efficiency of wEDM

[Doug]

This is a continuation of a thread in the General Board about using a wEDM to manufacture Douglas parts. This thread covers the non-Douglas portion of the pros & cons of wEDM, efficiencies, and reliability.

Link to previous thread:
https://www.douglasmotorcycles.net/index.php?topic=6670.msg25076#msg25076

Re: 2 3/4 Kickstarters
« Reply #3 on: 09Aug17 at 08:53 »

Hi Doug,

Not sure why you think they are high maintenance? My experience has been the complete opposite, its and extremely efficient and reliable technology as there is a very minimal load on any moving parts during the cutting process. Also, this machine is quite unique on the consumables side. It only uses about 6 meters of wire at a time that's wound on a drum at the back of the machine. It runs the same 6 metres of Wire slowly back and forward, reusing it over and over again giving you about 40 hours of cutting out of the one 6 meter length of the Wire. All other CNC Wirecutter I have used, the Wire passes through the job once and its dropped into a bin at the back of the machine. As the machine runs untended I'm not worried if it runs all night.

Cheers, Glenn

Glenn,

Based on as I said having run four and been responsible for authorizing and writing the checks for supplies, service, and maintenance.

These were for used machines dating to c1997 and a new one I bought for the shop I was managing in 2010. We ran them two shifts, however it was a prototype shop environment and not a mold & die shop; so while we ran them every workday (and occasionally overnight) we were not cutting big slabs of steel for days on end. Utilization was about 30%, as I preferred to have a machine waiting around for a machinist rather than a guy waiting in a queue to get in a wEDM.

Now you already know about wEDM, but I am going to go into more detail in case anyone else is reading along.

The manufactures I have used are Charmilles and a Mitsubishi (new). The reusing of the wire for several hours is something I have not heard of, nor could I find much about the model you have. The ones I have used or seen run the wire through once. I have heard of some shops on roughing passes reuse the wire a second time. This is on machines that did not crimp or chop the wire on exit. Given the amperage used for roughing, they reported a high rate of wire breakages and so confined the practice to daytime hours when machinist were around. The alternative was to reduce the amps so the wire still had sufficient tensile strength for the two passes. Of course, the cutting speed was reduced and for most shops the time was more valuable than the saving in wire. As for wire, we used the most popular size (in our locale, anyway) which was 0.25mm. I would say we ran this 99% of the time. For fine work we would use 0.1, generally to get the smaller corner radius. The new machine would reliable auto-thread and work with 0.05mm though we never had occasion to go that small or the exotic realm of even small wires sizes for mirco-wEDM. Stocking diamond wire guides for each size wire and model machine was expensive, at about $800-1200 a set, so we just kept the two sizes on hand and it met out needs. I suspect that your machine is using a larger wire in order to be able to keep reusing it over and over. I realize that does not mean it is less accurate, it just means the minimum inside corner radius is larger. The kerf will gradually decrease as the wire is eroded away, but I imaging the manufacturer recommend putting in a fresh wire when you run the finish skim passes, if the finish and accuracy requires it. Sometimes the rough pass off the machine is good enough.

But it is still slow, whether you are standing there watching it or not. Speed is a factor of how much amperage you can put through the wire and how a high frequency you can generate as erosion is only one spark at a time. Excluding your machine, every other one is spooling off wire and using electricity during that time. Milling is still faster. While some of my guys did tend to use the wEDM and an electric band saw, it was considered a high-value process. It could do things the milling machine couldn't. Hardness of the work piece was not an issue; it just had to be conductive. Cutting forces were low (some electromotive forces, and pressure from flushing) meant fixturing and work holding could be light. Some parts were too delicate to grab securely for milling, but could be lightly gripped for wEDM and cut. While the wEDM would have smaller servo drives and lacked a main spindle motor for a milling cutter, they still consumed more electricity than the medium sized CNC milling machine centers handling work pieces of comparable sizes. While I did not monitor power consumption, the wEDM were fused at half again higher amperages, based on their kVA ratings on their data plates.

And it did not end there. We had all of the machines on a Preventative Maintenance plan (it was a requirement for us). This was partly our in-house maintenance, and for the wEDM, a PM annual by the manufacture. That was about $4k per machine and included all the wear items (excluding the diamond guides), wire feed belts, the bearing exposed to water and prone to deteriorate, and a few days of the factory technician's time to tear the machine down partially and reassemble. Of course we could have had our own guy do the work, or hired an independent machine repair firm, but the parts still came from the OEM and it was a package deal with the tech support. So there was not much point in shopping around; they still gouged you for the parts and you were paying someone to do the work. The advantage with the OEM is we could get them to often throw in a few extra parts for free. Also you could pump them for information while on site. Of course we could call them too, but not quite the same as cornering them by the machine and asking them why the 'xyz' would do the 'abc' when you were using the 'klm'.

Outside PM services did not include filtration or water treatment, which we took care of of. When I started using wEDM in 1992, the typical production wEDM were submerged work pieces. Then the 'spray' models dominated. Then as the generators improved and the amperage got higher the trend swung back to the submerged machines. All of course require conditioning and filtration of the flushing water, thought the rest of the tank if a submerged model could be unfiltered. Each machine had a bank of paper filter cartridges. These we changed twice a year and were about $200 to $400US per machine. We also drained and cleaned out the tanks as part of the twice annual service. The filters and the water have dissolved metals in them which could be absorbed through the skin, so disposal was considered a hazardous material and cost extra. We were able to go twice a year because we kept the tanks chilled 24/7 (they were tied into the buildings HVAC via a chill water heat exchange, and saved us the space and maintaining of individual chiller units) and we have a UV lamp in the flush water circuit to keep the organic growth down. We were considering adding an ozone generator too, but never got around to it.

Accurate work requires maintaining the conductivity of the water in the kerf, so we monitored the micro-Siemens of the water in the flushing tank in real time. This was conditioned by the deionization resin canisters that the flushing water was run through. The reality was you just needed to check the monitor once a day as an indication of the health of the DI resin, and make sure nothing had unexpectedly failed. These we changed twice a year at about $600 a bottle, and we were buying 'reconditioned' tanks that we about half new media rather than brand new. We also had a deionization bottle that we used to pre-treat the water that we topped up the tanks daily (sometime twice if used a lot) to offset evaporation. This took some of the load off the bottle attached to each machine and we noticed made them last much longer. That bottle got changed twice a year as well (same cost as the one on each machine), but was servicing three machines.

The machine I bought new in 2010 was trouble free, and so were the older ones because of the PM attention they received. Even so, you could tell on the older machines by the way it ran when it was coming due on a schedualed PM. Usually the auto-threader would stop working reliably, as they were the most sensitive to build up and deposits. Then getting the correct burn when cutting tapers (they all had U&V axis for taper cutting) would start to get unreliable and you would have to tweak the default power settings to compensate. We kept them clean, but you still had to periodically take the upper and lower heads apart and clean out the insides. Even with the PM on the older machines the circuit board would start failing. Particularly the board for the generator as it was switching the highest amps and frequency. Once they got over ten years old, about every three years one board or another would have to get sent out for a rebuild. We kept spare boards in stock salvaged form identical models, to minimize down time. We had a few milling and turning centers that were far older that did not have those kinds of problem. Replacing CRT tubes, yes, circuit boards, very rarely. Something about the amps, frequency, and 'noise' in and around the wEDM was hard on electronics. We did not send the boards back to the manufacturer, it would almost be cheaper to buy another used machine! But even so to have a power board rebuilt by a reputable outside firm was about $5k.

So my experience was they were much more expensive to keep. Absolutely necessary to have for the work we were doing, but milling was hands-down cheaper. And we did have some very good mills (and machinists) so we did not have to use the wEDM except where milling just couldn't cut it (sorry for the pun). That was pretty rare as besides the 'conventional' milling centers we had several 5-axis mills and 8 and 11- axis turn-mill centers. Sometimes it was a job that could be milled quicker, but we wEDM cut it anyway because the setup was simpler/easier/faster, or we were not throwing as much material away in work holding. Not that cost of material was a big factor for us, but sometime you only have so big a lump of metal in the stock room and you cannot wait to order in a larger piece.

So when I looked at CNC machinery for my own workshop, I had no problem with buying a second-hand lathe (that I knew the history of) and figuring I could get many more years use out of it without major expense. Ditto if I bought a CNC mill (new), that if I took good care of it, I would likely avoid expensive service calls. Yes, I could and would do the repairs myself, but doing that work for free does not make the machine trouble free! It would still cost me time, if not money. But when it came to the wEDM, most of the industrial scale machines that are readily available and modestly priced, I remember how much effort we put into keeping our older models reliable with extensive PM programs. Even with that, and on the new machine that didn't need replacement parts (yet) the consumables, supplies, and support infrastructure was a lot of money. That does not stop me from wanting one. But there are a lot of things I would like to have that I can manage without. Now if I did see one of the smaller Brothers wEDM machines, in a late model, and a keen price...

-Doug


[fix typos. 10Aug17 -Doug]