What's new
What's new

How to scale trodes for over burn??

HSM_CHIEF

Aluminum
Joined
Aug 2, 2017
Hey everyone I currently use solid edge primarily for my modeling and fusion 360 for my cam and a little modeling I’m proficient in both but I was just wondering how does everyone go about compensating for orbit on an electrode? Say I want the electrode .010 a side bigger for orbiting is there a way to manipulate the model so its finish electrode size? I know I can just run a negative stock to leave under my stock to leave tab but that only goes to 1/2 of the tool dia.... I often have to use tool down to .010 in dia and I’m not able to run a .010 a side offset I can only run a .005 which takes a lot longer to burn... I was just wondering if anyone else knows a way to do this other then making a completly new model! And no scaling does not work!! Thanks everyone!
 
Hi HSM_Chief:
I shell a copy of the model with the amount of orbit + overburn I want, turn it into a dumb solid, bring it into an assembly with the original model, mate the two together than subtract the shell from the original model with the cavity cut feature.
Now I have a true trode model I can just put HSM toolpaths on with no lying about the cutter size or running negative stock allowance.
It almost always works just fine but sometimes Solidworks will shit the bed when trying to calculate the Boolean subtraction.
When that happens usually I can make it work by modeling a big lump onto the outside of the shelled model to avoid the dreaded "zero thickness " condition that makes Solidworks get all pissy.

It's a bit of farting about but it's the best way I've been able to discover so far and I've been with SW since SW98.
If someone has come up with a better way; I'm all ears!!
For all I know SW maybe has a utility or there's an add-on that does this brilliantly already.
Also I don't know if Solid Edge or Fusion can do the shell or the subtraction...I've never run either.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com
 
Hi HSM_Chief:
I shell a copy of the model with the amount of orbit + overburn I want, turn it into a dumb solid, bring it into an assembly with the original model, mate the two together than subtract the shell from the original model with the cavity cut feature.
Now I have a true trode model I can just put HSM toolpaths on with no lying about the cutter size or running negative stock allowance.
It almost always works just fine but sometimes Solidworks will shit the bed when trying to calculate the Boolean subtraction.
When that happens usually I can make it work by modeling a big lump onto the outside of the shelled model to avoid the dreaded "zero thickness " condition that makes Solidworks get all pissy.

It's a bit of farting about but it's the best way I've been able to discover so far and I've been with SW since SW98.
If someone has come up with a better way; I'm all ears!!
For all I know SW maybe has a utility or there's an add-on that does this brilliantly already.
Also I don't know if Solid Edge or Fusion can do the shell or the subtraction...I've never run either.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com

Thanks Marcus, what’s the point of doing the shell tho? I’m not really familiar with the shell command I rarely use it. How do you make that compensate for your overburn and orbit? Can you just offset faces?. I guess I’m just not really sure what you mean... but I’m interested in how you do this tho because sometimes it’s a real pain in the ass running a negative stock allowance or lying about my tool dia... we’ve been fighting this stuff at our shop for a while... I mean we make it work it’s just a pain. Once again thanks again!
 
Hi again HSM_CHIEF:
The shell operation on the first copy of the model creates a solid that represents the overburn plus the orbit.
It is the amount the electrode must be undersized by, and it is a very quick and easy way to get the same offset everywhere.
You can use it to cut that amount away from either your original part model, or even just a lump you've modeled to represent your electrode blank.
I have to scoot for a bit but I'll post a pictorial later on today so you can see how I do it.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com
 
Hi again HSM_CHIEF:
The shell operation on the first copy of the model creates a solid that represents the overburn plus the orbit.
It is the amount the electrode must be undersized by, and it is a very quick and easy way to get the same offset everywhere.
You can use it to cut that amount away from either your original part model, or even just a lump you've modeled to represent your electrode blank.
I have to scoot for a bit but I'll post a pictorial later on today so you can see how I do it.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com

Sounds good! I really appreciate it!! Thank you
 
OK here it is.
There are two versions of how to do this that I use and I've touched on both.
The first is to make the shell cut onto a model of the part.
I've added a lump onto the model first, then constructed my shell mated them together and cut the shell into the part model to get my final trode.
So here's my model:
part section.JPG

