Three axes to grind?
Our grinding industry may also be making a last stand, some feel. The only metalworking segment to lower prices in 1982, they are suffering, no doubt about it. They are accused of living in the stone age by their more nasty detractors. Very often small and family owned, some of these companies may soon be steam-rollered by the McDonalds of metalworking. Only their customers will decide.
I'm a small businessman too, and getting smaller all the time. I'm worried about the grinding industry--I've got a lot of friends there I would like to help stay in business. Since you people out there are going to make the final decision, I thought you might like to get both sides of this story. Vendor's viewpoint
Charlie Gallmeyer is vice president, marketing, of Gallmeyer & Livingston Co, Grand Rapids, MI, and facing the biggest dilemma of the small domestic grinding company today--to CNC or not to CNC? They specialize in large high-precision surface grinders, a small segment of a fairly small industry.
As Charlie told me recently, "In our long history, we have always had fairly limited R&D budgets, yet we've had to go off on a lot of wild-goose chases--air bearings, hydrostatic ways, wheel-balancing gimmicks, cylindrical-column ways, etc. We've got at least five 40-yr veterans here who've been tantalized too many times with 'first of its kind' ploys that never panned out.
"This is very risky business! Machine tools are facing high technological change at a time of very little growth. The risks are great and the rewards do not promise to be commensurate.
"A sophisticated control can more than double the cost of a machine. One of our biggest decisions this year is whether to offer three-axes CNC programmability in a creep-feed surface machine. Is there enough of a market to justify our R&D expense?
"We recently solved a creep-feed grinding application with single-axis downfeed control a lot cheaper than computer-based controls could have done it. It had all the sophistication the customer wanted. A lot of surface grinders are used to make one-of-a-kind parts. An average-skilled operator can do that job faster manually than anyone could program the machine just to make one part.
"This may sound reactionary, but a lot of applications don't need all the razzle-dazzle that a lot of people think they need. We've seen a trendiness here that's not all bottom-line calculations--a feeling that 'we've got to have it, no matter what it costs.'
"Take conventional reciprocating surface grinding. Do you really care how far the part overruns the wheel--to ten thousandths of an inch--when you just need a flat surface? That's why we're putting our programmable-control efforts here on downfeed control.
"Yet people come to us and say, 'Everything else we have is three-axes. That's the way we program. That's what we want.' We're afraid this is just another flash in the pan, and that people will come around to our point of view--that this is overkill--after we've spent $500,000 developing a product they find they don't need. And then they will blame us to boot because we're the grinding experts!"
Thus Charlie feels three scoops of CNC is nonsense for reciprocating grinders when one should do it, but for creep-feed surface grinding, yes, three axes does have potential. The question is when do you guys really want it: today, next year, or several years from now? Aerospace rebuts
To help Charlie find out what's really on his customers' minds, I interviewed three of the people he's turned away for three-axes CNC surface grinder quotes. First I talked with Fred Wilson at Rockwell International, Downey, CA. He was the most critical of the three on the response he's been getting from the grinding industry.
"Our grinding requirements range from very simple flat sheets, ground to enhance the finish, to some very complicated, sophisticated, and convoluted pieces that also must be ground to hold finish tolerances. We're finding that milling is getting so good that we don't need to grind to hold physical tolerances any more. But we have yet to find a way to mill an 8-rms finish. That still has to be ground.
"We are looking for a three-axes machine so we could use lower-skilled operators to set soft stops from the control panel. We need user-friendly controls. There are fewer and fewer "real" machinists in industry today. All those skills that used to reside with the operator, we must now rely on the machine to provide.
"So we need controls that are either CNC, with a skilled machinist/programmer providing the expertise via a program, or an interrogative manual-data-input (MDI) control that is user-friendly and can be addressed in conversational machine-shop language. You can't take a three-dimensional grinding problem and split it up--one axis CNC-level downfeed control and the other two for the operator.
"I'd love to find a grinder that could go both ways--MDI for the simple parts and CNC for the complex. And I have not found one yet. I've found milling machines that can go both ways, but nobody's making a grinder that can--either domestic or foreign.
"The closest so far is one in Germany, but they use so many bells and whistles to do the MDI work that they almost destroy the value of the MDI. It gets too complicated.
"As far as the grinding industry goes today, I don't think the majority of them know what day it is. Most are still operating in the 1920s. Yes, they certainly know basic grinding, but they haven't kept up with what's going on in the world around them.
"For example, I was really surprised at how few grinding manufacturers have adapted to creep-feed grinding techniques and shocked at how long the Europeans have been doing creep-feed grinding--not how much they do today, but how long they have been doing it. This shows how long we've had our head under our wing, ignoring what's going on in the world!
"I feel that creep-feed technology is still almost unknown in this country. Milling is taking jobs away from grinding because we can mill to 0.0002", yet with the latest continuous-dress creep feed technology, grinding could take those jobs back because you can remove metal faster than you can mill it.
