What's new
What's new

New Machine Day- Makino - EDAF2

Hi Rob:
With your background and your eagerness to learn, you are going to do just fine in this new domain.
It's actually not all that different from any other subtractive metalworking process: you are using a slightly different tool with slightly different constraints and advantages, but ultimately you still are getting rid of the material you don't want so your part will emerge from the raw block you started out with.

So when you invoke this process for your projects, think in terms of those advantages and constraints, and then plan your process just like you would if you were milling it or turning it or grinding it.

Bud Guitrau's book is an excellent resource to help you understand what those advantages and constraints are...many of them are straightforward and obvious (especially once they've been pointed out to you), but some are subtle and will be experience driven.
So you have to accept fuckups as part of your education...embrace them as the price of learning.
Once you've been at this for a while, it will be second nature to you...it's truly not that hard.
A good bit of it is learning what the limits are...and with a top of the line machine like yours, those limits are pretty incredible, so you can do stuff easily that would be impossibly hard if you didn't have the toy...I use mine routinely (but not often enough to justify a really good one) to make stuff that other shops just can't do..

Welcome to the club!

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 
Last edited:
1st test burn today!! Marcus, I see why you stressed good work holding, I hadn't thought to buy a quality vice for it yet my Kurt's are a bit too big, so I used my crapO, Amazon/China special "shop fox" vice ($58 and I got screwed...) which in my defense I bought only to hold parts in place for laser marking. The suction clearing cycle + poor vice allowed the test block to move just a bit so we couldn't reburn it after the 1st burn-target .250" depth, we yielded .248" and got the targeted finish.

I love the Hyper-I control, its very, very intuitive. Many places offer both one click to the exact page in the manual & a video clip for help. It's really awesome to be able to see the graphics and answer simple questions to make a program that works in a few minutes with no experience.


We'll be training tomorrow and might receive our 1 uninstalled option- 10 port, M-code/solenoid actuated suction block which should let us run some longer back burns with multiple electrodes unattended.

I can't wait to get my setup right & carbide & CuW. I think we're going to burn some simple stuff instead of cutting it just for practice and fun.
 

Attachments

  • burning2.jpg
    burning2.jpg
    145.1 KB · Views: 17
  • burning1.jpg
    burning1.jpg
    173.4 KB · Views: 16
  • test-block.jpg
    test-block.jpg
    125.7 KB · Views: 19
  • test-block2.jpg
    test-block2.jpg
    98.3 KB · Views: 18

Happy 4th of July weekend Americans!!!
We burned our first part, 2 counter bored 3/8 holes into a sine vise build from A2 - 58Rc.

Now I really understand what you guys were warning me about. I didn't realize how much recast doesn't get cleaned into the filters, nor did I get how fast graphite electrodes wear out. I made one poco3 counter bore for 2 holes, but needed another one for the finish pass. We have a new policy- if we make an electrode for anything- make 5 of them.

Marcus, thank you for stopping me wasting more time on that dumbass fixture. Version 2 coming soon!!
 
Hi again Rob:
So you now have wet feet in the magical world of sinker EDM.
Congratulations!

When you plan your jobs, a few things are worthwhile to note:
When the electrodes involve more than the simplest of operations, a strategy most people use for sinker EDM is to try to make as few of those expensive and time consuming electrodes as possible.
Two ways to achieve that:
1) Stage the trodes so you rough all your burns with one trode, then run through all the burns with a semifinishing trode, then run a separate trode for each finishing burn if they are super fussy.
On the next set of burns, use the same rougher, one of the original finishers as your semifinisher, and then new finishers for each separate burn.
Now toss your rougher and use your original semifinisher as your new rougher etc etc.
That way each trode makes a few burns before it's tossed or re-cut.
You judge when to toss a trode by the corner wear and you do so by determining when the semifinish trode can't clean up all but the last teeny bit of the corners anymore without showing corner wear too.
You can often get quite a few burns out of a trode by using it for roughing and semifinishing when it can't function as a finisher anymore.
So the corollary to that is that you make all the trodes identical and orbit them varying amounts to rough, and then to semifinish and finish.

2) Try to burn as much as possible on "No Wear" settings.
Just about every modern machine has those, and it is a specific combination of spark generator settings that preserves the trode.
It is slower than the most efficient burn settings, so you have to judge whether it's cheaper to make more trodes or cheaper to let the machine run longer.
Your job criteria will determine which to choose.

Another thing to note:
Often, when making trodes you get a better outcome breaking the trodes up so each is simpler to make, rather than getting greedy about trying to cram every damned feature onto a single trode.
Here's a mold cavity that shows this:
DSCN5523.JPG

You can see it's pretty feature rich, so it was done with three trodes.
Here they are all together:
DSCN5524.JPG

Here's the first:
dscn5525.jpg

Here's the second:
DSCN5526.JPG

And here's the third:
DSCN5527.JPG

So you can see that by breaking up the trodes, each is much more manageable to make.
To give you an idea of the complexity , the first and third trodes required both wire EDM and CNC milling to make them.
It was a lot of screwing around, (around 6 hours for the worst one) but it did the job.
Three trodes total were used for each burn...you can see what a nightmare it would have been to put all the features onto a single trode and then try to make four of them identically. (Separate finisher for each cavity, so four trodes of each)
Two identical cavities (Hotside and Coldside) were burned using those trodes and using the strategy I described so all trodes got more than one use except the first rougher and the last finisher.
The roughers were shot after one burn, so the semifinishers from the first burn were used as roughers for the second burn and the finishers from the first burn became semifinishers for the second burn.

So the sequence is:
Mount and align the first trode
Rough one burn
Change to the second trode
Semi the same burn.
Rough the second burn with the semi trode from the first burn
Change to the third trode.
Go back to the first burn and finish it.
Go to the second burn and semi it with the finisher from the first burn.
Change to the fourth trode
Finish the second burn.

This minimizes the number of trodes you use and the number of indexes of both trode and burn position you have to do.
After a while it becomes second nature to plan burns this way.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 
Last edited:
Hi Bob,

Regarding what Marcus already posted in message #12 -- the importance of climate control can not be understated if/when trying to cut (and/or measure) to "tenth" or sub-tenth accuracies.

I was in the wire edm business for a number of years, and I picked up a LOT of work from other wire edm shops because of my setup -- Wire machine and ALL measuring equipment) was in a ~400 sq.ft., carpeted, insulated, climate-controlled room that was held to 68F plus/minus less than 1 degree. All the heating/cooling airflow was controlled in a manner that no air would blow directly onto the wire machine or any of the measuring equipment. All the lighting was "soft" and indirect so no bright lights were pointed at the machine or the measuring tools. Cotton gloves when holding/measuring parts, and also never holding a very low mass part by hand. Over and above the 120 gallons of water in the wire machine, I had six 40-gallon plastic barrels full of water in the room as "heat sinks". (also, if someone dropped by to measure a part, I could drop it in one of the barrels of water for a minute and then get to measuring it).

This may sound crazy or extreme, but since plus/minus tolerance are virtually always given in absolutes (rather than in percentage of the size of the part), this fanatic control of temperature and handling can really pay off on larger, low-mass / high CTE parts.

fwiw...
PM
 
Thank you Marcus & precisionmetal. This is definitely a learning process... I really didn't understand how quickly the graphite trodes wear. That's some really beautiful work Marcus!! Those trodes would be a nightmare to machine in a single piece, funny how breaking it up into multiple ops actually saves time vs the one & done mill/turn mentality.

The "copper" trodes in your picture, Are those CuW? It seems like the material is available in 70%-30%, 60%-40% or 75/25. What do you prefer to use?
I'm waiting on some tooling items and will make a few others. How do you feel about ER32 collets to hold round or trodes cut from round bar?

Makino is back to install our Mcode flushing option, its a bigger job then it sounds. Once that's complete we'll get our training.

When your programming orbits using the same sized trodes - how does one decide how "close" to tolerance to orbit the rough vs semi vs finish trode? I've only used a rough/finish so far... I'm hoping the art of EDM becomes a bit clearer with the training.

precisionmetal- We're a temperature disaster right now. We have a ton of work to get us to "temperature controlled". I'm aiming at +/- 5deg F. Currently our Winter to Summer max 60-85 F range. We need AC badly, summers are the worst for consistency right now (70-85F) 15 deg swings during the day. Winter is OK we can hold 60-65. The QC lab is always good but we only measure there. Your setup sounds amazing, hopefully we'll get close to what you built.

Thanks
Rob
 
Hi again Rob:
The electrode material I use pretty much exclusively is a copper alloy called tellurium copper or Telco for short.
It is a free machining grade...the tellurium in it provides that.

I use it so I don't have to deal with the incredible mess graphite makes if you don't have proper dust control.
It's also MUCH nicer to wire EDM cut than graphite is.

I avoid copper tungsten except for very small burns where I need excellent wear resistance in my trode.
It's expensive as Hell, and it's abrasive enough that it's more of a challenge to machine.
When you have a big complex trode with lots of 3D surfacing on it, just changing out all the cutters you need is nuts...here's an example:
DSCN5529.JPG

A block of copper tungsten this size would be at least 500 bucks, and you can imagine how many 3/64" and 1/32" diameter extended length bullnose and ball cutters it would take to mill the overall shape and all the little radii when the cutters can go a square inch or so before they won't cut anymore.
So yeah, copper tungsten is nice stuff, but not for general burning.
If you can handle it, graphite is good for general burning.
Copper is not as good (wears faster) but at least I don't look like a coal miner at the end of the day, and my shop doesn't look like a skank pit.

With regard to collets...whatever you need to hang onto your stock.
Bigger collet holders take up more vertical room which is always at a premium on a sinker, so the smallest you can get away with is good.

With regard to planning your trode size and your orbits...0.005" per side is common for sizing the trode as I alluded to in a previous post.
Regarding how much to allow for overburn, you have the benefit of a modern control (you LUCKY man!)
That means, the control will find the best way to get the cavity to the size you want at the finish you want.
It will burn in such a way as to get you there as efficiently as possible, calculating the roughing orbit, the semifinishing orbit and the finishing orbit that it needs so all the roughness from the first burns are removed in the last burns.
I have to plan all that manually.

The rule of thumb with sinker EDM is that you rough to size as close as you dare to go...just enough undersize to clean up in the finishing burns.
Sinker EDM is paradoxical in that the trode wears much more aggressively during finish burning, and finish burning is very slow.
So your control will set the orbit to rough as much as possible and vary the orbit so by the time you're to depth, you're also to size and your finish is correct.
The only thing you put a new trode into the burn for, is to pick out the corners where the rougher wore away, and this can take several trodes on a fussy burn with sharp corners.

While I'm on the subject of corners...when you orbit a trode, (even a sharp cornered one) you will make radii in all the corners that are at least as big as the orbit, and usually bigger because of that pesky trode wear.
With some shapes you can vector your last trode into the corners to sharpen them...your Makino guys will show you how to do that and when you can.
The details of how to plan for that are a whole 'nother subject, but be sure to pay attention when you get your training.
It's not complicated but it is occasionally counterintuitive.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 
Lol, Marcus- I'm barely getting my fancy machine to make holes to size and your doing amazing parts manually. I'd love to see what you could do with this thing. Thank you for Bud's book- really helpful!!!!
Makino installed the 10 mcode flushing option, we're short a few pieces of sheet metal but they'll get here eventually, function over form for the time being. Working on our 10 port flushing pot and the funky electrodes on our .0002" tolerance parts.

We made our 1st test part- its a blind hole slightly oversized with an .008" radius, tool steel 65Rc.

A little progress everyday

Here's the control running...
 

Attachments

  • part1.jpg
    part1.jpg
    129.7 KB · Views: 14
  • machine2.jpg
    machine2.jpg
    106.7 KB · Views: 14
  • machine1.jpg
    machine1.jpg
    156.1 KB · Views: 14
Hi again Rob:
A thing to remember; the magnificence of the finished part is really all about the magnificence of the electrode(s).
Make a fancy trode...burn a fancy part.
With few exceptions, this is how it works and the examples I showed are like that too.
The machine just burns...it doesn't care what the trode looks like and with a modern machine it doesn't care in what direction you want to burn .
Here's an example of that...the star shaped trode is vectored around the tube that's standing vertically in the tank and the burns are all sideways burns.
DSCN1629.JPG

DSCN1628.JPG

With your background in 5 axis milling you are exceptionally well positioned to make complex trodes with considerably less pain than many or most.

The place where you will need experience is learning to take advantage of things this tech can do that is hard or impossible with other tech.
For example, you need pay far less attention to the rigidity of the setup...you can hang a trode out a ridiculous amount and you can do the same with the workpiece and still make an accurate burn.
I've countersunk screw holes onto the inside faces of a pair of opposed flanges 6" apart that some idiot designer specified so it was impossible to build any other way...totally unglamorous to look at until you understand how un-manufacturable it is without a sinker.
The electrode was hanging off the end of an "L" shaped arm so it could get between the flanges...it looked super flimsy but it did the job just fine.
Similarly I've burned 0.015" through holes into a diaphragm right next to a vertical wall two inches tall...no other way to make it.

Also, you can capitalize on the C axis with a bit of creativity...indexing the trode or rolling the trode or bolting the trode in the tank and putting the job in the trode holder and rotating it.
You can engrave onto a flat trode and then roll the workpiece past it, or you can roll a cylindrical trode around a cylindrical stub and burn a pattern onto it in places you could never engrave with a cutter.
You can put a ball shaped trode on the end of a stoutish wire and burn holes around corners, you can connect the bottoms of vertical blind bores with a cross passage...these are all jobs that have come across my desk in decades of running the sinker.

So think a bit outside the box compared to how you would tackle a job with chip making technology...your creativity will define the limits of what you can do.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 
Last edited:
Making good parts!! I really like this control, its easy to use, well organized by project then programs. The machine is able to very tightly control both size and finish to .0001". I thought this was cool graphics showing a finish pass taking an additional .0002".


Marcus- I did make a substantial mistake, one of those you don't know what you don't know types. It appears that all the discharge dressing or in machine dressing techniques including the fancy wire option require rpm. My spindle is 10rpm max which doesn't allow me to discharge dress round shapes in the machine. Makino shows 300rpm minimum and up to 1000rpm preferred in the techniques book. It's very costly to replace the brand new spindle with a 1000rpm unit. So I'm looking for a solution which doesn't involve replacing a brand new spindle or buying a new wire machine...:scratchchin:
Does anyone make a low runout device which will spin a trode mounted in a 3r macro pallet? Doesn't need to be controlled by the machine, just spin accurately and let me remove a few thousands from the trode diameter.
Alternatively- I made a pallet holder for the Integrex, hoping some very gentle turning/cutting will yield more concentric trodes.

I'll test cut a Telco trode on the pallet tomorrow. EDM is awesome!!!

Thanks!!
 

Attachments

  • punch1s.jpg
    punch1s.jpg
    190.7 KB · Views: 7
  • punch2s.jpg
    punch2s.jpg
    168 KB · Views: 7
Hi Rob:
There is (or perhaps was) an attachment available to spin a trode called a "Rotobore" made by Everest Industries in California.
It has a head that can be adjusted within a tenth for both radial and axial runout.
It is a purely manual unit.
I have one for my Hansvedt sinker.

The biggest downside is that it takes up a LOT of height and it needs to be mounted to the ram so its axis points straight down.
Every time you dismount it, you disturb that alignment, which means that you will make a conical hole even with a cylindrical electrode if you don't get it perfect when you re-mount it.
That makes it a bodge, and when you're working to tenths it's an unacceptable bodge.