I add a lump to the bottom to make my trode blank:
model with lump.JPG

I take a second instance of my model and shell it to the thickness for overburn and orbit:
(this view is a section view so you can see how thin the shell is if you look closely at the edges of the model)
overburn shell.JPG

I align my shell onto the trode blank:
shell superimposed on part with lump.JPG

Last I cut the shell into the modified part model:
PM won't let me add more images to this post so next post I'll show the rest.
That last attachment is an oblique view of the part...PM went kind of goofy on me and won't let me move it back where it belongs or delete it.
MC
 

Attachments

  • part 1.JPG
    part 1.JPG
    24.7 KB · Views: 89
Last edited:
Here's the next bit from the previous post; the shell cut into the part model: the shell is the red part obviously.
(I've sectioned the shell so you can see it's relationship to the trode better)
sectioned shell on trode.JPG


Here's another view:
Boolean cut shell to part model.JPG


And here is the trode sitting in the cavity showing the overburn and orbit gap.
Obviously this is a trode for a spherical orbit (the orbit in this example is 0.010":
cavity with trode.JPG
The next post will show the variation where I make a lump for my trode and then cut my shell into that lump.
This way is better for some trodes; usually the simpler ones.
Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com
 
Last edited:
here's the last post of this set

OK, the other way:
Make the shell as you did before.
I posted originally that the shell gets turned into a parasolid then back into a dumb Solidworks part.
That's because the conversion breaks the links between the part model and the shell model better than any other way I know and sometimes SW has a hissy fit about cutting one configuration of the part into another configuration of the part and calling the result a third configuration of the part so I like to be safe and make that hissy fit impossible.

I should have added that comment into the first method since it applies there not here...my apologies but I was having a senior moment at the time because PM was doing weird shit, and I got distracted.

So in this version I model a trode blank:
trode blank.JPG


Here it is with the shell sitting in it:
part in trode blank.JPG


I make a conventional cut so the shell cut will separate the trode model into two bodies:
rough cut.JPG


Last I make my cavity cut to make the trode shape and just pick the body I wish to keep:
trode model from trode blank.JPG


My trode sitting in my cavity:
cavity with trode.JPG
So that's it folks; two ways to break out trodes that almost always do the job for me less painfully than offsetting surfaces and figuring out what's what.
Remember though these trodes are for SPHERICAL ORBIT.
If you want to make trodes for XY orbit or other variants, you have to modify the technique to suit, and that typically involves modifying the shell.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com
 
Last edited:
OK, the other way:
Make the shell as you did before.
I posted originally that the shell gets turned into a parasolid then back into a dumb Solidworks part.
That's because the conversion breaks the links between the part model and the shell model better than any other way I know and sometimes SW has a hissy fit about cutting one configuration of the part into another configuration of the part and calling the result a third configuration of the part so I like to be safe and make that hissy fit impossible.

I should have added that comment into the first method since it applies there not here...my apologies but I was having a senior moment at the time because PM was doing weird shit, and I got distracted.

So in this version I model a trode blank:
View attachment 216666


Here it is with the shell sitting in it:
View attachment 216667


I make a conventional cut so the shell cut will separate the trode model into two bodies:
View attachment 216668


Last I make my cavity cut to make the trode shape and just pick the body I wish to keep:
View attachment 216669


My trode sitting in my cavity:
View attachment 216670
So that's it folks; two ways to break out trodes that almost always do the job for me less painfully than offsetting surfaces and figuring out what's what.
Remember though these trodes are for SPHERICAL ORBIT.
If you want to make trodes for XY orbit or other variants, you have to modify the technique to suit, and that typically involves modifying the shell.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com
Thanks for this Marcus!!! I understand now! I’ll def give this a shot on my next electrode! The stuff I’m doing is quite a bit more complex than the picture you posted but I think the same concept still applies! I’ll give it a try and let you know how I do! Once again thanks for all the time!
 
Hey everyone I currently use solid edge primarily for my modeling and fusion 360 for my cam and a little modeling I’m proficient in both but I was just wondering how does everyone go about compensating for orbit on an electrode? Say I want the electrode .010 a side bigger for orbiting is there a way to manipulate the model so its finish electrode size? I know I can just run a negative stock to leave under my stock to leave tab but that only goes to 1/2 of the tool dia.... I often have to use tool down to .010 in dia and I’m not able to run a .010 a side offset I can only run a .005 which takes a lot longer to burn... I was just wondering if anyone else knows a way to do this other then making a completly new model! And no scaling does not work!! Thanks everyone!

Are you doing round orbit? Spherical or some other complex orbit?

A long time ago we used to manipulate models but we got away from that method because we were able to do it faster in the operations or with cutters. We also use a lot of templates and journals to automate as much as possible. If you set up some operation templates with various commonly used operations/settings you can program electrodes, or any part, very fast.

For medium and large electrodes we use round orbit and we generally never change the electrode models which saves us a huge amount of time. For programming these electrodes we use bullnosed cutters and specify the the diameter undersized so it has the result of applying round overburn in the XY plane. Some shops do not like this method because, depending on the EDM's capability, there can be a margin of error on large draft walls such as 10-45 degrees or more of draft. This is mainly due to many EDM's finishing patterns which will raise up slightly for final orbiting after the floors have been finished. Our machines raise up .002-.003" but they can be programmed to stay down if need be. Not always the best for burn conditions but it can work. Another factor for us is our molds are typically medium and large tools so there are typically a bit larger part tolerances so use use that to our advantage.

When we program for spherical orbit we always use negative stock in the operations and cut all surfaces, including staging surfaces. We use this orbit pattern on for smallish electrodes, which require a high degree of accuracy, because it can add a significant amount of burn time for larger trodes with a lot of surface area.

A good example using our round orbit work flow is to program a large trode for .010"/s orbit. I have my electrode model which I had subtracted from my cavity and it's in its own prt file, then I import a set of pre-made operations from my library which are close to what I need, select the body for part geometry and then start generating operations. In parallel if possible. NX is very good at tool path containment so I rarely create any geometry at all. Once in a while I might draw some lines or arcs for trimming but it is rare.

Here is another a big time saver...if you ever have to program multiple overburns for the same electrode, such as a roughing trode and a finishing trode, you are going to save even more time. Take my previous example using round orbit and lets make a .020"/side rougher or perhaps a fine .005"/s finisher. What I do in NX is simply import a set of tools with the appropriate diameters, copy the existing operations and file the copied operations under the new tools then regenerate all the new operation in parallel. Super fast even on large electrodes. If you are doing multiple overburns for a trode with spherical orbit then you'll have to do this in the operations but, again, not a huge deal if you are using templates. In NX I can specify different geometry groups with various part stocks so when I copy and paste operations they will automatically inherit stock as well as some other settings of my choosing. (can you tell I really like Siemens NX? LOL)

In summary always try to use all of your software's capability so you don't create any more geometry than you have to. This goes for models as well as any curves, surfaces and solids for tool path containment/extensions. Do whatever is fastest for you but I encourge everyone to keep an open mind on what other people are doing.

NX 11.0.2
NX12
 
Hi Qwan:
I know exactly what you're talking about and I used to do it exactly the same way.
Interestingly I started out where you are now and moved away to my current method, whereas you started out where I am now and moved away to your current method.
So each of us found something unsatisfactory about our starting points.

In my case it came about when I was forced to shift from Mastercam to HSMWorks (my ex business partner HATES Mastercam and LOVES HSMWorks and I've come to mostly agree with him).
Because it's comparatively easy to make model changes in Solidworks and then just run whatever HSMWorks outputs when the toolpaths are rebuilt; it's become the default way for me.

When it was Mastercam it was a royal pain in the ass to bring in a new model and try to rebuild all the operations, and it was far easier to simply do as you are doing, but you could never tell graphically what you were going to get and you couldn't interrogate the model to check what you cut; you had to infer the actual trode dimensions from what you knew about how you lied to the CAM software.

So overall the new way is mostly working well for me, but it depends on getting a good shell cut, and Solidworks can make that a real pain in the ass at times.

When Solidworks gives me the big finger I go back to how you do it and how I used to do it and that works well for me too.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com
 
Are you doing round orbit? Spherical or some other complex orbit?

A long time ago we used to manipulate models but we got away from that method because we were able to do it faster in the operations or with cutters. We also use a lot of templates and journals to automate as much as possible. If you set up some operation templates with various commonly used operations/settings you can program electrodes, or any part, very fast.

For medium and large electrodes we use round orbit and we generally never change the electrode models which saves us a huge amount of time. For programming these electrodes we use bullnosed cutters and specify the the diameter undersized so it has the result of applying round overburn in the XY plane. Some shops do not like this method because, depending on the EDM's capability, there can be a margin of error on large draft walls such as 10-45 degrees or more of draft. This is mainly due to many EDM's finishing patterns which will raise up slightly for final orbiting after the floors have been finished. Our machines raise up .002-.003" but they can be programmed to stay down if need be. Not always the best for burn conditions but it can work. Another factor for us is our molds are typically medium and large tools so there are typically a bit larger part tolerances so use use that to our advantage.

When we program for spherical orbit we always use negative stock in the operations and cut all surfaces, including staging surfaces. We use this orbit pattern on for smallish electrodes, which require a high degree of accuracy, because it can add a significant amount of burn time for larger trodes with a lot of surface area.

A good example using our round orbit work flow is to program a large trode for .010"/s orbit. I have my electrode model which I had subtracted from my cavity and it's in its own prt file, then I import a set of pre-made operations from my library which are close to what I need, select the body for part geometry and then start generating operations. In parallel if possible. NX is very good at tool path containment so I rarely create any geometry at all. Once in a while I might draw some lines or arcs for trimming but it is rare.

Here is another a big time saver...if you ever have to program multiple overburns for the same electrode, such as a roughing trode and a finishing trode, you are going to save even more time. Take my previous example using round orbit and lets make a .020"/side rougher or perhaps a fine .005"/s finisher. What I do in NX is simply import a set of tools with the appropriate diameters, copy the existing operations and file the copied operations under the new tools then regenerate all the new operation in parallel. Super fast even on large electrodes. If you are doing multiple overburns for a trode with spherical orbit then you'll have to do this in the operations but, again, not a huge deal if you are using templates. In NX I can specify different geometry groups with various part stocks so when I copy and paste operations they will automatically inherit stock as well as some other settings of my choosing. (can you tell I really like Siemens NX? LOL)

In summary always try to use all of your software's capability so you don't create any more geometry than you have to. This goes for models as well as any curves, surfaces and solids for tool path containment/extensions. Do whatever is fastest for you but I encourge everyone to keep an open mind on what other people are doing.

NX 11.0.2
NX12
Quan this is pretty much how I’ve been going about making my trodes in the past but my problem is the such small tooling I can’t run a large over burn with a negative stock allowance.... so In some cases I do need to create the model to finished electrode size... but since we’re on the topic and implmex and you seem to know a lot more about the EDM process then me.. I have a question say I make an electrode with a true .010 a side 3D over burn or “spherical” and say that it’s a tapered pocket and my pocket is supposed to be .100 deep if I run a true 3D tool offset my pocket depth with come out .110? And we have to grind off the electrode when I’m done machining it? This makes no sense to me to run it true 3D with a Z offset like that? Our highs on our parts must come out right within +-.001 in most cases. We are running a sodick edm. Is grinding the electrode off when I’m done pointless? Or can we just run an XY offset my maintain my desired depth then burn the electrode 3D?
 
Hi again HSM_CHIEF:
This can be a bit tricky to visualize, but here's how it works.

The first fundamental thing to remember is that the amount the electrode will orbit is NOT the same as the amount the electrode must be undersized; that is because what we usually call the "Orbit" is actually the orbit plus the overburn.
The overburn of course varies with the power we put into the electrode, so on a modern machine with an adaptive generator the actual machine orbit varies as the generator changes it's power settings while it's cutting and the machine recognizes the overburn amount and adjusts the orbit on the fly to keep from overcutting or undercutting the cavity.

In the kind of spherical orbit I'm familiar with (an X-Y orbit of constantly diminishing diameter), the actual spherical part is done only at the very end of the cycle, and the tip of the trode ends up only the overburn distance away from the floor of the cavity, while it also ends up at the overburn plus the orbit distance from each vertical sidewall, and some fraction of the overburn plus orbit distance away from all the angled faces of the cavity.

So by the very nature of the spherical orbit, the electrode is deeper in the cavity than it would be if it was placed in it's "nominal" position (equal gap all round between cavity surfaces and trode surfaces) once it reaches the very end of its path, and it's deeper by the amount of the orbit and shy of the cavity floor by the overburn amount.

