Materials Science · Units, loads, stress and strain

Stiffness and strength answer different questions

Explain how elastic modulus, component geometry and a specified strength limit affect different decisions. Derive a simple rod model and test both an extension requirement and a yield condition.

Useful background: Materials, questions and levels of evidence. Each worked case explains the quantities it uses.

Back to the chapter

After reading, you can

Distinguish material modulus, component stiffness and a specified strength endpoint.
Derive axial rod extension from stress, strain and a linear-elastic material relation.
Predict separately how length, area and modulus affect extension and stress.
Apply deformation and strength criteria together, recognising when the elastic model ceases to apply.

Two different questions about a support

A support can bend too far without breaking. Stiffness concerns deformation under load; strength concerns reaching a specified limit such as yield or maximum engineering stress. A material with high strength need not have proportionally high elastic modulus.

Imagine a tie rod that must keep two parts aligned while it is pulled. One question is how far it extends under the working load. Another is whether it reaches a material limit. A rod can remain unbroken and entirely elastic yet extend too much for the assembly to function.

Elastic deformation is recoverable on unloading. Yielding involves plastic deformation, with an irreversible part under the specified test conditions. A reported strength must identify its endpoint: yield stress and the maximum engineering stress in a tensile test are different quantities. Neither word supplies the elastic slope by itself.

Modulus connects stress to elastic strain

For a small, uniaxial, linear-elastic response, write σ = Eε. Here σ is axial stress, ε is axial strain and E is Young’s modulus. The relation says that doubling stress doubles strain within that regime. Because strain is dimensionless, E has stress units, such as Pa or GPa.

At a given stress, a larger E gives less elastic strain: ε = σ/E. On a graph of stress vertically against strain horizontally, E is the slope of the straight elastic region. It describes a material response in the stated direction and conditions; it is not the force needed to move an entire component by one millimetre.

Recoverable does not always mean proportional. Some materials have nonlinear elastic or time-dependent responses, for which one constant E over the whole loading history is inadequate. Here the linear relation is an explicit model, used where the assumed material behaviour is acceptable.

The material and the shape

For a uniform slender rod in small uniaxial linear-elastic deformation, extension is FL/(EA). E is Young’s modulus, L length and A area. Increasing area or decreasing length makes that rod stiffer even if E stays unchanged. The relation requires the stated geometry and elastic regime; it does not predict arbitrary bending or failure.

The equation follows from definitions already introduced. The average axial engineering stress is σ = F/A, where A is the original cross-sectional area. Engineering strain is ε = δ/L, where δ is the extension and L is the original length between the rod ends being assessed. Substituting them into σ = Eε gives F/A = Eδ/L. Rearranging gives δ = FL/(EA).

Component stiffness in this model is k = F/δ = EA/L. Its units are force per length, such as N/mm; modulus E has units N/mm². The area and length turn a material relation into a component relation. Unit checking makes the distinction visible: EA/L has units (N/mm²) × mm²/mm = N/mm.

Assume a straight rod with constant area and effective modulus, axial tension, negligible self-weight and small deformation, away from grips or other stress concentrations. Use a bending, contact or instability model when those effects matter. In particular, an axial extension formula does not check buckling of a compressed slender member.

Changing one input in the same linear-elastic axial model while holding the others fixed. These are model predictions, not universal component rules.
Change	Elastic extension δ	Average axial stress F/A
Double E	Halves	Unchanged
Double A	Halves	Halves
Double L	Doubles	Unchanged
Double F	Doubles, if still elastic	Doubles

A larger strength is not a smaller extension

Consider two illustrative rods with the same original length L = 500 mm and area A = 20 mm². Rod A has E = 50 GPa and a stipulated axial yield stress of 300 MPa. Rod B has E = 100 GPa and a stipulated yield stress of 200 MPa. These values define a teaching comparison; they are not measurements or recommended material grades.

At F = 2000 N, both have the same average stress, 100 MPa, because their loads and areas match. This is below both stipulated yield stresses. The higher-modulus rod B nevertheless extends less: 0.5 mm instead of 1.0 mm. If the assembly permits at most 0.8 mm extension, rod A misses that requirement despite having the higher yield stress.

The distinction is between a demand and a resistance. The loading and model predict stress and extension; a stated yield condition and an allowable extension provide two different comparisons. Passing one does not imply passing the other. The worked case below shows the calculation in full.

Illustrative matched-geometry comparison at F = 2000 N, with L = 500 mm and A = 20 mm² for both rods.
Quantity	Rod A	Rod B
Young’s modulus E	50 GPa	100 GPa
Stipulated yield stress	300 MPa	200 MPa
Axial stiffness k	2000 N/mm	4000 N/mm
Average stress	100 MPa	100 MPa
Elastic extension	1.0 mm	0.5 mm
Meets 0.8 mm extension limit?	No	Yes

Illustrative linear-elastic force versus extension curves for the two rods with length 500 mm and area 20 mm². Rod A is a solid line of slope 2000 N/mm and reaches 2000 N at 1.0 mm. Rod B is dashed, has slope 4000 N/mm and reaches 2000 N at 0.5 mm. A dotted vertical extension limit is at 0.8 mm. Both lines stop at the worked load, not at yield.

Check the regime before trusting the comparison

If a calculation predicts a stress above the applicable yield condition, the assumption of wholly elastic response no longer describes that state. Continuing the straight elastic line does not calculate the ensuing plastic deformation. A quoted yield or tensile strength also does not, by itself, assess cracks, repeated loading, long-term deformation or a joint.

