What's new
What's new

Reasonable angles to cut on wire EDM?

Fal Grunt

Titanium
Joined
Aug 5, 2010
Location
Medina OH
I have used LOTS of wire work, but have never run one myself. Leaves a bit of a knowledge gap. Previous employer the wire guys could cut me just about ANYTHING. They never protested, at least not to me.

Recently I have been working with a new *potential* customer with a pretty tricky tool to make. Issue is the required angles that the customer requires. I went to a new/old edm shop, new supplier for me and my business, but old, as in used by several shops I have worked at and at least one friend for over 30 years.

While taking a walk through the shop, the GM showed me some fancy new EDM that burns 3x as fast and that would cut up to a 45 degree angle.

Then when I sent in prints for quoting, they initially no quoted the job due to the angles. The angles are 10 and 25 degrees.

The tolerances are tight, and are compounded by the angles, but this is where my knowledge is lacking.

Working with the customers some today to see if we can redesign the tool and reduce those angles. From the wire work perspective, what would be a good angle to work towards? From the opposite end, I am working with the tool designer to reduce the angle as much from their end.
 
It's not really a simple answer. There is much to it. Angle will vary with UV offset combined with height. The angle created by offsetting in one axis 0.020" at a height of 0.250" is 4.57°. But if you increase the height to 3.0", the angle decreases immensely to 0.38°.

So, if the overall part is 4.0" high, you can easily see how achieving even 2° can get painful, quickly.

And that is just one aspect...

One might need different guides to get some angle work accomplished. Different flush cups...

And then there's the issue of getting the settings to work well for all cuts within the set up.

It can go on and on...
 
Moist I have worked with were good to around 15-20 degrees. It all depends on the flush cups, machine u,v movement, and tank size. The part dimensions can affect the amount of angle that can be cut. Theres always multiple setups and cutting vertical.
 
Hi Fal Grunt:
Zahnrad Kopf and plastikdreams have hit the highlights but there are some other details too that are worth knowing:

1) The machine you have will greatly influence what kinds of tapers you can cut and importantly what kinds of tapers you can cut ACCURATELY.
The old tilt head AGIE's were the best by far...precisionmetal who still contributes on this forum had one and knows the details far better than I but the gist of it was that tilting the wire guides and the flush cups meant that the wire never had to wrap around the guides so the wire was always running free through the guides and that allowed large tapers to be cut with the same precision as zero tapers.
Only these machines could do this.

2) Once you try to cut accurate tapers with a conventional machine, all things that don't matter too much when cutting zero tapers all of a sudden matter a lot.
Wire tension, flushing pressure. the height of the upper head, the hardness of the wire, the condition of the guides...all will play an outsized role in determining what you get, and the disappointment you feel when you first clock in your part on the sine plate and find out it's nowhere near what you programmed is a right royal pain in the ass.

Of course as the taper angle gets worse, the problems multiply, in addition to what Zahnrad Kopf described about the limits of the UV axis movements and the interference of flush cups and wire guides both ZK and plastikdreams alluded to.

So yeah, the machine makers CLAIM you can cut 45 degrees, and you can, but doing it accurately is a whole 'nother set of challenges.

One of the best ways I've found to cut accurate angles is to mount my parts either on a sine plate or on my rotary axis, but obviously you can only do certain geometries that way.
There is a rotary setup that's like a five axis mill...Jauch Schmieder made one and I think Hirschmann did too, however they are breathtakingly pricey so they are rare as hen's teeth.
But they'll do any taper you want (on small parts only), and dead nuts accurately.

So typically with a lump bolted to the platen and a 2" tall workpiece on my conventional CHMER wire EDM I can cut a 15 degree taper and it'll be within about 0.005" total deviation when I put it on a sine plate but do no test cuts and fiddling to get it better.
0.005" deviation in 2 inches is pretty fucking rough by toolmakers standards!
If I need it better I need to fiddle.
If I need it within half a thou I can get there with some pain.
If I need it dead nuts within tenths, it's a crap shoot.

I'll betcha precision metal could get within a tenth or two without even breaking a sweat on his old tilt head AGIE...that's the difference not curving the wire around the guides makes.
The old AGIE's were truly magnificent for taper cutting, but they were apparently the Devil's very own to realign if you bumped the upper head even a tiny little bit.
So nobody builds them anymore...pity really.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
I have used LOTS of wire work, but have never run one myself. Leaves a bit of a knowledge gap. Previous employer the wire guys could cut me just about ANYTHING. They never protested, at least not to me.

Recently I have been working with a new *potential* customer with a pretty tricky tool to make. Issue is the required angles that the customer requires. I went to a new/old edm shop, new supplier for me and my business, but old, as in used by several shops I have worked at and at least one friend for over 30 years.

While taking a walk through the shop, the GM showed me some fancy new EDM that burns 3x as fast and that would cut up to a 45 degree angle.

Then when I sent in prints for quoting, they initially no quoted the job due to the angles. The angles are 10 and 25 degrees.

The tolerances are tight, and are compounded by the angles, but this is where my knowledge is lacking.

