Machining weldments

[Paul Erland]

Hey folks,
I would like to use weldments for a project but not sure how to prepare them for machining. I believe the process is called normalizing but maybe that's not the right technical term. As I understand it, I need to soften the welds and relieve the stresses in the metal from the welding process by heating the part and allowing it to slowly cool. For steel, how hot does the part need to be? I'm assuming cherry red. Does the part need to stay at temp for some minimum amount of time?

Tried internet searches for weldments and normalizing but didn't really find what I was looking for. Any information or links is appreciated.

Thanks, Paul

 

[Mike N]

What type of material are you welding?

Mild steel will not require any treatment.

Cast Iron & Tool Steel should preheated to prevent cracking.

Alloy steel should be anealled if you plan on machining the welded area.

 

[Paul Erland]

Mild steel would be the material but wouldn't the bending stresses and hardness of the welds still be a problem?

 

[ProdEng]

If you try a search for "stress relieve fabrication" you will find lots to read. Distortion of a fabrication is possible during machining and depends partly on how much material you remove.

 

[Deanofid]

Mild steel would be the material but wouldn't the bending stresses and hardness of the welds still be a problem?

If you use common welding rods, like 60xx or 7018, the welds won't be hard to machine. MIG welds from mild wire won't be any trouble either. Clean the flux off well before you machine.
Not sure what you mean by "bending stresses".

 

[Paul Erland]

By bending stress I refer to the pull the welds put on the parts. My experience with welding is that the cooling weld tends to pull on the parts which can cause the parts to move towards each other. Weld two parts at 90 degrees and after welding you usually find the parts at slightly less than 90 degrees when it has cooled. This stress is always in the metal like a spring. I figured that machining would allow this stress to distort the part as material is removed.

 

[Entropy455]

Assume you “T” together two pieces of steel. You grab some 1/16” stick electrode, and place a 3/16" fillet weld down the length of one side. The part will likely bend over a bit at the weld, forming a bent tee, unless you tack the backside prior to running the fillet.

Here’s why: steel has a thermal coefficient of expansion of about 6 millionths of an inch, per inch, per degree F.

That means that a 3/16” fillet bead (0.265” across the diagonal) will shrink by 0.0042” as the weld cools from a steel solidification temperature of about 2700 degrees F, down to a room temperature of 70 degrees F. Four thousandths might not seem like a lot, but when it’s placed in the root of a weld, it’s giant. This is why steel parts bend out of shape when you weld them.

If your parts have sufficient rigidity (or if you tack the backside), they will not visibly deform, however significant residual stress remains in the part. When you attempt to machine a weldment for close tolerances, it’s not uncommon to see surfaces grow (or shrink) to new dimensions, as adjacent surfaces are machined.

To stabilize the part for machining, one can accomplish a “post-weld stress-relief”.

This is normally accomplished by heating and maintaining the weldment between 1100 to 1200 degrees F. This temperature is cold enough to prevent appreciable changes to the lattice grain structure, but hot enough to permit the grain boundaries to slip (or creep) in a slow and controlled fashion – which is the mode of stress relieving. It's normaly baked for about 12 hours, then slow cooled.

Any shop that can heat treat and temper steel, is capable of accomplishing a stress-relief heat-treatment. It should be one of the less expensive processes offered.

 

[Jasonb]

If you can get the part upto red heat and hold it there for a while and then allow to cool as slowly as possible then it can be done at home, really depends on the size of the part and what heating arrangements you have. The same method can be used to get rid of hard spots in iron castings.

This is a good example and there are several other fabrications being treated later in the thread.

J

 

[steamer]

I THINK Paul is talking about internal stress put up by welding and fabrication process

Once you start cutting the weldment will start to move around. Stress Relieving is the correct term

The most "popular" method is with heat, If I recall to 1175F for 1 hour for every inch of thickness followed by slow cooling.

This is not annealing, which would be much hotter. This would cause a lot of distortion in a weldment because the parts would probably sag in the furnas

The other method involves shaking the weldment at or near its natural frequency.....I've seen it done....but not recently.

The old school way is park the part outside for a few months in the weather...works well on iron,,,,no so well on steel due to oxidation.

Normalizing , to me anyway, implies stress relieving followed by a sub zero treatment.....that's how I've used it anyway. It imparts long term stability to the part...ie it doesn't move around as much with age.


Dave

 

[Maryak]

Hi Guys,

In my experience, machining weldments, done by stick, (MMA), is a pain in the butt especially if it involves trying to blend them to the mating parts such as a Flywheel and hub. The weld will machine OK as will the 2 originals but the boundaries of the weld are usually hard and cause the tool to either rub and lose shape or jump away at this point.

I guess that IMHO Yes they need to be heated up to a bright red and then allowed to cool slowly, either in the furnace or buried under lime for 4 hours minimum.

Best Regards
Bob

 

[Entropy455]

Any steel that has more than about 0.28% carbon is going to harden up in the heat affected zone when SMAW (stick) welded. The penetration is good, but the heat input is simply too high if you intend to machine the component without stress relieving.

I’ve found that TIG welding with 70S-2 or 70S-6 rod will minimize the hardening effect within the heat effective zone (to the point where you "can" get away with not stress-relieving). The 70S weld beads are easier to machine also.

FWIW, proper stress relieving is normally accomplished over many hours within an industrial oven - not over a few minutes with a torch in the garage. It must stay between 1100 and 1200 degrees F for several hours. Parts requiring high tollarance will stay in the oven for 12 hours or more. . .

