Extruding the Hard Alloys: Billet Temperature, Exit Speed and the Bearing That Limits Both
A 7075 shape and a 6063 shape are made on the same kind of press with the same kind of die, and almost nothing about the run is the same. The hard alloys — 2xxx, 5xxx and 7xxx — have a short hot-working range, a high flow stress at the die, and a tendency to melt at the grain boundaries at temperatures a 6xxx billet would run at all day. What changes is the billet temperature, the exit speed, the way the bearing is cut, and the fact that the metal is quenched and aged after the press rather than at it. This page is about those four things. The wall-thickness limits and the extrudability ratings that decide whether a shape belongs on a given press at all are on the circumscribing circle diameter and minimum wall page; this page starts after that decision has been made.
Why the window is narrow
The three families fail for different reasons, and the reason tells you which variable to hold.
The 2xxx alloys are AlCuMg, discovered in 1911 and still the backbone of airframe extrusions through 2014 and 2024. Silicon in 2014 is there to make the alloy respond to artificial aging, which natural aging alone cannot reach; 2024 is the high-strength naturally aged replacement for 2017, and its artificially aged tempers are reached by strain hardening the quenched material. Neither is forgiving at the die, but both have a usable range of roughly 420 to 450 °C of billet temperature.
The 5xxx alloys have a high deformation resistance across the whole working range, so they need high billet temperatures — and the higher the temperature, the closer the run sits to incipient melting. Magnesium is held at 5 to 6% maximum for the extrudable grades because more Mg makes the alloy harder to extrude and more prone to intergranular and stress-corrosion cracking.
The 7xxx alloys are the tightest. The hot-working range is very short because the deformation resistance is high and incipient melting starts at a relatively low temperature. Add copper and zirconium and the maximum extrusion speed falls further; the magnesium content moves it more than the zinc does. That is why the same table that gives 5083 an exit speed of 2 to 6 m/min allows 7075 only 0.8 to 2 m/min.
Billet temperature and exit speed, by alloy
The published starting values are below. The container temperature follows the billet, not the other way round: the container is set about 50 °C below the billet temperature, whichever scale the press is working in.
| Alloy | Type | Billet temperature, °C | Exit speed, m/min |
|---|---|---|---|
| 2014, 2024 | Heat treatable | 420 to 450 | 1.5 to 3.5 |
| 5083, 5086, 5456 | Non-heat-treatable | 440 to 450 | 2 to 6 |
| 7001 | Heat treatable | 370 to 415 | 0.5 to 1.5 |
| 7075, 7079 | Heat treatable | 300 to 460 | 0.8 to 2 |
| 7049, 7150, 7178 | Heat treatable | 300 to 440 | 0.8 to 1.8 |
Billet temperature and exit speed for hard alloys, after Laue and Stenger. The values are for guidance only: the final shape and the extrusion ratio set the real numbers, and it is often necessary to start lower than the table allows. Source values also given in °F and ft/min.
Two things in that table are worth reading twice. The first is that 5083 and 5456 run hotter and faster than 7075 despite being harder to deform in the press. Heat-treatable is not the same as difficult: the 5xxx alloys accept a high temperature because their solidus is high, while 7075 does not because its solidus is not. The second is the width of the 7075 band — 300 to 460 °C covers everything from a thin, wide, hard-to-fill shape at the bottom to a thick simple shape at the top, and the number that matters on the day is the exit temperature, which is the billet temperature plus the heat the deformation adds.
The bearing length is a temperature control
Extrusion ratio is fixed by the shape and the container, so temperature and speed are the only two process variables the press actually has. On a hard alloy the die adds a third, and it is the one most often left at whatever the die was cut with.
Bearing length controls metal flow and friction, and friction generates heat exactly where the surface is being made — the outer skin, a few tenths of a millimetre from the land. The longer the bearing at a fixed ram speed, the more shear deformation and the higher the local temperature rise. Push that past the alloy’s limit and the results are predictable and specific:
- Surface cracks, starting as fine transverse tears on the faces that carry the most bearing
- Peripheral coarse grain, where the surface layer recrystallizes because it was hotter than the core
- Die drag, from aluminum adhering to the bearing and tearing the surface as it passes
The recommendation that follows is to optimize bearing length for the extrusion ratio rather than inherit it from the drawing, because a bearing that is too long costs speed twice: once because the press cannot push faster without overheating, and once because the surface it produces will not pass inspection. The bearing is also the cheapest die change to make, which is why it is the first place a correction starts — the order of corrections is set out on the die correction page.
The oxide in the container costs more than the die
Aluminum adheres to steel. The billet’s oxide skin does not slide on the liner; friction holds it back, and the dummy pad shears the surface metal off into the container as it advances. After upset the new billet’s oxide mixes with the layer already in the container, and the layer grows billet by billet until the amount reaching the extrusion is no longer acceptable. On an aerospace shape, that is the failure mode that matters, because entrapped oxide becomes a surface or subsurface blister, or a stringer inside the section.
The practices that hold it in check are procedural, not metallurgical: run a clean-out block of the correct size every 15 to 20 billets and at every alloy change, use a scalped billet, and keep a correct temperature differential between billet and container. That last one is counter-intuitive. A large differential arrests the oxide on the container wall, where the clean-out block can take it out; a small differential lets the skin separate and flow into the extrusion instead. The geometry of the dummy pad and the clean-out block, and the clearances that make them work, are on the container liner and dummy block page.
