Precision ground tooling costs?

[WillieL]

What is it about precision ground tooling that makes it so outrageously expensive?

I know things like surface and center-less grinders don't come cheap, but does it really cost $100 to produce a 9 piece HSS pilot bushing set for a reamer that is the diameter of a drinking straw? 

[Brass_Machine]

So I am guessing here... to accurately grind some requires more skill for setup etc...

Or more truthfully, because they can?

Eric

[Tony Wells]

Well, that's only about $11 each. And you're right, grinders don't come cheap. Especially the CNC grinders likely used to produce such tooling. CNC shops need to generate about $100/HR to pay for themselves and a modest profit. Operator wages are not cheap either.

[WillieL]

Well, that's only about $11 each.

Hi Tony,

You know - that's a good point. I didn't think of it that way. I just see a few "Chiclets" in the palm of my hand and think, this is $100???   
Thinking about the individual cost, it's not so bad.

I don't know. I guess I was looking at them like diamonds. You buy a diamond from a jeweler and it costs $3000. If you want sell that same diamond back to the jeweler, it's suddenly worth $50.
 

[Lew_Merrick_PE]

Willie -- Stop and think about it.  Reamer bushings are made to a runout concentricity of .0003 inches.  How much time does it take to set-up for that level of accuracy?  How much time does it take to set-up to inspect that level of accuracy?  Now divide that by your quantity and multiply by a reasonable hourly shop rate.  You can purchase a reasonably good, accurate to .0005 dial indicator for something on the order of $125.  Now take a look at a dial indicator that will give you reasonably good, accurate to .00005 value.  This should tell you something...

[WillieL]

Lew -- I'm not a machinist, and never will be. I really don't know how long it takes to set up a CNC grinder that I am assuming turns out these parts by the dozens per run. I can't see these being produced one at a time. Correct me if I'm wrong.
I'm just imagining an automated machine spitting those things out all day long. I've never seen any being made so I may be way off base here.

You can purchase a reasonably good, accurate to .0005 dial indicator for something on the order of $125.  Now take a look at a dial indicator that will give you reasonably good, accurate to .00005 value.  This should tell you something...

I hope that last figure was a typo, considering the two Interapid DTI's I have are only good to 4 decimal places and cost me much more than $125.

And yes, I can see QC inspections adding quite a lot of cost to the bottom line too, so $11 each doesn't look so bad any more. I guess I'll just go to my corner and be quiet now.   
 

[Lew_Merrick_PE]

Willie -- The thing is that +/-.001 represents 100X the tool life or set-up of something that is +/- .0001.  In other words, somebody has to tweak the set-up 100X more often (round numbers) to make something to +/- .0001 than they do to make something to +/- .001.  Additionally, the maintenance for a machine resolving accurately to +/- .0001 is (about) 50X more expensive than that needed for a +/- .001 resolution machine.

To give you an idea, the machine I helped to build in the early-1970's that makes socket head cap screws by cold forming/working bar stock is a + .0000/- .0004 machine.  It will make (approximately) 250,000 screws (depending on size) between adjustments.  Each machine line (4/machine) can make up to 1.2 million screws per shift per week.  Another machine I helped design and build about 25 years ago machines parts that are +/- .0001.  It needs to be adjusted (approximately) every 100 parts.  It can make (about) 20,000 parts per shift per week.  (The parts in question are part of a common medical device.)  The SHCS machine removes almost no metal.  The medical device actually removes about 60% of the input metal.  Cutters wear faster than forming rollers.  Grinding wheels wear even faster.

[WillieL]

Ok then. Thank you for explaining that further Lew. I've never been around that kind of machinery and had no idea they needed to be adjusted that often in a hands-on manner. Being a home shop kind of guy, I was thinking they simply loaded up a machine with raw material and hit the "start" button.   

But I suppose we haven't made it that far yet in current technology, eh?
Well - I can dream can't I?  lol.

Unfortunately I have to live in a world of cheap manual machines and looser tolerances. But I enjoy being able to use them just the same. So please bear with me folks if when I ask some silly questions at times. I've still got a lot to learn. But I'll gladly take all the education I can get.   