Clearly this distance that the trode goes "deeper than nominal" depends on the size of the spark gap, which of course depends on the part finish desired.
So the electrode must be at least the same height as the finished cavity is deep and needs to be cleared at the parting surface so it can't hack into the parting surface when it's gone below nominal.

Where you set nominal depends on your preferences and the way your machine calculates what it is supposed to do.
There are a couple of ways this can be done:

The old-fashioned way is to touch the tip of the trode on a reference surface and tell the machine to go down the depth of the cavity minus the overburn.
This is how it was universally done in the days before CNC EDM.

A more common and more "Modern" way is to establish the position of every trode holder to a common datum in the trode milling machine and set up the EDM so it knows where that datum was set.
When the toolpath is programmed in the EDM, it will allow for the extra depth it must go beyond the datum to end up with a proper depth of cavity, and it will calculate that allowance based on the finish you tell it you want to end up with.

So long as the trode is necked back so it can't goober the parting edges of the cavity, it doesn't matter how much longer the trode is cut. but both machines must know where the datum is and it cannot be allowed to change.

There are obvious benefits and drawbacks to each approach; which you pick depends on what is easier for your workflow, your machine and is the least confusing to you.

The "old" approach has the benefit of making recutting the trode easier...when you recut the trode you obviously booger the relationship between the trode datum and the clamping face of the trode holder...with the "old" way you don't care because you will reestablish the proper relationship between workpiece and trode when you do your touch off with the trode tip just before you begin burning again.

The "new" way has the benefit of allowing you to cycle trodes through the milling machine and the EDM machine without having to keep track of the datums; they never change, but you waste more trode material, because you cannot recut trodes without boogering that datum relationship.

Intermediate between the two extremes of course is to keep track and then adjust the EDM machine's datum after every trode recut by re-inputting it's value rather than doing a re-touch...it comes to almost the same thing but if the trode is very fragile you risk it every time you do a manual touch, especially if you can't keep the reference surfaces spotlessly clean.

So courses for horses; the main point is that the trode plus the neck down land must be longer than the cavity is deep, so grinding off the end of the trode harms you if it was cut properly to start with.
Your cavity will be distorted when the trode is all the way down.

Measuring the trode can be deceptive; that's why I like going to the pain of having a true trode model; I can mate it in the cavity, section the lot in my CAD program and see if it's going to be OK at all extremes of the orbit envelope.

On a last note, if you look critically at the trode model in my example you'll see it's not correct if you ever wanted to actually burn with that trode.
It's not cleared quite enough at the parting surface...it's going to wipe out the cavity edges when it finishes its orbit.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com
 
Quan this is pretty much how I’ve been going about making my trodes in the past but my problem is the such small tooling I can’t run a large over burn with a negative stock allowance.... so In some cases I do need to create the model to finished electrode size... but since we’re on the topic and implmex and you seem to know a lot more about the EDM process then me.. I have a question say I make an electrode with a true .010 a side 3D over burn or “spherical” and say that it’s a tapered pocket and my pocket is supposed to be .100 deep if I run a true 3D tool offset my pocket depth with come out .110? And we have to grind off the electrode when I’m done machining it? This makes no sense to me to run it true 3D with a Z offset like that? Our highs on our parts must come out right within +-.001 in most cases. We are running a sodick edm. Is grinding the electrode off when I’m done pointless? Or can we just run an XY offset my maintain my desired depth then burn the electrode 3D?

For discussion's sake we'll ignore talking about the spark gap and just concentrate on the overburn because it all comes out in the end...
Another way to look at it is when you cut an electrode for spherical orbit you are removing the same amount from all surfaces. All the Z levels will be correct because when the electrode will be down to final depth. As for staging, all of our Makino EDM sinkers stage from whatever face we chose and we don't have to compensate for the overburn but I have heard of some sinkers where the operator does have to compensate for it.

You could even design the staging compensation into your trode design but whatever way you decide to use the important thing is to have the programmer and the edm operator know about it, agree on it, make it part of your standard operating procedure and update your procedure document. Having everyone in the know and on the same page is so important! When parts get to the edm stage they are very expensive!
 








 
Back
Top