I encourage you to go back to Makino and let them work out a solution...I'm unconvinced you cannot discharge dress a trode without a high speed C axis.
It may not be as efficient as running it at the proper speed, but burning is burning, and you may well be able to make it work together with Makino's help.

On a last note; did they not KNOW what you were trying to accomplish when they sold you the machine?
Surely, if they did know and advised you poorly, they bear at least some responsibility for the pickle you now find yourself in.

When you approached them, how did you present your manufacturing problem to them?

If you determine they do share some blame, I would expect them to be very motivated to put your concerns to bed and help make it right.

In any event, I encourage you to reach out to them...if it is done respectfully and without throwing a tantrum, you may well be able to negotiate a solution that doesn't kill you financially.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 
Thanks Marcus! If you haven't had a laugh today check out my concentric trode making setup... I can't believe it worked, that mess left me .0002" out which didn't seem to matter burning CuW on WC- maybe high electrode wear worked in my favor. I'll shorten that stack maybe cut the error in half.

The spindle thing is 100% my fault, I just didn't know enough to ask about it and didn't consider the challenge electrode concentricity can present. I ASSumed that 10rpm would be enough to dress trodes, maybe it can work, I'm going to try it. All the "difficult" parts we need to produce were submitted w the PO, Makino trained me to get them all done. They did their job. I'll try some groveling, perhaps the Makino machining Gods will shine on me.

I burned some WC today using a flush pot.
Burned a .2" diameter hole through .710" of tungsten carbide in 3h 17min, using a programmed 10 VDI. It looks like I have a bunch of recast in the hole. The oil was flowing up so it makes sense the top end of the hole would collect some material. Do you think I should turn up the volume of flushing fluid? I was surprised how "dirty" the carbide machined.

I know the machine is good to bring the hole in as close as I want to get, but how much would you leave on it for polishing/recast/haz removal?

Thanks
Rob
 

Attachments

  • sketchy.jpg
    sketchy.jpg
    291.5 KB · Views: 4
  • carbide_hole.jpg
    carbide_hole.jpg
    437.4 KB · Views: 4
  • Rough.jpg
    Rough.jpg
    180.5 KB · Views: 4
Hi again Rob:
I would frame the pitch to Makino as follows:
Find out what it would have cost to factory install it and be prepared to pony up for the factory installed price, since it's unreasonable to expect to get it for nothing.
Ask them if they'd cover the cost of having the tech come in and get it mounted and running.

After all, they were aware you needed to be able to make your shapes to those tight tolerances.
Discharge dressing is an expected part of that (or at least a need that could be reasonably anticipated), so they really should have questioned your C axis choice aggressively.

It's not like you are a cheapass or stupid customer, (at least it certainly doesn't appear that way to me), so a reasonable pitch on their part for this most useful option, was unlikely to be met with brainless, stubborn resistance on your part.

So it can be argued they missed the boat here.

A compromise like I described above is very reasonable...much of their cost would not really be a "hard" cost.
They have to pay the tech his salary whether he's at your place or somewhere else, so their biggest hard expense is likely going to be the cost of getting him into your plant and supporting him when he is there.


You came to them as a novice seeking their expertise and recommendations...they did not give you ideal ones, so to my mind they do have some skin in the game.

So it is not "100% your fault"...far from it.
You didn't know, you never represented that you did know, but THEY did know or should have done.

I encourage you to knock on their door.
Frame it as a problem to be overcome and invite them to find a solution...that's their job.
Whether it's finding the best settings to enable you to fully use what you currently have, or biting the bitter pill and offering you a path forward to upgrade, is something you should pursue and they should work with you on.

The worst they can do is sniffily refuse... but I've always had the impression they were a respectable company offering a superb product with good service, and I'm sure they're very motivated to continue to be seen as reasonable and fair.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 








 
Back
Top