Bore of Tube shrinking during weld

[sammcc]

I have a tube with a pre drilled hole in it and a collar slid over that. I am then fitting a 1.125" outside diameter tube into this as per the illustration. The tube obviously shrinks during welding which means we need to ream it out afterwards as another tube slides inside it. What is best practice to reduce the shrinking.
At the moment we place collapsable bungs up the tube to help reduce it.

 

[moltenmetal]

I am interested if others have managed it. A copper plug right at the point of welding might be effective as a "chill" (you didn't mention the material), but I would imagine you would need to drive or press it out afterward and you might need a new one each time... Shrinkage can be controlled by limiting heat input and interpass temperatures, by using automated welding techniques, and of course by changing the material to one that shrinks less.

 

[dhengr]

Sammcc:
But, you probably won’t completely eliminate that kind of deformation and shrinkage caused by the welding. It’s the nature of the beast when you are welding. So you better resign yourself to having to do some final machining or reaming of the bore. If you do a hundred of them and measure them all, very carefully, you might determine that you can over-bore them a bit in the first place, and minimize the reaming after the welding. But then you’ll probably have a single or a couple slight high spots in the bore.

 

[tbuelna]

An expanding plug/collet fixture seems like a good approach to controlling distortion of the tube ID. Are you saying this approach does not result in a tube ID with acceptable roundness/size after welding? I think the out-of-roundness after welding is mostly due to the varying shape of the collar cross section at the weld joint. Changing the sequence used for the welds might also help.

 

[sammcc]

Thanks for the replies, At the moment we use an expandable collet , I have attached a picture of it but the tubes still have to be reamed out after this has been removed. Due to the shape of the frame these have to be done manually using an air drill (atom).
If I were to measure the distortion would it be proportional in the sense that I could counterbore the tubes before welding to take that into account or would they still distort maybe more because of a slightly reduced wall thickness.

Also, The expandable collets we use now are made from brass, would copper act as a better heat sink?

 

[EdStainless]

When you weld the tube heats, and expands. But you have it inside of a fitting so it can't expand outward, the metal has to inward. It is the nature of physics.
I suggest that you build a fixture or guide to assist with reaming after the weld. I don't think that you will be able to compensate by pre-counter-boaring unless you are willing for this section to be loose.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[dhengr]

Sammcc:
I don’t understand how finely that fit is intended to be, and you haven’t chosen to tell us. If the next pipe just has to fit in the i.d. easily and it is o.k. that it ride on a couple high spots or high ridges on the i.d., then you could probably do some of them, measure the change in the i.d., and finally learn what the oversize bore has to be. During this learning process you would ream them to make them work, after measuring. Finally, you might be able to avoid the reaming operation from what you’ve learned. I don’t think it’s a heat sink (cooling) issue. The welding operation is inducing a circumferential residual stress which is squeezing the pipe to slightly smaller dia., and the more so the nearer you are to the weld, and you can’t find a mandrel strong enough to prevent this. You could expand the pipe, putting it in circumferential tension to counteract this shrinkage from the welding. That expansion dimension should be about what I’m telling you to measure, but you are measuring the shrinkage you want to overbore to compensate for. This shrinkage will depend on the heat input during welding, the pipe dia. and wall thickness, and the configuration of the weld. So, a thicker pipe with all else being the same, will shrink less. A lighter, less heat input, weld will cause less shrinkage, all else being the same. This is kinda like the material movements during a shrink-fitting operation.

 

[sammcc]

Thanks for the replies, I have more of these to make so I will figure out what the distortion is and see if I can counterbore to counter act it and also build a fixtue to allow machine reaming as a back up.

 

[Compositepro]

You can also "ballize" the tubing. That is, push a suitably sized, and lubricated ball through the tube to expand it to the desired diameter.

 

[drawoh]

sammcc,

Is your tube shrinking, or is it warping?

Could you heat everything up before you weld? Weld warping is caused by differential cooling.

--
JHG

 

[sammcc]

Compositepro- I have never seen this done before, do you have any examples?

Drawoh it is mainly shrinkage after welding, there is some distortion which is minimal.

 

[Compositepro]

https://www.precisionballs.com/Ball_Sizing.php