Tips for milling electrodes

[Webby]

From plastav:

 

I want to share tips for milling electrodes.

A lot of people have trouble with milling electrodes with flat end mills

and negative surface offset , because it is currently impossible with

Mastercam with 3d toolpathes.

There are couple of simple ways to do so without fooling mastercam with

diameter of tool and radius.

 

1. If you have a solid model it is always possible to make a shell from

it with a thickness to a needed offset.

After that to make a boolean remove and the result will be a model with

a needed negative offset which can be milled with any tool with an

offset 0.

 

2. The old wireframe toolpaths like swept2d or ruled can be milled

with a zero radius tool with a negative offset and this may be a

solution for many simple cases.

 

3. When you don't have and can not make a solid and the second way is

not suitable, you can make the offseted surfaces.

I do it this way:

 

A. copy the target layer on the separate layer and make them one color.

 

B. make new active empty layer for the resulted surfaces with a

different color.

 

C. Analize->surfaces->set norms to set all the surface normals in one

direction.

 

D Create ->Surface->offset to offset them .

If the result is not good after undo change the offset value to the

opposite and proceed once more.

 

You can check a result visually.

Sometimes some entities may be in the opposite direction anyway.

You can correct them making surface offset once more with the double

opposite amount .

The result needs sometimes manual trimming of some surfaces.

Now it can be milled with any tool with an offset0.

 

Sorry for the possible grammar mistakes.

[Webby]

Received this response to a Tech Tip area post, so I've moved the Thread here for discussion.

 

quote:

---

I would like to reply to the 'tips for making electrodes' by Plastav. While its true that shelling out a solid will decrease the 'core' part of your solid, it will also increase the cavity part of it. I had been very discouraged with trying to find a way to make trodes when I happened on this tip, and jumped in head first. Unfortunately the only dimensions I checked after shelling my solid were the 'core' dimensions. I made a total of 25 trodes and ended up scrapping my block...

---

Discuss.

 

[ 04-16-2003, 03:29 PM: Message edited by: Webmaster ]

[Tony]

quote:

---

Unfortunately the only dimensions I checked after shelling my solid were the 'core' dimensions.

---

GUILTY

 

I Personally leave the models intact.

Then Use negative stock with Bull nose cutters

And or lie about the cutter dia.

 

(I have a Carbon Tool library set up)

It Works for me...

 

I did copy the Post when plastav first

put it up So that I could check out

his process someday.

 

[ 04-16-2003, 04:07 PM: Message edited by: Tony ]

[Roger Peterson]

The offseting of surfaces works ok, surface trimming can be an issue, the shell works much better. Depending on the geometry you don't neccesarily want to go negative in the z just the x and y. Doing a scale XYZ can work well and splitting up your toolpaths into steep and shallow portions can also work. Unfortunately I haven't found any one thing that works every time for everything. If you can machine your whole trode with surface toolpaths you can use negative stock to leave.

 

[ 04-18-2003, 10:28 AM: Message edited by: Roger ]

[Anbu Lingappan - Moldplus SA]

Hi

 

Creating offset surfaces on complex models may not work that well. Could turn into a nightmare on some parts.

 

Best recommended method is the use negative stock offset while machining - since you can control XY separate from Z, very important.

 

With Moldplus Electrode Maker, we discussed this matter several times and after much feedback from customers and resellers, realized that it is more reliable to have users apply the spark gap / negative offset while machining.

 

With Solids, it may be a bit easier to apply the negative scale since you have control of separate XY and Z control. But then again, advisable to check first people who have done this.

 

Please refer to the previous discussions on electrode making on this forum. They go into this topic in great detail.

 

Here is 1 link to refer to:

 

http://www.emastercam.com/cgi-bin/ultimate...ic;f=1;t=003096

 

Regards

 

Anbu

[mold100]

If anyone is having any questionable difficulty with trode machining techniques try a search we have covered some pretty helpful info on this topic several times It is helpful reading especially to newer programmers or people who are looking for new methods.

 

[ 04-17-2003, 10:56 AM: Message edited by: mold100 ]

[Fig sign]

I want to reply .