Butt thickness, and what happens at the back end
In direct extrusion the center of the billet moves faster than the periphery, so the dead-metal zone runs down the outside of the billet and, after about two thirds of the billet has gone, the surface layer turns inward toward the die. Everything that was on the surface — Al2O3, MgAl2O4 — goes with it. Seen on a transverse section, the result is an annular ring of oxide, the internal pipe. It gets worse as container friction rises and as the container runs colder than the billet, because a cold outer layer has a higher flow stress and the center then outruns it further.
The butt is the control. A butt that is thin relative to the billet diameter lets radial flow dominate, and the back end of the extrusion is left with axial holes or funnels — piping — that can run a considerable distance forward from the back face. Heavy extrusions at low ratio are the worst case, because the radial flow is stronger. Macro-etching the back end is how the piping is found before a customer finds it in machining.
Two related defects are worth knowing by name. Internal voids in a 7016 bumper extrusion were traced to feed-in from the butt: material from the depleted, liquated zone next to the billet surface had been drawn into the section, and the voids appeared well away from the blister that first drew attention. And a 356 mm 7075 billet produced a duplex structure at the butt end on a standard practice, which is a reminder that the back end of a hard-alloy run is a different material from the front.
Quench and age: the part that happens after the press
6xxx is quenched at the press and aged in a batch oven. The hard alloys are not: solution heat treatment for 7075 and 2024 is done in a separate temperature- and time-controlled furnace with moisture control, followed by a controlled water or polymer glycol quench, and only then by stretching, sawing and aging. The cooling rate is a specification, not a preference — for 7075 and 2024 it should not be less than 800 °F (445 °C) per second through the 750 to 550 °F (400 to 285 °C) range.
| Alloy | Solution temperature | After quench | Aging temperature | Aging time | Final temper |
|---|---|---|---|---|---|
| 2014 | 500 °C (935 °F) | T4 | 160 °C (320 °F) | 18 h | T6 |
| 2024 | 495 °C (920 °F) | T3 | 190 °C (375 °F) | 12 h | T81 |
| 7001 | 465 °C (870 °F) | W | 120 °C (250 °F) | 24 h | T6 |
| 7075 | 465 °C (870 °F) | W | 120 °C (250 °F) | 24 h | T6 |
Typical heat treatment practices for hard alloys. Soak time in the solution furnace is set by section thickness; temperatures and aging times move with the specification and the supplier’s practice.
What is being dissolved is the hardening phase: CuAl2 and CuMgAl2 in the 2xxx, Mg2Al3 in the 5xxx, and MgZn2 with the copper phases in the 7xxx. Quench too slowly and the phase comes out as coarse precipitate at the grain boundaries instead of staying in solution, which is where the loss of tensile strength and the susceptibility to intergranular corrosion after aging both come from.
Blisters: hydrogen, and where it enters
On thin-gage 7075, a blister is a bubble on the surface that appears either at the press or, more often, during solution heat treatment. It is hydrogen. Hydrogen is the only gas appreciably soluble in solid or molten aluminum, it enters during melting, casting, heat treatment, cleaning and pickling, and it migrates to inclusions and high-stress sites where it recombines into molecular hydrogen that cannot diffuse out. Heat the part and that trapped gas expands the cavity under it.
Four controls decide how bad it gets. Heat treatment time, because the deposition accumulates. Temperature, because mobility rises with it and a furnace that overshoots into the incipient-melting range will blister parts that were fine on arrival. Grain boundary area, because a fine grain spreads the same amount of hydrogen over more boundary and produces many small voids instead of a few large ones. And alloy chemistry, because high-magnesium alloys have a less continuous oxide film and pick up hydrogen more readily.
Two shop-floor findings from the source are worth carrying. A volatile fluoride compound — sodium fluoborate is the one named — introduced into the furnace as a drying agent reduced blistering in both furnace types tested, and the type of furnace made no difference either way. And a protective corrosion inhibitor applied at the mill, on extrusions in the O temper, minimized blister formation during solution heat treatment; the same material supplied in F temper blistered more readily than in O.
Where this leaves a die steel order
Hard alloys punish the die in a way 6xxx does not: the window is narrower, the bearing has to be shorter and better finished, and the pressure is higher. The die material is still chromium hot work steel, AISI H13 and its European equivalent 1.2344, and it has to be supplied in a condition that lets a bearing be re-cut after the first trial. The working hardness and temper are on the H13 tool steel page, the ESR programme for extrusion tooling on the 1.2344 ESR H13 supply page, the bearing itself on the die bearing length page, the wear that follows a long run on the die wear page, the stack and the hollow-die options on the die types and tooling page, and the grade comparison across the range on the best tool steels for extrusion dies page. The temperature window itself, series by series, is on the extrusion temperature page.
Before you set a press from this page
These are published starting values and the reasons behind them, not a schedule for a specific press. Billet temperature, exit speed and bearing length are set by the shape, the extrusion ratio, the container and the alloy lot, and the only number that transfers between shops is the exit temperature limit. Confirm every figure against the alloy specification and your own first-article run before committing a billet.
Source, Aluminum Extrusion Technology, P. K. Saha, ASM International, 2000, chapter 7, extrusion of hard alloys, pages 188 to 203.
