SVX Engine cooling "Again & Again"

[bazza]

Seems the big radiator worked a treat for the EJ207.

We had a EG33 casualty in another car... autopsy being performed now.

[SVXRide]

Originally Posted by bazza

Seems the big radiator worked a treat for the EJ207.

We had a EG33 casualty in another car... autopsy being performed now.

Bazza,

To what extent, if any, was the cooling system modified in the recently departed?

Cheers,
Bill

[Tapani]

Is anyone interested in solving this ? Not a real issue for myself, but never the less an interesting problem.

I am still leaning on cavitation.... the root cause is naturally unknown.

http://www.waterpumpu.com/news-blog-...-and-solutions

Marisa - you could talk to the Automotive Engineering division guys? 20 years ago they used to have after work sessions to kick around interesting stuff....

/T

[Tapani]

I couldn't resist myself.... searched the net a bit yesterday and gave the whole issue a few more thoughts.

I am not saying this is the whole truth, but for those who want to rev the engine above 5500 rpm for extended times it's worth considering..... and challenging .

High heat load/soak under extreme ambients, good midrange, tall gearing and automatic only and zero tolerance for durability issues define the design envelope for a road car like this. It was never ment to be driven like an STi...

The stock system works well and as designed. The flow rate of the mechanical pump is app. linearily proportional to the pump speed upto the point where the cavitation starts. Pump capacity and gearing is chosen based on low speed requirements and cavitation at extended high speed operation is a fact. The location of the thermostat is based on sound facts (- at least for the road car).

All this suits me personally fine - especially with the mildly blown engine, which probably utilizes some of the included capacity margin in the system at it is..... I am happy to shift at 5500 - 5800 rpm under track conditions anyway - also from lubrication stability point of view..... and as said earlier - I do not want to break the car intentionally.... it's a grand tourer and a summer time daily driver for me..... with the occasional visit to track days mainly to socialize with the crowd - and get a feel for the handling on track.

Interesting threads/links I found:

http://www.sau.com.au/forums/topic/3...mp-flow-rates/

https://www.youtube.com/watch?v=K_w3gcvA87I

Also, see the attachement. A lot of discussion can also be found related to the temperature / flow differences of the LH and RH banks what Tony found - not related to the EG33, but to GM LSX engines.

Pure ethlyne glycol has a huge margin for cavitation and flashing resistance compared to a normal coolant mix - especially close to operating (and pump inlet) temperature - see the charts in the attchment. We know it has a lower specific heat capacity and thus is not as good a media for heat transfer as pure water is at a given flow rate, but I trust it's been shown that there's some margin in the system capacity as a whole at lower engine speeds so it might be a good idea to try this out - see if the start of the cavitation is shifted to higher revs.

The other route is to start playing around with impellers like the gyus in the Nissan thread do - they have the luxury of being able to choose from three OEM pumps with different capacities and to fiddle with the drive ratios - it's driven by the aux belt. The different impellers in all Subaru engines is uncharted territory.... at least to me.

So, who has a good stock setup and feels this is a big enough problem to start searching for the edges of the design envelope one change at a time?

Remember that when my car spat coolant at track conditions - and there only - it was a HG issue.....

Kind regards,

/Tapani

[Tapani]

A friend bought a Mustang with the Coyote engine.

I just had a peek under her hood.... the stang, I mean....

I really like the idea of the cooling system. See attchement.

/T

-edit: Check out the photo. The copper plate in the thermostat blocks the bypass flow when the thermostat opens. It is a three way valve.

[EdWindows]

I'd still like to know how this problem was solved, but 112 pages is a lot to read through.
So far I've seen 3 things:
-increase water pump inlet from 28mm to 40mm?
-move thermostat to top radiator hose
-fit a crossflow radiator

Only the 3rd seems like a viable solution to cooling, but it may not fix the cavitation issue.
EDIT: found this link that lists the causes and general solutions to prevent cavitation.
http://www.mcnallyinstitute.com/19-html/19-03.html

[Tapani]

I am beating this old horse.... apologies for that.

I found an interesting deatil in the latest 1.6 VAG TDI engine. The rate of the mechanical water pump is variable using a sliding collar around the impeller (to restrict the flow). It's done to speed up the heat up process and to minimize pumping losses when high capacity is not needed.

The only real hit I found was in a long term test article....

A centrifugal adjuster for the EG33, eh ??

/T

[92 SVX]

Well thought I would post this.

larger tstat opening, 50mm OD or 2" relocate Tstate to the return hose, use either a crossflow rad (tanks on the sides) that has the cap on the left side, so the pump pressure can not overcome the rad caps pressure spring and dump coolant, or have the rad outlet relocated to straight across from the tstat cover/water pump either rad used needs 50mm/2" outlet and hose.

