Remove warp from aluminium plate?

[aneevuser]

I have a sheet of 1050 grade, 6 mm aluminium, dimensions 250 mm x 175 mm.

After drilling, it seems to have acquired a warp. I've attached images. The line at which the deviation from flat starts is marked in black on the plan view image.

When the sheet is clamped at one edge, the maximum deflection at the other is about 2 mm.

So the obvious question: is there any way that I can remove this warp without specialist tools? - I don't want to have to mill it, for example.

[T3sl4co1l]

Turn it over and slap it gently with a hammer until it lies flat against your platen.  Flip and repeat until acceptably flat.

Mind that heavy hits deform the material, pushing it out from the struck location -- you want just enough force to bend out the flex, not so much that the material is displaced.  Continued heavy hammering causes stretching, which is great for making 3D shapes, and, obviously, what you need to avoid here.   Example: hammering in the middle causes the material to stretch there, making a domed shape.  Hammering the edges causes it to stretch there, making a kinked / saddle shape.  Hammering evenly over the entire surface, balances these out of course, stretching and thinning the whole sheet.

If you're actually in such a situation, i.e. it doesn't lay flat after correcting the bend, it seems to kink itself into shapes, then it may need to be stretched in just the right places to counter that.  This is a bit nontrivial and best developed as a skill, so, not really in scope here.

And at which point, you're maybe better off heating it to annealing temp (400C or so), to relieve the stress, and then just tap it flat.  But that's probably also the last thing you want to do on already extremely soft 1050 aluminum... it'll be little better than metallic putty after annealing(!).

Tim

[aneevuser]

Turn it over and slap it gently with a hammer until it lies flat against your platen.  Flip and repeat until acceptably flat.

Yes, that's the obvious approach but I didn't want to try it without good reason, given the relative softness of the plate - seems a bit too easy to go horribly wrong.

I have tried bending it against the warp with G clamps, but that does nothing, AFAICS - simply rebounds after pressure is removed.

And at which point, you're maybe better off heating it to annealing temp (400C or so), to relieve the stress, and then just tap it flat.  But that's probably also the last thing you want to do on already extremely soft 1050 aluminum... it'll be little better than metallic putty after annealing(!).

I don't want to try annealing it, not least as I can find no entirely authoratitive info. on how to un-anneal the stuff, or if it is even genuinely possible without specialist ovens (lots of waffle about "time hardening" or something, and work hardening by hammering, neither of which appeal).

Would it help if I heat it to sub-anneal temperatures while bashing it with a hammer, or is that pointless?

[Ian.M]

Would it help if I heat it to sub-anneal temperatures while bashing it with a hammer, or is that pointless?

Pointless - aluminum is 'hot short' so working it hot vastly increases the risk of cracking.  Also holding many aluminum alloys at elevated temperature can result in grain growth, with increased intergranular precipitates, also increasing the risk of cracking

[T3sl4co1l]

Yeah exactly, you can't harden it again, it's not a hardening alloy -- you could perhaps do that with 6061 or what have you, where you could anneal, quench from annealing temp, then age in the 200C ballpark for some hours (exact time/temp depends on alloy and condition) for a, T5 I think temper.  But 1xxx, 3003, etc. don't have the alloying elements that allow this kind of hardening cycle to work -- your only option is to hammer/roll/etc. until work hardening gets it there.  Which presumably, your plate already is, so it's as good as it's going to get.

That it doesn't take a set from being clamped flat, is encouraging -- it's not dead soft.  Having strength, means it has to be bent past flat, to get into the plastic deformation range, before it will take a set.  It's... not so useful of a property to you right now, but it's also a necessary property in general.

Which is kinda where hammering comes in, because hitting it tends to over-bend it (material curves up from the anvil).  So, hit it a few times against the curve, and at some point it should come flat.  Start with light taps, and get harder until it's taking a set.  Do it evenly over the curve, get it flatter; turn it over, correct the parts you over-corrected, etc.  Also, use a smooth clean hammer, don't want to mar the surface too much.  Can use plastic, cardboard, leather, etc. to soften and spread out the blow too.  Soft-blow hammer, rubber mallet, etc.

Also, the fact that it's work hardened, means there's internal stress in it, too -- this is probably not what caused the bend in the first place, not from just drilling anyway; maybe you dropped it, or cranked down on it too hard while drilling, or something... well, whatever, it is how it is now.  But this internal property is important when, for example, milling into the surface: relieving some of the stress from one side (by literally removing the material there) causes the other side to take over, and it warps towards or away from the cut.  Again, likely not what's happened here, but important to know when doing stuff like that.

