Matching small diameter piston and cylinder to ~.0001"

Hi guys...

I'm looking at making some small diameter (~.375" bore) pistons and cylinders. They will be ground and honed, respectively- probably to a few tenths (hopefully) and then the task will be to measure them all up and make matched sets within about .0001". Now, we'll CNC turn them, heat treat, send out for grind and DLC coating as appropriate... I'm sure that I could have the coaters / grinders measure them and baggie them with the sizes marked on them. However, what's going to be the best way to measure this stuff to confirm it? I probably will make a few hundred of these sets per year, so if I'm going to gauge them in house, I don't want to spend more then a couple of grand on the tools.

Our shop is not WELL climate controlled- it varies from 62 to 75 deg F, but being they are very small, I could measure them in my office which is climate controlled. Or, I could wait for a time of day when the shop and parts ARE 68degF, and measure them then. Night time would work fine.

The pistons I don't think would be too bad- I probably just need a really good set of calibrated mics. The cylinders though are tougher. Doing some research, I keep coming back to air gauges. It looks like they are available used somewhat- quite a few on ebay- and then I would just need an air pin made, and a ring gauge that size...

The other option is to try to find a metrology place local and farm the work out. Receive them back in labeled baggies and just sort them.The fit of these parts is critical and the failure rate must be very low, so trusting the grinder / platers is not an option, at least not at first.

What do you guys think is the best approach in this situation?

I think that a lot could be done with calibrated ring and pin/plug gauges. Mostly since that's how they gotta go together in use.
I would not trust measuring with a mic the pistons to .0001 simply because of the different processes involved in making it, possible out of roundness and what not, but definitely is a good quick means of sorting them closely for the next step.

I would consider a Sunnen Hone and a Sunnen Bore Gage.

PG Bore - Sunnen

Gene

peter.blais said:

I'm looking at making some small diameter (~.375" bore) pistons and cylinders. They will be ground and honed, respectively- probably to a few tenths (hopefully) and then the task will be to measure them all up and make matched sets within about .0001".

The pistons I don't think would be too bad- I probably just need a really good set of calibrated mics. The cylinders though are tougher. Doing some research, I keep coming back to air gauges.

Click to expand...

You have two problems to deal with: the average diameters, and the out-of-roundness. I've never used an air gauge myself, so what I'm about to say may be wrong, but it seems to me that won't help you with the latter problem. If the cylinder is out of round, the gauge could be too close to the wall for some of the circumference, and too far for the rest, and flow the same amount of air as if it had the perfect clearance all around.

If this were my problem to solve, I would look into a Talyrond (or Formscan, Formtester, Roundtest, Rondcom from other manufacturers) for testing the roundness to the tolerance you demand. I actually own a Talyrond myself that I lucked into, as well as a home-made version based on optical bench components I assembled myself that I used before the Talyrond came along (the latter is good only to ~0.0002"). Even with these instruments, measuring roundness at the level of a ten-thou. isn't trivial. However, I can't imagine how it would be done any other way.

As for measuring average diameter at this level, that's relatively straightforward (although, again, not trivial). For the pistons, my Mahr 1" and Brown & Sharp 4" bench micrometers both easily can do that. The latter has a gauge amplifier that can measure differences from a standard that are quite a bit smaller than the 0.0001" you need.

The bores are a little trickier. Resolution isn't the issue, but absolute accuracy is. But, a variety of bore micrometers with ring gauges have been made by various manufacturers in recent years (although some are no longer available new), including Brown & Sharpe, Starrett, Diatest, Mahr, and Mitutoyo, that will give you the 0.0001". I have a Diatest stand that operates sort of like a miniature arbor press, holding the gauge stationary and raising the table to insert the bore of a cylinder around the probe of the gauge. I don't think you will achieve the tolerance you require without such a stand because of the problem of holding the gauge precisely perpendicular. Also, the~1" "throw" of this stand is large enough, and controllable enough, that I can determine the taper of any bore over that depth to a height resolution that is easily better than 1 mm (I could put a dial indicator on the platform to give better resolution if ever I needed it).

The Diatest-type gauges use a relatively thin ring of material to measure the bore, so they have a length resolution that's probably ~0.01". That is, if there are any variations in bore over a length scale of 0.01", the gauge will show it. Other types of gauges use a split cylinder to do the measuring. I have a Starrett that works like this. Without going out to the cabinet to measure it, I'll say its length is more like 5 mm, so it averages out any variations in bore over this length scale.

Ah- we actually use a sunnen gage already on bigger stuff all the time- automotive connecting rods- and anything that happens to be in that size range. You can definitely tell out of round on those, easily, and they don't need much operator skill...

This one's about right- graduated in .00005", but that's also a big big scale. Just need to find one like this, but one size smaller.

Sunnen Precision Hole Gage Model PG-800-E | eBay

We could always use both mics and ring gages on the OD's of the pistons to check for roundness and fit...

To be assured of the accuracy of the measurement of any dimension, the instrument should have a rated accuracy ten times better than the tolerance being measured.
'
When you get into very tight tolerances, this becomes impractical. The technology simply does not exist at any reasonable cost.