Working with the customers some today to see if we can redesign the tool and reduce those angles. From the wire work perspective, what would be a good angle to work towards? From the opposite end, I am working with the tool designer to reduce the angle as much from their end.

I agree with Zahnrad Kopf 100%, there is no simple answer to that with all the variables in 4axis machining. I like to keep it no taller than 4.000in and no more than 10 degrees taper for precision. There are cases where I've gone taller but still keep the taper down. Sometimes it can require to mount the work piece on a 5 or so degree angle to take out some of the taper.
 
Another potential factor is if you need to have a 'land' area at the top or bottom of the taper. Like relief in a die block, but needing to control the land thickness with some degree of accuracy. That gets to be a real pain in the ass!
 
I appreciate all the great feedback, thank you!

I realize my description of the work is fairly vague, I would love nothing more than to show the print and ask for input. Unfortunately that is not an option.

To explain it further, in hopes of maybe getting a little more detailed, specific responses, the workpiece is 1/2" thick. The 25 degrees relief is flanked by the two 10 degree faces. The intersection of these faces have tight tolerance radius, and the face, is .060" wide. Plus nothing minus .001" tolerance.

The wire shop quoting the work, said they will put the tool in on the 25 degree angle, but their concern is the two 10 degree faces and holding the plus nothing minus .001 tolerance.
 
That's going to be an expensive part.

Here's the thing - as my old Friend Marcus correctly points out - setting it up and doing it are one thing. Make no mistake; I am in no way minimizing that effort.

However... verifying that with Sine Bar/Plate or CMM and holding those dimensions and tolerances? That is a whole other thing. It can get frustrating very quickly. I say that as someone that has done their fair share of 4ax wire work over the years now, and also as someone that has been doing a lot of 4ax wire work lately. A LOT. And the parts I am doing can not be accommodated with fixtures/set ups. The whole darn part is full of 4ax.

It takes an immense amount of effort and constant attention to achieve and then maintain accuracies like that.

It's easy to write. Not so easy to do in real life.

Good luck.
 
Maybe I am just lucky, or the machine was dialed in before my time, but I just checked a part we did this morning with a .067dia x .10" land with a .591dia at the top. It's an oddball angle, but with gage pins and our micro-vu I am fairly (80-90%?) confident the top diameter is within a few thou. Would I want to sign off that, no I would not for the record... just throwing out a real world example. Our machines are Sodick submersibles. I am not overly familiar with them so can't be more specific
 
Hi All:
Going back to the accuracy thing; wire work is often presumed to be dead nuts accurate and in many instances that is true, depending of course on what your definition of "dead nuts accurate" is.
What gets bandied about in our world is split tenths, so 0.00005" which is as much of a metrology challenge as it is a cutting challenge and as we all know, at those kinds of tolerances, the heat from a stray fart at the other end of the shop will fuck up the job.

So in that world, plus nothing minus 0.001" is often sniffingly referred to as "boilermaker tolerances", until, of course, someone is asked to actually HOLD those tolerances on a complex feature after which everybody immediately backs up ten steps.

When it's a taper job and the tapers are big tapers as these are, I would expect to throw quite a few away before I got a good one, and I'd be sneaking up on my features wherever I could.

With the feature you're describing, Fal Grunt, you have to try to imagine just how the Hell is anyone even going to measure whether it's good or not?
Verifying the face width of a small radius on a compound taper is an extraordinary metrology challenge, that you're going to be hard pressed to do in process.
Even with a high end CMM, you will be inferring what you've got from the points you can actually interrogate.
So at the very least you need really really repeatable fixtures so you can take the job off the machine, interrogate it , put it back on the machine and whizz off another tenth or tweak the angle a few arc seconds over in the desired direction.

If you can still find someone who runs one of those tilt head AGIEs you'd probably have the best chance at it, but it's still going to take a few kicks at the can to get a good one.
On that machine you'd lay down a big lump of steel, program to nominal, cut one, toss it if it's no good, correct your code and cut another, all in one setup.

Too bad precisionmetal is out of the game...he would have been the perfect guy to take on the challenge.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
Thank you all again for your responses!

I completely agree that the inspection part is nearly as difficult as the making part.

The customer receiving, can inspect the part. I am not sure about the wire EDM shop.

The customer walked me through their processes and inspection. They have a, I don't know WHAT it is called. But it measures radii, forms, etc, with a spec'd 50 millionths accuracy. It is a cross between a optical comparator and a camera? When they receive a tool, they lay it out on the machine.

So if you were doing this job, what would make you more comfortable? 12 degree, flanked by 5 and 5? Obviously the closer to straight the better, but I assume there is a point of diminishing returns?
 
We did quite a few tapers here in my shop and always checked the results over rolls. On occasion where there were multiple angles, we would set the stock up on a sine bar so we didn't have to offset quite to far.
 
As a guy who has fixed hundreds of those Agie machines, i usually planned on 2 full days to redo geometry and that was if i didn't have to work on the Z axis. When it was right it was amazing what you could do with the machine. no special guides, no additional calibrations, no corrections, just good old fashioned mechanical accuracy. Plus it was cool to be able to auto thread on an angle!
 
