Titivating a Wire EDM Machine

[awemawson]

Sorry about this chaps but I'm starting another re-build thread that might go on for the odd page or two 

I bought this machine on a whim back in October 2011. It was basically working but was in dire need of some TLC. No preventative maintenance had been done in years, various panels and covers were missing, and although working, the Fanuc 6 CNC control was hard to use due to most of it's indicator bulbs being blown, and several key buttons either missing or broken.

[awemawson]

So what is a "Wire EDM Machine" 

Well the EDM stands for Electro Discharge Machining - otherwise known as spark erosion. In this machines case, the work piece is attacked with a thin brass wire (0.25 mm diameter) and the electronics creates a controlled discharge between them. A gap is maintained across which the discharge occurs explosively dislodging minute bits of the work piece and also the wire. For this reason the wire is 'use once only' and runs off a feed spool, via several cunning mechanisms, and ends up on a take up spool ready to go to the scrap man. The gap between the work piece and the wire is 'flushed' with de-ionised water to clear the removed material, maintain electrical isolation and very importantly, cool the wire.

Mean while the CNC controller moves the work piece by servo action in X & Y taking all sorts of parameters into consideration to keep accuracy of the part.

A second set of servos can move the wire in 'U & V' tilting the wire up to 20 degrees to make three dimensional cuts.

So in concept you can imagine it as a band saw that uses an extremely thin blade, and can cut tool steel up to 200 mm thick

[awemawson]

So the wire starts life nice and fresh and new on the feed spool, which is gently biased in reverse to stop it 'running on'. It then passes through a set of tensioning rollers and brakes, and a 'wire break sensor' eventually being presented to the lower wire guide, which is a diamond with a hole in it. As the wire is electrically 'live' all the parts have to be insulated from the body of the machine.

[awemawson]

Near the lower guide there is (or rather SHOULD be - more on this later) a sliding contact that energises the wire. This is a fat disk of Tungsten Carbide over which the wire passes. This contact is duplicated near the upper wire guide so that current is fed into the wire both from below and above the work piece.

Both wire guides have co-axial water feeds, the flow of which is manually controlled to form a thin 'string' of water surrounding the wire

[awemawson]

So having passed through the work piece and also through the upper guide, the wire goes through another set of tensioning rollers and on to the take up spool

[awemawson]

The 'table' on which the work is positioned is an open centered rectangle giving the machine a working envelope of  450 x 650 x 200 mm, (and up to 300 kGs  )

This table is moved by servo control from the Fanuc 6 control and attached to the side of it is a 'dry run plotter' that draws a representation of the part on paper without actually cutting metal

[chipenter]

You don't do things by halfs Andrew do you ?

[awemawson]

"You don't do things by halfs Andrew do you ?"

Well Jeff - if you're going to do something you may as well do it properly 


The machine works - make good parts but is pretty slow - really a 'lights out and leave it' sort of operation.

So far I have got the Fanuc 6 and its operation panel sorted out, managing to beg borrow and steal compatible parts. And I have upgraded its memory so it can store larger programs.

Memory on the Fanuc 6 is 'Bubble Memory', a technology that came and went over a very short space of a few years:

http://en.wikipedia.org/wiki/Bubble_memory

I have also replaced the tiny bearings in all the wire guide pulleys - a fun exercise as the pulleys in the main are very delicate ceramic with press fit stainess steel bearings.

I've also re-made most but not all the missing tin work panels

[awemawson]

So what's triggered this flurry of postings ?

Well glancing through eBay as I'm prone to do, I came across some of the lower contact electrodes and thought it was about time I sorted mine out.

The machine was originally an 'Automatic Wire Feed' variant - ie you positioned the table so that a hole lined up with the wire guides, and it would thread itself. Very handy indeed as threading 0.25 mm brass wire through equally small wire guides standing on your head to get at them is a bit of a pain. Sadly this feature proved to be unreliable and at some time was removed from it.

In the process I suspect that the lower assembly was replaced with one from a similar but different machine and there is no provision for the lower tungsten carbide contact to be mounted. The relevant wire is fixed to the insulated centre of the lower pulley and does nothing at all.

The net result of this is that the machine can only operate at half the current it is designed for without melting the wire.

