help with tight tolerance cylindrical part

[LeoC]

I work in a product design house and we have 50 engineers always challenging us with some tough jobs. This part is a little out of our expertise here in our machine shop and we were looking for some advice to see if this part is doable and suggestions to make it easier. If you happen to be interesting in making the part, there is a contact on the pdf or you can PM me here.

 

Thanks for the help, LeoC

outer-cyl.PDF

[jlw™]

no bid and good luck

 

undercut dim missing too

[Guest]

AT those tolerances whoever does take it is going to charge you through the nose cost X 10x's

 

2" deep ±.0001  lol

[Newbeeee™]

THAT is wrapped up in $100 dollar bills

I love the statement - All dimensions are provisional, subject to discussion.

Form Fit and Function. They're either needed or they're not

[#Rekd™]

Advice would be to lower the tolerances where possible to reduce the cost.

[compressor sam]

if the tolerances cannot be changed, this would need to be ground in a temperature controlled enviroment.

[Matthew Hajicek - Singularity]

You should talk to a gauge making company.  No one else is likely to have the equipment.

[Newbeeee™]

if the tolerances cannot be changed, this would need to be ground in a temperature controlled enviroment.

I would suggest beat it out, stress relieve it through heat treat, grind/hone the bore, then mandrel off the bore to grind the OD's or do them between centres (you'd need a small centre In the small end and chamfer ground in the bore)

To hone the bore, you'll need a wider undercut or tight tolerance apply over only say 3/4 the length of the bore. But the bore will have to be straight, because you only have 6 microns concentricity for the OD.

Air gauging to measure the part.

As Sam says - temp and humidity controlled all the way.

 

What is this part for?

[Watcher]

Good luck holding these tolerances on stainless... 316/410 is a pain in the xxxx to hold tolerances..

 

Also they don´t mention surface finishes... To hold that 0.0001" tolerances you need to have a tight surface finish and some GD&T like straightness, cylindricity, etc... Don´t think one can hold those using carbide... Ship control in stainless with low depths of cut is poor and will scrap the part...

[LeoC]

Thanks for the responses so far. As I mentioned earlier, we are a product design house and we have a machine shop in house. I work in the machine shop with two other guys. So we make all kinds of things, scientific instruments, medical equipment, consumer devices, just about anything, always something different so can be interesting at times. This part has a mating piece that goes inside. Talking to the engineer more, the main concern we have is the clearance between the two parts so they can be match fitted per se. So if the ID goes too big or small, the mating part can be changed just so long as the clearance between the two is what we require. We appreciate the feed back and knowledge of the members of this forum.

 

LeoC

[Newbeeee™]

Leo,

What is going inside this part? What material? Is it designed as an actuator (sliding in and out) or???

 

Be very careful of galling with stainless steel and close fits.

[compressor sam]

I would suggest beat it out, stress relieve it through heat treat, grind/hone the bore, then mandrel off the bore to grind the OD's or do them between centres (you'd need a small centre In the small end and chamfer ground in the bore)

To hone the bore, you'll need a wider undercut or tight tolerance apply over only say 3/4 the length of the bore. But the bore will have to be straight, because you only have 6 microns concentricity for the OD.

Air gauging to measure the part.

As Sam says - temp and humidity controlled all the way.

 

What is this part for?

We put an enquiry out for air gauges. These are most definitely not cheap !!!

I visited a company last year who had purchased a small range of air gauges 100,000 pound later !!!

[Newbeeee™]

We put an enquiry out for air gauges. These are most definitely not cheap !!!

I visited a company last year who had purchased a small range of air gauges 100,000 pound later !!!

Sam - it isn't cheap, but how else can you sensibly measure to these tolerances?

This is why I initially said Form Fit and Function. Reality check required.

It's no good designing a part with tolerances like this and checking them with a digital mic, and kidding yourself you're on the money!

[mkd]

guessing a first year engineer came up with this one.

 hint; typing lots of zeros doesn't make you a brilliant engineer. Finding a way without it, does.

 

If you are looking for fit @ a nomilal diameter the proper drafting methodology would be to use a class of fit.

[jeff]

A .002" slip fit on the I.D. but has to be +/- .0001"

I have never understood logic like this.

[compressor sam]

Sam - it isn't cheap, but how else can you sensibly measure to these tolerances?

This is why I initially said Form Fit and Function. Reality check required.

