Avoiding warping while machining 1018

dksoba · Aug 7, 2015

I need to machine a plate of cold rolled 1018. It's 1/2"x6"x27" and will be machined down to 0.3937" (10mm). Thickness isn't critical, but flatness is kind of critical. It should probably be within +/- 0.005" over the 27" length.

Based on the searching I've done and my experience machining cold rolled stainless, I know (read: assume) that this material is going to warp due to removing internal stresses.

Here's my strategy:

1) Use epoxy to hold the plate to a fixture plate (with minimal stresses due to work holding), and then machine away a little (how much... not sure).
2) Flip the plate over, use epoxy again, and machine a little off the other side.
3) Repeat 1 and 2 until I get close (maybe 0.01") to the final thickness, then switch to a vacuum fixture for the last facing operation.

Does this have a shot of working? How many times would I have to go through 1 and 2? Is there an easier way?

My customer specified that the yield strength should be around 50,000 psi, and is willing to switch steels as long as that constraint is met. I wanted to go with A36 hot rolled (because I assumed with hot rolled I'd have less of a problem, but please let me know if I'm right or not on that) but the yield strength is a bit lower.

Matt

Atomkinder · Aug 7, 2015

Hot rolled is always going to yield a better result, and that's my first suggestion. Second is go ahead use 3-4 vises. Gluing it is going to introduce all sorts of cutting force issues I think, and won't necessarily give you the result you're hoping for.

Spencer in NH · Aug 7, 2015

Have the material stress-relieved by a competent heat-treating shop.

dksoba · Aug 7, 2015

Or what about making it from 4140? This won't warp will it?

Spencer in NH · Aug 7, 2015

Any material that is extruded, rolled or forged will have internal stresses. A cast material, if specifically processed, can be stress free. Hot rolled steel is rolled while hot (!), and therefore has less stress in the stock. The material composition matters less than the manufacturing process.

Or you can have the stock stress-relieved.

Derek Smalls · Aug 7, 2015

Atomkinder said:
Hot rolled is always going to yield a better result, and that's my first suggestion. Second is go ahead use 3-4 vises. Gluing it is going to introduce all sorts of cutting force issues I think, and won't necessarily give you the result you're hoping for.

That's the way I would go about it. I've done this sort of thing several times and the multiple vises work great. Gluing will create all sorts of problems most likely. Part could very well spring while being machined and come partly or completely UN glued. That would make a real mess! You may even need to do a straightening op before your done, just depends on what that particular piece wants to do. The stresses are more concentrated on the outer most surface, that is to say, the deeper you get into it, the less stress you will be dealing with. At first, until both sides are clean, try to keep the surfaces parallel, and remove same amount from each side, even if you need to cut steps to do this. In other words, if when you get to the second side the plate is high in the middle(which it will be) don't machine it flat, take the same amount off everywhere, leaving it high in the middle. If you take .030 at the ends, back off in the middle so you only take .030. This will keep things balanced and ultimately it will stay as flat as possible. Then go back to the first side and try to get it reasonably flat.
You really should be using hot rolled if possible.

paracongo308 · Aug 7, 2015

You take .045 off one side and unclamp and you will have a slight bow, then flip over clamp down and taking the bow out, cut the 2nd side .045 Now once you unclamp part the bow will be removed. Used to work with 1018 making huge headed keys up to 3 inch square 18 inch long for some very large gears & shafts. The key was to 50/50 the material removed. The keys would be machined in a 24" G&E shaper and they would come out pretty damn flat. After you rough out your .045 per side, repeat with final .007-.008 per side.

scadvice · Aug 7, 2015

Shop tools used to carry that thickness (10mm) but I think it was only 150mm wide. 36" long, as I recall...
Do a search usshoptools...

You could also find a double disc grinder who can grind a plate for you. Machining it can be a nightmare even stress relieved. Also check with TCI, I could be wrong, but I understand they are doing ferrous materials now. They are close to you.

dstryr · Aug 8, 2015

Buy 3/4 plate and have it ground to size. Cutting that will blow...

Can it be done? Sure. Do you want to do it? Not really.

adama · Aug 8, 2015

Why can you not just start with a bit of 10mm plate? If you get the pickled and oiled laser grade material you should be good to go.

Even if its a bit bowed, pressing it flat is easy enough to get into your tolerance band, certainly a sub 10 minute job on a manual shop press

Mike1974 · Aug 8, 2015

dstryr said:
Buy 3/4 plate and have it ground to size. Cutting that will blow...

Can it be done? Sure. Do you want to do it? Not really.

I second that. Blanchard it top and bottom to remove the scale and relieve some stress, send out for stress relieve op at a heat treat place, blanchard to final size, should be relatively easy. I had to blanchard some crs bigger than that, its a bit of a pita, but better than trying to mill it.

edit: possibly even skip the stress relieving....

DMF_TomB · Aug 8, 2015

have metal stress relieved or annealed this will lower warpage considerably.
.
you always rough machine as evenly from both sides as possible and rechuck for finish cut. that is let the part assume any shape it wants not beating it flat in the vise and take a light finish cut with vise at low clamping pressure.
.
what good is forcing a curved part flat then machine it more but when unclamped it springs curved again.
.
large castings are held on airy points. deliberately not on the ends or fully supported. they warp and are machined flat. finish cuts are often .017" then .002 then .001 , etc. held correctly they are supported but are free to assume the shape they want. there is a science to it. thin pieces are more tricky
.
https://en.wikipedia.org/wiki/Airy_points
.

hanermo · Aug 8, 2015

Your strategy is good and will work, if you use a proper glue.
Imho.

Likely, any of the metal filled epoxies would work.
They can resist 10x more forces than machining it flat will need.
About 1000 kg/sq cm is common, 3478 locktite/hysol for example.

You may need twice per surface, perhaps just twice per first surface.
Its not a tight spec.

Machining strategy may be important.
Heat may affect local errors, or not, if you use coolant.

DMF_TomB · Aug 8, 2015

rechucking or loosening vise and letting part spring the way it wants and reclamping lightly for a finish cut is a standard technique taught to new machinist.
.
using glue would be slower and when removed from glue can still easily warp. long pieces are usually held in 2 to 4 vises all in alignment side by side

desim · Aug 10, 2015

Had a tread about making straight edges. The customer had spec'ed 1018 and we made some that were ok, but the 1018 warped up to 0.021" over 96", Talked him into A36 and he sent an e-mail back these were the best straight edges they had received in 20 years. The A-36 straight edges had one piece that was warped 0.003" but the rest of the pieces were less. A-36 has a lot less stress than 1018 as a rule.

dksoba · Aug 10, 2015

desim said:
Had a tread about making straight edges. The customer had spec'ed 1018 and we made some that were ok, but the 1018 warped up to 0.021" over 96", Talked him into A36 and he sent an e-mail back these were the best straight edges they had received in 20 years. The A-36 straight edges had one piece that was warped 0.003" but the rest of the pieces were less. A-36 has a lot less stress than 1018 as a rule.

A36 (according to wikipedia) has a yield strength of 36,000 psi, which is too low per my customers requirements. I will try to talk them into it.

Also, what options do I have for hot rolled or stress relieved materials in the 1/2" x 6" x 26" range? Where do I get them quickly? My customer mentioned 4140 which it seems I can get in a normalized state. According to AISI 4140 Alloy Steel (UNS G41400), it looks like 4140 has a yield strength of 60,200 psi, which is better than 1018. However, it appears I can only get 1/2" x 6" x 24" long pieces (from McMaster-Carr), which isn't long enough for my needs.

Matt

TDegenhart · Aug 10, 2015

Why not buy tooling plate?

Tom

Pattnmaker · Aug 10, 2015

See if you can get 44W which is a Canadian spec for structural steel similar to A36 but it has a higher yield 44,000 psi which is closer to your 50,000

desim · Aug 10, 2015

We work with annealed 4140 and it is a very stable material, more expensive, but stable and machines well.

dksoba · Aug 10, 2015

desim said:
We work with annealed 4140 and it is a very stable material, more expensive, but stable and machines well.

Where does one buy this?

Matt

Avoiding warping while machining 1018

Hot Rolled

Titanium

Stainless

Hot Rolled

Stainless

Stainless

Banned

Titanium

Diamond

Diamond

Diamond

Diamond

Titanium

Diamond

Aluminum

Hot Rolled

Diamond

Stainless

Aluminum

Hot Rolled

Similar threads