This is a bit of a duplicate post from my last entry under the "anyone with epoxy experience". I felt as if there was enough new information that I should re-examine whether there really is a hole in the primary runner.



I've been cleaning the offending part with acetone and for the life of me, I can't see how this thing is causing a coolant leak. There's no visible hole, light doesn't shine through it, and acetone pooled around the spot in question does not leak through.



So the facts are:

1) car burns coolant but only seems to do so under boost conditions above .75kg/cm2

2) car does not seem to burn coolant under .7kg/cm2 or under vacuum conditions

3) coolant residue was found in the uderside of the oil filler cap

4) if I boost above .75kg/cm2 and the engine "swallows" a bit of coolant, I can rev it, drive around a bit and the coolant will clean out and the car operates normally. 100% normal under vacuum conditions. Vacuum levels are just fine, about 375mm to 410mm at 1100 rpm

4) there seems to be a pitted spot in the front rotor's primary intake runner

5) we can press on it with force and it does not open up or give way



After I cleaned it, I had the idea of sealing off the ports. They did this at the shop too, using hands to block off the ports and using compressed air. A thin sheet of paper, stuck in the coolant passage below the port fluttered when air was applied to the port.



I decided to duplicate that experiment but I used thin latex gloves and towels to seal the ports. The soft latex provided a good seal. I also plugged the small port inside the injector bung and proceeded to blow as hard as I could (no jokes please) into the top of the injector port. Enough force was applied that eventually the glove/towel combo was pushed out. No air seemed to leak into the coolant passage.



Since it only leaks when boost is applied, I would think that the coolant would leak under vacuum conditions as well as under boost. The coolant system is under far greater pressure, 16psi to be exact, so theory holds it would leak all the time, no?



I'm about ready to run back over to Walmart and purchase a bit of playdough to really seal everything off but I'm almost ready to bet my left nut the "pit" in question is not causing the leak. I'm footing the bill for the whole fiasco since the shop has determined that my port job was responsible for the coolant leak. I'm now wondering if it's a convenient scapegoat for not assembling/checking the engine properly.



Where are common areas that leak coolant? Besides problems with porting, afik, they can leak around the spark plug area. We tested that. Nothing. Now, if the plates weren't checked for flatness and there was one small area which was uneven, higher boost levels would expose this weakness wouldn't it? If the bolts weren't tensioned properly, unevenly or too tight, problems could be created, correct?



Thing is, if the engine were expanding laterally, it seems like oil/coolant would leak OUTSIDE the motor as well, not just inside.

rear rotor primary runner. Same circular plug at the bottom of the runner, again, looking through the primary injector hole.

interesting........i think i remember hearing about this problem from pepe loco..... something about the casting style on original cosmo plates.... he used all new stuff on his 20B

Any evidence of seepage past coolant O rings? Since the condition is boost sensitive, I would suspect something in the intake system like any warm up devices with coolant hoses leading to them. Although, since coolant is under the oil fill cap, coolant is also in the engine oiling system. So where could the oiling system and water jacket interact? Maybe a cracked coolant passage in an endplate that is alongside an oil passage?

It'd have to be a spot that allows coolant into the combustion chamber too. I didn't see any evidence of seepage. We used stock Mazda o-rings and they come apart when pulling the engine apart. I'm not enough of an expert to tell however. I didn't see any evidence of pinching or a broken seal.

https://www.nopistons.com/forums/pub...1047683561.gif https://www.nopistons.com/forums/pub...1047683561.gif It is weird that is only leak on boost, it should leak all the time. I take it you didn't do the coolant pressure test before you tore the motor down.

To me it seems that one of the plates is warped, did you reuse the old plates? But then it would leak all the time WTF https://www.nopistons.com/forums/pub...1047683561.gif

those patterns are normal on the castings, you definitely do not have a hole there in the port runner, but you could improve that porting a slight bit. Nice pics BTW.



I suspect O ring failure, triple check that possibility, make sure the groove is intact that supports the O rings, the back side in some areas gets thin and can break.



If you can rule out O ring failure, check the housings for cracks, and especially important check the freeze plugs for problems. It's really the only other place the coolant/oil can interact.

forgot to mention, I have pretty much all emissions devices removed, everything under the UIM. No coolant lines to the throttle body either.

Digital Nikon D1H and a 60mm macro w/ off camera flash cord comes in handy sometimes :-) I put a smooth layer of epoxy over the pitted area. We'll see how it looks in the morning. If it looks like crap, I'll sand/port it out.



I'll double check the rest of the plates. Thanks for the head's up pengaru.

Ok, I checked the intermediate plate again around the 0-ring groove for anything abnormal. Again, when the pulled the motor apart, the 0-rings seemed to be intact but they came apart (as I was told they should) when we pulled the plates off the housings.



What I noticed after the inspection was a bit of carbon deposited IN the rotor groove (circled in red) and a bit more black discoloration just south of that region on the face of the plate (yellow arrow). They cleaned everything within a day or two as well so if there was any other evidence, it was washed away. Dunno. Does the black just inside the 0-ring groove signal failure on the 0-ring's part?