SVX Engine cooling "Again & Again"

[Dessertrunner]

Hi Joe its great to get your input. I think it would be good to get this sorted out so its clear for everyone to understand.

At 7,500rpm the coolant can move 153 kw of heat at 3,000rpm it will only move 61 kw of heat. If you slow the pump you won't get rid of enough heat.

I am a bit short on time to debate this to far so I will come back to it tonight.
Tony

[Trevor]

Originally Posted by svxistentialist

Don't want to muddy the waters or the coolant but the points Trevor made in his lengthy post are correct imo.

A free-er flowing and highly efficient radiator will only be able to dump extra engine generated heat if the coolant is being forced through it at the optimum rate for the whole system.

Too slow and the excess heat from the engine will not be pumped [transferred if you will] fast enough to the highly efficient rad.

Too fast, and the rate of heat transfer in BTUs as Tony notes will not be optimised.

What seems to be happening in our [generalised] case is that high pump speeds are causing higher than acceptable pressures in the system. These high pressures are a direct result of flow restrictions that are more likely to be in the engine block than in the radiator complex, though I have run no tests to prove that. It is just my opinion based on current reports.

Trevor I think is advocating reducing the coolant flow to match what the engine normally runs at, rather than the very high speeds of coolant flow that consistent speeds over 4500 rpm might generate. I see great merit in this approach.

However I will admit I would be happier to see tests identify where the restrictions are in the block, and if possible relieve them to allow a faster flow. Building up pressures so that the rad cap loses coolant is bad news.

I'm not offering answers here, just saying I think Trevor's thinking on the problem is correct as I see it.

Joe

Very special thanks Joe.

At long, long last a smidgeon of common sense has been recorded within a thread which has become ridiculous.

However you suggest further testing with regard to the engine block. Most of the thread has been taken up with this angle of approach, and several alterations have been carried out to no avail. The cost and complications involved with testing and modifications to date, far exceed that involved with my simple suggestion, which has stupidly been continually resisted and ignored.

[Dessertrunner]

Originally Posted by Trevor

Very special thanks Joe.

At long, long last a smidgeon of common sense has been recorded within a thread which has become ridiculous.

Trevor take a chill pill or I will spit the dummy at your comment. Matt and my car doesn't overheat. So we have acheived a lot.
Tony

[Trevor]

Originally Posted by Dessertrunner

Trevor take a chill pill or I will spit the dummy at your comment. Matt and my car doesn't overheat. So we have acheived a lot.
Tony

Nice one Tony.

No specific advice has been posted to cover a universal fix. There are obviously several still seeking a solution. Rather than take a chill pill in accordance with your very direct instruction, as a worthwhile alternative, I request that you please kindly assist those still in need. Meantime Spitting out the dummy you continue sucking, could prove a worthwhile move.

[oab_au]

Originally Posted by Dessertrunner

Hi Joe,
Given that now I see the reason that one is better then the other I feel like a total idot that needs councling.

Side tank radiators on the condiation that the radiator cap is on the same side as the suction. What I learn't with Steve's and it applies to our radiator to a lessor extent was that at around 7,000 rpm our water pumps will pump to 24psi. At that pressure when the coolent hits the radiator there is a head pressure in the top tank, in Steves a case it is above 10 psi so the cap blows and releases the coolant.
What I was trying to do to fix his issue was to drop the pump flow which was wrong instead we need to get his cap on the suction side in that way his radiator is pressurized instead of operating in vacuum.

Think about it with our standard cars we rev the motor above 6k and we suddenly get major head pressure in the top tank which results in the engine blowing all the cooant out the overflow because top tank pressure exceeds cap pressure.

In Steves case he system was cooling the engine it was just that he was blowing coolant out the overflow which ment that in the end (long race) he had none left or the pump could no longer pump due to lack of coolant. If you follow me a bit further, coolant is blown out the overflow which mean an overall vacuum developing in the system where does that vacuum turn up its at the pump suction untill it goes so high the pump can no longer pump and the system dies. That is why I saw -10psi at his suction.
Well Joe hope that answers your question also you should have a look at side tank radiators the majoity have the cap on the out let side.

Have a great day.
Tony

Tony I think that all of the testing that you have done has established one common thing, and this last run of Steve's points it up again. The trouble he had in the last run, of not overheating for 180km, only needing a water top-up, and the pressure testing that you have done on his radiator, show pressure differences as being the problem.

As the pump speed increases, the flow increases, to increase the pressure differences across the system, causing the cap to blow pressure, and the pump's inlet pressure to drop to cause boiling at the impeller.

I think it would be worth a try to fit a pressure operated valve to join the pumps outlet, back around to the inlet. This way all is standard till the pumps speed increases to the point, that the pressures would cause the problem. The increased pressure across the pump opens the valve to connect the inlet to the outlet, reducing the pressure difference to prevent the troubles that it causes.