 

[Davo J]

Only a guess, but I think the CNC grinders would be set up to take wheel wear into account, after all they would have been doing them for a while and would be able to calculate wear.
But someone would need to be checking every so often for quality control to make sure they where on size. Who knows they might just have go and no go gauges and a shop helper doing it. (in China it would be, LOL)

That type of stuff is a hole new world we can only dream about having, LOL

Dave

[Tony Wells]

Actually, production machines found in the modern world of machinery can have automated methods of measuring, sometimes using laser micrometers, and can make offsets as needed to compensate for tooling wear. I expect that products such as the pilots we started off discussing would not be run at the sort of volume to justify such equipment though, but it's possible. What happens in a typical quality program is a sampling is used where at predetermined intervals, a sample part is picked out to inspect and if needed, all parts between that part and the previous accepted part are sequestered until a 100% inspection can sort out that lot. It could be 10, 50, 100 or more between inspections. Parts that fail may wind up in "economy" grade tooling, be reworked to specification if it is valuable enough to justify, or possibly reworked to a different size. Let's say it fails by being undersized by 0.001. It could be ground to the next smaller standard size.....probably on a manual machine, unless the volume of rejects justifies another CNC.

I'm not saying for a fact that these pilots are handled this way, just suggesting the possibility in general terms where products such as that are manufactured.

[Lew_Merrick_PE]

Tony,

As one of the people who developed in-line inspection technologies for the USAF/NASA Factory of the Future program in the 1980's, I am quite aware of both the abilities and the cost of doing such things -- as well as the limits of accuracy such techniques entail.  I actually own four different 1/4 wave (32 nm) optical flats (and have been using them in gage work since 1974).  I was a member of the team that developed active-interferometry for the National Bureau of Standards in 1982-1984.  Things have improved immensely since then, but it is still a process using costly equipment requiring no small amount of time to set-up and calibrate.  Break-even quantities are still fairly large (though I admit to being more than a decade out of the loop on such calculations).

One of the more interesting experiments I helped perform was to check machinist's ability to measure dimensions to an accuracy of less than .001 inch.  If they used calibrated tool-room micrometers, their accuracy ran in the .0005 range.  However, if they used their own micrometers, their accuracy ran in the .0002 range.  We ran this test using 150 different machinists of more than 20 years of experience.  I am not sure what conclusions to draw from this (other than suggesting that feel is more important than many think), but I thought the results were interesting...

[Tony Wells]

Lew, I've read some whitepapers on such research, and having spent considerable time in QC/QA, I tend to agree with the analysis. Many people underestimate the value of the human role in measurements. Although we never will be able to hand-measure to the sub-micron level, what we can do is more than sufficient for most production work.

[Lew_Merrick_PE]

Tony,

Among the weird and strange things I have done, I designed and set-up production on unit knives for various Special Operation's groups.  These knives were delivered with an extremely sharp edge.  This was ground on a CNC milling machine using four grades of diamond grinding wheels.  The as it comes off of the machine edge is sufficient to give a shave as smooth as the best straight-razor shave you have ever had.  This process added nearly $40 (@ $60/hour) in manufacturing cost to the knives.  True accuracy takes more time than most people think.

[ieezitin]

Hi.

To quote Lew  “True accuracy takes more time than most people think.”   I 100% agree it does if your setting up a lot, if your needs and requirements change a lot, but not where you set it and forget it comes into play.  Tool and die makers have been making jigs for 200 years and banging out parts with exceptional accuracy, history shows after a peak prices plummet.  In the last 20 years with the  aid of  computer it’s got out of the world efficient and faster.

The original poster is on track why does it cost $100 for 4 items I can clasp in my hand. It’s already been said (because they can) large up-to-date manufacturing facilities are sparse even in a global world. Every tom dick and harry makes machine tools and consumables but very few can produce the first class 100% reliable products where the need is vast.

It all depends on perspective, how they and you look at it. If its portrayed as a commodity and sold as such the price can be controlled. Diamonds are an example, there are plenty in the ground and not that hard to harvest, in fact I bet there budget on security in protecting them comes close as to digging them up.
Just my 50 pence worth.      Anthony.