automating weld of fitting to a small shell

[RNDguy]

We want to automate the process of welding internally threaded fittings of 3/4" to 1.5" OD to shells of 9" to 12" dia and of 1/8" thickness using MIG. The material of both the fittings and shell is mild steel. Any help/info/tips in this regard will be helpful. Is this even worth pursuing? Thanks in advance.

 

[Goahead]

To decide if a process is worth pursuing one should compare the costs of automated processes vs. the manual one. With increasing production quantities there is usually a break point over which the automated process is more economic.
You could probably discuss Mig applications with Ed Craig from

http://www.weldreality.com

You may wish to consider application of friction welding instead of Mig for very high production rates.

Regards,

http://www.welding-advisers.com/

 

[BillPSU]

This sounds like a hydraulic/pneumatic cylinder application.

Automating a MIG process means you need some means of cleaning the nozzle and cutting off the stick out along with motion of the gun. A robot would be what I would recommend along with the part feeding/removal mechanisms. The fittings will need to be radiused and orientated properly to assure good fitup.

I would actually recommend not MIG but a stud welding process and drilling the through hole after welding. It is much eastier to automate and faster.

 

[GenB]

You will have to tack the nozzles and then migweld with automation.
robotics may be the best aplication.
genb

 

[GenB]

Please keep posting on progress,
I am in the process of automating similar work.
GenB

 

[RNDguy]

Thanks for all the replies. GenB, one of the difficulties we are realising is that because the fittings make the shape of a saddle where they meet the shell, good tracking of the weld may be essential for sound welding. Does this appear to be within the limits of current proven technology with respect to machine vision etc.? How about simply building a fixture which may do a combination of turntable and up-and-down motion on a barrel cam to replicate motion along the "saddle"? The latter has been suggested to us.
There may be enough experience out there regarding this problem, which by its nature seems commonplace enough.
BillPSU, very perceptive indeed, this is a pneumatic application at 150 psi. Sounds like you have been down this street, have you guys done this automation?

 

[metengr]

RNDguy;
Have you looked at an auto-GTAW welding process? There are many small TIG torches that can be fixtured to an arm for automation and accurate tracking. I don't see where you need high deposition welding processes to justify using MIG.

 

[MikeHalloran]

I might:

- emboss flats in the shells to put the weld in a single plane.

- pierce the shells at the same time.

- machine a step in the proximal face of the fitting, so it would be partially self- fixturing.

- add a relief groove in the o.d. of the fitting, so the weld would be between a flange and a shell of similar thickness.





Mike Halloran
Pembroke Pines, FL, USA

 

[GenB]

Depending on the production pcs, robotics can be the best aplication,
you program it to do the exact weld,
if you want to spend the money on a robot,
or if you want to outsource as well,
let us know and i will post a web site to follow.
genblr

 

[DAVIDSTECKER]

You might want to look into "Stud" or "Spin" welding your boss in place or something like the "Flowdrill" process in which the metal is re-fromed into a boss that can be tapped.

 

[AutoSolutions]

RNDguy

I used to work for a custom machine builder that specialized in automated welding for just this sort of thing. We did everything from integrating purchased articulated arm robots to custom built specialized robots.

I've seen what you describe accomplished for the fittings on hydraulic cylinders with both fixed automation and an articulated arm robot. The right solution certainly depends on the desired throughput.

The fixed automation application revolved the torch around a verical axis, while riding up and down on a cam to maintain the correct height. The operator tacked the part on the cylinder and monitored the weld.

At the other end of the spectrum, you could feed fitting parts automatically with a vibratory bowl feeder and automatically place and weld them.

 

[steveh]

Good question. Have you considered capacitive discharge welding? We do a lot of robotic MIG welding where I work but we have found CD is especially good at the type of thing you are talking about.
You can monitor current and displacement during the process and maintaining the equipment is cheaper and more trouble-free than MIG. If you're interested, try looking at companies like Impuls (Germany) or Banner (US).

 

[BillPSU]

I've manufactured hydraulic cylinders using a spud welder with a rise and fall cam follower are which followed the saddle of the part. This was a hand loaded operation which could be automated in the loading and unloading of the parts. The difficulty is keeping the welding nozzle clean in an automated operation as the nozzle on a spud welding would have be cleaned manually. If this amount of manual intervention is allowable then you could have one built. The machine that I was using was a Bancroft which is still in business and the model is the Welda-Round.

You will have to watch the burn back to make sure there is no crater at the start or stop. You will see weld distortion inside the barrel of the cylinder and the area around the weld will shrink. We were using .25" wall tubing 4" OD. We saw so much shrink we honed/skived-burnished after welding. We couldn't install the piston past the port if pre-honed material was used.

 

[RNDguy]

All of you have given some very good advice indeed, I want to thank all of you for the general enthusiasm, interest and help in this regard. Consensus appears to imply that it is a do-able thing.
To respond to some of the suggestions:
- We are looking at MIG because this seems to be the preferred route for welding automation of this kind, and I am told the higher deposition rate can give a bigger weld size thus countering the lack of precision somewhat inherent in this type of welding and its tracking.
- Our own experience with Capacitive Discharge welding for weld studs has a mixed opinion. In a volume process, operators can tend to overlook critical settings, and factors like plant electrical load variations are beyond control which also affect the issue. The application is critical from the standpoint of high pressure pneumatics.
- There is a school of thought in-house which wants to do fixed automation as a first step. If any one can recommend a few good companies in this regard, then I will truly appreciate it. Thanks again for a wonderful discussion.

 

[MintJulep]

Do you really need to weld?

Investigate vacuum furnace brazing.