[92 SVX]

Now all that being said couple things I have read and have been thinking of, there has been lots of back and forth on which cooling will be better, slowing the flow speed to do multiple things, such as reducing pressure and to allow the coolant mix more time to run across the heads to absorb more heat.

I thought that was the first thing that was thought of and disproved, trevors idea of drilling the impellers so they could not build as much pressure/speed, while the theory was sound in practice not as easy, and yes tapani I did see your pic, that may or may not work because of the cover over the vanes drilling through that cover may cause it to cavitate. also at what point is too much water flow going to disprove what shotgun slade said about the overall mass of water will cool more then slowing down the water?

And I like the electric pump run that at what ever speed is determined to have the perfect water flow/pressure no matter what rpm the engine is at, I just dont know if its practical for everyday driven cars?

And final thoughts incorporating both YT's and Tony's ideas, as well as several other people including boxer's, run 2 coolant hoses off each side like YT did except one out by the ac and one by power steering, have tstat housings wielded into the top of the rad like boxer's (he had one but same idea) and of course drill holes in the tstat but even more importantly have a bypass as part of that wielded in housing, this bypass will pass water into/through the rad which of course means back down to the suction side of the pump. then you can block of/use the bypass ports for other applications, such as turbo or oil cooler.

[rjwoods77]

Something to add to the discussion. Look at the reviews for my installation notes...

http://www.rallysportdirect.com/part...r-pump-housing

[bishop]

I was looking at one of those and planning to add one to my build whenever that gets started.

[shotgunslade]

It has been a long time since this thread has been amended, but I have some new and relevant data. The build on my '94 Emerald Green was getting to be 9 years old, so I thought I would have the belts changed (10 year recommended interval). The shop said that while they were at it, they should go ahead and change the water pump. The fact that the thermostat on my car had been removed from the water pump and a remote thermostat had been added slipped my mind. Drove it home after the work and it was bad old days all over again, but worse. Right after the first time I took it over 5000 rpm, the temperature needle very rapidly moved from slightly below horizontal to vertical, past the red zone. Pulled over, let it cool a bit, drove it home, trying to keep it about 3000 rpm. Still, frequent blips into the red, then subsiding. Got home, overflow tank is full of coolant. Called the shop, scheduled a visit to remove the thermostat. (I don't have a lift and I'm getting old). They removed it. Back to being nice. No overheating spikes. Proving the hypothesis that at least part of the spike overheating problem is cavitation at the inlet to the water pump, caused by the flow restriction of the thermostat. I'm a retired mechancial (HVAC) engineer, and our standard for pump installation is 5 diameters of straight pipe at the suction side or the pump or an inlet diffuser to even out flow into the pump and reduced turbulence. Here are teh standards for pump configuration:

The suction pipe should never be smaller than the suction nozzle of the pump and in most cases it should be at least one size larger. Suction pipes should be as short and as straight as possible. Suction pipe velocities should be in the 1.0 – 1.5 metre per second range, unless suction conditions are unusually good. Higher velocities will increase the friction loss and can result in trouble some air and vapour separation. This is further complicated when elbows or tees are located adjacent to the pump suction nozzle. In that case uneven flow patterns or vapour separation keeps the liquid from evenly filling the impeller. This upsets hydraulic balance leading to noise, vibration, possible cavitation and excessive shaft deflection. Cavitation, erosion damage, shaft breakage or premature bearing failure etc. may result.

Note that the guidance is given in Britspell for all our Southern hemisphere friends.

[oab_au]

Thanks for the report Dan, the thermostat was the cause of the cavitation and moving it to the top solved that bit, but overheating still persisted at high rpms, this was found to be caused by the water holes from the block to the crossover manifold, these holes in the block are too big at 30 mm, this allows all the pumps flow, to flow through one side, allowing the water in the other side to slow and boil around the exhaust pockets, reducing these holes to 25 mm (as the 4 cylinder engines are) prevents it from all flowing through one side, to maintain flow through both sides.

Of course the best fix is to do the two outlets as you, and the EZ engines have.
This is in Ozzie speak.

[shotgunslade]

Harvey

Good to hear from you. Hope you’re hanging in there. Had my 70th last year and finally retired. No more track time for me. Just take my hot rod out early Sunday morning into the foothills of the Pocono Mountains just north of Princeton NJ and scare the sheep and cows.

[oab_au]

Yes Dan I am 10 up on you, but still at it, do the same on the mountain roads, but it's not the same as it used to be out west, the police take a very serious view now days on high speed touring, and I need my licence.