PDA

View Full Version : sphere suggestions



cadwiz
03-29-2007, 08:15 AM
I need to create the two hemispherical shapes below out of titanium. Once created they will be welded together to form a hollow sphere with the tube exit. Diameter is 2.5" and wall thickness is on the order of 1/16", maybe less. Want this to be light with minimal thermal mass. Need to create another, slightly larger sphere to encase this one to form a vacuum insulating jacket. The completed sphere needs to be able to hold up to around 300psi. I've got a radius attachment I built for the lathe but it's only good to 2". I could mill on a RT with a boring head at 45 degree tilt. I don't think either of these methods would work due to the wall thickness and lack of rigidity. What about a die/hydraulic press? Is there a way to form this without creating distortion wrinkles? Any other innovative ways to create this?

http://i33.photobucket.com/albums/d70/cadwiz/ARCSphere.jpg

Any suggestions would be appreciated.
Thanks, Cadwiz

sandiapaul
03-29-2007, 08:50 AM
Metal spinning comes to mind...but I have no idea how Ti would spin...
here are a few pics and instruction:
http://www.coe.ufrj.br/~acmq/spinning/

Evan
03-29-2007, 09:36 AM
Sounds like a job for explosive forming.

http://archive.metalformingmagazine.com/1997/01/7mfjan5.htm

TGTool
03-29-2007, 10:38 AM
Turn the OD. Cast it in low melting point allow. Machine the ID.

cadwiz
03-29-2007, 10:43 AM
What's explosive forming? This needs to be something that's relatively easy to do in a shop (either mine or at work). I'm prototyping this for work and can't afford to tie up our plant machinist on development. If successsful we'll need to fabricate several of these for a physical property testing instrument that we're developing. Were you kidding or for real? Explosives in my home shop probably wouldn't go over very well with the local authorities (already had a neighbor blow his hand off with a homemade cannon that went awry, not my cannon by the way). However, if there is such a method perhaps outsourcing would be a possibility.

Cadwiz

cadwiz
03-29-2007, 10:53 AM
OK, I didn't notice the attached link. Explosive metal forming. Hmmm, sounds interesting and apparently proven for thin wall titanium. I'll look into it further. Still willing to consider other, more traditional, techniques as I would like to tackle this problem myself. Would go a long way in justifying why I've got all the crap in my shop to SWMBO.

Cadwiz

Rusty Marlin
03-29-2007, 11:43 AM
Take a look at chemical milling too.
But I think Evan has the best idea.

john hobdeclipe
03-29-2007, 11:51 AM
Spinning would probably be the quickest and cheapest method. And relatively quiet, too. I remember reading that CP titanium forms about like stainless steel, but I've not yet tried spinning either one. Here is a link to Terry Tynan's spinning website, which includes a forum where you might learn some more about spinning.
http://metalspinningworkshop.com/

And here is a discussion about explosive forming, with some further links to explode...oops I mean explore.

http://www.sculpture.net/community/showthread.php?t=2324&highlight=explosive

madman
03-29-2007, 12:22 PM
External Mas cam cut points and also internal. Lots of work. Perhaps cnc lathe. Get a friend to make it on there cnc or use cad program and develop cut points and file sand to finish form???

Ausserdog
03-29-2007, 02:03 PM
Another option is hydroforming. Pretty neat and has been done at home by others. Used to have a link (can't lay my hands on it right now) where a guy built a custom expansion pipe for his motorcycle out of sheet metal. He used a small home made hydraulic pump and water after welding the parts together.

Swarf&Sparks
03-29-2007, 03:15 PM
"Cast it in low melting point allow"
Has someone missed the point here, or have I?
Low melt Ti?

Todd Tolhurst
03-29-2007, 03:17 PM
The low melting point alloy is used to form a fixture to hold the workpiece in the lathe so the interior can be turned.

cadwiz
03-29-2007, 03:46 PM
I was wondering what that meant. Makes a lot more sense now.
Cadwiz

pcarpenter
03-29-2007, 05:23 PM
So why do you need a titanium toilet valve ball?:D

DR
03-29-2007, 08:51 PM
Send the job out to somebody who specializes in this sort of thing, spinning, explosion forming, hydroforming, etc.

This is for your job, right? Not a home hobby project. Let the boss pay the bill.

BTW, I believe a lot of titanium forming is done hot. What's the alloy? CP titanium isn't bad stuff to work with, some of the high strength stuff is a nightmare.

Ries
03-29-2007, 10:57 PM
I would agree that the job, as you described it, is not a quick and easy, or cheap, home shop job.

I would have em spun. For one or two, you may have to go with stock sizes of tooling. Most spinners have hemisphere dies in a lot of sizes.

Stamping em with a press might work, but tooling is a lot more expensive, and suited to runs of thousands of parts.

The explosive forming option might not be as expensive as you would think- I have a friend who just bought a new pump for his fish vacuum cleaner- he runs a tender in alaska, and has a 12" or so hose connected to a 25hp vacuum, and he sucks the fish (herring and salmon both) out of fishermens boats and right onto his sorting tables with this enormous vacuum. He needed a new pump recently, and ending up having one built custom in Mexico, of all places- about 9 grand, and it is made from Stainless and Aluminum that are explosively bonded together.
Given all the other stuff that went into the 9 grand of pump, I cant imagine the bonded metal cost too awfully much- maybe a grand or two, for custom made, thick plate (1" or so) explosively welded somewhere in New England to order- not a stock product.

You do realize that anything made from titanium basically will cost you 2 to 3 times what it would in stainless, and anything in stainless usually runs 5 times, minimum of what it costs to make in mild steel?

Which means expensive.

Evan
03-29-2007, 11:12 PM
A lot depends on the alloy. The high strength alloys of titanium I have worked with could never be spun to that shape. They have a nasty tendency to tear when stretched. That also goes for ordinary die forming processes. It isn't very well suited as a metal removal job because of the wall thickness and strength requirement.

That size sphere has a surface area of 19.6 sq inches. At 300 psi the total load from internal pressure is almost 3 tons.

wierdscience
03-29-2007, 11:38 PM
Does it absolutely have to be Ti?

DR
03-30-2007, 12:13 AM
I just looked up titanium in the Metals Handbook.

It seems the 6AL-4V alloy (I think that's a common one) can be die formed and spun among other methods. Not at home though, for deep parts they use high temps, 800 to 1200F.

Spinning is done with a torch directed at the part to maintain heat.

The CP (chemical pure) can be treated similarly to stainless, much easier to deal with than the alloys.

abn
03-30-2007, 02:59 AM
You're probably referencing the Eurospares site (Run by Michael Moore who is a poster here).

http://www.eurospares.com/frame8.htm



Another option is hydroforming. Pretty neat and has been done at home by others. Used to have a link (can't lay my hands on it right now) where a guy built a custom expansion pipe for his motorcycle out of sheet metal. He used a small home made hydraulic pump and water after welding the parts together.

DR
03-30-2007, 11:30 AM
You're probably referencing the Eurospares site (Run by Michael Moore who is a poster here).

http://www.eurospares.com/frame8.htm


That's a really cool method of forming parts. For large parts relatively low pressures still give enough pressure to form.

To form the hemisphere in question here I think we'd be talking about pressures upwards of 15,000psi+. Not a simple home shop job to seal up a containment arrangement with pressures like that.

BobWarfield
03-30-2007, 01:27 PM
I would make up a sort of collet or pot chuck to go in your lathe that has a hemispherical depression. Cut your half spheres to shape on the outside in one lathe setup, and then put them into the lathe with the pot chuck's hemispherical back side to support the wall while you turn the inside. Put a flange on the part that is heavy enough to support and hold the works while you're turning out the inside. The flange should be heavy enough not to crush, but the flange's shape will facilitate that. You'll Tig around the flange circumference of the two pieces to seal it up.

You should be able to make the little inlet fitting as part of the whole thing as well.

Seems like an ideal CNC job as well given you want spheres. You're going to need to get your speeds, feeds, and cutters just right. You want as little cutting force as possible on the thin metal, so sharp positive rake tooling is called for. CNC will make it easier to get repeatable feeds and speeds once you figure out the optimal values.

Best,

BW

Ausserdog
03-30-2007, 01:36 PM
abn,

Yep that's the one. Thanks for the help!

DR
03-31-2007, 12:00 AM
Okay, let's assume he gets the hemispheres done one way or another....

How's he going to weld them together? Seems to me, Ti needs to be welded in a vacuum chamber, am I right?

Have we gotten beyond the home shop capabilities of Cadwiz yet?

Swarf&Sparks
03-31-2007, 07:46 AM
"Ti needs to be welded in a vacuum chamber, am I right?"
Nope, Ti welds quite easily with TIG and straight argon.

cadwiz
03-31-2007, 08:17 AM
Thanks for the suggestions, keep'em coming. We're looking into both spinning and explosive forming just to see what may be feasible.

Just a little background: We actually have a source for a preformed Ti sphere with the outlet tube. The problem is it's very ex$pensive, pre-welded, and requires a minimum $10k order! We need to insert a small agitator and shaft through the hole before it's welded and the company won't supply just the pieces. Ti and minimal wall thickness is required for low thermal mass and non-reactivity. This application is for a fairly low-cost, reusable, reaction vessel that is instrumented for Accelerated Rate Calorimety. We've done it at low volume in glass for up to 200 psi using cheap bottles. Glass is non-reactive but sometimes unpredictable as you can guess. Most times we can go up to 200 and sometimes it fails prematurely. I think the reason is induced stress points created by attaching the head. Looking for something more reliable with higher pressure rating and spherical geometry is the ticket. The system as a whole is totally custom built and developed in-house so it's not something you can buy off the shelf. We're investigating all options but it would be nice if we can find something that can be readily fabricated in-house.

Thanks again and keep the suggestions coming,
Cadwiz

DR
03-31-2007, 08:26 AM
"Ti needs to be welded in a vacuum chamber, am I right?"
Nope, Ti welds quite easily with TIG and straight argon.

I found this info:

"The techniques and equipment used in welding titanium are similar to those required for other high-performance materials, such as stainless steels or nickel-base alloys. Titanium, however, demands greater attention to cleanliness and to the use of auxiliary inert gas shielding than these materials. Molten titanium weld metal must be totally protected from contamination by air. Also, hot heat-affected zones and root side of titanium welds must be shielded until temperatures drop below 800F (427C)."

DR
03-31-2007, 08:35 AM
........
Just a little background: We actually have a source for a preformed Ti sphere with the outlet tube. The problem is it's very ex$pensive, pre-welded, and requires a minimum $10k order! We need to insert a small agitator and shaft through the hole before it's welded and the company won't supply just the pieces. ......

Not sure how to interpret that....does it mean the process to form the hemispheres is extremely difficult, $10K worth of difficulty, or these guys consider an order under that amount to be a nuisance?

Have you checked with some of the larger spinning/forming houses that do aerospace work??

Rustybolt
03-31-2007, 09:02 AM
I found this info:

"The techniques and equipment used in welding titanium are similar to those required for other high-performance materials, such as stainless steels or nickel-base alloys. Titanium, however, demands greater attention to cleanliness and to the use of auxiliary inert gas shielding than these materials. Molten titanium weld metal must be totally protected from contamination by air. Also, hot heat-affected zones and root side of titanium welds must be shielded until temperatures drop below 800F (427C)."



Some places use a sealed box with an open top filled with argon. Some make a sandblasting cabinet type of arrangement, with gloves. For small jobs there are oversize ceramic or pyrex nozzels for the tig gun.

Evan
03-31-2007, 09:43 AM
If glass is acceptable then you should talk to a scientific glass supplier. Properly made from borosilicate glass and annealed it will withstand the pressure and will be cheaper to have custom fabricated than titanium.

cadwiz
03-31-2007, 07:19 PM
Guess I didn't quite convey the whole picture. We have two of the most unbelievable scientific glass blowers around, amazing what they can do. What we've done before was with annealed borosilicate glass. They even tried a different type of glass. The original gameplan was to use easily obtainable glass bottles as the reaction chamber, then dispose/replace as needed. The problem seems to be with the attachment of the upper head with all the instrumentation. Been using a threaded arrangement, threaded by our in-house machinist. I think the issue is induced stress when the head is screwed on. I'm coming in on the tail end of a lot of previous work so forgive me if I don't know all the details. The jump up to titanium and the spherical shape is to go above what has worked in the past (next jump in performance). This application is to investigate elevated temperature/pressure behaviour prior to throwing a large mass of chemicals in a big reactor so it's worth the effort. Just trying to find a low cost solution. Glass isn't the answer at the pressures we need to go to. Got to find a way to make titanium work.
Cadwiz

nheng
03-31-2007, 08:31 PM
Cadwiz, You might want to look into MIM, metal injection molding. The metal powder is sintered into a solid. We've looked at small precision parts made from stainless steel using this process and they are quite impressive. When I think powdered metals, I think of cheap crap that falls apart ... not the case here.

Metal injection molding has a size limitation and your parts might be at the limit which was a several inch cube (but that was a couple of years ago).

A polymer binder is added to fine (around 25 micron) powder to allow injection molding. After molding, the green part is sintered at high temp (2600 F) and the binder vaporizes.

Not sure if titanium is on anyones list but it might be worth a few calls.

added - here ya go, titanium: http://www.praxistechnologyinc.com/

Den

Dr. Rob
04-01-2007, 11:28 AM
I didn't bother to read the details of your application.

If you want a titanium sphere with even wall thickness, you make it from sheet metal.

Cut out a disc of titanium sheet. Buy a 2 1/2" ball bearing. Go find a piece of pipe, ID 2 1/2" plus a bit.

Put the disc on the pipe, ball bearing on top of the disc and press it through.

It makes a hemisphere. So make two.

True in lathe, tig weld with gas shield both sides.

Drill / bore for your thru-tube and weld that too.

Done. It takes less time than reading this thread.


Alternative: Press in rubber instead of tube. Alt 2: Make a radiused press tool instead of buying a ball.

.

DR
04-01-2007, 11:47 AM
I didn't bother to read the details of your application.

If you want a titanium sphere with even wall thickness, you make it from sheet metal.

Cut out a disc of titanium sheet. Buy a 2 1/2" ball bearing. Go find a piece of pipe, ID 2 1/2" plus a bit.

Put the disc on the pipe, ball bearing on top of the disc and press it through.

It makes a hemisphere. So make two.

True in lathe, tig weld with gas shield both sides.

Drill / bore for your thru-tube and weld that too.

Done. It takes less time than reading this thread.


Alternative: Press in rubber instead of tube. Alt 2: Make a radiused press tool instead of buying a ball.

.

Great idea.......

This is one of those "why didn't I think of that" deals.

The main problem is, it won't work. There's the issue of severe wrinkling as the part is formed.

In the stamping/forming industry old timers sometimes suggest trying a forming setup with a piece of paper. In this case, imagine the wrinkles you'd get doing this with paper.

Once it's established the paper will wrinkle, then you go on to devise ways to stop the wrinkling and form a smooth part. One way is to use blankholding devices to prevent the wrinkling as the part forms (now we're getting into an area of complications).

Ti is difficult material to form to relatively deep shapes and prevent wrinkles.

Dr. Rob
04-02-2007, 07:02 PM
Yes it does work. I just did it.

The use of paper may be wise, but paper lacks the fundamental characteristic of malleability.

Tried the tube trick, it works but was tricky (not really: I just spun a lip on the blank to anchor it to the tube before pressing.)

Tried a rubber press trick, it was better. See exceptionally lousy quality pics below. Those are hockey pucks.

It was easy. No wrinkles.

http://i35.photobucket.com/albums/d165/DrRob101/Hemisph%20Press/HPIM0378.jpg

http://i35.photobucket.com/albums/d165/DrRob101/Hemisph%20Press/HPIM0382.jpg

scott96088
04-02-2007, 07:25 PM
Yes it does work. I just did it.

The use of paper may be wise, but paper lacks the fundamental characteristic of malleability.

Tried the tube trick, it works but was tricky (not really: I just spun a lip on the blank to anchor it to the tube before pressing.)

Tried a rubber press trick, it was better. See exceptionally lousy quality pics below. Those are hockey pucks.

It was easy. No wrinkles.

http://i35.photobucket.com/albums/d165/DrRob101/Hemisph%20Press/HPIM0378.jpg


http://i35.photobucket.com/albums/d165/DrRob101/Hemisph%20Press/HPIM0382.jpg


Dr. Rob the outcome of your press job is easy to see works very well. I have a couple questions.

1. Is that Ti. that you pressed in your photo and if so how thick?

2. If not what is it and how thick?

3. How much press force do you figure it took to press?

Scott

Evan
04-02-2007, 08:52 PM
A few observations:

The metal was being stretched only, so no wrinkles.

That degree of stretching will only work with material that is thick enough to begin with and won't work at all with many Ti alloys because they will tear.

The sphere needed is considerably larger and must be from thinner material. It is probably beyond the normal drawing capability of any alloy of Ti.

[added]

Pure unalloyed Ti is unusually malleable with an elongation at break of 54%. High strength alloys have elongation specs from a high of 20% to as low as 3%.

bob_s
04-02-2007, 10:02 PM
So heat the Ti as it is being formed

Evan
04-02-2007, 10:28 PM
That will work if having the Ti in the annealed state is acceptable. It may not be.

[added]

Of course there will be a HAZ from welding and most of the properties can be restored by heat treating and aging.

DR
04-03-2007, 07:42 AM
........
1)If you want a titanium sphere with even wall thickness, you make it from sheet metal.

2)Cut out a disc of titanium sheet. Buy a 2 1/2" ball bearing. Go find a piece of pipe, ID 2 1/2" plus a bit.

3_Put the disc on the pipe, ball bearing on top of the disc and press it through.

It makes a hemisphere. So make two.

True in lathe, tig weld with gas shield both sides.

Drill / bore for your thru-tube and weld that too.

Done. It takes less time than reading this thread.


4)Alternative: Press in rubber instead of tube. Alt 2: Make a radiused press tool instead of buying a ball.

.


Dr Rob,

I numbered your points above for ease of responding.

Thanks for the test pictures, what material, surely not Ti?

1) Agreed, you would start with sheet metal, for practical purposes in drawing we can consider the wall thickness will not change significantly. Obviously, your test pieces are thinned a great deal.

2) your test hemisphere appears close to 1" diameter. In successful drawing with simple setups there're important relationships of part diameter to metal thickness, your material looks relatively thick with respect to part diameter making drawing easier than with the required 2-1/2" diameter, .060" thickness ratio specified.

You started with a square blank, that's not an ideal blank to draw a round object. Your blank doesn't appear distorted, showing no signs of distortion around the edges, leading to the conclusion this was almost pure stretching as opposed to drawing.

3) You didn't press it through. Doing that with a tough material like Ti is where the wrinkles would develop.


4) It isn't clear how the hockey puck was used to do the "rubber" pressing. (I come from an area where hockey isn't played so I have no first hand knowledge of the hardness of a puck, I would have guessed they're very hard)

In conclusion, you did a nice test and came up with interesting pieces. You've shown how easy it is to stretch-form hemispheres in what appears to be soft aluminum.

Bob G
04-03-2007, 10:48 AM
I did a Google search for [draw quality titanium sheet] and it appears that there are alloys available. For small quanities all you need is a punch, a die ring and a pressure ring. You will need some press capibility but the shape you are working on is almost perfect for a draw die.
Good Luck
Bob G

DR
04-03-2007, 12:28 PM
.......For small quanities all you need is a punch, a die ring and a pressure ring. You will need some press capibility but the shape you are working on is almost perfect for a draw die.
........



That's a bit of a simplification........

The punch would be the dome shape, the die ring would be the mating lower part. The pressure ring is a blankholder which applies clamping force to constrain the blank as it's drawn into the die. Without the pressure ring the material would form wrinkles as it's drawn.

The pressure ring has to exert enough force to keep the blank flat and to allow the blank to compress smaller into itself as the round blank's diameter gets smaller during the drawing process. Tough materials like Ti don't like to compress easily.

In a production setting double acting presses are used, one ram does the drawing, another allows variable force on the pressure ring during the process.

There's a basic problem with drawing domes. The dome shaped punch contacts the blank in the center. The blank is constrained under force around its perimeter by the pressure ring. This leaves the area from the blank center where the punch first engages it and the die's inner diameter uncontrolled.

Bad things can happen in this uncontrolled region, tears and unwanted stretching. This is why many materials are easier to draw as straight sided cups than domes.

Domes can be drawn in multiple die sets, starting first with a shallow cup draw, progressing to the dome as the final draw.

My point is, conceptually this is a simple part to make. In practice it may not be so simple.



On edit: Dome shapes and tapered side parts might be better candidates for spinning than drawing. Straight sided parts are generally better drawn than spun.

Dr. Rob
04-03-2007, 05:14 PM
The post and process were simplified in order to make a point. The original poster asked for an easy doable way to make an experimental part. I wasn't intending to make the part perfectly and then write a thesis about it.

No, material isn't titanium. It is scrap I picked up from the floor. Probably mild steel. Thickness is around 1 mm. It happened to be squarish. Yes, ball is approximately 1" diameter.

The hockey puck with the hole in it is the material holding pressure pad. The puck and ball are of roughly equal thickness. As the ram comes down and pushes the ball into the blank, it also compresses the rubber puck and clamps the material flat. Very convenient, and no mating or matching die parts necessary. That rusty round column thing (= die) just happened to have a round hole in it with a diameter of about 27 mm.

The edges of the hole are provided with a friendly radius, and plentiful lubrication is provided to allow blank to slide and be drawn.

You can not see it in the pics, but yes, the blank is noticeably distorted. Thickness at bottom of crater, I have no idea.

Obviously this whole thing can be improved upon, but the point was to provide the guy with an answer. An action rather than theories.

Maybe the whole thing would be better if pressed into a female die with rubber.

Or just spin the thing.

Now you know.

Doc.

.