What's new
What's new

Burn depth control question

William Ward

Cast Iron
Joined
Mar 28, 2003
Location
Upland, California,USA
First off let me say thank you again to Marcus and Nobby for the help with the rotary dressing a couple of months ago. :D:D
I have a new question about depth control and maybe finding a better approach to what I am trying to do.

I am burning a straight sub-gate that is .550 +\- deep. (Some tools are a bit less, some are as much as .200 more.) The gate itself is .015 dia. with a .002 land between the cavity and the sub-runner or runner drop.
The problem I'm having is controlling the depth of the runner drop so that the .002 land does not come out too long or too short as either of those conditions affect the gate vestige on the part. If the land is too long the vestige is too big from the first shot and if the land is too short the gate will wear oversize in short order which also creates a large vestige.


Runner_Drop.jpg



Currently I am burning the gate from the cavity side as deep as I can(.150+/-). Then I am turning the part over and burning the runner drop from the runner side. I am using one rough electrode to get as much material out as I can and to connect the gate with the runner drop more sooner than later so I can then use the gate to flush from the bottom on the finish burn.

I'm using Copper-Tungsten for both gate and runner electrodes and the work material is CPM 10V at 62 Rc. IonoPlus dielectric.

The rotary dressing solved my location issues as far as keeping the drop and gate concentric but the depth issue is kicking our A!%.

I've tried reaching in with a .016 carbide pin from the runner side to measure the depth but I'm not getting consistent enough results to have much faith in that approach.
 
Last edited:
I had one of those 2 AM brain flashes:ack2: that the issue might be hydraulic pressure against the electrode but I have no experiance with that in EDM flushing.

If anyone has thoughts on this I'm open to all of it.
Thanks
 
Hi William:
These are the kinds of jobs that'll make you tear your hair out, so you have my sympathy.
Measuring something like this is the whole challenge and it can be the Devil's very own when you're sitting in the bottom of the work tank, the features are tiny and you need to be super consistent.

There are two basic ways to go about it:
One is to experiment with the process until you have it totally nailed down and then do it exactly the same way every time; relying on the consistency of the process to infer what you are going to get.
The other way is to try to interrogate the developing part features and make adjustments on the fly.

Way number one is how a lot of people do it, and with the precision of most modern high end sinkers you can do pretty well but your upfront investment in experiments is very high, and if you don't do a lot of them, it's hard to justify.

Way number two is a big challenge to get meaningful targets against which you can measure.
There are two basic ways to do that too:
First, you can try to interrogate the feature you're making.
Dental polysiloxane rubber and a good shadowgraph is one way, and it will get you into the range of +/- 0.001" or so but it gets a bit uncertain if you need to control it tighter.

Second, is to interrogate the electrode after the burn and find out what it looks like compared to how it looked when you started the burn, both for geometry and for position of the features you're most concerned about.
Touch probing it or inspecting it optically is the way forward, and most default to touch probing it while it's still on the machine because it's easy and you don't need to buy anything or dismount the trode.

If you can rig a measuring scope on the machine, you can use that too, but the key is to not remove the trode or the workpiece from the machine unless you absolutely have to.
When you're working very tight tolerances, the repeatability of your electrode mount becomes an important issue; 3R or Erowa or Hirschmann are pretty damned good, but they're not perfect.

I doubt very much that hydraulic pressure is the culprit in your case: the reason I am unconvinced is that the surface area of the trode is so small, the flushing pressure would have to be enormous to create enough force to deflect something unless the trodes are exceptionally long and skinny.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com
 
The end of the drop is hemispherical. That indicates to me that I can get accurate measurements over a precision steel ball, then retrieve the ball with a magnet. I'd find a ball bearing that is a loose fit. If too loose, lay the situation at the mouth of the chamfer in CAD so you know exactly where the top of the ball is relative to the gate land.

Measure with an AGD indicator. Long Island Indicator Co. has small threaded chucks, with either inch or metric threads to match your brand of indicator, that allow you to adapt the indicator to a sewing needle, or in this case, a piece of pin stock with the end ground flat so you know you are picking up the top of the ball consistently. If you have trouble getting the ball out of the hole (and turning up the flushing pressure isn't an option :-) silver solder the ball to the end of the pin and keep it for inspecting these set-ups.

Dennis

On edit, here's a link to the pin vise contact point: 133A : MTC Contact Points for Dial Indicators
 
Last edited:
Thanks for the insights guys.
I already have a modified pin-vise for holding long thin gage pins or in this case carbide rods in an indicator.
I also like the possibilities of the ball for giving a surface to check against. I'll model it up and see what I need there.
We have checked a couple using Repro-Rubber but it's a bit too subjective as Marcus suggests.
Thanks again for the help.
I'll keep coming back for more if I can't crack this.
 