"But where can you buy those kind of machines in this country? And if you go overseas for it, you will pay a lot of extra money for the machine. And I hate to put that much distance between our guys and the parts and service supplies we would need for such a critical machine." Complex extrusion grinding
A second viewpoint came from a production engineer at a leading aluminum extrusion plant in the Midwest. "We make extrusion mandrels that go inside helicopter blade shapes as they are extruded, and we need a grinder with three-axes control. We have been successful in finding one supplier who is building us a grinder based on one of their standard models. It took about a year of investigating to find anyone willing to do that for us.
"Our part is not symmetrical along the part centerline. We have other machines to make round mandrels. Most vendors wanted to sell us a standard machine. None really wanted to go through the trouble of programming. If you look at a rotary indexer on a screw machine, that's all we wanted--a precision grinder with a rotary indexer and three-axes control. But only one of the 20 companies we contacted was willing to go through the trouble of working out the programming.
"We make these parts today by gluing templates onto the end of the steel, and running the grinder up and down the part until we start knocking the blue off the template, and then index it a half degree and start grinding away again. We needed something that will index itself, keep track of the X and Y axes, and keep the wheel trued up. There was no standard machine to do this, and that's what everybody wanted to sell last year. Hell, I would like to sell only standard extrusions too, if I could, but I can't!
"Another problem was that we might make two or three mandrels of each shape, and that would be it for a couple years. So to go back and make a replacement, we needed something that's easy to set up, and that's also why we needed CNC. Now, typically, it takes 6 weeks to make a large complex-shaped mandrel. And if it's a very important shape, we would have to build a third mandrel as a backup-backup, even though it will probably never be used, With CNC, we will skip that extra backup, because we know we can grind a backup in one week, whenever we break a mandrel. And if it is not a critical part, we can eliminate the need for any backup at all.
"So there's big savings in inventory. We have the ability to make some 10,000 shapes with only a couple days' notice, so we have a lot of tooling sitting around. The more we can reduce that, the better off we are.
"I'm not really dissatisfied with the response we got from the grinding industry. What we were asking for was really off the wall. We get lots of interest in any of our more conventional grinding problems. Everybody is pounding on our door to solve those!
"Without that third rotary axis, I would have needed a guy out there turning the part by hand! Or use a wheel-former approach, reforming the wheel for every mandrel and every different-pitched area on it. You could go through a whole stack of wheels in one big hurry!
"The labor angle was also important. We have one specialist who likes to work the 3 to 11 shift and that's one reason it takes 6 weeks/mandrel. He does all the work on them, and he's good at what he does, but he can have a couple of off days and mess up a mandrel.
"Also, with CNC, we can count on the second one being exactly the same as the first, although repeatability is a secondary benefit. Our main goal was to reduce grinding time with CNC from 6 weeks to 1 week.
"I was willing to build the machine myself if I had to. I knew where to get all the components. We have a fairly good CAD staff with the latest McAuto system, so we could have done it. It would have taken an extra year or so, and tied me up for months on end to get it set up and going, but we would have, if it had come to that!" Knee-joint reaction
A third viewpoint came from a project engineer with a very specialized application, "We were looking for a machine to grind cam lobes on orthopedic knee-joint replacements--a better way of doing it. This was a low-volume product, and it didn't justify a really elaborate machine. We were thinking that maybe a CNC surface grinder in the creep-feed mode would do the job best.
"Of all the people we contacted, only four or five responded, and not all of them proposed surface grinders. The more favorable ideas were cylindrical grinders, and we agree that we need the flexibility to grind other parts besides this one.
"And we are not looking down our nose at anyone who declined to quote, because that happens to us quite often with our oddball applications.
"I think the US grinding industry today is giving us good products, but they seem too hesitant in going into computer controls. A lot of them are doing a good job in magazine loading, through-feeding, or in-feeding, etc. But as tolerances tighten up, or we have to hold one part diameter concentric to another diameter, nobody really offers what we need here.
"We are thinking about surface grinders because we feel with CNC we would benefit from being able to move the table, rise, move over, etc. We have about 30 CNC machining and turning centers, but have no experience with CNC grinding. Our typical batch sizes are low -- we would consider 100 pieces a large production run, and for the machine we are seeking, we might make only 800 pieces a year. So we need the capability to do many other part types to justify a new machine. But that is exactly the advantage of CNC.
"We're thinking of creep feed because this knee part has a 6"-long stem on the bottom side that, we feel, could be creep-feed ground in one pass. We are now doing these parts, without the stem, on a modified cam grinder that used to grind Chevrolet cams. The only problem with this approach is that we can't clear the long stem. And this triggered our search. If it takes a couple of years to find what we want, that's fine, we can afford to keep on looking.
"I can appreciate the position of the small company like G&L who makes a good grinder, but has a very limited R&D budget. Unless they have something that really fits in well with our requirements, it would be cost-prohibitive for them to get into it." What about you?
So how do you feel about all this? Should Charlie and the grinding fraternity be moving more rapidly in the three-axes direction? If you would like to see three-axes CNC offered on a reciprocating grinder, circle E36, or to vote for three-axes CNC on a creep-feed surface grinder, circle E37. For literature on the surface grinders Gallmeyer & Livingston has right now, circle E44.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Joe Doaks Machine Systems|
|Publication:||Tooling & Production|
|Date:||May 1, 1984|
|Previous Article:||No short cuts to new technology.|
|Next Article:||Rotary dressing: new technology, new choices.|