Hey guys...just had a discussion with Rick at Rotary Performance about why my freshly built engine had developed a coolant leak so quickly (< 1 year).

So the motor was rebuilt with fresh seals, new rotor housings, resurfaced everything, streetport, 3mm seals and Evans coolant and zero pressure. A few months after the rebuild, it developed a minor coolant leak, which was made more evident when using a pressurized cap. It killed me because the motor otherwise was running perfectly with good compression (8.8 +- 0.1 on rear rotor, 8.6 on front rotor).

So upon teardown, Rick noticed that the motor was essentially flawless, all the seals were great, surfaces looked great, no cracks, and was finally able to track down the problem when measuring the rotor housings. The coolant leak was in the front rotor in the combustion chamber area. The rear rotor housing was perfect (79.99mm all around), but the front housing had warped (79.99 at the top, 79.95 towards the middle, and 79.94 at the bottom), and the only unusual thing he noticed besides that was some gunkiness of the coolant around the same area.

That being said, the only reasonable assumption we could come up with was that it had something to do with the Evans coolant running at 0psi and that their might have been a coolant flow issue by using the Evans, which is more viscous, and running no pressure.

Does anybody else have any possible explanations? I'd already decided to go back to standard water/coolant even before this discovery, but the possibility that running 0psi creating flow issues was an interesting thought.

 

hmmm

what was the highest temp that you got with Evans?

I REALLY have doubts that viscous fluid could warp metal. High temps are a more likely culprit.

 

The temperature never got above 95 Celsius prior to the problem occurring from what I can recall.

What was odd was that from what I understood, overheating conditions usually result in the rear housing being warped first, not the front. The warped metal was almost certainly caused by high temperatures, but most likely very localized high temperatures in the front rotor.

I'm just thinking out loud here, but the engineers designed the cooling passages with smaller and larger passages in certain areas to try to best distribute heat around the entire engine (i.e. there are smaller coolant passages around the intake area and larger passages around the combustion area), and I'm wondering if using a thicker coolant with zero pressure alters the flow characteristics sufficiently to create localized hot spots. Of course this is what Evans was designed to avoid, and I'm very aware that many people have used it with only positive results.

I personally loved the Evans given the nice safety margin it provided in case of overheating (no boil-over conditions), but in the end, I realize that the incremental cost and risk isn't worth the benefit.

 

Your choice of coolant does not appear to be at issue here. Many of us have run Evans without issue at zero pressure; zero pressure is a misnomer anyway because the pump puts up pressure or there would be no flow.

I disagree with your conclusions that there is a risk running the Evans, I would say there is risk dealing with a shop that has this concern as there is no engineering basis for it. I hate to even utter this but many shops will allow a customer to believe unfounded causes for incidents just to satiate the customers need to know something. Your situation sounds like an air pocket but could have been due to many other things such as a fouled cooling jacket. Look somewhere other than the Evans for this incident unless you discover some new data, who knows, maybe they had a quality problem with that batch, heat it up with a thermometer and see what happens under controlled conditions.

 

Yeah I was thinking the same thing, and Rick admitted that this conclusion was simply a guess since he couldn't locate any other possible problem.

I should mention that there were some tuning sessions in between, and it's possible that during the runs, there might have been temporary lean spots and whatnot.

Still, in the end, I realize the extra cost of the Evans (I've spent nearly $200 bucks on that stuff over the past several years) didn't make up for the benefit. Rick and Steve Kan, both reputable builders/tuners didn't deny the possible benefit of Evans, but they both mentioned to me that they do not recommend its use.

They stated that these motors, when properly built and tuned, run just fine with standard coolant and pressure and that they were engineered for that purpose anyway.

 

I don't know jack about materials, but could this be due to a manufacturing flaw?

As far as Rick goes, the man is a damn perfectionist when it comes to engines....and if he DID make a mistake, he would certainly come clean about it.

 

I'm no expert here, but I would think that coolant-related issues would allow both the front and rear to overheat at a similar rate, IF both rotor housings were generating the same amount of heat. Since only one housing warped, I would assume that only one housing was seeing increased temps. Is there any way to determine if the front rotor was running hotter than the rear?


On the other hand, it is possible that the different viscosity of the Evans is creating flow/turbulence issues when used with the stock water pump. Ideally, you want some turbulence (but not so much that the fluid vaporizes), so that as much coolant comes into contact with the metal surfaces as possible.


Finally, I still refuse to believe that Evans transfers more heat than water without proof. All you would need to do is measure coolant temps on both the hot AND cold side of the radiator, with Evans and traditional water-based coolant. The fluid that yields the largest differences in temperatures is transferring the most heat.

-s-

 

I think this is a good line of thinking. Although I've never torn down an engine with warped housings to know, are there other symptoms of a local overheat?

I have to wonder if the gunky coolant is normal for any kind of coolant leak that pushes exhaust gas into the coolant, and is a result of the leak, not a clue to the reason.

Interesting point. I think the phenomena that causes this is cavitation, not merely turbulence. The more turbulent the flow (without cavitating), the better from a heat transfer point of view.

Unfortunately you'd have to assume that the volume flow rate is the same, which it may not be.

I would continue investigating the water pump and flow conditions, but I don't have any suggestions how to investigate further.

Dave

 

Thanks everybody for your input! Like I said before, I know that many people use Evans with no problems at all. At this point, there's really nothing I can blame it on except that perhaps there was an air bubble that developed at some point (possibly during tuning?) that resulted in the localized hot spot.

My next damn engine (my fourth since I bought the car) better be damn near indestructible and run for 200k miles.

=]

 

And regarding what Dave said regarding volume flow rate, I would think that viscosity would affect flow rate so the heat transfer radiator test probably wouldn't be a fair comparison.

 

Both my prior motors all warped on the front housing near the top where you would expect an air pocket to be. Both were on EG and water.

I have 9K on a Malloy reman running evans npg+ at 7psi and AST w/o any issue. Oh yeah, in the Phoenix summer, I'm seeing 120-128C regularly under boost.

Even after a hard summer, I had Steve Kan tuned my car only to find out I have fuel problem above 5K but it still made 264 whp and 280 ft lbs at that low rpm.

 

One risk of running zero/evans is creating a vacuum in lower rad hose at high revs, due to pressure drop across rad, and partial temporary collape of hose.

With no ast, run 7 psi cap.

With any ast this is not a problem with zero-cap, IF a 2mm pill is put in the hose to the filler neck. If old stock ast, throw away. If new stock ast must be used, same pill and drill out restriction in lower nipple.

 

Where the heck did you guys find a 7psi cap? I remember looking for one all over the place...