It's no good designing a part with tolerances like this and checking them with a digital mic, and kidding yourself you're on the money!

absolutely true, this part requires major investment, and if it is a one off, they will be scared off by the price, that is probably the best way to get the designers to look at their design again, P.S. i was not having a dig, i was horrified when we got the quote for air gauges ourselves, my boss`s face was a picture !

[LeoC]

I agree, it really should be classified in the "class fit" format as all we are interested in is the clearance between the two pieces, not to an exact size. I know it is easy to make fun of the engineers but in this case, this has to be this tight to perform a certain function. This was designed by some Doctor's so yes, it can be challenging at times.

 

So far we are going to try turning the part as close as possible and then send it to an ID grinder that says they can do it, we'll see. I'll let you know how it turns out.

 

LeoC

 

Here is a link to where I work so you can get a better idea of the stuff we do here. The photos of the machine shop area are quite old and need to be updated.

 

 

http://novoengineering.com/

[Newbeeee™]

Leo - turn some spares and give them to the grinders.

Tell them you want all back, including their setting pieces that will be ground over size etc.

Then you can test assemble the parts and actually see if a tenth or two bigger than drawing actually makes a difference or not.

Theory is one thing, practical is another...

[Colin Gilchrist]

I know a couple real Tool and Die Makers that could hit these tolerances in their sleep. They routinely hold 20-50 Millionths for dimensional accuracy. It doesn't come cheap, as a Moore Jig Grinder runs somewhere around $250-400K, for a manual machine and the various precision attachments needed. But tolerances like these are certainly achievable.

 

For the uninitiated; I highly recommend the book "The Foundations of Mechanical Accuracy" by Wayne R. Moore. It will change your perception of what is possible in the world of Manufacturing. It is one of the most valuable books on my shelf, which is incredible considering it was published in 1970, well before the modern area of CNC equipment.

 

http://mooretool.com/publications.html

[Newbeeee™]

Colin,

I'm not saying it's unachievable. All I'm asking is is it necessary. Very few applications require tight tolerances like this, but without knowing, yes this could be one of them.

 

I see parts detailed regularly with tight limits and a lot of the time it's default CAD system tolerances and when questioned, the engineer can open them right up with no issues.

Problem is when it's too late and the part has been qualified and gone into production. No one will change a drawing then because of re-qualification fees. The part has then been 'designed expensive' for the rest of its product life...

We (UK) suffer from high pressure to get parts made in the far east to save money.

If parts were designed for manufacture in the first place, 99 times out of 100 there is no issue as the part can be cost-effectively made here, and the far east doesn't even get mentioned.

[jeff]

and a lot of the time it's default CAD system tolerances

 

I see this a LOT on Autocad drawings. And many times the dimension was altered but not the geometry to match. I've had to re-draw so many parts it's insane.

[Quickmike]

I have that book by Wayne Moore. Unbelievable the tolerances that were being held.

Pictures in the book of them scraping. Thermal blankets on the castings in the area not being scraped. So the scrapers body heat has less of an effect on the casting. Indirect lights to reduce local light "energy" ie.(heat energy) and stratification. Trying to keep the floor and the ceiling the same temperature. I think some of the finish scraping is done in rooms sunk into the ground to help with this.

All being done back in the late 50's I think.

[Colin Gilchrist]

I have that book by Wayne Moore. Unbelievable the tolerances that were being held.

Pictures in the book of them scraping. Thermal blankets on the castings in the area not being scraped. So the scrapers body heat has less of an effect on the casting. Indirect lights to reduce local light "energy" ie.(heat energy) and stratification. Trying to keep the floor and the ceiling the same temperature. I think some of the finish scraping is done in rooms sunk into the ground to help with this.

All being done back in the late 50's I think.

 

That is exactly what I was referring to. The methodology it takes to not only machine, but to measure with extreme accuracy. You have to look at the entire production system in order to hold those tolerances. They built their machines with hand scrapped and lapped ways, with tolerances that were the tightest. Ball bearings that were held to 10 millionths in size and roundness. Ways measured to 25 millionths. All in Bridgeport Connecticut.

[LeoC]

Unfortunately, this part does need to hold those tolerances in order to do what it needs to do.

 

LeoC

 

http://novoengineering.com/

[Mjölnir]

Unfortunately, this part does need to hold those tolerances in order to do what it needs to do.

 

LeoC

 

http://novoengineering.com/

 

suggest give this place a jingle; that's where we'd probably go for a part with those tolerances.

 

http://www.bbprecision.com