help on sheet forming process!

well Mr Simmons' recognition is nice, in fact its one of parts of an industrial turbine.(somehow its a secret , so keep it just secret here ;-) )

Mr Simmons , can you please help me just a bit more , I just want to know how GE make such parts , even if you yourself don't know much , maybe you can introduce someone else to me who has experience in this area.

And I could see about $3000 in tooling with a piece part price of $1500 for labor as long as the process would produce parts to the OP's requirements.

It is totally unclear whether there are strict tolerances >> in fact we have called some , but no one accept the risk of it. >>

what is unclear?! as I said , its a turbine part , so tolerances are strict , if you need it , the deviation in diameter is not more than 0.15 mm in 1860 mm , in inch :0.006 inch in 73 inch is it clear now?

Holy cow. Forget rolling, then. In fact, forget any kind of forming, unless it's followed by machining. And as for machining, you're pushing beyond the limits for machining a weldment that's going to move around as the weld is machined off and it relieves the stresses. Because a stainless steel weldment of that size is going to be loaded with stresses, unless it's thoroughly annealed first.

In other words, this is a multi-step job no matter how you look at it. I think you're looking at roll-forming two pieces (the flange, if it's attached to the cylindrical section at 45 deg., as it appears to be, will have to be a quarter-circle cut out of a large plate, which is why I pegged the material price for that piece at over $3,000); welding them; annealing the whole works to relieve all stresses; and then turning it to final shape. And after you unclamp it from the VTL faceplate, I'll bet it will still move more than 0.15 mm in 1860 mm. It will have to be re-clamped and finish-machined in a separate operation. All of which means that if you truly need 15 mm thickness you're going to have to start with something thicker than that.

If these are the specs you've given to possible vendors, I'm not surprised they won't bid on it. The final tolerance is a killer. It would be easier if you left it as a complete ring, because the springing from stress relief would be better constrained.

-- Ed Huntress

My contact there has long since retired, so the best I can do is speculate based on what I saw while working on various projects and extrapolating to your part. The processes are not unusual or unique - I presume the value here is the confirmation that I saw parts similar to yours fabricated in this manner.

-Roll the cylindrical segment from a rectangular piece of plate

-Roll the conical segment from a semicircular piece of plate

-Weld the segments together

-Stress relieve

-Finish machine

The black art in this sequence is the practical, hands-on knowledge required to control welding distortion, and to predict how much the rough weldment must be oversized to allow machining to size to accommodate the distortion that can't be controlled.

I re-thought this because something has been nagging at me. Because you said "sheet" I've been thinking in terms of plate. This isn't a plate-rolling job. It's a ring-forging job.

It could be ring-forged and then machined. You'd have to start with a complete ring but you'd have a fairly low-stress part after machining. Then you might be able to achieve your tolerances.

Search on "ring forging." I think you'll find what you're looking for. You might even make your projected cost.

Sorry I floundered around on this. I usually can identify the process needed for doing odd jobs, after decades of reporting on practically the whole metalworking field, but sometimes we trap ourselves into one line of thinking when we should be looking at another. Once I thought of it as a complete ring, it jarred my thinking.

This job is right up the alley for ring-forging.

-- Ed Huntress

I like it.

I think a part like this would cost a LOT more than $1500 per part.

The way it would probably be done in industry these days is to get somebody like Scot Forge

roll forge you a blank. This would give you an oversize ring, without welds, that had the grain forged continuosly around the circumference for strength.

Then, you send it to a machine shop that has a good sized VTL or VBM.

I was in a factory in Italy a couple of years ago that built big machines, and they were making similar sized and shaped parts in this way- a roll forged ring, machined on a vertical lathe about the size of a two car garage. They had a whole row of these machines at work.

The material cost is one thing- but the real cost is in the equipment capable of doing this size of work to these tolerances, and the experience and infrastructure to support these machines.

This aint no garage job.

Like I said- If I seriously needed one of these, I would pick the brains of companies that do it everyday, as job shops, and I would start with Scot.

Shantia said "$16,000 is just fine," so it should be OK, if he's making more than one. He did mention that making a full ring and splitting it would be OK, too. However, he said at one early point that it had to be cold-formed, while this process is hot-forming. But he may have been thinking only about casting as the alternative.

I agree -- especially if he's going to make more than a couple of them. The first one will cost a bundle.

Hahaha! And you want to roll-form that or spin it!? Forget it. Simply forget it. As soon as you cut it, it will spring in *any* direction. You just can start with a forged part or a ticker rolled part, stress-relief it. Pre-machine it, stress-relief it again and then finish it.

Nick