​‌​​‌​

EN AW 5083 / UNS A95083 / AMS 4056

​‌​​‌​

EN AW 5083 (UNS A95083) is a type of aluminium alloy known in particular for its high corrosion resistance, good weldability and excellent mechanical properties. The alloy belongs to the aluminium-magnesium (Al-Mg) group.

5083 O aluminium plate has a low strength level, and the alloy is not recommended for applications above 65 °C.​‌​​‌​

EN AW 5083 is generally an alloy with good machinability. There are nevertheless some points to observe during machining.

Turning and milling: Cutting speed — medium cutting speeds are recommended. Cutting tools — carbide inserts or hardened steel tooling should be preferred. Cooling — a good cutting fluid should be used during machining, because heating of the material during machining can adversely affect the quality of the work.​‌​​‌​

Weldability: EN AW 5083 aluminium alloy offers excellent weldability and can readily be welded by the TIG and MIG methods in particular.

Points to observe during welding: Suitable welding consumables should be used so that thermal damage does not occur during welding. Attention should be paid to the cooling rate after welding; controlled cooling prevents internal stresses forming in the material.​‌​​‌​

Heat treatment: Heat treatment can be applied, but low temperatures should generally be preferred during hot working. High temperatures can adversely affect the mechanical properties of the alloy.

​‌​​‌​

What EN AW-5083 Is — and Why 65 °C Is the Most Important Number on This Alloy

EN AW-5083 (chemical symbol Al Mg4.5 Mn0.7 / material number 3.3547 / old DIN name AlMg4.5Mn / US equivalent A95083 / AFNOR A-G4.5MC / BS N8) is a non-heat-treatable aluminium-magnesium alloy. Nominally it carries 4.0–4.9 % Mg, 0.4–1.0 % Mn and 0.05–0.25 % Cr. It has the highest strength of the non-heat-treatable aluminium alloys, which is why it is the world’s standard material for ships, tanks, tankers and armour plate.​‌​​‌​

5083 in two sentences: it is excellent in seawater, it keeps its strength after welding, and it gets stronger at cryogenic temperatures. In exchange, it must not be placed in continuous service above 65 °C — and that is not a preference, it is a metallurgical prohibition. Magnesium is supersaturated in aluminium even at room temperature; at warm temperatures it precipitates at the grain boundaries as β phase (Al₃Mg₂ / Mg₂Al₃) and forms a continuous network. That network is anodic: the material becomes open to intergranular corrosion and stress corrosion cracking. This is called sensitisation, and there is no way back once it happens in service.

This page gives that topic the most space, because most real-world 5083 failures come from it — not from welding defects, not from thickness calculations, not from material quality. Below we take apart what the H116 and H321 tempers actually are, what ASTM G66 and ASTM G67 measure, why ASTM B928 exists at all, and what a classification society certificate really guarantees.​‌​​‌​

Standards by Product Form

​‌​​‌​

​‌​​‌​

5083 is one of the rare aluminium alloys that is fully covered in both the European and the American standard systems. Marine use then adds a separate layer of standards on top.

H116 and H321 — the Marine Tempers, and What the Certificate Actually Guarantees

​‌​​‌​

​‌​​‌​

This is the most important commercial topic on 5083. H116 and H321 are not ordinary hardness grades; they are tempers with a specified corrosion performance.

ASTM G66 and ASTM G67 — what the tests measure

ASTM G67 (NAMLT — Nitric Acid Mass Loss Test): the specimen is held in concentrated nitric acid at 30 °C for 24 hours. The acid preferentially dissolves the magnesium-rich β phase (Al₃Mg₂) rather than the aluminium matrix. If a continuous β network exists at the grain boundaries, grains fall away one by one and the mass loss jumps. Interpretation: ~1–15 mg/cm² = resistant; ~25–75 mg/cm² = susceptible (a continuous grain-boundary precipitate network). The acceptance threshold commonly used in marine specifications is 15 mg/cm². ASTM G66 (ASSET): a visual assessment of exfoliation corrosion susceptibility of 5xxx alloys; the result is not a mass but a rating. What ASTM B928 says is this: 5xxx products with nominal magnesium of 3 % or more, in the H116 and H321 tempers, shall show no evidence of exfoliation corrosion in G66 and/or an acceptable mass loss in G67.​‌​​‌​