Tim

[jpanhalt]

Find someone with slip rollers who knows how to use them.  6mm may be a bit thick for some, but not impossible given the alloy.

[mathsquid]

What if you put it between two pieces of plywood (to protect the surface) and parked a car or truck on it overnight? I'm not sure if that would be effective or not.

[free_electron]

a word of advice ..

[andy3055]

Whatever you do, don't use a hammer. Use a mallet preferably a wooden one.

[CatalinaWOW]

This is like many other tasks.  Some folks seem to be born with the knack for correcting these kind of issues.  Others can develop the skill with practice.  And fortunately only a few find it forever beyond them.  Most people are not in the first category, so if this workpiece is important to you practice on something else a few times to develop some feel for what is required.  It is not that hard to make it irrecoverably much worse.

[beanflying]

Chances are whatever you do to reverse that bend caused by the saw cut you will add reversed stress into a bit of material still containing residual stresses. This sort of Aluminium is formed by rolling to a Temper in one direction and it is not normalised in anyway after as far as I know or remember.

If you have more material cut the same blank at 90 degrees and run the cut and leave it. I would bet you get a very different result including little to zero warp.

Other than that you will need likely to look at a different grade of Aluminium. You might get lucky locally and find an offcut of Tooling Plate which is typically Cast. There was some bits floating around on evilbay US a while ago.

This Search or similar Mic-6 ATP-5 etc https://www.ebay.com/sch/i.html?_from=R40&_trksid=p2334524.m570.l1313&_nkw=%28mic-6%2C+atp-5%29&_sacat=0&LH_TitleDesc=0&_odkw=mic-6&_osacat=0

[aneevuser]

That it doesn't take a set from being clamped flat, is encouraging -- it's not dead soft.  Having strength, means it has to be bent past flat, to get into the plastic deformation range, before it will take a set.  It's... not so useful of a property to you right now, but it's also a necessary property in general.

I may try more agressive clamping - I would prefer not to have to hammer it.

Also, the fact that it's work hardened, means there's internal stress in it, too -- this is probably not what caused the bend in the first place, not from just drilling anyway; maybe you dropped it, or cranked down on it too hard while drilling, or something... well, whatever, it is how it is now.
Tim

Well, the warp formed after I drilled it, and it hasn't been dropped - the holes are countersunk on the other side, so more material has been removed than you can see from the image.

Regardless, the bloody thing has deformed, so the cause is not overly important right now.

[aneevuser]

Chances are whatever you do to reverse that bend caused by the saw cut you will add reversed stress into a bit of material still containing residual stresses. This sort of Aluminium is formed by rolling to a Temper in one direction and it is not normalised in anyway after as far as I know or remember.

I didn't cut this - I bought it cut-to-size, and it was fine on arrival. The warp has been caused by drilling, AFAICS - no warp before drilling, warp after drilling.

Other than that you will need likely to look at a different grade of Aluminium. You might get lucky locally and find an offcut of Tooling Plate which is typically Cast. There was some bits floating around on evilbay US a while ago.

Is tooling plate less likely to have residual internal stresses? If so, may be useful to know about for future reference.

[aneevuser]

This is like many other tasks.  Some folks seem to be born with the knack for correcting these kind of issues.

I suspect that I fall into the Father Ted category.

[aneevuser]

a word of advice ..

That's one of the most useful documentaries on metalwork that I've seen. Not sure that I've got one of those Irish hammers though - I suspect that they make all the difference to the end result.

[beanflying]

Chances are whatever you do to reverse that bend caused by the saw cut you will add reversed stress into a bit of material still containing residual stresses. This sort of Aluminium is formed by rolling to a Temper in one direction and it is not normalised in anyway after as far as I know or remember.

I didn't cut this - I bought it cut-to-size, and it was fine on arrival. The warp has been caused by drilling, AFAICS - no warp before drilling, warp after drilling.

Other than that you will need likely to look at a different grade of Aluminium. You might get lucky locally and find an offcut of Tooling Plate which is typically Cast. There was some bits floating around on evilbay US a while ago.

Is tooling plate less likely to have residual internal stresses? If so, may be useful to know about for future reference.

Yes as it is cast rather than mechanically worked into shape via a roller or forced extrusion process. There is some in the UK on evilbay marketed for 3D printer beds too eBay auction: #265311533227

The long cut will be why it warped the stress relief from the holes will be minor in comparison.

By all means try reversing the bend and holding it too, unlikely you will break it but also it is unlikely you will get it 'flat' and have it stay that way. You could also try a long soak of a few hours in your home oven to try and reduce the stresses (normalize) it but generally you need a higher temperature typically 300-400C the downside to this is you will also change the temper (generally soften it).

