Electrode manufacture

[Andy_Wallis]

During programming of electrode using minus stock in the parameters page. Using sharp cutters or cutters that have a smaller radius than the spark gap being used. Mastercam takes and age to regenerate the toolpath (Sometimes still working out when I come in the morning from the night before!). i.e. using a 0.3 spark gap for roughing electrode and trying to use a sharp cutter or a cutter with a small radius to corner out.

I think Mastercam struggles does anyone have any advice on speeding this up? Or making the process more user friendly?

I have tried scaling the electrode and cutting to zero stock to leave but surfaces just overlap.

Mastercam UK are not much help when I have asked the question.

Delcam & Vero seem to be OK when working out which is why we are leaning towards leaving Mastercam.

Thanks in advance for any advice.

[Teh Bear without Brains kinda idiot]

I  have seen something like this 

Mostly with surface contour toolpath

When tool dia is close to the minimum diameter in contour

In other words tool has not much room for movement

I  ve seen it 20 years ago but sometimes see it even now

 If it is so I know  how to deal with it

 

BR

 

[Andy_Wallis]

Thanks for the reply BR,

How do you fix it?

I am in the same boat as you - this has been as issue as long as I used Mastercam - 15years +.

Andy

[pullo]

I modify the tool diameter and radius of the tool.  if your spark gap is 0.3mm then D6 ball becomes D5.7 R2.85 and D6 sharp is D5.7 R0 .   The neat thing about this is your

Z depths correspond to what you have on the screen. I have another electrode holder which I can use to tilt the axis of the electrode 90. I can then rotate the electrode 90 degs allowing me to machine the electrodes in five planes  planes altogether.  When I machine the electrodes  in the tileted position I add a shift away from the electode on my previous Z-axis which is equal to half my spark gap (0.15 in this case.

 

Gracjan

 

 

 

[pullo]

I have been now doing electrodes like this for the past 20 yrs.  

1. the Z depths are the same as in the 3d -file.

2. toolpath calculation is the same as otherwise (fast)

3. I have made a main program which handles all the shifts when machining "sideways" (on Heidenhain). I just jot down the Z depth of a tool  which always drops down  to the same absolute height  ( the reference point of my electrode holder ) and use  that as an input for all the shifts. So every electrode height can be different and the sideways machining happens on the fly. 

4 . I have to admit here that theoretically "cheating" diameters when the corner radius is 0 is not correct, but since the most I do is 0.3mm  it has no bearing on the outcome. My tool range is 50 mm to 0.2mm (that would be 0.001 for a spark of 0.2 )

Gracjan

[Andy_Wallis]

Thanks for your response Gracjan.

I understand what you saying about the cutter diameter & radius - makes perfect sense. But what puzzling me is the spark gap on Z plane if I have used a 6mm R1 to finish a part then that becomes 5.7mm 0.7R, the vertical faces would be correct but the flatter faces would be incorrect unless the offset of the tool was adjusted to suit the spark gap. Unless I'm miss understanding what your telling me???

On a simple shape like the one provided in your photos it makes perfect sense. 

The attachment is what I program & cut regularly. 

Any help is appreciated.

[George Vassilev]

If you changing the rad of the tool , might be better to programme with Tip compensation to  center not the tip of the tool

[pullo]

Ok  let's go thru this methodically. 

1. I have about 800 molds behind me so the method is proven  (proof 1)

2. beside the point , but I must say the electrode detail in yours pics is very  nice.

3 in pic 0 you can see the part I am machining. So there is a taper and a round shape and some horizontal surface . its a parallel old school toolpath. (pic  0 )

4. pic 1 is the toolpaths looking from front.

5. pic2 is the negative stock toolpath shifted in  Z (moved) by the radial undercut - identical toolpath so identical electrodes !   There is just a vertical shift  , which is very convinient in my  "cheating method " , as what you measure on screen (z) must be identical to what you measure  on the machine. 

Gracjan

 

 

[pullo]

BTW  Andy,  D6 R1 used in cutting an elctrode  with an undercut of D0.3 is D5.7 (D6-  D0.3 = D5.7 ), R0.85 (R1-R0.15 = R0.85)  mm NOT  D5.7 R0.7 .  

 

I did the same trick of back plotting toolpaths and shifting the negative stock to with the same shape as in Pic 1 but now the tool was D6 R0  and D5.7 R0.  Theoretically , the largest error is when the surface normal is 45 degrees  with  an undercutting of 0.05 mm and going down to zero as you approach  horizontal or vertical.  Horizontal and vertical cuts are identical for both toolpaths.

Now  , as you know if you cut anything with a sharp tool , there might be as much as 0.2-.3 mm material left where the normal vector is 45 degrees.  That's because the tool does not have enough time to cut or the sharp edge has fallen off.  

So to sum it all up , You must machine with rounded tools  for dimensional precision and follow  up with a sharp tool to clear up the internal sharp corners  where the rounded tool 

left some material.  

 

Gracjan

[Andy_Wallis]

Thats awesome Gracjan!

Thanks very much for that I will give it a go!

I cannot thank you enough.

I will let you know how I get on

Andy

[pullo]

I've given it a go  already . Just apply the experience

Gracjan

[Compaq007]

On 7/13/2017 at 9:11 AM, pullo said:

So to sum it all up , You must machine with rounded tools  for dimensional precision and follow  up with a sharp tool to clear up the internal sharp corners  where the rounded tool 

Good piont, we use ball end mills only to finish 3D feature and square endmills for vetical walls..

[Compaq007]

Pullo's method works great for finishing electrodes. For roughing electrodes just add your spark gap difference to the stock to leave, no need to shift the Z. We make electrodes 24-7 and it works great.

[Compaq007]

If you have additional processes such as plating or polishing you can actually apply the same concept two leave additional polish stock or plating stock.

[pullo]

On 8/12/2017 at 2:26 AM, Cartman007 said:

Pullo's method works great for finishing electrodes. For roughing electrodes just add your spark gap difference to the stock to leave, no need to shift the Z. We make electrodes 24-7 and it works great.

I did not understand that .  Adding positive Stock To Leave would leave you with a bigger electrode, for roughing you need an even smaller electrode. In my method, I don't go to the  "dark side" i.e. I never use negative stock to leave as the toolpath  calculations basically stop to work (for all practical reasons ) . So for roughing I would just use a tool with an even bigger difference between the programmed and real life tool diameter.

So I reiterate , there is no use of stock to leave whatsoever , just playing with the program tool diameter and radius.

Two ways to go about taking out programming errors  concerned with tool diameter.

1. make several tool db files named Undercut 0_2 , Undercut 0_4 , etc. and preset all the tool diameters once and then  forget the whole thing , just use the right tool db file for the

right job (rough or finish)

2 As my tool magazine is always the same, I know what are the real tool diameters. Now if I program the electrodes  using smaller tool diameters , I can always recalculate the

diameter difference in my post processor and post it out .  If for example the tool diameter difference is not the same for every op , that's an error that can be automatically flagged.  So you are free to set the tool diameters on the fly , but you know that Big Brother is Watching You thru your post.

I have several built in checks into my posts , which flag (actually stop the postprocessing ) several my errors.

1.  The stock to leave for all finish toolpaths must be zero.

2. the stock to leave for all roughing toolpaths must be greater than zero.

3. for a three axis machine the WCS must be identical to the Toolplane 

4. for a 5-axis machine the WCS of all the ops in a program must be identical.

5. for a 5-axis machine the origin of the WCS and the toolplane must be identical.

6.  Toolpath tranformation method for a 3-axis machine must be set to Coordinate.

and about four more which I see as illogical in toolpath programming but won't mention them here as they do not apply to electrode machining.

 

Gracjan

[Compaq007]

On 8/13/2017 at 10:57 PM, pullo said:

I did not understand that .  Adding positive Stock To Leave would leave you with a bigger electrode, for roughing you need an even smaller electrode. In my method, I don't go to the  "dark side" i.e. I never use negative stock to leave as the toolpath  calculations basically stop to work (for all practical reasons ) . So for roughing I would just use a tool with an even bigger difference between the programmed and real life tool diameter.

Gracjan,

I should have clarify that the stock to leave amount is always negative "towards the dark side" to obtain an smaller electrode, as you mention, and it is done only for the roughing electrodes and for super-roughing for which the spark gap is  -0.010" to -0.015".

As per the calculations: I have never noticed a issue, operations regenerate the same. I have to say that our spark gap never exceeds -0.015 and the smallest tool we use is 0.25mm (0.00984") . The only issue we could run into is the tool disappearing but it has yet to happen. We make an average 9k to 10K electrodes per year.

We do have tool libraries with smaller diameters (spark gap included in the tool diameter) which we use only when the spark-gap in NOT included in the model, essentially we are lying to MasterCam telling it that the tool diameter is smaller to make up for the lack of spark gap.  

BTW....I like those safety flags-checks in the post-processor.. number 2 will always keep you from scrapping graphite.

any way, keep making dust, cheers! 

 

[pullo]

I probably make  only about a few hundred electrodes a year , so I would gladly hear a few more pointers from you  Cartman007

 

Gracjan

[Watcher]

You should look at Moldplus add-on to automate electrode creation and geometry manipulation. It would increase your productivity significantly. 

JM2C 

[pullo]

Avid user since it's inception  in the nineties, it's been there for a long time....

Gracjan

[Compaq007]

Our electro designer is unigraphix power user .

[Compaq007]

GracJan,  I have a question for you sr... or any body really...

What process do you use to clean your electrodes before sending them off to final inspection or EDM?

Sometimes our graphite electrodes are soaked in EDM fluid and we use an ultrasonic tank filled with acetone to wash-off a any residue before sending them off to CMM .  This process really gets the trodes really clean but due to safety concerns we are looking to replace the acetone for another non flammable liquid .

 

I was wondering if you use anything different that can help us.

 

Any input is much appreciated.  

[Newbeeee™]

There's a few water based cleaners that work pretty well (talking general machined component degreasing here - not trodes) but in my experience they work a lot better with heat.

Whether heat will affect the trode???

[Compaq007]

thank you newbeeee for the suggestion, I am all ears.

 we have a solution that we use for degreasing but I am not sure that it is water base.

I don't think heat would be  a problem, but any thin film or residue left behind would, the CMM would pick up on that.