What a class certificate guarantees — and what it does not

IT GUARANTEES: that the plate, at the moment of delivery, met the chemical, mechanical and corrosion acceptance criteria, and that the producer and mill are approved by the relevant classification society. IT DOES NOT GUARANTEE: that the material will not sensitise during its service life. That distinction is vital. The G67 test measures today’s grain-boundary condition. US defence research measured a G67 mass loss of 19–25 mg/cm² on 40–50-year-old 5083 armour plate, and found a continuous, magnesium-rich phase 10–15 nm thick at the grain boundaries — purely from long-term ageing at ambient temperature. In other words the material passed the test at delivery and crossed the threshold decades later. A certificate is a starting condition, not a lifetime warranty.​‌​​‌​

Temperature Limits — the 65 °C Rule

This is the most important section on this page and it should be read without softening.​‌​​‌​

The concrete applications that are forbidden​‌​​‌​

Hot water tanks and boilers. Structures near steam lines. Heated process tanks. Panels exposed to heat in exhaust areas and engine rooms. Dark, closed volumes that heat in the sun. High-temperature desalination equipment. We have seen commercial pages claiming that “5083 is suitable for 60–80 °C seawater desalination”; that statement directly contradicts the 65 °C rule and is not supported by any verified source. For these applications, consider a heat-treatable 6xxx (EN AW-6082) or a lower-magnesium alloy instead of a 5xxx.

Product Forms With NO Standard, or Unverified​‌​​‌​

Chemical Composition​‌​​‌​

Mechanical Properties​‌​​‌​

​‌​​‌​

The values below are standard minima (where a range is given for Rm, the lower figure is the minimum and the upper figure is a cap). Hardness values are typical and for information only.

Physical Properties​‌​​‌​

A design warning that comes from the expansion coefficient: 5083 expands roughly twice as much as steel. If an aluminium superstructure is bolted to a steel hull, temperature differences create stress at the joint — critical for both fatigue and galvanic corrosion. At aluminium-to-steel transitions, an explosion-bonded transition joint is the standard shipbuilding solution.​‌​​‌​

Heat Treatment and Thermal Stability

There is NO strengthening heat treatment for 5083. Heat treatment on this alloy exists only to soften or to stabilize.​‌​​‌​

Welding​‌​​‌​

5083 welds beautifully — and that is one of the reasons the alloy exists. Manufacturer rating tables give TIG 2, MIG 2, resistance welding 2 (on a scale where 1 = very good); gas welding is weaker at 3–4. One supplier uses the phrase “excellent weldability” directly for TIG and MIG.

Post-weld strength — the honest table

This is 5083’s strongest selling point and also its most overstated topic. The correct statement is: in the heat-affected zone the strain hardening introduced by cold work is lost and the material locally returns to the O (annealed) level. So if you weld an H321 plate, the strength at the edge of the weld is not H321 strength but O strength: the yield drops from ~215 MPa to the ~125 MPa band. So why does everyone say it “keeps its strength after welding”? Because on 5083 even the O condition is high — higher than the full temper of many alloys. Compare: a heat-treatable 6xxx also drops when welded, but it drops further and recovering it requires re-heat-treatment — impossible on a welded ship hull. Design rule: calculate a welded 5083 structure with O/H111 values, not H321 values. The high H321 figures apply only away from the welds.​‌​​‌​

Machining

5083 is not “hard”, it is “sticky”. The manufacturer machinability rating is 2 (good) in the cold-worked condition and 3 (moderate) when soft annealed; another supplier simply calls it “fair”. The problem is not cutting speed — it is that the chip will not break.​‌​​‌​

Corrosion — Where It Excels, and WHERE IT FAILS​‌​​‌​

Where it excels​‌​​‌​

In the manufacturer rating tables 5083 scores 1 (very good) in normal atmosphere and weather and 1 (very good) in seawater — the best score obtainable in the aluminium family. The reason is simple: magnesium makes aluminium’s oxide film more stable in chloride environments, and in 5083 copper is deliberately held below 0.10 %. Seawater, salt air, industrial atmosphere, most industrial chemicals — 5083 works uncoated in all of them. It is suitable for food contact per DIN EN 602. Anodic protection is good too: protective anodising rating 2.

