Cuts carbide better than butter?
Well, I'm here to tell you that you better rethink. I sat in on an interesting display of enthusiasm by a young man who seems to be sitting on a gold mine. He was updating our editors on the latest EDM developments, and pointing out that he can sell you machines that cut carbide surfaces so fine they can hardly be measured.
If you've got as much to do with tooling as I have, you will need to read this and get up to speed with me; that is, if you want to compete with me in the future! Taking the measure of EDM
That man is Randy Bormann, national product manager, Agietron Corp, Addison, IL. According to Randy, "People still think of EDM as a tap-buster, rather than as a productive part of the toolroom.
"And they don't understand EDM'd surfaces measured in rms. Specifying an EMD'd surface in rms yields a finish that looks far worse than it really is. When an operator looks at our cut samples to decide which he needs, he will pick a level 9 (which is 300 rms) because he thinks it is 32 rms. People assume that an rms 32, because it is a very rough ground surface, will also be a rough EDM surface.
"Many people also don't understand what recast means," he continues. "We used to say it was best defined as equivalent to te overcut per side. If the rough-cut gap between a workpiece and electrode was 0.0001"/side, you would get a recast layer 0.0001" deep, for example. Some of the melted material, before it has a chance to pull off the surface and be flushed away by the dielectric, recasts itself on the base material.
"Because this metal resolidifies on the surface, it becomes far more brittle than the parent material. In short, the faster we cut, the higher the current required, the worse the finish, and the thicker the recast layer. The main point to remember is that surface finish is controllable.
"So we try to limit amperage, and over the past ten years we have gotten a more finite control over the low-amperage ranges, and the lower the amperage, the less recast we can offer." The challenge of cutting carbide
How'd you get into cutting carbide? "We sold our first wire-cut machine in the US in 1969, and people have wondered ever since what else they could do with them. For example, they needed to section several hundred-dollars' worth of carbide using several thousand-dollars' worth of diamond grinding wheels. So they asked us if they could EDM this carbide, and we said it was possible, since it conducts electricity. But they soon discovered that when EDM-cut carbide tooling was placed under intense pressures in stamping or forming operations, the carbide would flake.
"We have been successfully cutting carbide for about five years, using some precautions. It cuts much slower than steel, because it is much more dense and far less conductive. We take a full cut to rough out the material, and then a trim cut to clean the surface. And we taught people how to plan for the limited amount of flaking that occurs." Searching for understanding
"Despite this success, we felt it was high time to study exactly what is going on with EDM'd carbide. We were impressed by the clarity of some photos taken by scanning electron microscopes, and felt that an SEM would be useful to look at EDMhd carbide particles.
"In the past, using toolroom microscopes, we had noted areas along the cut surface where, for no apparent reason, holes would appear--in some areas far worse than in others. We wanted to compare capacitive-discharge EDM'd surfaces with pulsed-generator-produced surfaces.
"The former uses a first-generation generator. The EDM control system discharges capacitor-stored electrical energy based on the frontal-gap voltage, which varies directly with the distance between the wire and the workpiece (which is constantly changing). The problem is that sometimes the controller senses the right voltage gap for discharge before the capacitance is at full power, and you get major variations in discharge energy, even with microsecond timing.
"On the other hand, the pulsed generator looks at only one thing--stored energy--and makes each discharge exactly the same in energy content. Every on-time period is exactly the same duration, only the off-time periods vary. The result is a much more consistent finish.
"This explains the random holes on the capacitance-discharge EDM'd surface. The power supply stored too much energy waiting for the proper frontal-gap voltage, and then discharged a tremendous amount of energy that takes out a big hunk of material.
"Carbide is essentially a bunch of tungsten particles bound together by cobalt. If some of the cobalt is removed at a cut edge, some of the tungsten particles fall off. What is happening at an EDM'd surface is actually an electrochemical machining (ECM) process. The higher the voltage levels, the higher the cobalt depletion. Thus, we can confidently say that the finer the limits we put on process voltages, the better the finish on the resultant carbide part. Of course, this limits the speeds we can achieve. We also can get much sharper edges and geometries with pulsed-EDM'd surfaces."
So they decided to spend a little money on independent metallurgical research. "When we looked at the results of metallurgical edge studies using an SEM," Randy explains, "we saw that the tungsten particles, normally very sharp and distinct, were all rounded off. We can now definitely say that with carbide, there is no recast layer. It must be called a 'remelt' layer. We don't remove metal and recast it to the surface. The material is so resistive, it justs melts and lays down on the surface. The SEM photos show all these rounded particles--at a total depth of about 0.001", and the remelt area is about 0.0002"." 23 percent less cavities?
"There are still some holes in the material. For the capacitance-discharge cut area, we see that cobalt is still there, but its concentration is variable. In the pulsed sample, the concentration is consistant from the remelt area right down into the parent material. There is still some flaking of tungsten in the remelt zone, but it is very consistent. We don't lose the full 0.0002", but only between 50 millionths and 0.0001"."
But what about those holes you were seeing? "We have since found that since carbide is a sintered material, it sometimes has areas that are not up to the standards the material manufacturers warrant. We have had discussions with a coupleof major carbide manufacturers who have seen this report and accept its legitimacy. We are now testing various kinds of carbide to see if one reacts any better in EDM situations than another, and if so, whether we could recommend it.
"The main benefit of our latest technology is eliminating the need to plan on flaking or the need for secondary operations to meet finish requirements. We can now guarantee finer EDM'd surfaces than have been possible before and eliminate most secondary operations, except those necessary from a cosmetic standpoint.
So what about the carbide-grinding diehards? "Sure, you can grind very accurate punches and dies, but you cannot grind and control the consistency of the clearance between those punches and dies as well, as as easily, as you can on a wire-cut machine. The longer life we're seeing is not necessary because the surface quality is better, but because the clearance between male and female part is the same all the way around. Even where the dies are 50 millionths off, they are off all the way around, not just in one spot.
"We have competitive situations where we and two other suppliers do test cuts for a customer who evaluates them and finds that all three had the same surface finish. This is because our surface finish is now at the point where a standard profilometer is no longer valid--its tip just isn't fine enough to measure the deepest valley. Thus our sample and Brands X and Y all look the same. If they could get down into the deepest areas on X and Y, they could see that they are not the same. Really, the only way to measure surface finish today is to etch the top surface and examine it with an SEM. But who's willing to take those steps?" Who needs surfaces that good?
Do people really need these levels of surface quality? "Some say they want pockets to hold oil," Randy replies. "They put grease or oil on their punches, but I don't think that's functional anymore. Fineblanking dies have zero clearance and oil or grease dose absolutely no good in those applications.
"Yet, the question is valid about whether our EDM'd surfaces are so fine that the market is not interested. I don't think so. For example, we can give the compacting-die maker a surface he couldn't get before without a secondary operation.
"Engineers today have fallen in love with four-place decimals. These tolerances are very hard to achieve unless you can work on dies in the very fine finish ranges. But I admit that not everyone needs this." Sittin' on a gold mine?
Randy is very enthused about the market potential for EDM. "I predict that in the next five years, EDM will become the most popular selling price of tool-room equpment in the US. The tool-and-die shop that does not have a wire-cut EDM will not stay in business! And the moldmaking shop that doesn't replace existing conventional EDM with CNC EDM will be noncompetitive! The driving force will be tool quality.
"As far as we can tell, the total in-use US wire-cut EDM population is 4000 machines. I've been told that that numbering, the US number should be two to three times the Japanese number. To be seven times the Japanese number. To be seven times less is incredible to me!
"EDM is still in its infancy in this country. The Japanese and Europeans have made EDM a productive part of the shop. They design for EDM, and look for reasons to take advantage of EDM.
"We have a long list of companies ith wire-EDM machines doing nothing but outside job work. I love to find people having $8000/yr of wire-EDM work done on the outside, and show them how much more they could be doing with their own machine.
"An average starting investment is about $100,000, ranging up to the most expensive at $250,000. One customer of ours billed $24,000 the first month he had our wire-cut machine, which was roughly six months of payments on the machine. His next five months were free--all profit! It's a very productive machine, particularly with tapering capability. The only real limitation is that it must be a through cut.
"On the labor side, CNC has simplified the operation tremendously. The learning curve on our CNC wire machine is about three months now. After that, the operator can handle 95 percent of the jobs he'll ever face. A conventional EDM machine takes much more experience and has a longer learning curve. It simply takes more moxie.
"CNC has changed conventional EDM from a 80 percent attended, 20 percent unattended operation to the opposite 80-20 relationship. When you add toolchangers, automatic control of the power supply, automatic flush controls, etc, you can eat away at the remaining 20 percent."
So if you'd like to be more competitive in carbide cutting and hear more from Agietron, circle E1.
|Printer friendly Cite/link Email Feedback|
|Publication:||Tooling & Production|
|Date:||Jun 1, 1984|
|Previous Article:||Indirect labor - the bane of automation.|
|Next Article:||DROs: a closer look.|