[aneevuser]

Yes as it is cast rather than mechanically worked into shape via a roller or forced extrusion process. There is some in the UK on evilbay marketed for 3D printer beds too eBay auction: #265311533227

Thanks. That is very useful to know. These residual stresses in formed metal seem to be quite a problem - I'm no great metal worker, and am quite surprised by how tricky it is to work some materials.

The long cut will be why it warped the stress relief from the holes will be minor in comparison.

I'm not sure what you mean by "long cut" - I haven't cut this, I have only drilled it.

By all means try reversing the bend and holding it too, unlikely you will break it but also it is unlikely you will get it 'flat' and have it stay that way. You could also try a long soak of a few hours in your home oven to try and reduce the stresses (normalize) it but generally you need a higher temperature typically 300-400C the downside to this is you will also change the temper (generally soften it).

I think that I shall avoid any kind of heat treatment - seems to easy to screw alu. up by doing that. If my hammering goes horribly wrong, I may treat this piece as a learning experience and see what happens when I heat it.

And re: tooling plate - it looks like there's a hefty premium over standard rolled plate.

[jpanhalt]

I think "the long cut" may apply to the sheared edge in your top photo.  Although, the curl looks to be in both directions.  One sees curl like that when relatively narrow strips are sheared from larger plates.  Are you sure there was no curl before drilling?  Was it fully supported while drilling?

As mentioned slip rolling would be my first choice as it is easiest to control.  If you have a press, my second choice would be to try to bend back between two supports.  I would put a 3rd piece under it to limit the amount of deflection.  Assume you will need to go past flat for the bend to take.  I prefer to bend less and repeat than to straighten from the other direction.  I tend to go about 5° past vertical for 90° bends starting with flat stock.

[beanflying]

The obvious shear marks on the photo above. When you get to thicker sheets like this a shear really becomes an issue as it is not a simple knife moving vertically down against a lower blade but the two are at a slight angle so it is a shallow scissor type of action which tend to create a curling force. There is ways to minimise the deformation and stresses but when you shear metal but over a certain thickness best thing to do is look at another cut method. This will vary a lot depending on the shear as it is quite possible to shear stock of 25-30mm with the really heavy ones. Preferably I guess you would use some form of cold saw over a certain thickness but this comes at a time cost. So Bandsaws, Cold Cutting saws or Waterjet are some of the alternates all of these put very low stresses into the metal across a longer length.

Typically as your part was fairly small it likely came off the back side of the shear meaning it wouldn't have been supported by the clamping jaws of the shear either. That is just speculation without knowing your source but it would be uncommon to poke the large part of the sheet out the back and keep the small bit to hold onto.

[aneevuser]

OK, with much more vigorous clamping than I tried before, I have eliminated most of the deflection. I've attached a couple of images for comparison.

It's not perfect, but good enough for my purposes, and I don't want to go any further as I seem to be introducing a small amount of deflection along the long axis, as I reduce it along the short one.

Anyway, thanks to all for the suggestions and info.

[aneevuser]

I think "the long cut" may apply to the sheared edge in your top photo.  Although, the curl looks to be in both directions.  One sees curl like that when relatively narrow strips are sheared from larger plates.  Are you sure there was no curl before drilling?  Was it fully supported while drilling?

No, it was fine before drilling, and it was indeed supported fully. So it was my drilling that caused the problem, but I can't tell you why.

[andy3055]

Deforming happens in this kind of work due to heat. Possible causes can be blunt bits and too much force. If you pour water or coolent used for this sort of work, it should stay flat. That is in addition to clamping sufficiently. Also keep the clamping on till it cools down.

[aneevuser]

Deforming happens in this kind of work due to heat. Possible causes can be blunt bits and too much force.

That's a useful observation. I suspect that if heat is the problem, then it was due to the countersinking - I countersunk an M8 hole without coolant, and I have a feeling that I noticed the warp after I'd done that.

If you pour water or coolent used for this sort of work, it should stay flat. That is in addition to clamping sufficiently. Also keep the clamping on till it cools down.

Yes, I tend to be a bit lax with coolant - need to pay more attention in future. I rarely drill plate though, and heat doesn't seem to be much of a problem with box section.

[ndzied1]

Stresses in sheet and plate is very common due to the "memory" effect mentioned in this article. A plate that looks flat often warps after some machining operation (like drilling) that releases some of the stress in one place that was balancing out the plate. Now with unbalanced stresses, it warps. So it's probably more correct to say that the drilling released memory stresses in the plate rather than the drilling being a direct cause of the warpage.

https://www.ryerson.com/resource/the-gauge/temper-pass-vs-stretcher-level