WHERE IT FAILS — sensitisation​‌​​‌​

This is 5083’s only real weakness and the most important paragraph on this page. The alloy carries 4.0–4.9 % Mg; aluminium’s room-temperature solubility for magnesium is far below that. The material is therefore supersaturated and thermodynamically unstable from birth. Given enough temperature and enough time, the magnesium precipitates at the grain boundaries as β phase (Al₃Mg₂ / Mg₂Al₃). That phase is far more anodic than the aluminium matrix: in contact with a chloride electrolyte it dissolves preferentially along the grain boundaries.

Other failure points

Galvanic corrosion. 5083 is active relative to steel, stainless steel, copper and bronze, and when coupled to them 5083 is the side that is sacrificed. In shipbuilding, an explosion-bonded transition joint or an insulated bolted connection is used between an aluminium superstructure and a steel hull. If stainless fasteners are used, insulating bushes and washers are mandatory. Mercury and mercury compounds. Absolutely forbidden for 5xxx. Strong alkalis. Above pH 9 the oxide film dissolves and the aluminium is attacked rapidly. Waters containing copper ions. Copper plates out on the aluminium surface and forms micro cathodes. Decorative anodising. Although the protective anodising rating is 2 (good), the decorative anodising rating is 4 (poor) and the paint/coating rating is 4 (poor) — 5083 is not an architectural visible-surface alloy. For that, 5754 is far more suitable (anodising rating 1).​‌​​‌​

Frequently Asked Questions

Should I buy H116 or H321? My supplier offers both and one is more expensive.​‌​​‌​

For most projects it does not matter, and the price difference is usually stock-driven rather than technical. Both are marine tempers with specified corrosion performance for 5xxx alloys at 3 % magnesium and above; both are tested to ASTM G66 and/or G67; both are approved by classification societies such as DNV, Lloyd’s Register, ABS and Bureau Veritas. The difference lies in the production route: H321 explicitly includes a stabilization treatment, while H116 is not a route but a performance definition — the producer chooses the route as long as the product passes the test. The real decision criterion is this: take whichever your project specification or classification society asks for; if the specification accepts both, take what is in stock. The actual danger is elsewhere: if a supplier offers you H32 or H34, they are offering material with no corrosion test requirement — that is where the price difference comes from, and it is not an acceptable saving on a marine structure.

We are building a hot water tank. Is 5083 suitable? You say its seawater resistance is excellent.​‌​​‌​

No. This is the most expensive mistake you can make with 5083. Its seawater resistance is excellent — in cold seawater. As the temperature rises, a completely different mechanism takes over: sensitisation. In an alloy with 4.5 % magnesium, warm temperatures precipitate β phase (Al₃Mg₂) at the grain boundaries; that phase is anodic and the material opens to intergranular corrosion and stress corrosion cracking. Several independent sources use the same sentence: “it is not recommended for use in temperatures in excess of 65 °C”; the Australian standard AS 1734 states it as a requirement. And the process is irreversible — once a structure has sensitised, it cannot be corrected in the field. What to do instead: for hot service consider a lower-magnesium alloy or a heat-treatable 6xxx (EN AW-6082), and have the design confirmed by a materials engineer. Note: we have seen commercial pages claiming “60–80 °C desalination” for 5083; that statement directly contradicts the 65 °C rule and is not supported by any verified source.

Can we use 4043 as filler? It is what we have and it flows better.​‌​​‌​

No. This rule has no exceptions. 4043 is an Al-Si filler designed for the 6xxx series. When it meets a base metal containing more than 3 % magnesium, Mg₂Si forms in the weld metal; the result is a brittle weld with low ductility that is open to corrosion. The correct fillers for 5083 are ER5183 (SG-AlMg4.5Mn), ER5356 (SG-AlMg5) and ER5087; one source also lists ER5556. ER5183 is the first choice because its chemistry is closest to the base metal and it gives the highest weld metal strength; ER5356 is more common and easier to obtain. The rule: 5xxx base metal takes 5xxx filler. Flow is welder comfort; a brittle weld is a structural risk.