1200 degrees is where the steel is just starting to show a dull dark red.

Once you hit orange, you’re in the austenitic zone, and you’ve completely changed the lattice structure of the steel (dissolved all of the carbon). At this point, unless you can slow-cool the weldment at a rate not to exceed 100 degrees F per hour, you are going to have more residual stress than if you just left it alone.

 

[steamer]

From various sources....google is your friend.

"...STRESS RELIEVING: (300 - 1400°F)

Thermal Stress Relieving is generally applied to metallic materials that have been cold-worked, formed, machined, flame-cut, or weld-fabricated to reduce residual stresses for dimensional stability or reduced risk of premature failure in service.

ASME Section I: (900 - 1300°F)
Stress relieve/Post-Weld Heat Treat in accordance with ASME Section I, PW-39 for carbon steel weldments....."

 

[steamer]

A link to vibratory stress relieving

http://www.vsr2.net/


and another link from the Weld Journal

http://www.aws.org/wj/sept01/cullison.html

 

[Entropy455]

If you wish to cut the oven time down, you can lightly tap the steel adjacent to the welds with a hammer, and the impulse events will encourage the grain boundaries to slip.

On that note: stress relieving carbon steel below 1000 degrees is virtually ineffective, without some sort of additional mechanical help.

Compare it to cooking a turkey in the oven. Too hot, or too cold, and you’ll have an undesired outcome.

 

[Mike N]

Mild steel would be the material but wouldn't the bending stresses and hardness of the welds still be a problem?

Not if you use mild steel welding rods!

 

[Deanofid]

On that note: stress relieving carbon steel below 1000 degrees is virtually ineffective, without some sort of additional mechanical help.

He's not using carbon steel. He told us the material he's using in the 3rd post.
All he needs to do is weld with 60xx or 7018 or a mild wire in a MIG. He can stress relieve it to good effect with a ball pein hammer.

Paul Erland, make your weldments with equal weld beads in "symmetrical" fashion as much as possible. In other words, if you well on one side of something, try to put a similar bead on the other side of that something. Mild steel doesn't build up a lot of stress from welding, but the weld itself does make some.

On one machine job I had, we made hundreds of large clamps that were weldments to make the assembly, then machined over a fair part of the exposed surfaces.
Right off my welding table, I had a guy with a rose bud who went around the hot weld beads a few times, then beat the thing with a large ball tipped scaler. The next day, after these things had cooled, I put them on the lathe and faced the 10" flange and brought it to diameter. We never had any warping problems (post machining), since the pieces were made from mild steel, using a mild rod.
I'll mention that I was a welder first, and machinist second. We made so many of this kind of thing in that shop and had no trouble with warping, nor with the machining of the welds themselves.

I don't know exactly what you are making. Maybe if you tell us, and mention how large it is, you can get some advise that doesn't wander to all aspects of machine and welding work.

 

[steamer]

He's not using carbon steel. He told us the material he's using in the 3rd post.
All he needs to do is weld with 60xx or 7018 or a mild wire in a MIG. He can stress relieve it to good effect with a ball pein hammer.

Paul Erland, make your weldments with equal weld beads in "symmetrical" fashion as much as possible. In other words, if you well on one side of something, try to put a similar bead on the other side of that something. Mild steel doesn't build up a lot of stress from welding, but the weld itself does make some.

On one machine job I had, we made hundreds of large clamps that were weldments to make the assembly, then machined over a fair part of the exposed surfaces.
Right off my welding table, I had a guy with a rose bud who went around the hot weld beads a few times, then beat the thing with a large ball tipped scaler. The next day, after these things had cooled, I put them on the lathe and faced the 10" flange and brought it to diameter. We never had any warping problems (post machining), since the pieces were made from mild steel, using a mild rod.
I'll mention that I was a welder first, and machinist second. We made so many of this kind of thing in that shop and had no trouble with warping, nor with the machining of the welds themselves.

I don't know exactly what you are making. Maybe if you tell us, and mention how large it is, you can get some advise that doesn't wander to all aspects of machine and welding work.

Good points Dean

Dave

 

[Entropy455]

Mild steel and carbon steel are the same thing.

http://simple.wikipedia.org/wiki/Carbon_steel


Using a rosebud and a ball tipped scaler falls within the “additional mechanical help” category of low temperature stress relieving. As with most things in life, there’s more than one way to do things. I’ve designed close-tolerance industrial machine foundation components that were fabricated from weldments. Our machinists were unable to obtain satisfactory parallelism unless the pieces were fully stress-relieved at 1200 degrees, for a full 12 hours, prior to machining.

 

[Ken I]

Welding introduces stress - period !

Stresses introduced throughout the assembly but typically highest in or close to the weldments.

Machine away any stressed material and your job will change shape.

Heat up your (stressed) job and your job will change shape.

How much you need to worry about it depends on the accuracy of the finished part.

If you are going to be doing machining of the weldments and accuracy is an issue then a full destressing or normalising is required.

If its a garden gate - who cares.

Ken



what you want.

 

[steamer]

Hi

Although it is nice to wax poetic regarding process....guilty as charged....it doesn't provide much assistance to the guy in his shop with little more than a torch.

Now that we all know how it "should" be done, perhaps some pragmatic useful advice is in order for the Home shop guys who don't have a controlled furnas.......and let's keep it real.



Dave