Please understand that resolution and accuracy are not the same thing. A device with .00005" resolution will typically have a rated accuracy twice that or larger.

- Leigh

We have the PG-800 model here, and it goes down to .370. With the extra arm attachments, you can get as high as 3.000 diameter. Might not be a bad investment if you would be able to use all that range.

I split off the entire off-topic exchange regarding statistics into its own thread.

NO discussion of statistics will be permitted in this thread.

- Leigh

For that bore, you need an air gauge. It will accurately and reliably measure the diameter to 0.000007" and R&R @10% or better on a +/- 0.0001" tolerance. It will definitely detect out of round (ovality) in the bore to a large degree. For absolute roundness, straightness etc measurements you need a form gauge in a lab environment. These measure to 0.0000007" and will pass a 10% R&R on a tolerance in the millionths.

Allowing the part to fluctuate in ambient air temperatures and "catching it" when the ambient temperature is correct will not yield reliable results. The parts for tolerances this tight must acclimate to the correct temperature over time. The acclimation time depends on the mass of the parts to be measured, section thicknesses and the thermal coefficient of the material being measured. For a small aluminum part with thin walls such as a piston, 24 hours may be enough time. For a steel part with thick walls, it may be a week, a month or even more.

+1 here on Tony's post #9.

I am familiar with the process of making valve lifters. The also require very close tolerances and this is achieved with air gages. Mostly automated in the manufacturing environment with manual gages for double checking parts at different stages of Mfg. As stated above, air gages are very accurate.

Big B

Tonytn36 said:

For that bore, you need an air gauge. .

Click to expand...

Just to be different-LVDT gaging, 'cause I have it, and don't have air gaging. LVDT will do either OD or ID, and has sufficient resolution and accuracy for this application. Everything else Tony said is right.

This is a nice gaging opportunity, because we actually know what the usage of the part is. For fitting two parts together, MMC is the key issue. So, to make a good decision, you need to find the smallest ID in the ID part and find the largest OD on the OD part, and match them.

As far as gaging, I would think air gaging would be the easiest option for more operators. You need to rotate the part and report the dimensions above. The only thing is you need to know what the lobing is - perhaps with a roundness checker - to make sure you use the correct number of jets. Tri-lobe - 3 jets, bi-lobe, 2 jets.

Using a micrometer is OK if you have bi-lobal condition, but erroneous (very much so at tight tolerances) if tri-lobal (which is most common.) It also is the most prone to bias error.

I wonder if we aren't getting the carse before the hort. Is the OP the end user? Does he need absolute dimensions or grading for selective fits? Quantities? Are there prospect for more and/or larger orders?

There's lots of ways to gage the parts in quantity but the equipment is expensive and you may not have large enough numbers to justify/amortize its purchase. I suggest you sub the job to someone having the equipment delivering to you parts graded, certified, whatever in foam pockted boards, individual bags etc. The pistons are easy: they can be graded with comparator gaging equipment by roll them across the anvil for diameter at any nimber of points using a jo block stack as an on-site reference.

But the bores pose a special problem I don't trust dial indicating bore gages to 0.0001" without a calibrated ring gage as an on site reference. A stack of jo blocks with extended jaws are plain inadequate as a reference. You need a reference dia in the form of a ring. That said, IF the bores are in parts of substantial section, and IF the bore is a hard durable material, and IF a few can be graded out as a true cylinder then possible one or more of the finished parts can be sent to a cal lab for calibration, tagged with the actual size to 0.000010", and used for a gaging reference.

Regardless mechanical movement gaging gets iffy in this resolution no matter whan Federal Mahr says. I STRONGLY recommend LVDT cartridge and gage amp apparatus.

Something like this:

FEDERAL EMD-632B ANALOG METRIC AMP WITH EHE-1119 CARTRIDGE GAGE HEAD | eBay

The above link is just an example. I propose it knowing air gage apparatus is widely useds in identical applications but I have little experience with it.

There are many makers of electronic gaging equipment but buying new can be quite expensive. Older generation is just as accurate when calibrated as the new but it often lacks the "hooks" to integrate it into computerized data collection.

Forrest Addy said:

I suggest you sub the job to someone having the equipment delivering to you parts graded, certified, whatever in foam pockted boards, individual bags etc.

Click to expand...

I don't mean to pick on Forrest, but since joining this site a few months ago I've been struck by how common the response "outsource it" is to a question about how to make/measure something. At the same time, woven through threads are lamentations on how bad it will be for future capabilities that the U.S. has outsourced its manufacturing base overseas. It's only one very logical step from deciding it's cheaper to outsource a particular task, to outsource the entire factory overseas. Yes, it's always possible to outsource a job, and many times it makes economic sense to do so. But, I would think on this forum that the default answer to a question "How do I ___?" wouldn't be to suggest paying someone else to do it. Again, I don't mean to beat up Forrest over this, because it's a response that seems to be within the majority of threads asking a question.

