11.5 Material Selection for Engineering Design
Key Takeaways
- NCEES lists material selection as its own Materials sub-topic, so it is examined as a decision process rather than a property lookup.
- Specific strength is strength divided by density and specific stiffness is modulus divided by density; these govern weight-critical design.
- For a beam of fixed geometry minimizing mass at a given stiffness, the material index to maximize is E^(1/2)/rho, which is why wood and composites outperform steel per unit weight.
- Aluminum and steel have almost identical specific stiffness, so substituting aluminum for steel saves weight only by using a deeper or thicker section.
- A single limiting requirement such as maximum service temperature or chloride exposure eliminates whole material classes before any optimization begins.
11.5 Material Selection for Engineering Design
NCEES lists "Material selection" as a Materials sub-topic in its own right. It is not a property-recall item — it asks you to compare candidates against a stated requirement, which means normalizing properties and knowing which normalization the requirement implies.
The Selection Process
| Step | Question | Output |
|---|---|---|
| 1. Translate | What must the part do, and what must it not do? | Function, constraints, objective, free variables |
| 2. Screen | Which materials fail a hard constraint outright? | Reduced candidate list |
| 3. Rank | Which surviving candidate best serves the objective? | Material index |
| 4. Document | What supporting information decides between close finishers? | Cost, availability, joinability, code approval |
Screening comes before ranking, and it is where most real selections are decided. A single hard constraint — maximum service temperature, chloride exposure, food contact, code listing, electrical isolation — eliminates entire classes before any optimization begins. An exam item that specifies a service condition is usually testing screening, not optimization.
Normalizing Properties: Specific Strength and Stiffness
Absolute properties mislead whenever weight matters:
| Material | $\rho$ (kg/m³) | $E$ (GPa) | $\sigma_y$ (MPa) | $E/\rho$ (MPa·m³/kg) | $\sigma_y/\rho$ (kPa·m³/kg) |
|---|---|---|---|---|---|
| Structural steel (A36) | 7{,}850 | 200 | 250 | 25.5 | 31.8 |
| High-strength steel | 7{,}850 | 200 | 1{,}000 | 25.5 | 127 |
| Aluminum 6061-T6 | 2{,}700 | 69 | 276 | 25.6 | 102 |
| Titanium Ti-6Al-4V | 4{,}430 | 114 | 880 | 25.7 | 199 |
| Magnesium AZ91 | 1{,}810 | 45 | 160 | 24.9 | 88 |
| Carbon fiber composite | 1{,}600 | 150 | 1{,}500 | 93.8 | 938 |
| Douglas fir (along grain) | 500 | 13 | 50 | 26.0 | 100 |
| Alumina ceramic | 3{,}900 | 370 | 300 (compressive ~2{,}500) | 94.9 | 77 |
| Nylon 6/6 | 1{,}140 | 3 | 80 | 2.6 | 70 |
The striking result in that table: steel, aluminum, titanium, magnesium, and wood all have specific stiffness $E/\rho$ near 25–26. This is not coincidence — for metals it reflects the similar stiffness of atomic bonds per unit mass. Substituting aluminum for steel does not save weight in a stiffness-limited part unless the section geometry also changes, because the two have essentially identical $E/\rho$. Aluminum wins in practice by allowing a thicker or deeper section at equal weight, which raises $I$ — and $I$ scales with $h^3$.
Only carbon fiber composites and ceramics break out of the pack, at roughly 3.7× the specific stiffness of metals.
Material Indices
When the geometry is free to change, the property combination to maximize depends on the loading mode:
| Design objective | Free variable | Index to maximize |
|---|---|---|
| Minimum mass tie rod at given stiffness | Area | $E/\rho$ |
| Minimum mass beam at given stiffness | Section depth | $E^{1/2}/\rho$ |
| Minimum mass plate/panel at given stiffness | Thickness | $E^{1/3}/\rho$ |
| Minimum mass beam at given strength | Section depth | $\sigma_f^{2/3}/\rho$ |
| Maximum elastic energy storage (springs) | — | $\sigma_f^2/E$ |
| Best thermal shock resistance | — | $\sigma_f k/(E\alpha)$ |
The fractional exponents come from the fact that a beam's stiffness depends on $I \propto h^4$ while its mass depends on $h^2$, so allowing depth to change rewards low density more heavily than high modulus.