Hi again William:
One of the difficulties you're going to knock up against is that your measurement will always be indirect; in other words, you'll measure what you can and then infer what you actually care about from what you could measure.
Dropping a ball into the subgate is like that; you need to KNOW your spherical gate shape is actually spherical and of the correct size; if it's not you're hooped (especially if it's too small or egg shaped and not truly hemispherical).
If it's not round you're also screwed, but I'm assuming you're rotating the trode as you burn, so not round would be pretty hard to achieve.



If you are going to do a measurement of this sort, you're probably better off with a short carbide or HSS pin that has the proper taper ground on the end of it, so it'll drop into the tip of the gate.
Make a little tapered bushing for it that'll drop into the main gate body near the runner so the pin is well aligned with the axis of the gate when you drop it in.
Now you can drop gauge the top of the pin, and that'll tell you where the taper is positioned.
Since the two things you really really care about with the gate is the consistency of the diameter and the length of the land, this is probably the most precise way of getting you to that goal.
It's a variant on Modelman's excellent plan, but it doesn't rely quite so much on all things being perfect with the rest of the gate geometry.
Get all the heights within a gnat's ass of each other and I think you'll have it beat!

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com
 
Last edited:
Thanks again Marcus.
I think you have hit it. I haven't been rotating the trode on these burns. Hadn't really thought about it. Looking at the electrode wear it would certainly help the geometry.
I like the tapered pin and guide bushing idea a lot. Between that and trode rotation I should be able to dial this in.
You are absolutely correct about inferring dimensions from what can actually be measured. I will probably still use the casting rubber and comparator method along side the more direct measurement method until I'm comfortable that my process and inspection is repeatable. We have 250 tools with this gate design so this isn't going away anytime soon.
 
Bump to the top for progress report.

I've had a couple of detours into other projects for the last 4 months so I hadn't spent much time on this until the first of the year.
Based on suggestions from Brian at Makino I have split the runner drop into two 'trodes. Well three actually. 1st is a rougher to open the through hole from the back of the part to the gate. 2nd is 3 identical 'trodes to finish the long taper. 3rd is also 3 'trodes to finish the steeper taper and establish the .002 long gate land. Measuring the gate land is still a major challenge but now that the burning process is under control I'll concentrate on how to measure it.
Thanks again to everyone for the help
 
Hi William:
Glad to hear you have made progress with this project; I was kind of wondering what had become of it.
Since you were able to enlist Brian's help, may I assume you get to play on a nice new Makino now, instead of the Mits you were using?
If so I will be envious!
If not kudos to Brian for offering his knowledge and advice anyway, even without a sale...the mark of a true gentleman!!

On another note, with your new protocol; is it going to be consistent enough to allow you to just dress the trodes to the numbers and burn to the numbers as I suggested in post #3, or will you still be forced to measure each gate as you make it and tweak the last tenths?

Keep us posted; this is the kind of project we all can learn from.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
Marcus,
No I don't have a nice shiny new Makino to play with sitting next to the V22 Graphite. I still want one though. From what I have seen sinker controls have come a long way in the last 15 years.
Especially in regards to small deep features.
And Brian, well he is just a great guy who is willing to help with ideas and being another set of brain cells. I seem to be short of those some days.

As for consistency. I think at some point I will be able to burn to the numbers but I need to burn many more to see how stable it will become.

What I have noticed in turning and burning to dress electrodes is that it works very well on small electrodes. The current gate electrode is .012 x.300 . I burn them from a 1/4 inch copper-tungsten blank that is turned to .03 x .325. Burning/dressing takes less than an hour and gives me an excellent finish on the electrode. The short taper electrodes are turned from the same blanks as the gate electrodes but are .04 x .700 long. Dressing consists of dressing the taper and then dressing the end to establish the minor diameter and give a sharp corner on the end of the electrode.

In the future I will be making the gate electrodes from worn/used short taper electrodes and I will make the short taper electrodes from long taper 'trodes.

Something else that I got from Mits application support that is new for me is editing the generated power setting to remove all but the finishing E-packs. If i generate settings for 3 'trodes with .01 sq. in. area there might be 6 or 8 E-packs that start as aggressively as the control thinks a .01 sq. in. area can handle and then as the burn progresses it uses progressively less power to get the finish required. By editing out all but the 1 or 2 least aggressive settings I get much better final finish and far better depth control.

I gave up on dressing the long taper electrodes on the machine. It just takes way too long. What I did instead was to make a Hardinge tapered back plate to mount a 3R Macro chuck to so I can turn the long taper electrodes on the lathe in the holders they will be used in.

I will do my best to keep you posted on further developments.
 
Last edited:








 
Back
Top