Flush balancing

I have an S7 mold core with 2 inch deep, .030 wide rib features that form 14 chambers in the trode, 14 standing cores on the workpiece. I will pressure flush thru the spindle. The electrodes are poco C3 due to a thin, outer perimeter rib, .020" wide rib that extends .050 beyond the rest of the bottom of the electrode that is easily damaged. Likely it doesn’t really need to be copper impregnated, Poco3 would be fine, perhaps better. I believe it (the copper in the electrode) may be contributing to the build up of a globular-looking recast on the bottom of the deeper, outer-most perimeter rib, with a rougher, globular-like texture ion its floor than the other floor areas. The side wall surfaces look as expected other than some build up of tars/debris on some of the outer chamber walls, appearing like they aren’t getting enough flushing. ,
Jump time is high, Jump speed is high, cycle time is low. Poweon the troublesomer is conservative with an off time multiplication of 2. Ive tried both graphite and copper electrode settings and the copper settings are slower, but dont melt the copper out of the trode and produce better finishes on the floor of the troublesome outer perimeter rib…though it is slower.

The 7 features (chambers) on the inner most area of the electrode and are quite small…about 1/10th the perimeter rib length of the outer features chambers. The flush enters the center of a plenum chamber above and separated from the electrode, and flows thru individual holes to feed dielectric into each respective chamber. With the chambers arranged where the inner features flushed debris must travel into and across the outer details burn area before dropping into the final, deeper rib around the perimeter to exit. The flush flow could be thought of as burning the center detail, then when jump occurs, carrying the debris laden flush into/across the outer perimeters detail where it will pick up more debris, then finally be pushed out across the last barrier, a .050" deeper rib than all other feature, then exit by being pumped/pushed up and out of the deep perimeter rib into the tank.

Using an EDAF3 Makino with high speed jump for the first pass, and that works well enough with it’s strong pumping action. The finish passes are most troublesome as the outer features walls collect tar/debris build up and create slightly sunken pockets and leave pitting on the walls. The floor of the outer perimeter rib has the globular looking recast. There is some pitting along the outer group of chambers wall features, like debris and secondary arcing is building up and then the edm arcs thru it, making the little pits…seems inadequate/improperly balanced flushing volume to me.

I’d like to know how to correctly determine the area of each hole that feeds dielectric from the plenum to the individual trode chambers to balance the flow of dielectric throughout the trodes features and promote its flow to the outside in a balanced uniform manner. Is it a direct ratio, based on the area of each chamber? Do engineers use flow dynamic software to determine best volume balance, prevent dead zones with the least amount of flush pressure?

Thinking a bit further along the timeline of a single Jump cycle, the biggest problem may occur as the multi-chambered electrode retracts…and to large degree how rapidly it retracts. Now the small, but over fed by clean dielectric volume to the inner chambers is momentarily at a higher relative pressure compared to compared to the osuter chambers and the outer perimeter rib.

The system will try to equalize and higher pressure will flow into lower pressure, with the residue from the inner chambers being pushed up into the outer chambers, causing secondary EDM’ing and a dirtier burn because the walls of the lower pressure chambers are being exposed to contaminated dielectric at a time when they need to be washed off by clean dielectric, moving in one direction, never reversing, outward since we are pressure flushing.
It might be possible that dielectric flow reversal could occur more severely if retraction velocity were fast and long enough to temporarily draw previously expelled debris back into the electrode along its outer perimeter. The system will always try to equalize it’s internal pressures, but it’s not instant. The more equal the flow rates are to each chamber the less temporary variation will occur, improved flushing and a positive directional flow without reversals. should provide superior processing times and less fouling of workpiece and electrode surfaces.

So how do we best balance plenum fed chambers? How do we determine how much area each hole or holes should be relative to the chamber volume it’s feeding?

a few things come to mind…
arcing comes from overpowering the trode and over plating it in a “no wear” setting…or it comes from a poorly flushed condition
without having an actual picture, it sounds like this “Chambered” trode also needs to be vented to let gas out of the inner chambers…I will usually run a small flush drill…015" to .020"…high up in the void where gas can
get trapped…you need it to vent to the outside of the trode.
keep your jumps slow and long…use Model data 55…start at process 8 or 9 with .004" or .005" Overburn per side…let the trode wear a bit…it might take 4 passes…Poco 3 or Copper impregnated graphite will both work well…
Kirskies is right on the pressure…slow it down

Thanks for responding Paul. All the trodes 13 chambers have around 3/32" diameter holes pressure feeding them from a common plenum from the top of the trode, pushing dielectric out the bottom of the chambers, then to the outside . I believe the rougher finish on the bottom may be due to the control putting in a time limit on machining time beginning with cond 12. I took the time limits out of all the settings. I just used the setting offered by Project at first but switched over to a 154 (low wear 2%) beginning with cond 6, burning thru 14. Going to finish at a 14 condition at least… A 15 would be great, being a 2" deep, difficult to polish area. there are also openings on the sides of some chambers on 3 of the four sides. I wish I could show a photo of the trode/workpiece , but corporate says no sharing of photos.
I burn with at least 2x off time multiplier on the rougher settings and progress up too 5x by the time I get to 14 but its not a clean burn after condition 12. Risky as any pitting/arcing into the long side walls of the chambers can scrap them because it cant be polished out or burnt bigger. I could try it normal wear, or reverse polarity, but Id prefer not to…though What I’d like may not matter too much!
My jump speed is not high, I could drop it down a bit. Flush pressure is .07 Mpa. Cycle is very short as the least amount of sustained burn is kicking the shorted spark over 40% at one point in the circular orbit pattern.