Comparing alloy families under service conditions

A family name narrows the search

Steel, aluminium, titanium, copper and nickel alloy families offer different combinations of density, transport, processing and environmental response. Within each family, composition and treatment can vary widely. A typical table is a screening aid rather than a certificate for a chosen stock.

A welded bracket illustrates the difficulty. A low-density stock looks attractive, but its joint and nearby material acquire a new thermal history. The highest stock tensile strength need not describe the finished leg, and the lightest leg need not come from the lowest-density candidate if their areas may differ. We will first learn what real grade labels tell us, then work a separate invented finished-route comparison.

A family groups materials by their principal metallic constituent or a related composition range. A grade narrows the composition and other specified attributes. Condition adds the processing and structural state: for example, annealed, cold-worked, or solution-treated and aged. Product form, thickness and test direction can narrow the applicable data further. These labels answer different questions; none should disappear when a table is copied.

Alloy composition, phases and phase diagrams explains why the same overall composition can have different phase arrangements. Grains, dislocations and strengthening shows how solutes, particles and stored defects resist slip. Here those mechanisms help us ask which catalogue state can survive a proposed manufacturing route; we do not repeat their detailed strength models.

Read the property together with its qualifier

A minimum is a stated lower bound for the identified scope; a maximum is an upper bound. A typical value describes representative behaviour and is not automatically either bound. A range also needs its meaning: it might span conditions or observations, or specify an acceptance interval. Do not silently turn a range midpoint into a measured value for the chosen piece.

One historical 2019 table for extruded 6082-T6/T6511, in its 5–20 mm wall-thickness row, lists minimum 0.2% offset yield strength 260 MPa and minimum tensile strength 310 MPa. Yield and tensile strength are different endpoints. The same document warns that strength is reduced in the weld region. Neither minimum can simply be carried over to the finished welded region, and neither is a component allowable stress.

The 2000 Ti–6Al–4V sheet lists Young’s modulus 107–122 GPa at 20 °C, explicitly dependent on texture and heat treatment; its tensile table separately identifies product, condition and test direction. The 2007 alloy 718 bulletin labels its physical-constant tables typical and unsuitable for specification purposes. A numerical value without those labels has lost important information.

The Cu-ETP sheet labels density 8.92 g/cm³ and electrical conductivity 58 MS/m as typical annealed values at 20 °C. Its separate mechanical table is organised by temper. In the R220 row, the proof-stress entry is ≤140 MPa, despite a general column heading containing “Minimum”. The explicit inequality is an upper limit, not a guaranteed lower strength. Reading the cell, heading and notes together avoids reversing the evidence.

Keep source date and data status as well. These manufacturer publications give concrete examples of how data are labelled; they are not certificates for present stock or statements of current standard requirements. For an actual purchase or qualification, obtain the applicable specification and evidence for the intended product and finished route.

Select a state for a job

Specify loads, temperature, environment, joining, manufacturing and life requirements. Compare relevant grade-and-condition data, including uncertainty and availability. A corrosion-resistant alloy in one environment may be unsuitable in another; a high-strength state may complicate forming or welding.

The heat-affected zone is material near a weld whose structure changes through heating and cooling without melting. Precipitates can dissolve or coarsen; a cold-worked structure can recover; other alloys can undergo phase changes. The weld metal has its own solidification and composition history. Residual stress, defects and a changed surface can matter too. The relevant result depends on the actual process and material, rather than a universal percentage loss.

For the following original teaching problem, imagine three proposed finished manufacturing routes A, B and C for a welded bracket. Every number in their dossier is fictional, including density and modulus: these rows are not measured 6082, 316L or Ti–6Al–4V data and are not inferred from the real examples above. Each row supplies a uniform straight leg’s response in the same axial direction at 25 °C. A and B describe their respective stock after joining and surface preparation, without a subsequent strengthening treatment; C describes its stock after joining, a stipulated material-specific final treatment and surface preparation. All rows apply after those complete sequences, not to an intermediate stock state. Linear elasticity is stipulated over the model’s accepted range.

S is a supplied nominal axial screening limit for that leg, not a catalogue yield stress, an uncertainty allowance or a certified design stress. It is independently provided for the finished-route model; this lesson does not derive it from a tensile minimum. The dossier also stipulates whether the required stock form and proposed route are available within this fictional project window. Availability is separate evidence, not a property of an alloy family.

All candidates have an unmodelled joint and complete-bracket exposure assessment still pending. In particular, no local weld stress, fatigue, bending or life model is supplied. The calculation can select a leg candidate for further work, not qualify the welded bracket.

Size only the uniform axial leg

Replace one straight leg by a uniform member in axial tension, with original length L = 100 mm and force F = 1000 N. Its original area A may be chosen up to 10.0 mm². Require an extension δ no larger than 0.200 mm and a leg mass no larger than 3.00 g. The applied load is centred; the model ignores self-weight and end effects and includes no bending. End fittings, weld geometry and the rest of the bracket are excluded from its length, mass and compliance.

