Hi All:
This all makes sense...when you make the water less conductive (more resistive), you strip ions out of it by running it through the DI bottle...the same ions that can promote galvanic corrosion.
So (simplistically) less ions, less opportunity for rusting.
So far, so good, but there are a couple of things to consider that some of you have already touched on.
First, why did you have to strip out more ions from your water than normal... or did you actually do that?
Some have pointed out that conductivity probes and the circuits they send signals to can stop working or malfunction.
A simple and crude test is to short across the legs of the probe with a screwdriver...resistivity should fall to zero (conductivity rises to "infinity").
That will tell you if the probe is basically working but it won't tell you if it's working correctly...for that you need to do as nlancaster recommends in post #10.
Second, you have to ask how the extra ions got in there in the first place and why they are localized to the un-ground areas of the part and only in spots.
I immediately suspect something in the heat treat cycle.
If at any time the parts are immersed in molten salt during HT you will get this.
If a furnace is contaminated you will get this.
So it would be prudent to investigate the conditions during HT...it may be you can permanently solve your problem with a simple change to the pre-wire handling protocol.
Last to Luke Kerby's comments in post #9:
He is quite correct, the water conductivity will affect the dimensions of the part and here's basically how it works:
The wire senses the gap voltage between wire and workpiece to hold the proper distance between wire and workpiece before it discharges.
Conductivity too high and the discharge happens too early so the kerf around the wire is bigger.
Conductivity too low, and the wire has to get too close to the workpiece before a discharge is triggered...the kerf is too small, cut debris can't escape and the wire starts breaking
You can compensate for these effects by changing the wire offset values in the program...but that's a risky band aid to apply because it depends on maintaining the wrong conductivity to keep the workpieces coming out the correct size so all your wire offsets will be wrong for all jobs you run on the machine until you change the conductivity back to normal.
Also with a failing resin pack (conductivity increasing), the parts will start to vary in size until you notice something's wrong...you cannot hold size on critical features.
If you're already near the upper conductivity limit for an effective spark gap, when the gap continues getting bigger and bigger, the cut speed gets slower and slower as the sparks become less and less effective at blasting away a bit of the workpiece. (of course that's true regardless what conductivity you started with but starting at the upper limit makes it show up sooner and worse as the resin degrades)
So I believe it is worthwhile to check all this out if you want to be able to rely on the intrinsic accuracy of the wire cut process.
If you don't have to care about microns for your application, you can get away with it as is.
Cheers
Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining