Printer Friendly

Sizing up in-process gaging.

A window of opportunity is opening for in-process gaging. Its driven by these key needs: to push part tolerances beyond tenths, get immediate feedback when the process wanders from the ideal, and boost quality and reduce scrap.

In-process control, its proponents argue, can do these things. The challenge is convincing potential users that in-process measurement really works in harsh machining environments. There are still a lot of nonbelievers.

In-process gages are not cheap, ranging from $3000 for the most elementary manual system to $40,000 for the most elaborate. They're not usually for low lot sizes or the everyday job-shop situation. The ideal application is automotive-high production lots, tightening tolerances, and a desire to take operator judgment out of the loop to boost the uniformity of critical parts.

In grinding, where in-process gaging has been available for decades, less than a majority of machine purchasers today choose this option, probably a lot less. The estimates range from as low as 10% to as high as 50% in Germany and Japan where it has been better received. Says Aldo Vrh, national manager, Marposs Corp, Auburn Hills, MI, the leading producer of in-process gaging, "Why would anyone want an open-loop control when he can use an in-process gage to close that loop? I ask myself that question many times."

In-process gages, says Vrh, help the operator optimize the grinding cycle. "The result is higher precision, productivity, and process reliability. Nearly any grinder can be converted to in-process control. In automatic applications, the amplifier/controller produces signals to the machine's CNC based on actual part size to change feed, begin dwell, or initiate retract movements of the wheel. In manual applications, the operator reads part size continuously on the amplifier display and adjusts the machine to get final part size."

What can I measure?

In-process gages can adapt to a wide range of diameters for multiple cuts, Vrh points out. "Our wide-range measuring head can automatically reset to any diameter within its 20-mm (0.8") range, and we have a gage that makes absolute measurements over a 100-mm range (4") without recalibration." The latter has an internal glass scale for basic macromeasurement and an LVDT sensing element for the fine-tuning micromeasurement.

Marposs (and others) also offer gages with shock-absorbing devices to measure interrupted-cut parts, such as gear profiles. This is particularly useful in gear-pump manufacture to tighten the clearance between gear teeth and housing to boost pump pressure.

Calibration can be quick, a matter of seconds. With a special quick-release gage, you insert a master part, release the finger clutches, move the fingers to the right clamping position, relock them, and the gage is ready to run. The final fine-tune zeroing is done by the electronics.

In Marposs' case, all gages are based on inductive or LVDT sensors that can readily provide measurement accuracies of 1 micron, and when necessary, repeatabilities down to 0.1 micron (0.000 004").

Capacitive challenger

Leading rival to Marposs, both here and in Europe, is Movomatic. Until two years ago, Movomatic had been content to have its gages hitchhike into this country on advanced grinding and superfinishing machines from Switzerland, Germany, Britain, Italy, and Spain. Now, the company senses that the US market is ready to expand its use of in-process measurement, says James R Osborne, general manager, Movomatic USA, Greenville, RI.

Movomatic gages are based on a proprietary capacitive transducer. Whether capacitive or inductive, the sensor is always well shielded from the grinding environment and linked by parallel mechanical fingers that do the actual touching of part OD. To provide for the inevitable crashing of the gage, breakaway mechanisms assure that a shear pin or the fingers themselves snap or bend before the more expensive sensor inside is damaged.

Osborne feels strongly his capacitive transducers are superior to inductive sensors (unaffected by ambient temperature change and magnetic chucks, for example, he says). "We are quite happy to quote repeatability to 4 millionths on all our gages," he says, basing much of this on the gage's parallelogram finger movement and integral coolant flow for temperature stability. In critical applications in the 0.1 to 0.2-micron range, Marposs also insists on coolant flow to stabilize their gage and a thorough evaluation of the specific application.

Both companies agree that grinding parts to an accuracy of 0.0002" or 0.0003" is no longer a major challenge for a good grinder. But to split tenths, you must consider both gage quality, and its environment. You must examine temperature factors, the type of grinding wheel, the type of material, rpm, feeds, speeds, etc. The grinder can't compensate itself for process variation-that's why you use in-process gaging.