Worked Example: Beam Material for Minimum Weight
A stiffness-limited beam of fixed length may have any cross-section depth. Compare steel, aluminum, and Douglas fir using $E^{1/2}/\rho$.
Steel: $\dfrac{\sqrt{200{,}000\ \text{MPa}}}{7{,}850} = \dfrac{447.2}{7{,}850} = 0.0570$
Aluminum: $\dfrac{\sqrt{69{,}000}}{2{,}700} = \dfrac{262.7}{2{,}700} = 0.0973$
Douglas fir: $\dfrac{\sqrt{13{,}000}}{500} = \dfrac{114.0}{500} = 0.228$
Wood wins by a factor of 4 over steel and 2.3 over aluminum. This is the quantitative reason wood floor joists and aircraft spars (historically) and modern engineered lumber remain competitive: for a stiffness-limited beam whose depth is free, low density matters more than high modulus.
Note how the ranking reverses against the $E/\rho$ table, where all three were tied near 25. Same three materials, same stiffness objective — but allowing the depth to change alters which index applies and therefore which material wins. Reading which variable is free is the entire skill.
Screening Constraints That Decide Real Selections
| Constraint | Eliminates |
|---|---|
| Service above ~150 °C | Most thermoplastics |
| Service above ~500 °C | Aluminum, magnesium alloys |
| Seawater / chloride exposure | Carbon steel, 304 stainless in stagnant service |
| Electrical insulation required | All metals |
| Cyclic loading with a required infinite life | Aluminum (no true endurance limit) |
| Impact loading at low temperature | Ceramics, and ferritic steels below their transition temperature |
| Food or potable water contact | Lead-bearing alloys, many plasticized polymers |
| Optical transparency | Metals, most ceramics |
The aluminum fatigue constraint is a favorite exam item. Steel exhibits a true endurance limit — a stress below which life is effectively infinite. Aluminum does not: its S-N curve keeps descending, so every aluminum part under cyclic load has a finite life and must be designed to a specified number of cycles. This is why aircraft structures have mandated inspection intervals and retirement lives, and it is a screening constraint no amount of specific-strength advantage can overcome.
Worked Example: Selecting a Pressure Vessel Material
A vessel must hold a chloride-bearing process fluid at 200 °C, resist cyclic pressure for $10^7$ cycles, and minimize cost.
Screen:
- 200 °C eliminates thermoplastics and aluminum alloys (creep and strength loss)
- Chlorides at temperature eliminate 304 stainless (chloride stress corrosion cracking of austenitic grades)
- $10^7$ cycles with a required infinite life favors a material with a genuine endurance limit → steel family
- Ceramics are eliminated by their brittleness under cyclic pressure and thermal gradients
Surviving candidates: carbon steel with a corrosion-resistant lining or cladding, duplex stainless steel (much better chloride SCC resistance than austenitic grades), or a nickel alloy.
Rank on cost: clad carbon steel is typically the least expensive, duplex intermediate, nickel alloy highest.
Select clad carbon steel, with duplex as the alternative if cladding integrity or inspection access is a concern.
Note that no material index was needed. Four screening constraints reduced the entire materials universe to three candidates, and cost decided among them. That is how most real selections proceed, and why screening deserves the attention that ranking usually gets.
An engineer replaces a steel tie rod with an aluminum tie rod of the same stiffness. Approximately what weight change results?
For a stiffness-limited beam whose cross-section depth may be freely chosen, which material index should be maximized to minimize mass?
Why is aluminum generally unsuitable for a component that must survive an indefinitely large number of load cycles?
A component must operate at 250 degrees C in a chloride-bearing environment. Which screening step is most defensible?