The average engineering stress is σ = F/A. For a uniform small-strain linear-elastic member, ε = σ/E and δ = εL, so δ = FL/(EA). The extension is a length, while strain ε is a ratio. With N and mm, E must be in N/mm² = MPa: 70.0 GPa is 70000 MPa. These equations describe the same axial member, not a formula for the bending deflection of a shaped bracket.

The stress screen F/A ≤ S requires A ≥ F/S. Independently the extension screen requires A ≥ FL/(Eδ_max). The required area is the larger of those two lower bounds. Do not average them or add them: the same section must satisfy both inequalities. Finally check that it fits the 10.0 mm² area limit and the mass requirement.

Mass is density times volume: m = ρAL/1000 when ρ is in g/cm³ and A and L are in mm² and mm. The factor 1000 converts mm³ to cm³. At fixed area and length the denser candidate is heavier; at separately sized areas that ordering can change. Density and properties relative to mass develops the underlying comparison. Here we keep the actual two area constraints and the manufacturing route rather than ranking a single specific-property ratio.

For A, the stress bound is 1000/120 = 25/3 ≈ 8.333 mm²; the extension bound is 100000/(70000 × 0.200) = 50/7 ≈ 7.143 mm². Stress governs. B instead needs 50/11 ≈ 4.545 mm² from stress and 2.500 mm² from extension; stress governs again. C needs 10/3 ≈ 3.333 mm² from stress and 50/11 ≈ 4.545 mm² from extension, so extension governs. The first figure displays those separate bounds.

Use the unrounded bounds when evaluating a limiting calculation. The next table reports mathematical minimum areas, not rounded manufacturing dimensions. In particular, rounding 25/3 down does not exactly satisfy a 120 MPa limit. A selected section of 8.40 mm² for A gives a simple explicit check with some distance above that minimum.

A provisional choice includes what remains unresolved

At A = 8.40 mm², route A has stress 1000/8.40 ≈ 119.05 MPa, extension 100000/(70000 × 8.40) ≈ 0.17007 mm and mass 2.80 × 8.40 × 100/1000 = 2.352 g. It meets the supplied numerical and availability screens: area, stress, extension and mass fit their limits, and the required form and route are stipulated available. That makes A a candidate to continue investigating, not an approved bracket.

C is lighter at its separately calculated minimum even though its density exceeds A’s. Its higher modulus permits a smaller extension-limited area, while A is stress-limited. But C’s required form and route are unavailable in the fictional project window. B’s larger density gives a leg mass over the limit despite its smaller required area. No additional ranking by reputation changes either rejection.

The complete choice is still unresolved until the actual exposure, joint and whole shape are assessed. A chemical change can require different corrosion evidence; a hotter or longer duty can change the relevant state and deformation route. The electrical or thermal conductivity decisive for a conductor or heat path is not a measured requirement in this leg problem. It cannot be treated as interchangeable with low density or high tensile strength.

To carry the candidate forward, identify an actual grade, product and finished process, obtain matching property and variability evidence, and evaluate the joint and complete geometry under the real loading and exposure. A local uniform-leg screen does not resolve a weld toe stress concentration, residual stress or fatigue life. Thermal treatment and evolving microstructure helps interpret the history; it does not supply the missing bracket qualification.

Change the question before reusing the ranking

Tighten only δ_max from 0.200 to 0.0500 mm while keeping the fictional load, length, area cap and 3.00 g mass cap. Extension-area bounds increase fourfold: A needs about 28.571 mm², B 10.0 mm² and C about 18.182 mm². All are now extension-controlled. Their corresponding masses are 8.00, 7.80 and about 8.182 g. None passes the unchanged mass requirement, even if C becomes available; A and C also exceed the permitted area.

A fixed-section comparison asks something else. If all use A = 8.40 mm² and L = 100 mm, their masses are 2.352, 6.552 and 3.780 g. A is lighter than C at the same volume, whereas C was lighter at its own smaller required area. Those results are compatible: the separately sized legs did not occupy equal volumes. In the fixed-section comparison, only A meets 3.00 g.

Changing the manufacturing state is a third change. Suppose someone substitutes an unrelated fictional unwelded-stock strength of 300 MPa for route A’s finished-state S = 120 MPa. That incorrect calculation makes extension set A = 50/7 mm², with mass 2.00 g. But its actual finished-model stress is 140 MPa, above 120 MPa. The attractive result failed because the property column described the wrong state, not because the density arithmetic was wrong.

Requirements, dimensions that may vary, property definitions and finished condition all belong with a conclusion. If a requirement changes and no candidate passes, report that result and reconsider the design or evidence. Do not relax an unstated condition or assign a new material value just to obtain a winner.