This thread simply passed beside me and only today I found this thread.

 

Solid shell works !

I use it in an everyday manner and never scrapped a part .

But it has it`s own limitations (not always possible to make solid shell because parasolid kernel has it`s own limitations ).

For me the method was so simple, logic and obvious that I simply not entered deep into details.

Well I always make an solid model of an electrode and apply shell to it.

Naturally the common sence tells to implement it inward .

It doesn`t matter if it is a cavity or core the shell thickness must be the spark`s size.

The electrode working surfaces are the inner surfases of the shelled model and boolian remove needed only for separating them (you can get surfaces from solid and delete outer surfases or smthn else if you don`t like boolian remove ).

And that`s all.

But after making the geometry I ALWAYS CHECK IT !

All of us are humans and everyone can make mistake .

Now about core and cavity stuff!

quote:

---

While its true that shelling out a solid will decrease the 'core' part of your solid, it will also increase the cavity part of it.

---

That`s true and so must be !

But excuse me I understood that it is obvious

that you implement shelling ON THE ELECTRODE SOLID MODEL and not on parts of the mold.

I am the man who loves simple, safe things and never would advice smthn that I PERSONALLY HADN`T CHECKED HUNDREDS OF TIMES !

I invite everybody to throw away prejudice and try on some simple models this technique before getting to the conclusion that it is bad .

[kwolf]

I think the amount of gouge is maximum on a surface where the slope angle is 45 degrees. This amount is 0.41x(stock to leave).

The minimum gouge amount which is zero occurs in vertical and horizontal surfaces.

This is due to algorithm when using negative stock with an End-Mill cutter. The minimum tip radius which produces gouge-free toolpath equals to stock to leave value.

It's better if MC allows us to use it on surfaces, since there are times when we want to use it on Horiz/Vert surfaces.

[Fig sign]

Kwolf !

IMHO it is not correct !

The gouge amount is 0.41 only in normal direction and it is 0.586 in x and y direction in case of 45 degrees and in x or y directions can be far more in stepper and shallow angles.

What I think is that to implement it MC must change the algorithm of tool path generation.

It can be IMHO smthn like to build toolpath for tool with surface offset 0 and to use algorithm like shelling to get it offseted .

And there is a matter of gouge checking and undercutting and so on .

May be Mastercam can make a new toolpath like surface finish contour offset minus like it made now possible undercutting.

quote:

---

It's better if MC allows us to use it on surfaces, since there are times when we want to use it on Horiz/Vert surfaces.

---

This is really dangerous and no software developer would allow a dangerous option with a chance of human fault .

[kwolf]

quote:

---

What I think is that to implement it MC must change the algorithm of tool path generation.

It can be IMHO smthn like to build toolpath for tool with surface offset 0 and to use algorithm like shelling to get it offseted .

---

Let us S="stock to leave". IMHO for positive stock it offsets the tool to outside by S and calculates the toolpath then move up the toolpath by S to compensate for real smaller cutter.

It offsets the tool to inside by S and calculates the toolpath then move down by S to compensate for reeal bigger cutter.

This way there's no need to calculate offset of surfaces but has this disadvantage that it fails for negative stock with End_mill cutter.

[Fig sign]

quote:

---

This way there's no need to calculate offset of surfaces but has this disadvantage that it fails for negative stock with End_mill cutter.

---

Correct!

Because of this the tool path must be offseted .

[kwolf]

quote:

---

Correct!

Because of this the tool path must be offseted

---

Remember the tool is offset and the toolpath is created on original surfaces then the toolpath is moved in z direction by a certain amount which equals to stock to leave.

There's only and only one configuration that this algorithm won't work and this config is when negative stock is greater then tip radius. Because offsetting a tip radius inside with a value greater than tip radius produces a self intersecting geometry.

Be sure none of the popular CAMs offsets the surface nor the toolpath.

I asked a colleague using PowerMILL( a very powerful CAM) about this and he said PowerMILL can't machine trode with negaive stock using End-Mill either.

HTH

[foolsh]

kwolf,but ug can,it just says the toolpath produced is not accurate!