Even where a simple calculation shows it would be cheaper to hire someone else to make the measurements, there are additional factors to consider. For example, doing it yourself develops in-house expertise that can be useful for future jobs. Also, process development can be significantly aided if the production and measurement equipment are a few feet apart. Think of trying to make a small run of rods on your non-CNC, non-DRO lathe, all of which have to be the same diameter to within 0.0005" total deviation, and doing this when you don't own a micrometer. Make a rod, send it out for measurement, get the result the next day, turn the cross feed dial a tiny bit and make the next one, send it out... Sure, it could be done, but spending the money to buy your own micrometer means you'll be able to make the parts with confidence the same day you start the job, not a week later.

Forrest Addy said:

I don't trust dial indicating bore gages to 0.0001" without a calibrated ring gage as an on site reference...

Regardless mechanical movement gaging gets iffy in this resolution no matter whan Federal Mahr says. I STRONGLY recommend LVDT cartridge and gage amp apparatus.

Click to expand...

I disagree on the iffyness at this resolution (although, not about the necessity of having a calibrated ring gauge). The Mahr Supramess (of which I own two) has a claimed resolution and accuracy 5x better than needed for this job. I also have three electronic gauges, so I'm not wedded to doing things mechanically. Even aside from electronic logging of data, electronic instruments do have advantages, such as greater sensitivity than is possible mechanically. However, in ranges where both work (e.g. the 0.0001" needed to answer the OP's question), I find mechanical to be just as good as electronic.

Forrest did not say one word about "overseas".

In fact you're the only person who has mentioned the concept in this entire thread.

I suggest you read before you write.

- Leigh

The real Leigh said:

Forrest did not say one word about "overseas".

In fact you're the only person who has mentioned the concept in this entire thread.

I suggest you read before you write.

Click to expand...

Nowhere did I write, nor even imply, that Forrest mentioned "overseas." It was me who said that if you're going to outsource a task from your company because it's cheaper, that decision is just one step away from deciding to outsource everything overseas. Last week I sent a package to the Middle East that was in the hands of the person I sent it to just three days after it left my desk. Because such rapid transport is a reality, if it's cheaper to have something done overseas, concern about time delay is no longer a consideration. If a competitor down the street gets the same outsourced job done overseas to the same quality and the same turn-around, but a lower price, he can underbid you every time. However, if you develop the capabilities to do the job in-house, and you were able to apply those capabilities and knowledge to generate new types of jobs, you can compete where he can't. These kinds of considerations, not jingoist attitudes toward outsourcing overseas or not, are what will determine long-term viability of companies who have overseas competition.

Back to what I actually did write. There's nothing in the OP's requirements that would seem to make it unreasonable, technically or financially, to do the necessary measurements in-house to the required accuracy. So, given that the OP asked "what is the best approach to this situation," and given that this is a forum about machining, I tried to answer his question in an earlier post, followed up with additional details in my most recent one, with suggestions of specific tools that would let him do the measurements in-house. Forrest suggested he outsource it. That's fine, but I again say that if he does outsource it, doing so has implications for developing the manufacturing process that will give him the necessary precision.

I guess use of the word "outsource" has gone from a simple business based decision over efficient use of available resources to namecalling leading to "divide and despise".

My suggestion to outsource a tricky inspection procedure assumed that any shop is running to efficently use all its available resoures in the generation of profit, expansion, and insidentally standing to attract new business. Towards this end the management and administration may not wish to invest in costly new equipment, train/hire expertise, allocate space, etc for what may be a one-time need. A well run shop may decide to send a job out to a specialist down the street, across town, whereever; a specialist with the equipment, expertise, and (importantly from a QA point of view) disinterest in teh failure rate.

Should the need for high precision inspection resources re-emerge then the management may juggle future needs proirities.

There's no more stigma to outsourcing that there is sending up a pinch hitter when your team is down in the eighth with the bases loaded.. Romney-izing a profitable company and sending it to India for a 1% economic gain projected over five years is most certainly a despicable form of money grubbing but farming work out to a domestic specialist may be simple but shrewd conservation of resources.

There are many reasons for and against any business decision. Avoidance of media buzzwords du jour should be far down in priority.

We agree on everything you wrote, including:

Forrest Addy said:

There are many reasons for and against any business decision. Avoidance of media buzzwords du jour should be far down in priority.

Click to expand...

opscimc said:

These kinds of considerations, not jingoist attitudes toward outsourcing overseas or not, are what will determine long-term viability of companies who have overseas competition.

Click to expand...

The moderator chastised me for using the word "overseas" when discussing outsourcing, as if that word were as forbidden as a discussion of Harbor Freight lathes. It is not the same. Also, not mentioning it does not make the issue go away.

I chastised you for attributing that to Forrest, which was not true of the post under discussion.

- Leigh

The real Leigh said:

I chastised you for attributing that to Forrest, which was not true of the post under discussion.

Click to expand...

Again, I did not attribute that to Forrest.

opscimc said:

Nowhere did I write, nor even imply, that Forrest mentioned "overseas."

Click to expand...

However, let's let this drop, and get back to metrology.