Worked examples

Try the reasoning

1. Does “stainless” promise resistance in every chemical environment?

   Show the answer and reasoning

   No.

   Resistance depends on grade, surface state and exposure. The low-carbon 316L example improves resistance to one intergranular route after welding; that is not evidence for every chemical, temperature, crevice or surface condition.

2. Why retain more than one candidate initially?

   Show the answer and reasoning

   Manufacture and service constraints may eliminate different options.

   A single-property ranking can discard useful alternatives prematurely. In the worked comparison, C is a lighter mathematical leg but unavailable, while A remains a provisional candidate. A changed project window can change that screen without changing density.

3. A sheet lists a historical minimum 260 MPa proof/yield endpoint for a specified 6082-T6 product row. Another lists Cu-ETP R220 proof stress ≤140 MPa. May you conclude that every specimen in the latter row meets a required proof stress of at least 135 MPa?

   Show the answer and reasoning

   No. The ≤140 MPa entry is an upper limit and establishes no 135 MPa lower bound.

   A material with proof stress below 135 MPa could satisfy that upper limit. The 6082 lower-bound entry has a different grade, endpoint scope and product condition; it neither repairs the copper inference nor qualifies a welded 6082 component.

4. A design needs both a specified finished mechanical endpoint and electrical conductivity. A calculation combines the Cu-ETP annealed 20 °C typical 58 MS/m entry with a different temper’s mechanical row. Does the sheet alone prove that one finished part provides both?

   Show the answer and reasoning

   No. The two entries identify different states, and the conductivity is typical rather than a guaranteed lower limit.

   Obtain matching transport and mechanical evidence for the actual finished temper, geometry and temperature. A table organised by several conditions cannot be collapsed into a synthetic best-of-each-column grade.

5. The mounting interface now requires every fictional leg to use 8.40 mm² over 100 mm. Compare A and C masses and their compliance with the original 3.00 g mass cap. Does the earlier C-versus-A mass ordering have to persist?

   Show the answer and reasoning

   A has mass 2.352 g and C 3.780 g. A passes the mass cap; C fails. Their earlier separately sized ordering need not persist.

   Both now occupy 0.840 cm³. Multiplying by 2.80 and 4.50 g/cm³ gives the masses. Earlier C occupied less volume at its own smaller required area; equal volume is a different comparison. Both have stress about 119.05 MPa; C extension is about 0.10823 mm, so its failure here is mass, not extension.

6. Return to the original δ_max = 0.200 mm screen. If the fictional project window changes so C becomes available, is C’s lower calculated minimum leg mass enough to qualify its welded bracket? Which stated rejection disappears?

   Show the answer and reasoning

   The availability rejection disappears. C becomes another candidate for further assessment, but no bracket qualification follows.

   At its exact minimum area 50/11 mm², C has stress 220 MPa, extension 0.200 mm and mass 45/22 ≈ 2.045 g. Joint, exposure, whole-shape behaviour and variability evidence are still missing. An actual chosen area should not be rounded below the required bound.

7. Only the fictional leg length doubles from 100 to 200 mm, with F = 1000 N and δ_max = 0.200 mm unchanged. For B, find the controlling area, mass and result against the 10.0 mm² and 3.00 g caps.

   Show the answer and reasoning

   B becomes extension-controlled at 5.00 mm². Its 7.80 g leg fits the area cap but fails the mass cap.

   The stress bound stays 50/11 ≈ 4.545 mm², while FL/(Eδ_max) = 200000/(200000 × 0.200) = 5.00 mm². Mass is 7.80 × 5.00 × 200/1000 = 7.80 g. Both the required area and the volume length changed; simply doubling the earlier minimum mass would use an undersized area.

8. A report replaces “A is a candidate for further joint and exposure assessment” with “aluminium is the best alloy family for welded brackets”. Identify two unsupported transfers.

   Show the answer and reasoning

   It transfers one fictional finished route to a whole family and a uniform axial-leg screen to complete welded brackets.

   Other grades, conditions, loads, shapes and exposure requirements can change the answer. The report also loses the stipulated availability window and unresolved joint/life evidence. The conclusion must retain the model and its evidence boundaries.

What to carry forward

• Compare actual grades and conditions.
• Select for the complete service and manufacturing route.
• Family, grade, processing condition, product geometry and test direction describe different parts of the evidence.
• A minimum, a maximum and a typical value cannot be substituted for one another; read the cell and its notes.
• Use matching finished-route properties after joining. A stock strength is neither an automatic post-weld value nor a component allowable.
• The same axial leg must meet both stress and extension bounds before mass and availability are screened.
• A bounded candidate choice can be useful while complete-bracket behaviour and environmental qualification remain unresolved.