electrodes

[Müřlıń®]

+1 to Roger on radii being a diffrent size.

 

+1 to Lucky on cost effectiveness. Especially in forging dies.

 

Most of the time I'll bet you can get away with one to two electrode with an orbitor, instead of the three to four on a RAM type machine.

 

I never got to use an orbiting EDM.....

 

I heard you could get some fantastic finishes...

like 10 RMS...with no craters....

 

 

Murlin

[Roger Peterson]

Henry, I was just trying to illustrate why fudging the diameter doesn't give accurate surfaces.

 

 

HTH

[Racer X]

I agree with Lucky. I have been building electrodes for the better part of 15 years and used to do it in the "lie about cutter size" method. Ever since being introduced to the negative stock feature, it is the only way to go. We do really complex organic shapes and shutoffs and have NEVER had a problem with them fitting. Back in the old days I was taught to always leave a little preload on them and let them work their way in. (The old "Tonnage is our friend" approach) Now I will leave a maximum of .0005 on a shutoff if I am worried about it but not very often. Most of our parts are nylon or acetal and I have no flash problems. I have found that depending on your edm operator (some really hate programming for spherical orbit, which they must do for this to work. Although, I don't know why they oppose it) you may run into some resistance from them. If you can get them to start programming that way your problems are over.

 

I am unable to download your file from the .ftp. However, if you want to email it to me I would be more than happy to take a look at it.

 

[ 12-17-2003, 01:40 PM: Message edited by: Moto GP ]

[Racer X]

Spec

 

I got your file and played with it. I created my own cutter paths (into the goofy method I would use to cut this). I have it down to two operations with the same cutter. .0001 scallop. My backplot time is at 35 minutes as opposed to the 1hr 5 minutes on your file. I opted not to go with a surface rough program because I find that it just causes the machine to do a bunch of airball time. I went full depth with a .01 stepover. Hopevully, this is a graphite electrode.

 

Also I turned filter on which converts spline moves into arc moves therefore making a smaller, more smooth, and faster running program. Filter only works in either 0 or 90 degree cutting directions, however. Not in the 45 degree direction that I like to use. Anyway, it is two passes with the same cutter (.500 ball) in different directions to remove the scallops. Check it out and see what you think.

 

File is the same name with "modified toolpaths" added

 

[ 12-16-2003, 08:04 PM: Message edited by: MotoGP ]

[Midwest]

+1 on the neg stock and shutoff's

 

quote:

---

I have found that depending on your edm operator (some really hate programming for spherical orbit, which they must do for this to work. Although, I don't know why) you may run into some resistance from them. If you can get them to start programming that way your problems are over.

---

We have a Mits. EDM and programming for a spherical is fast and easy. The only reason I can think of is they take longer to burn and sometimes don't flush as good as some orbits. Our mits has 2 different spherical orbits to choose from.

[foolsh]

i find that i can't use a flat endmill for sharp z depth at bottom of the electrode,becos you can't leave negative stock for flat endmill.that's a problem

[Midwest]

You have to lie about your EM size and lower your TLO by your overburn (per side). If you use neg stock with one cutter and alter the dia. of an other tool in the same program to get square corners in your trode.

 

[ 12-17-2003, 10:54 PM: Message edited by: Midwest ]

[Midwest]

Another option is to use a bullnose cutter with a radius the same size as your orbit on the EDM. You have to use the spherical orbit to do this.

[specprogrammer]

Wow, you guys are still going on this? There's alot of good info here. Moto GP, thanks for looking at my electrode. It's always good to see how someone else does things. We normaly use the "lying" method. I just wanted to see if anyone knew of an easier way. Thanks to all.

[mold100]

After 1000,s of years people still dont realize lying causes problems, in almost any sitchyation.

[Roger Peterson]

I agree with Mold100. Sometimes the problems caused by lying are insignificant but it does make inacurate geometry none the less.

[Fig sign]

quote:

---

After 1000,s of years people still dont realize lying causes problems, in almost any sitchyation.

---

+1000 !

 

If your goal is to get the edm electrode smaller for a spark ammount that`s what I do programming and making them for 7 years.

Never lying about tool diameter or offset .

Always using negative offset .

If your step is 10 mm and you taking negative offset the for both surfaces the step will remain 10 !!

I imply negative offset for all model all operations :

Facing ,pocketing ,contour (x,y offet and Z offset too ),wireframe operations like swep2d ,all surface finish operations !

If my model MUST BR FINISHED WITH END MILL WITH CORNER RADIUS 0 and you know there is a limitation that MC will not calculate the offset ammount less then corner radius there are many fast ways to overcome this :

Nowadays I prefere to built new model offseted to the spark ammount with the help of a solid ->shell command and mill it to offset 0

Or use other ways for milling that has no such limitation like contour with taper angle or swep2d toolpath

You can also offset surfaces but this is not a simple thing to do and you must many times make additional trimmings after that (timeconsuming )

 

You can find examples of files with explanations

covering most of what I am telling here in My Lazy Iskander dirty tricks thread .

 

Anyway everyone can do what he preferes ,if the results are OK.

But for accurate work with accuracy like

+- 0.01 mm lying is Dangerous ,IMHO

 

This is what I think and do and I stick to it .

 

HTH