Geometry offers distinct ways to change a component. Shortening this rod increases its stiffness without reducing F/A at fixed force and area. Increasing area both stiffens it and reduces that average stress. Substituting a higher-E material can reduce extension while leaving the same geometry’s F/A unchanged. Ask which input and which requirement the proposed change actually affects.

Use these relations to understand the separation of material response, geometry and endpoint. Later lessons develop measured elastic slopes, yield definitions, bending, fracture and fatigue. A complete component assessment must include whichever loading and failure mechanisms its use introduces.

Worked examples

Same geometry, two materials and two criteria

Use the stipulated rods above at 2000 N. Calculate stress, elastic strain and extension for each. Which satisfies an extension limit of 0.8 mm, and does that answer follow from their yield stresses?

Keep units consistent: 1 GPa = 1000 N/mm². Thus E_A = 50 000 N/mm² and E_B = 100 000 N/mm². Both have σ = F/A = 2000/20 = 100 N/mm² = 100 MPa.
For A, ε_A = σ/E_A = 100/50 000 = 0.002. Its extension is δ_A = ε_A L = 0.002 × 500 = 1.0 mm. For B, ε_B = 100/100 000 = 0.001 and δ_B = 0.5 mm.
Check the same result directly: δ_A = (2000 × 500)/(50 000 × 20) = 1.0 mm. Doubling E while holding F, L and A fixed halves extension, preserving the original comparison.
Compare demands with their matching criteria. The 100 MPa stress is below both stipulated yield stresses, so this calculation is consistent with the assumed elastic state. Only B is within 0.8 mm extension; A’s 300 MPa yield stress does not cure its excessive elastic extension.
The strains are 0.2 % and 0.1 %, consistent with the small-deformation assumption. The stiffnesses are k_A = 2000/1.0 = 2000 N/mm and k_B = 2000/0.5 = 4000 N/mm. They are not moduli in MPa.

Both are below the stipulated yield levels at this load. A extends 1.0 mm and misses the extension requirement; B extends 0.5 mm and meets it.

Change the component without changing its material

Keep rod A’s material, 2000 N force and 500 mm length, but increase its area from 20 to 40 mm². Compare the new result with instead halving the original rod’s length to 250 mm while keeping area 20 mm².

Doubling the area gives σ = 2000/40 = 50 MPa. Its stiffness becomes k = (50 000 × 40)/500 = 4000 N/mm, so δ = 2000/4000 = 0.5 mm.
Halving the original length also gives k = (50 000 × 20)/250 = 4000 N/mm and δ = 0.5 mm. However, σ remains 2000/20 = 100 MPa.
Both variants meet the illustrative 0.8 mm extension requirement and remain below the stipulated 300 MPa yield stress. They achieve the same axial stiffness by different geometric changes.
Neither change increased E or the stipulated material yield stress. They changed the way the same material was used. The shorter option is available only if the assembly can accept the changed span; geometry is a design constraint as well as a calculation input.

Both alternatives extend 0.5 mm. The wider rod carries 50 MPa; the shorter rod still carries 100 MPa. Equal component stiffness does not require equal stress or a different material modulus.

Try the reasoning

Two uniform rods of the same material and area have lengths 100 mm and 400 mm. Under the same admissible axial tensile force, compare their extension, component stiffness, strain and average stress. A learner calls the short rod’s material four times stiffer; correct the claim.
Show the answer and reasoning
The long rod extends four times as much and has one quarter the component stiffness. Their strain, average stress and material modulus are the same.
At fixed F, A and E, δ = FL/(EA) scales with L and k = EA/L scales inversely. But ε = δ/L = F/(EA) and σ = F/A do not depend on L in this model. The geometry changed, not the material modulus.
Raise the illustrative matched-geometry rods A and B to 3000 N. Recalculate extension and stress, and judge the same 0.8 mm and stipulated yield criteria.
Show the answer and reasoning
Both stresses are 150 MPa. A extends 1.5 mm and B 0.75 mm. Both remain below their stipulated yield stresses, but only B meets the extension limit.
The load is 1.5 times the worked load. With both assumed states still elastic, stress and extension scale by 1.5. A component can fail its functional extension requirement without reaching the specified material yield endpoint.
For rod A, a worksheet substitutes E = 50 into FL/(EA) while using N, mm and mm², and reports δ = 1000 mm. Find the error before treating this as a physical prediction.
Show the answer and reasoning
The number 50 was in GPa and needed conversion to 50 000 N/mm². The consistent result is 1.0 mm.
The worksheet mixed GPa with N/mm² and overstated extension by 1000. Its enormous strain would also contradict the assumed small-deformation regime. Unit consistency must be established before interpreting a result as a material response.
At 4500 N, rod B’s unchanged area gives a calculated average stress of 225 MPa, exceeding its stipulated yield stress of 200 MPa. A learner extends the elastic formula to obtain δ = 1.125 mm and calls it the actual extension. Assess the inference.
Show the answer and reasoning
The 1.125 mm value is an extrapolation of the elastic model, not an established post-yield extension.
4500/20 = 225 MPa flags that the assumed wholly elastic state is outside its stipulated domain. The elastic quotient 4500/4000 = 1.125 mm does not model plastic response. Additional material behaviour and an appropriate model are required.

What to carry forward

Young’s modulus describes elastic stress versus strain; component stiffness includes geometry.
A high specified strength does not guarantee small elastic deformation.
For the stated axial rod model, δ = FL/(EA) and k = EA/L; keep their units distinct.
Compare each predicted demand with its own criterion, and stop treating the elastic line as an actual response after its assumptions fail.

References

Teaching resources and research supporting this lesson’s explanations and examples.