
| Stretch of a 3 m bar at yield | 0.0025 x 3,000 mm | 7.5 mm |
| Elastic energy per m3 up to yield | half x 500 MPa x 0.0025 | 0.63 MJ per m3 |
| Strain at break: IS 1786 asks at least 12 per cent for Fe500 and 16 for Fe500D, so take 14 per cent, at an average stress of about 550 MPa | 0.14 x 550 MPa | about 77 MJ per m3 |
| Ratio of total to elastic | 77 / 0.63 | about 120 |
| Bar volume, 16 mm bar (201 mm2) x 3 m | 201 x 3,000 mm3 | 0.0006 m3 |
| Energy to break the whole bar | 77 MJ/m3 x 0.0006 m3 | about 46 kJ |
All results calculated. The elongation in the standard test is measured over a short length that includes the neck, so it overstates a long bar somewhat; treat 46 kJ as an upper estimate. Check: 46 kJ equals the work of lifting a 1,000 kg car body by about 4.7 m (46,000 / 9,800). A single steel bar can soak up that much before parting, which is what protects a building in an earthquake. A glass rod of the same size (E about 70 GPa, breaking near 50 MPa) absorbs only about 10 J (calculated).
Source: Kosky et al., Exploring Engineering (Academic Press, 2013), ch. 11, 'Materials Engineering'- Kosky et al., Exploring Engineering (Academic Press, 2013) · ch. 11, 'Materials Engineering'
- Kosky et al., Exploring Engineering (Academic Press, 2013) · ch. 12, 'Mechanical Engineering' (section 12.4.1, mechanical behaviour of materials)
2.3 Real materials are far weaker than their bonds allow, and strength is the craft of blocking defects
Take the force needed to pull apart two planes of atoms in perfect iron. Simple models put it at tens of gigapascals, a hundred times more than the 250 MPa of mild steel. Real metals are weaker because they are not perfect. Their crystals contain line defects, called dislocations, that let planes of atoms slip over one another at far lower stress. The metal yields not by pulling every bond at once but by moving these defects one row at a time.
Hence the whole craft of strong metals is to jam up dislocations. Dissolved carbon and alloying atoms stick to them. Tiny grains give them boundaries to crash into. Cold drawing a wire tangles them. The prestressing strand used in bridges is drawn from steel with 0.8 per cent carbon, then drawn again, reaching 1,860 MPa, over seven times the yield of mild steel. The price is paid in ductility and in how easily the steel can be welded, which is why such wires are never welded.
Brittle materials such as glass and concrete have a different problem: tiny cracks that act as built-in notches, a subject that returns in the last part. Ceramic strength depends on the largest flaw, and flaws are random, so one pane of glass breaks at 30 MPa and another at 80 MPa.
To compare materials for structures it helps to divide by density. Specific stiffness is E / density, in joules per kilogram, and specific strength is strength / density. The table is revealing. Steel and aluminium have almost the same specific stiffness, so for a given mass of metal in a given shape neither is stiffer. Aluminium helps because its low density lets the same mass be spread into a deeper, thicker section, and depth increases bending stiffness steeply, as the next part shows. Timber is respectable on strength per kilogram, concrete poor, and a carbon-fibre composite far ahead of everything, at a high cost.
| Material | E (GPa) | Strength (MPa) | Density (kg/m3) | E / density (MJ/kg) | Strength / density (kJ/kg) |
|---|---|---|---|---|---|
| Structural steel (E250) | 200 | 250 | 7,850 | 25.5 | 32 |
| Rebar steel (Fe500) | 200 | 500 | 7,850 | 25.5 | 64 |
| Prestressing strand | 195 | 1,860 | 7,850 | 24.8 | 237 |
| Aluminium alloy (6061-T6 type) | 69 | 275 | 2,700 | 25.6 | 102 |
| Concrete (M25, in compression) | 25 | 25 | 2,400 | 10.4 | 10 |
| Timber, along the grain (oak class) | 12 | 50 | 700 | 17.1 | 71 |
| Carbon-fibre composite (along fibres) | 134 | 1,000 | 1,550 | 86.5 | 645 |
| Nylon 6 | 3 | 70 | 1,140 | 2.6 | 61 |
| HDPE | 0.8 | 25 | 950 | 0.8 | 26 |
Strength is yield for the metals, compression for concrete and a typical working value along the grain for timber and the fibre composite; plastics are given at yield or break. Ratios calculated. Sources: steel E from IS 456 (200 kN/mm2); strand from IS 14268 and IS 1343; aluminium from Aluminum Association data for 6061-T6 (68.9 GPa, 276 MPa); oak from the USDA Wood Handbook; carbon fibre from the Toray T700S data sheet (134 GPa at 60 per cent fibre; 1,000 MPa is a working value, well below its 2,860 MPa test strength). Real products vary by 20 per cent or more. Check: steel and aluminium both come out near 25 MJ/kg.
Source: Kosky et al., Exploring Engineering (Academic Press, 2013), ch. 11, 'Materials Engineering'- Kosky et al., Exploring Engineering (Academic Press, 2013) · ch. 11, 'Materials Engineering'
- Kosky et al., Exploring Engineering (Academic Press, 2013) · ch. 12, 'Mechanical Engineering' (section 12.4.1, mechanical behaviour of materials)
See it in the table
Al AluminiumSt SteelTm TimberCheck your understanding
A designer wants a stiffer floor beam of the same shape. Which change helps?
Steel grades all have about the same E, so a stronger grade does not stiffen. Aluminium has a third of the modulus. A deeper section increases the second moment of area, which sets bending stiffness.What does the area under the whole stress-strain curve represent?
Stress times strain has the units of energy per volume, so the area under the curve is the energy needed to break a unit volume. Steel with a long plastic range has a very large area.A 3 m steel rod and a 3 m nylon rod of the same diameter carry the same pull. Roughly how do their stretches compare?
Stretch is proportional to 1/E. Steel at 200 GPa against nylon at about 3 GPa gives a ratio of about 67.