What's new
What's new

Machining intersecting radii/cylinder

Fal Grunt

Titanium
Joined
Aug 5, 2010
Location
Medina OH
Lots of awesome ideas!
If polymer is strong enough (30ksi, tops?) maybe cast out of Zinc? Some of the zinc alloys are pretty strong and zinc casting is easily an in-the-shop project when it needs to be.
Well, it's like most manufactured products. There is what works, there is what is best, and then that which is neither because of customers pre conceived notions.

Random anecdote that springs to mind. I worked for a big tier 1 automotive firm, and one of the projects I helped prototype was for a Ford transmission. I can't remember the name of the project now but it went into small Ford cars, like the focus, and I don't know the rest of the details, something like twin input shafts, and some gizmo wizardry. Anyway, Ford made them for only a few years I think, everyone universally complained. The car was an automatic, but designed to drive like a manual. The number one complaint was that when a customer let their foot off the brake the car did not roll forward like they expected. So while the car got better gas mileage and was "better" for several other irrelevant points, the customers did not like them because of preconceived notions. Which brings me to the point of this anecdote. My wife's Subaru. Those G^$%*(% engineers ruined the F^&*($#$ car because of preconceived notions. It has a CVT transmission, but starting with our model year (the car we test drove didn't do it) they programmed the car to REV UP and cut out to "simulate" a shift. Because people didn't like the fact it didn't SHIFT. Ruins, absolutely RUINS the point of a CVT. I'd wager the terrible mileage it gets is because of that nonsense. PERCEPTION RULES REASON.

I'll get off my soapbox now.

Late to the party but is there any reason you cannot make it in two pieces, one with a socket and the other with a post and press or bolt them together?
If you have a 4th axis you could use a slab cutter and rotate the part about one cylinder whilst cutting to the intersection line, then do the same with the other cylinder.
We've discussed doing two pieces, but the attachment is the tricky part, time consuming, which reads expensive.

If I had a 4th, this would be easier.
I'd look into MIM if it was me.

Of course you'd end up with a 1000-year supply of parts, but if you can bake the setup cost into say... the first year or so of sales, then it's all gravy after that! :D

PM
I haven't looked into MIM, would it be any more cost effective than casting? I'd have the same concerns with tooling cost. I don't know how well these parts will sell, and being a low dollar part, it is hard to justify dedicated tooling. The $8k for casting would be a lot of parts...

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
I look at it this way. If your looking at making a small, low dollar part, the markup had better be pretty good. So for example sake, say it's a $25 part. For me I'm going to try and make that part for $10-$15. I know some here will spit their coffee across their keyboard and exclaim it isn't worth turning on the machine for such a low margin. While casting that part might cost $3? the tooling at $8k would take 320 parts just to pay for the cost of the tool. In this case, I'm paying myself whatever portion of that $10-$15 is the production cost, minus tooling, material, etc. vs paying the casting place $8k for tooling, or however much MIM (dies?) would cost.

From the numbers I made up Friday night, I can get 250-300 pcs from 24' of bar stock. Based on moderate speeds and feeds, multiple setups, and some form of coining tool, I figure the parts will cost me about $18 ea. So I can setup and run 50 pieces for a cost of say $1000, with actual outlay of cash being maybe $300 for tooling, material, and soft jaws. If those 50 pieces take a year to sell, then I know I'm not going to invest in casting tooling or MIM or any other "production" tooling. If those 50 sell in a month (I'd have a stroke if they did) then I'll machine a batch of 100. Then 250. A batch of 250 pieces profit would almost cover half cost of casting tooling. Then I'll take a look at it again.
 

implmex

Titanium
Joined
Jun 23, 2002
Location
Vancouver BC Canada
Hi Fal Grunt:
Without intending to derail this thread, this is probably a good opportunity to talk about the reality of trying to MIM this in quantity in the USA.

First a small aside:
I did a fairly large MIM project in the early 2000's when there was still a local MIM vendor in my area.
I designed the parts.
I designed the mold
I built the mold.
I built the sintering support tooling.
They ran the parts.

All was straightforward and ran pretty uneventfully.

Fast forward to the project I showed in a previous post.
No way for love or money could I convince ANYONE in North America to take this on without a $250,000.00 initial commitment and a total projected project price of about 3/4 million dollars.

It was nuts...I tried...believe me I tried to do it in North America, but my original vendor had given up on MIM and sold off the gear to the National Research Council, and NOBODY among dozens of tries all through North America would do more than the first look before rejecting the project as unprofitable.

I and my customer did it in China for...get this... UNDER $25,000.00 and that was for four molds and all the sintering support tooling.

They did a pretty damn good job too, and in under 8 weeks!
The customer orders 5000 units at a time, so 20,000 parts per order, once a year.
These are low tolerance but high value dental parts, so his profit is immense, and he's a happy dude who does not even consider doing anything manufacturing related in North America anymore.
(He was once a stalwart...USA or Canadian suppliers only!)

So if you hope to do this in North America...good luck if your quantities are small and your budget is limited.
I've been getting the same crap lately for all my outsourcing...local is becoming hopeless.

As an example...a robotics project just went to China in April...lots of 5 axis milling... about 20 parts or so.
They turned it in under 3 weeks.
It's not brilliant, but it is more than acceptable and it was CHEAP.

A couple of those parts hadn't been fully engineered so I held them back, solved the unresolved design issues and then vendored them out locally to a couple of guys I know who are very talented machinists.
They're not finished yet, and the due date was 3 weeks ago.

So to drag this back to the OP's problem...if you hope to rely on local supply chains for some exotic process, think again.
You may get lucky, you may not.

Also, with respect to MIM as a candidate process...yes in theory...no in the real world unless you have a special "IN" somewhere with a vendor.
At 25 grand, this project was an absolute steal, but you have remarked that a capital investment of 8 grand is more than the project can legitimately risk, so, attractive as it is technically, in reality it's a fantasy.
The China molds were about 5 grand apiece, and the rest went to the sintering support tooling, which you need too.

Now a last thought:
Do you think you could resistance weld these together?
They're small and skinny enough you might get away with it.
Spot welders are cheap to buy.
The biggest downside is that the spot weld will look like Hell and will be rougher than a cat's ass too.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 
Last edited:

Radar987

Cast Iron
Joined
Dec 30, 2011
Location
CA
So I don't LIKE it, but I think it would function. The part still bears on the radii that contact the mating part. View attachment 366537

Looks like this part wouldn't be constrained in the mating bore. It would be able to float upwards.

Rather than cutting the flat all the way to the end of the cylindrical area, I'd slot a relief around the halfway point and use a small end mill to finish up the radius of the rectangular boss inside the relief.

That'll leave some of the cylindrical area fully intact. Does that make sense?
 

Fal Grunt

Titanium
Joined
Aug 5, 2010
Location
Medina OH
Implmex, thanks for sharing that experience. It sounds like MIM isn’t going to be a possibility.

I do know the company I made powdered metal tooling for (different but similar) was a huge operation and some of the tooling we made/repaired was for older products. In fact I got to rebuild a piece of tooling my grandfather originally built in the 60’s or 70’s. It made a ring gear (massive) for John Deere.

Radar, I’m not 100% sure I picture what your saying. I think I do? I looked at making a double angle tool to plunge into the radius to relieve that top edge, as the radius is more or less irrelevant on the bottom edges of tangency. (I think I said that right?)
 

Fal Grunt

Titanium
Joined
Aug 5, 2010
Location
Medina OH
Weve ordered investment castings to +/-.001 tolerences, pricing is going to depend on who yur going to. What is the material?
Everyone I found quoted +- .005, with one option for tighter tolerances. One shop said they could potentially do +- .003. With the tighter tolerance the tooling went from $8k to something like $12k.