5083 or 6082? Both give similar strength and both are weldable.​‌​​‌​

The decision is made on welding. The base metal numbers really are close: 6082-T6/T651 plate gives Rm 295–310 MPa · Rp0.2 240–260 MPa; 5083-H321 plate gives Rm ≥305 MPa · Rp0.2 ≥215 MPa. The difference appears after welding. 6082 is heat-treatable: welding dissolves the precipitates in the HAZ and that zone falls to the T4 level; recovering the original strength requires re-solution treatment and ageing — impossible on a welded ship hull. 5083 is not heat-treatable: welding returns the HAZ to the O (annealed) level, which is already high (Rp0.2 in the ~110–125 MPa band and Rm above ~270 MPa). Conclusion: 5083 for heavily welded plate structures (boats, tanks, hulls, armour); 6082 for lightly welded, profile-rich structures (chassis, frames, architectural load paths) (EN AW-6082). The third criterion is temperature: if continuous service exceeds 65 °C, 5083 is eliminated.

Common Datasheet Errors and Traps​‌​​‌​

1. “Not recommended for use in temperatures above 650 °C”. We found that sentence on a real supplier page. The correct figure is 65 °C. One extra zero moves the limit a factor of ten in the wrong direction. It is the most dangerous typographical error possible on this alloy and can influence real service decisions. 650 °C is effectively where aluminium melts.
2. “Service temperature 135–145 °C”. One manufacturer datasheet family prints the same generic row for every alloy, and gives 135–145 °C continuous / 180–190 °C short-term for 5083 too. That is WRONG for 5083 and directly contradicts the 65 °C rule. Do not trust generic template rows.
3. Confusing H32 with H321. One digit, two different worlds. H321 carries a corrosion test requirement; H32 does not. Buying H32 for a marine structure means not buying the assurance you think you bought.
4. “5083-T6”. No such material exists. 5083 is not heat-treatable. An offer using that designation does not know the alloy.
5. “It fully retains its strength after welding”. Incomplete. The correct statement: the HAZ returns to the O level, but 5083’s O level is already high. Calculate the welded structure with O/H111 values, not H321 values.
6. Mixing ASTM typical values with EN minima in one table. The commonly cited ASTM figures Rm ~317 MPa · Rp0.2 ~228 MPa are not the same thing as the EN 485-2 H321 minima (Rm ≥305 · Rp0.2 ≥215). State which system you are working in.
7. The H116 yield figure. One supplier sheet gives Rp0.2 min 195 MPa for H116, while the EN 485-2 derived H321 rows give a minimum of 215 MPa. We have seen both; thickness-tabulated EN 485-2 data for H116 could not be verified in this study. Confirm against the standard if you are writing a specification.
8. The annealing temperature contradiction. One manufacturer gives 380–420 °C, another 330–400 °C. We have seen both.
9. “Melting point 570 °C”. Alloys do not melt at a single point. The verified solidification range is 580–640 °C; a single figure is a simplification.
10. Expansion coefficient confusion. Sources give 24.2, 23.8 and 25 × 10⁻⁶ K⁻¹ for 20–100 °C. Never publish a coefficient without its temperature range.
11. Applying a post-weld stress-relief anneal. Do not. It puts the material precisely into the sensitisation band. On 5083, stress is managed by design and weld sequence, not by heat treatment.
12. Using 4043 filler. It produces brittle Mg₂Si. 5xxx base metal takes 5xxx filler.
13. Treating a class certificate as a lifetime guarantee. A certificate documents the condition at delivery. Sensitisation develops in service; the measured G67 mass loss of 19–25 mg/cm² on 40–50-year-old plate is the proof.
14. Using stainless fasteners without insulation. 5083 is active relative to stainless; in a salt environment the aluminium is the sacrificial side. Use insulating bushes and washers.
15. Expecting decorative anodising. Protective anodising rates 2 (good), decorative anodising 4 (poor) and paint/coating 4 (poor). For visible architectural surfaces 5754 is far more suitable.
16. Planning bends in a hard temper. In H34 drawn tube the elongation falls to 3 %. Use O/H111 anywhere bending is involved.
17. Treating G66 and G67 as the same test. G67 is a mass loss measurement (mg/cm²); G66 is a visual exfoliation rating. B928 cites them together.
18. Assuming a density of 2.7. 5083 is 2.66 g/cm³; on a large ship structure that difference runs into tonnes.

​‌​​‌​