Really millionths?

It's hard to imagine measuring to millionths in the machining environment with all its variables. Things bend, distort with temperature, swarf or chips get in the way, or films form on the spinning part. Can you really measure on the machine as well as in the lab?

"Yes," Osborne answers, emphatically, although admitting it is difficult to convince people this is true. (See box, View from a grinding expert). "However, if you have tremendous variations in that gage's working environment, then obviously there is going to be variation in the finished part, regardless of what any gage can do. You need gaging that is stable, reliable, and stays set once its calibrated. "Certainly, the most important variable is temperature fluctuation. We accommodate that by gage design, choice of finger materials, coolant flow, and our sensing transducer. As a result of this, there is no need to actually monitor temperature." Tip wear can be combatted by choosing diamond, sapphire, or carbide inserts, although hardened steel is standard and usually sufficient. Because tips tend to last for months or years, people tend to forget about gaging errors caused by this wear, Osborne warns. So, if you're trying to crack the 10-millionth mark, he advises, don't use old tips. They're not expensive, and are easy to replace. One environment too difficult to locate in-process measurement is centerless grinding. You can't measure the part in real time anyway because it is changing dimension as it goes through, and the grinder could not respond quickly enough to adjust for a single part. So, the exit dimension is used to establish trends, and a good candidate to do this post-process measurement is air gaging.

Turning is different

With turning (particularly hard turning) challenging grinding, in-process gaging is moving into turning. Unlike grinding measurement done in real time, turning measurement must be done when the part isn't being cut. The challenge here is finding available space, and swinging the gage into measuring position repeatably. Besides dodging cutting tools, the gaging difficulty here is the danger of trapping a chip in the gage. Says Movomatic's Osborne, "I have been approached by companies where one operator tends four or five turning machines and can't be constantly making adjustments. Our in-process gages can do that adjustment, and because these companies are achieving quite good RMS finishes, it makes our job even easier." What about the chip problem in the lathe situation? "Air nozzles can be placed to blow chips away. With CNC pushing surface finish to the limits, users can't afford chip accumulations that could cause scoring, so they're doing that for us anyway."

ID gaging?

Both Control Gaging Inc, Ann Arbor, MI, and Marposs offer in-process gages for ID grinding-slim sensing mechanisms for inserting inside the grinder's workhead. Measurement can be either simultaneous with grinding (if space between part ID and grinding wheel permit), or synchronized to follow along beside the wheel as it oscillates. Any droop effects of the cantilevered sensing mechanism, says Marposs' Vrh, are offsetting. "What you subtract from the upper probe, you add to the lower probe." This measurement option is offered on machines made by Cincinnati Milacron and Bryant, and used in the automotive, bearing, off-highway, and agricultural-equipment industries, he points out. Movomatic does not offer a gage for measuring ID grinding in real time, feeling that there's not enough space to maintain the precision of their OD grinding gages. "The accuracies they are getting can be achieved judiciously by a grinder without a gage," says Osborne. "We do do IDs, but not in real time. We come in through the back of the workhead with a pivoting mandrel, measure the ID of the surface, and then withdraw to allow the quill to enter. And, of course, we do IDs with the traditional air-gaging probes."

In-process vs post

There are some distinctions to be made between in-process gaging and the far broader field of post-process gaging. While acknowledging that in-process gaging has made great strides in conquering the hostile machine environment, Edmunds Gages, Farmington, CT, a manufacturer of post-process gaging, points to their recent installations of automatic post-process feedback systems at GM Saturn and Ford Motor.

"In-process gaging does not really measure part size," notes Edmunds' Don Maurer. "Part size is measured by an off-line gage and used to calibrate the in-process gage's set point. Thus, the machine's capability is a function of the accuracy of these off-line instruments.

"In contrast, automatic post-process feedback gaging can measure part size (and control quality) free of the machine's influence."