If you have a source for .001 castings, please let me know, I’d be interested in having them quote this part.

The three quotes I received were for 4140 and 8620.
 

implmex

Titanium
Joined
Jun 23, 2002
Location
Vancouver BC Canada
Hi again Fal Grunt:
I have a suspicion you're never going to find an investment caster who can do these to +/- 0.001" in steel and I say so for a couple of reasons.
First you have to injection mold your waxes very consistently, so you're starting to eat into your tolerance band right from the first moldmaking step and again from your first processing step (shooting and then ejecting the waxes).
Your second step ( spruing) is similarly a source of inconsistency, as is investing and burnout.
Then you have to pour with all of the inconsistency that implies.
Then you have to devest with some sort of grit blasting process and trim the parts accurately from the sprues.
With all of those processing steps together, you have to finish within 0.001" of nominal...that's a really tall order.
In fact, I'm impressed they will even guarantee +/- 0.005".

Does this all sound cheap to implement?
Does it sound like an intrinsically robust, tweakable, stable process to you?
Can you see the challenges in every step of developing the process well enough to make +/- 0.001" parts and then certify them?

I can't see this being successful for any reasonable money and I doubt it can be made successful at all if the hard goal is the tolerance band you describe.

So I'd abandon too much effort toward trying to cast these...I think the hurdles are unreasonably high (and expensive).
But you could maybe cast them and then coin them.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 

Fal Grunt

Titanium
Joined
Aug 5, 2010
Location
Medina OH
Hi again Fal Grunt:
I have a suspicion you're never going to find an investment caster who can do these to +/- 0.001" in steel and I say so for a couple of reasons.
First you have to injection mold your waxes very consistently, so you're starting to eat into your tolerance band right from the first moldmaking step and again from your first processing step (shooting and then ejecting the waxes).
Your second step ( spruing) is similarly a source of inconsistency, as is investing and burnout.
Then you have to pour with all of the inconsistency that implies.
Then you have to devest with some sort of grit blasting process and trim the parts accurately from the sprues.
With all of those processing steps together, you have to finish within 0.001" of nominal...that's a really tall order.
In fact, I'm impressed they will even guarantee +/- 0.005".

Does this all sound cheap to implement?
Does it sound like an intrinsically robust, tweakable, stable process to you?
Can you see the challenges in every step of developing the process well enough to make +/- 0.001" parts and then certify them?

I can't see this being successful for any reasonable money and I doubt it can be made successful at all if the hard goal is the tolerance band you describe.

So I'd abandon too much effort toward trying to cast these...I think the hurdles are unreasonably high (and expensive).
But you could maybe cast them and then coin them.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
I completely agree, I seriously doubt that without a highly refined process that it would be very difficult to achieve those tolerances. Again this is where with a low volume part it simply doesn’t make sense. That being said a friend of mine is a mold maker (plastic injection molding) who has cast several items himself. He showed me that I could purchase everything needed to do one off wax and do the molds myself. I still doubt that I could achieve anywhere near +-.001 tolerance.

I also reasoned that if going from a .005 to a .003 was a near 50% increase, the next step down would be that much again or more.
 

implmex

Titanium
Joined
Jun 23, 2002
Location
Vancouver BC Canada
Hi Fal Grunt:
Yep and you can build a dead nuts accurate perfect mold and still have the parts come out all over the map.
Just shooting and ejecting the waxes (or plastic burnout patterns) consistently is an art, never mind all the other steps.
Process inconsistency is your enemy and it lurks in all kinds of unexpected places with multi step processes as I'm sure you know well.
It can be the Devil's very own to get it all dialed in and keep it that way.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 
Last edited:

Garwood

Diamond
Joined
Oct 10, 2009
Location
Oregon
I would try pushing some 1018 into a tool with about 40 tons and keep adding heat until I get the look I want. Then carburize and tumble.
 








 
Top