So my intention is to re-design the layout of that area so that the contact can be in the wire path. Problem is, this machine (like me!) is rather long in the tooth (mid 1980's) and the 'stainless' support arm is badly corroded.

I've avoided delving in before due to the fact things are very likely to break being undone. But now I'm setting out hopefully to fix it and I'll be glad to have you all along with me as I do it. 

[hermetic]

I.m looking forward to the Journey Andrew. I gather from your description that this machine operates with the wire through the workpiece cutting a bit like a bandsaw?
Phil

[NormanV]

This is going to be interesting.
There was an article in 'Model Engineer' No 4432 15th June 2012 that told the story of using an EDM machine to cut a cylinder block for a 4"scale traction engine model.

[dsquire]

Hi Andrew

I thought that it was kind of quiet on your corner of the island lately. I guess that is because you were working behind closed doors bringing yet another machine back to life. You will have a good following on this I am sure. 

Cheers 

Don

[awemawson]

I.m looking forward to the Journey Andrew. I gather from your description that this machine operates with the wire through the workpiece cutting a bit like a bandsaw?
Phil

Phil, there is an excellent free downloadable book describing the process here:

http://reliableedm.com/Complete-EDM-Handbook.php

[awemawson]

So where to start 

Realistically I need to remove the entire lower arm assembly to be able not only to assess what is there, but also to be able to do whatever machining is needed to mount the contact and re-locate the guide pulley.

This is what we are dealing with:

[awemawson]

Now the arm is mounted onto the body of the machine using four largish cap screws - unless I can undo those then probably the whole exercise is a waste of time.

So I tentatively loosened then re-tightened each one - surprisingly they shifted quite easily. At this stage I presumed that there would be alignment dowels to accurately locate the arm.

[awemawson]

OK that hurdle jumped I had the confidence to start on the other end of the arm and strip off it's water feed and electrical connections

[awemawson]

So now I could withdraw the rather stiff and brittle water feed pipe and co-axial screened power feed wires which are threaded through the arm assembly.

Last remaining bit to take off was the 'wire break detector' which is simply a microswitch with a long wire actuator that has a ceramic tube threaded on it. This was one of the bits that I'd fixed when I got the machine. A previous owner had just cut it out of circuit when it failed 

[awemawson]

OK so now the arm bolts can be removed to free the arm. Easy enough and I was surprised to find that in fact there are no locating dowels for the flange - just a 'shelf' that it rests on.

[awemawson]

So what was revealed ?

Well as I expected rather a lot of corrosion - in fact it looks a horrid mess!

Have some pictures:

[awemawson]

Now none of the corrosion gives any structural issues, but no doubt undoing some of those cap head screws holding the blue insulator will be 'fun'

I decided, true to form, to dunk it in a bucket of warm citric acid solution for a few hours and see what happens. Never used it before with stainless steel.

[lordedmond]

Andrew

Now you know why they say water is the universal solvent ,deionised and dis tiled are the most active both will try to leach back the minerals removed from them. That why you should never drink distileled water it will leach back from your gut

Good luck with this major project

We have a edm machine at the club that uses shaped electrodes and paraffin as the electrolyte/flushing liquid.


Stuart

[awemawson]

Thanks for the good wishes.

Perhaps yours is a bit like this one ?

http://madmodder.net/index.php/topic,9319.0.html

[awemawson]

The Tungsten Carbide contact electrodes have just been delivered. I didn't go with the eBay ones, as although they were somewhat cheaper they weren't quite the same size as the originals. They would have worked fine I'm sure. However I reasoned that if I could source ones to the original spec at least in the future replacements would be available.

Worth noting that these contacts do not revolve - the wire slides over them and eventually wears a groove, whereupon you slacken their mount, turn a fresh face to the wire, and tighten up.

I've just taken a tentative look at the bucket of citric acid - quite a bit of progress but it needs more time, after all it's only had barely an hour and a half so far. At least the heads of the countersink cap screws that retain the blue insulator can now be distinguished from the rest of the mess 

[lordedmond]

Yep Andrew very simler not used it myself it came out of Ratclffe on Soar PS


Stuart

[awemawson]

I decided that it was worth soaking the entire arm in citric acid, so I transferred it to a tote box that was almost big enough (!)  - it'll need at least an over night soak if not a few days