Hi again Fal Grunt:
I just realized no one responded to your question in post #11.
So here goes:
First, there are two things to be aware of:
1) As soon as you tip the wire you take a hit in accuracy and you lose maybe roughly one order of magnitude; so if you have a machine that can routinely cut within a tenth, now you can expect to cut within a thou.
This occurs because of the mechanics of what you are doing when you displace the upper guide relative to the lower guide and all of the new sources of error that accrue as soon as you do that.
Within a certain range, it doesn't matter all that much whether you are running a tiny angle or just a small angle.
2) There is another hard jump in inaccuracy as you get toward the maximum angle the machine has the UV travel for and the guides and flush cups for.
This comes from a number of factors that compound as the angle gets much above 5 degrees.
One of the less obvious factors is the amount the upper head is pulled out of position by the offset forces that do not exist when the guides are aligned, and of course, the quality of the machine construction determines how this theoretical factor plays out in real life.
My old Sodick A320 had a relatively weak upper head arrangement and I began to have measurable head displacement above about 10 degrees, but it was an old machine when I got it.
I can't move my upper head a tenth on my much newer CHMER, but that machine is built like a brick shithouse by comparison to the Sodick, and it's lived a much gentler life too.

So the actual angle is not the driving constraint in that first accuracy hit, it's the fact that there is any angle at all.
Here's another not very obvious factor.
The wire tension and the wire hardness are super important.
With a hard wire you have to force it around the curve of the doughnut shaped wire guides and it takes on a sinusoidal shape that gets worse as the wire tension goes down.
With a soft wire it's a bitch to thread and it cannot take very much tension before it breaks.(but it still need enough to be straight, not sinusoidal between the guides)
Servomotor controlled wire tensioners are way more consistent than powder clutch wire tensioners, so if your tension is fluctuating, so is your taper angle and face flatness and this means that a servomotor machine is typically better than a powder clutch machine if the goal is to have the same taper all the way around the part.

As you dive into the weeds with all this arcane nonsense, you can rapidly begin to see just how hard it is to make an accurate taper...any taper, not just the big ones.
So all must accept that order of magnitude loss of routine accuracy; you cannot simply program it and expect it to be good.

Zahnrad Kopf alluded to just how much work it can be to get what you want whenever you tilt the wire even a bit...these are a few of the reasons why.
That means you can't hope to just "fix" it by reducing the taper angles...the screwaround factor will still be there even if the angle is reduced to 1 degree.

A skilled and knowledgeable guy who understands his machine well can work around these limitations to make an acceptable part, but it can be deceptively hard to do.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
Last edited:
Most newer machines have taper / angle software that will calculate the exact pivot point on the guides allowing for precise tapers up to 45 degrees with the proper guides and if there is enough U / V travel.
 
Hi Shane:
I understand and believe you that the software to calculate the UV displacement needed to cut accurate tapers has improved as you say, but the mechanics involved are still the same.
I went through this when I was considering which wire machine to buy, and I got lots of promises from machine makers (granted this is a fair while ago but still...).
None would GUARANTEE the level of precision they implied in their hype, and in a fit of candour, one AGIE guy spilled the beans about how customers who ran tilt head machines were badly disappointed when those old machines were retired and replaced with more modern plain UV displacement machines and they suddenly found they could no longer hold tolerance on bigger tapers without a whole lot more farting around, and in some instances couldn't hold it at all.

Just as Zahnrad Kopf has occasionally commented on, about the difference between the hype and the reality of cutting accurate gears with a hobbing attachment on the CNC lathe vs a proper gear hobber, the sales brochures and promo videos look awfully good, but when you pin the sales guys down everyone starts to squirm a bit.

Now I don't KNOW that I'm speaking truth here, but I'm pretty confident.
Granted my machine is a Chevy, not a Cadillac, but the influence of the mechanical principles involved does not go away just because the brand name on the machine is different.

Just for shits and giggles, try noodling the wire tension up and down while skimming a taper sometime, and watch the sparks move up and down the cut face as you do.
That's just one of the factors you're up against, and in the tenths world we work in, it's a major source of error all by itself.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
Marcus,

Thank you for all your input.

I have phone calls out to a couple shops to find a good fit for this part.

thanks
 
I don't think the taper angle even matters (in theory), it is the tolerance that is screwing the job up. Current job we regularly burn 4-5-8 degree tapers. That said, it doesn't matter too much for function, BUT per my previous post, it cuts pretty damn close.

On an older Charmilles I have burned 20deg tapers and that machine was not even a submersible! :eek: (burned pretty slow admittedly...) :leaving:

On a side note about holding tenths in a wire edm. I once burned a mold detail (I think) about 3" tall. It was out .0003" or so and the customer was not happy. I thought .0001" or so per inch was not bad, that is .001" over 10".
 
So our machines have " taper flex " software. With a measuring jig it measures the exact pivot point at specified angles. Along with Taper Guides and flush cups we can hold + / -.0005 on 4 axis burns up to 30 deg. I attached a pic showing the measurements for upper and lower guides and the angle it measured.


taper.jpg
 








 
Back
Top