Another company close to the definition of in-process gaging, but on the post-process side, is Valenite Gaging Systems, Madison Hts, MI, a division of GTE Valenite. Valenite's Paul Mueller agrees that measuring a part while it is being ground is a miserable environment to work in, particular ID grinding. Their idea of "in-process gaging" are the gaging stations they integrate into transfer lines for the auto industry. Although "in" the transfer-line process, they are technically post-process because they don't measure parts while they are being cut, but moments later.

"We do a lot of cylinder boring with our cutting tools, and at the following station use gage heads to measure these bores. That's still a pretty rotten, wet environment-- certainly not a clean room.

"Our Camset systems statistically process that measurement data with SPC to determine the trend in bore size, and then feed that back to the metalcutting stations where we have adjustable boring bars. So, this is a closed-loop system: gage the bores, analyze the data, and adjust tool position to hold bore size. Minimum adjustment is 10 millionths. We also do this for facing and turning operations where the adjustable tool is stationary."

Match grinding

GTE Valenite also does what Mueller calls "pre-process gaging" or match grinding. "For two parts that have to fit together, we first measure the bore, grind the stem to fit, and then measure the stem post-process to prove we did the right thing. If the gage has enough range, we can use the same gage for pre-and post-process gaging."

Although match-grinding technology is admittedly twenty years old, few people are doing it, Mueller admits. Both Marposs and Movomatic also offer match-grinding systems (first measure the bore in various ways, and then use in-process gaging to control the size of the ground OD).

Like others, Mueller is amazed at how few grinding machines are purchased with in-process gaging. "People either don't need precise size control, or they do not believe in measurement on the machine. Yet, on the opposite end of this spectrum-transfer-lines-there probably isn't an engine block or piston produced in this country in the past decade that doesn't use some form of on-machine gaging and size control. For connecting rods, maybe 50% use size control.

"When I see these systems working in the real world, I'm constantly impressed-that we can make equipment with this sophistication work in this environment!"

View from a grinding expert

Dr Stuart Salmon, noted grinding consultant and creep-feed expert, agrees that the in-process idea is gaining credibility here, but he sees its use mainly in a relatively narrow window between the limits of what you can do on a good machine without gaging and the micron area where you begin to question the reliability of in-process measurement. "It's not for super-precision work. For production work with relatively open tolerancing, its okay for anything less demanding than [+ or -]0.0005"."

Salmon's estimate of the percentage of grinders bought today with in-process gaging is less than 10%. "Mostly, you see this in high-production automotive work with long setup times because it takes a long time to get the in-process gage system to work. If you're doing medium to small batch quantities, you're better off taking the part out of the machine and putting it into a post-process gage to actually control the process based on that measurement."

The big question, he says, is why try to react to tiny fluctuations beyond the capability of the machine to respond. "I don't know of a grinder that can even move within a few millionths to respond to measurements of that degree. Parts are going to be vibrating with an amplitude larger than that. Then there's the thickness of coolant films, and the effects of wear on the touch-type probes. And when you get into millionths and looking at surface finish, there are peaks and valleys within that tolerance, and you have to ask what is the probe riding on?"

He also senses little difference between the acceptance of in-process gaging here and in Europe. "Where gaging is required in Europe's high-accuracy processes, it's usually a post-process gage feeding back correction information to the machine for the next part."
COPYRIGHT 1991 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:includes related article, experts opinion
Author:Sprow, Eugene
Publication:Tooling & Production
Date:Dec 1, 1991
Words:2229
Previous Article:Tool steels for tough tasks.
Next Article:Finding flexibility in automated assembly.
Topics:


Related Articles
Gaging systems take command.
Column gages - holding their own.
Tool training pays dividends.
Flexible systems go with the flow.
Quality at IMTS: making it right the first time.
The Taguchi fallacy - a way to go wrong in SPC.
Guidance on thread gaging.
Cellular machining with confidence.
Choosing the right gage.
Standard seeks to curb inspection vagueness.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters