[Apples]

Imagine a welded structure 4"x4" box section say 1/4" wall thickness. Welded up in a rectangle 3ftx3ft. eg a base and table for a CNC machine.

We all know that it will warp and bend as you weld it and just after when it cools. But after that what happens long term.

Say a day later, a week later etc will it still keep pulling and twisting and wanting to warp? Yes you can stress relieve, but that is now what I am asking.

Peter

[jsheerin]

Maybe someone with lots of welding experience can post what they've seen, but until then I'd say it's impossible to say for sure how much something will move. It would depend on how you welded it.

[lgalla]

Brazing puts less stress on steel as the temperature is around 450C which is well below the melting point of steel.The frame would not warp as much as welding.Brazing can be as strong as MIG or TIG.
Larry

[wcarrothers1]

If you avoide heating to much as you weld you will not have issues..

Pretty simple process. If you are welding 2 pieces together for example. You might be tempted to weld a beed complete from 1 end to the other, That will cause a lot of heat build up and will bend as it cools (become a pretzil) instead, weld several maybe 1" long beeds perhaps 12" apart down the length, Then come back and offset that with a 1" beed between the 2 at 6" from either..

same goes for joining ends to legs. Weld small beeds and come back adding more later not alowing it to get to hot (by not laying a beed all the way around the joining part at 1 time)..

Pretty simple process.. I'm no welding expert, but have seen plenty of welders put a couple beeds down, and say those will be stronger then one contigous beed will ever be. Think it has to do with the flex (or lack of flex) in a single beed vs how multiple beeds would perform.

When you see welders work, they will never follow eachother down a section they are working on. they will either work toward or away from eachother, again to minimize heat built up in 1 area.


b.

[Apples]

That is called "backstep". I'm not worried about the Heat Affectd Zone (HAZ) and the frame warping and twisting.

I am asking what happens later on in a days time after the whole thing has cooled back down to air temperature? Will it still want to slowly "wiggle" it self into a comfortable position. If so how long will this take?

[alkhemist]

Two assumptions:
A1. Assuming you're welding low carbon tube - so assume 50ksi yield strength of the base metal.
A2. You're probably welding with a 7018 (or comparable) electrode/wire - 70ksi yield strength weld metal {this is the most commonly use electrode/wire for general purpose structural welding of low carbon steel} .

As pointed out (and you acknowledge being aware of), it best to keep heat input to a minimum for the sake of thermal distortion

It is also best keep the amount of weld metal to a minimum to limit the amount of base metal yielding (which, I am reading, is the basis of your concern). The weldment, due to solidification shrinkage of the weld metal, will develop and exert full yield strength (70ksi) upon the base metal pieces. Thus we end up with 70ksi weld forces pulling of 50ksi yield strength base metal. If the weldment is too big in comparision to base metal thickness, the base will give first through distortion (best case) or cracking (worst case) [beware of using higher strength weld metals as the distortions can be dramatically increased - more (strength / weld size) is not always better].

It has been my experience with low carbon steels and compatible filler metals, that weld shrinkage strain relief, through base metal creeping or cracking, may not be noticeable within a 24 of welding (assuming cooling to an ambient room temperature air) and are normally complete within 48 hrs. After this time frame, what you've welded should be stable.

[cornbinder23]

Actually that is called stitch welding, backstep welding is when you make a series of tacks and start at the first tack, welding back to you, then moving to the next tack and welding back to you so on and so forth,
What you are trying to do is break up the harmonics associated with the heating and cooling cycle of the material, as it tends to cool at the same rate as the weld is layed down, which will definitley cause warping issues

Also stitch welding is recommended in high vibration processes

I am not sure what the materials inherent "creep" would be though
I dont think it would be an issue with the short spans on these machines although I have seen 16" I beams sag significantly over time and long spans,
but that is also quite a bit of mass


JTCUSTOMS

[Apples]

Okay I'd be welding channel section 6"x3" with 3/8 wall. In the shape of a square or rectangle about 2-3ft square. Small CNC frame etc. So let it sit for 48hrs and it should be stable, sounds good.

Yep it will just be normal mild steel and welded with ER70S-6.

I have seen 16" I beams sag significantly over time and long spans,
You think that those beams are strong and rigid, but when you see them bouncing around and bowing on the back of a semi trailer you realize that everything my gave give and sag to some degree.

[Fastest1]

Do you have a cnc mill working now? Is it mounted to a base?

[Apples]

Nope, this is the mill base I'm making. That is why I'm interested to know how much it may move.

[jsheerin]

Most of the stuff I've read about welded (or cast) parts moving over time without some kind of stress relieving treatment gives the time frames in years or thousands of years - not two days. That's what I thought you were asking about. Reading Slocum's Precision Machine Design last night (good bed time reading - puts you right to sleep), he was talking about stress relieving through heat treatment or vibration. He said if you let the frame sit around for a few thousand years, all the residual stresses would work themselves out and the frame wouldn't move any more. I've also read that seasonal temperature changes can cause a frame to warp when it hasn't been stress relieved. The question is still how much - if it's a tiny amount it might not be a problem for a router.

[alkhemist]

jsheerin: Agreed, unless one uses a stress relieving treatment, there will be very long term residual stress locked in to a welded fabrication. I didn't see this as what Apples' was asking. I took the question to be more practical – “if I welded it - how long before any appreciable/noticeable unexpected movement (due to welding induced stresses) will stop”. From almost two decades of personal experience with large scale steel fabrication of thick plates and shapes, I have found that the practical observable limit of gross weld stress induced distortion and/or cracking is within 48 of the last arc being completed. That is, again, not to say that all internal residual stresses are relieved, because as you correctly point out; they are not. However, after the 48hr window, the remaining residual stresses are very unlikely to cause any further measureable dimensional changes.

Apples: You commented: “I have seen 16” I beams sag significantly over time and long spans…”. The degree of “sag” is dependent upon a number of things including span and shape. I beams are much stronger in the “I” position (due to shape geometry) than in the “H” position, so one will expect see more side to side sway (sweep) under load. Due to the same principle of lever-action, less force is also required to induce sweep in a longer beam; than in a shorter beam. Steel being very elastic, will return to its original position (unless the force exerted is in excess of the steel’s yield strength – in which case a permanent bend will be set) once the force is removed.

In the case of your 3’ long channels (MC6x12’s it would appear), for a router application, I doubt if you will be loading them to a degree that you will see load induced deflection. If it is still a concern, the following link will provide you with a free excel spreadsheet tool that will do most of the engineering math for you (to get you started: values for an MC6x12 are: d=6.00”, bf=2.497”, tf=0.375”, tw=0.310”, Ix = 18.7, Zx=5.53, Sx=6.24, Iy=1.85, ry=0.724, J=0.155, cw=11.3 - for the rest; you’re on your own – this makes my head hurt).

View Document - American Institute of Steel Construction << beamanal13.xls - from steelTOOLS.org

[Apples]

Yeah well that is what I thought, that it took a hundred years give or take a few decades hahaha....to naturally stress relieve.

Well the idea is to make a small metal cutting mill out of steel. How do they stress relieve. I think they have to stick the whole weldment into and oven for a few hours don't they. Or they can go the other way and freeze it (like engine parts) in liquid nitrogen and that does something to it too, in terms or stress relieving it?

I do not think that I can just wave an oxy torch over it to stress relieve it?

[jsheerin]

Hold at 1100F for 1hr/inch of weld thickness is what I found. For cooling, change temperature at 400F/hr / max thickness of material in inches, with the rate to be between 400F/hr and 100F/hr. Something like that is what I plan to do for my mill.

[Apples]

Hmm.

Do they have to stress relieve cast iron structures as well? I'm trying to find out why they use cast iron instead of normal mild steel.

[jsheerin]

Yes, they do afaik. Cast iron has more internal damping than steel which is better to make machines from. It's easier to cast a big part than weld it up a lot of the time. Another advantage that might not be such a big deal anymore is that cast iron was easier to scrape.

[Apples]

I'm thinking weld up a frame then epoxy granite it. Or just bolt up bits of thick flat sections to make a frame then epoxy granite. Bolting will get rid of the heat and warping issue but many not be as strong. But the epoxy mix should help hold it together with a bit of clamping force as it sets, I would think.

[Benonymous]

I'm building my frame out of 50 x 75 x 4 mm box section. Mine is all bolted to preclude the problems of welding stresses. When I was discussing my build with a friend of mine, who is a very experienced machinist, I proposed milling the mounting surfaces for my rails. He said to measure them with a straightedge first and hopefully leave them as they came because milling the top off the section could easily make it twist or warp. The internal stresses would cause this he reckoned. Anyway, the box section was very straight and consistent so it doesn't need to be machined. I just thought it was interesting that milling some of the material off could cause a stress problem too.

[Apples]

Yeah I too have heard that about milling the box section. I am still toying with the idea of bolting it all up for that very reason as well.

Then you just use shims to level it up and get it straight.

[Boomslang]

I did some serious welding on my mild steel frame, and the end result is straight to within 0.2mm. After the bearings are attached and spoilboard cut I don't even think it will need shimming. (the frame)