The port on the pump's outlet could be used to connect it back through the pressure valve to the heater feeds, on the pump cover. It would take a bit more research to find the right pressure for the by-pass valve to open, and the amount of water that it recirculated.

Nothing that your research lab can't do, a?.

Harvey.

[Dessertrunner]

Hi Harvey good to get your input,
Steve and I spoke about the problem and he is doing a simlar thing to what you suggested (Great mind think alike). In the short term he has connected the back of the head hoses to the pump side fitting. We feel sure that will reduce the problem but I don't expect it will solve it completly.

Untill the weekend I didn't really appricate the issue of getting flow throught the radiator. I was straching my head thinking how the hell am I going to get coolant to SUCK through the radiator, also if you remember every time I tested the radiator in the start at higher revs coolant came out the overflow. Now we have the answers.

Was looking at a Ford radiator with side tanks and even it has the cap on the suction side. With out trying to sound to insulting to Subaru you got to wonder what the hell they were thinking when they did the cooling on these cars. The problem seems to be they assumed it would work.

Tony

[SilverSpear]

Originally Posted by Dessertrunner

Was looking at a Ford radiator with side tanks and even it has the cap on the suction side. With out trying to sound to insulting to Subaru you got to wonder what the hell they were thinking when they did the cooling on these cars. The problem seems to be they assumed it would work.

Tony

Tony, when my OEM radiator blew up (upper tanks cracked), I tried to find a replacement locally without luck.

So I started looking for an alternative and I bought a JDM Legacy rad (sorry dunno which year, the guy selling the part wrote on it "Legacy").

That rad had side tanks, was wider and a few mm thicker than the SVX's. I am sorry to say that cooling went worse than it was before. I couldn't speed uphills without the car overheating. This is from personal experience, same rad, different position tanks, OEM design wins.

[Trevor]

Originally Posted by oab_au

Tony I think that all of the testing that you have done has established one common thing, and this last run of Steve's points it up again. The trouble he had in the last run, of not overheating for 180km, only needing a water top-up, and the pressure testing that you have done on his radiator, show pressure differences as being the problem.

As the pump speed increases, the flow increases, to increase the pressure differences across the system, causing the cap to blow pressure, and the pump's inlet pressure to drop to cause boiling at the impeller.

I think it would be worth a try to fit a pressure operated valve to join the pumps outlet, back around to the inlet. This way all is standard till the pumps speed increases to the point, that the pressures would cause the problem. The increased pressure across the pump opens the valve to connect the inlet to the outlet, reducing the pressure difference to prevent the troubles that it causes.

The port on the pump's outlet could be used to connect it back through the pressure valve to the heater feeds, on the pump cover. It would take a bit more research to find the right pressure for the by-pass valve to open, and the amount of water that it recirculated.

Nothing that your research lab can't do, a?.

Harvey.

Harvey,

Yes, a pressure difference is the problem, i.e. exactly as I have continuously been pointing out throughout this thread. However the actual problem involves the difference in pressure, delivered by the pump at abnormal RPM, when compared with the optimum designed pressure applicable in respect of the radiator and overall flow path. As a result of the optimum flow rate of the radiator being exceeded, the top tank pressure becomes abnormal and the pressure cap releases coolant as a result of the excessive pressure, i.e. it does its job well.

By way of an answer to his question, Tony should accept what I previously advised and note that as a result of the resistance offered by the radiator, once the optimum flow rate is exceeded, the pump will become starved and a vacuum will develop at the pump inlet. Again the mugwump bird will be in full flight.

N.B. Fitting “a pressure operated valve to join the pumps outlet, back around to the inlet”, i.e. a pressure controlled bypass, is an excellent idea. However this does involve some mechanical complications, as well as the necessity of obtaining a suitable valve. This should preferably be adjustable, and components from a reducing valve could provide the means of making up something suitable. This should fall well within Tony’s area of skills and could be an interesting project.

P.S. It could be possible to make something up, using the guts of a radiator cap. This would provide a suitably calibrated relief valve and several variations would be available for experimentation.

[Trevor]

Originally Posted by Dessertrunner

Was looking at a Ford radiator with side tanks and even it has the cap on the suction side. With out trying to sound to insulting to Subaru you got to wonder what the hell they were thinking when they did the cooling on these cars. The problem seems to be they assumed it would work. Tony

The subaru engineers knew their design would work, provided the car was driven normally. They set optimum levels in respect of all components for good reason. They were instructed by management to produce a car at a certain price level. They designed the pump and cooling flow path accordingly.

Actually the Subaru engineers should be complimented. They were exactly accurate and were able to establish and work within very close tolerances when establishing optimum levels. The SVX was not designed as a desert racer.

[Dessertrunner]

SH_T "hell must be freezing over" Trevor agrees with Harvey I need to get that posting framed. Great to think you guys CAN agree.

In all seriousness I agree that we can put in a bypass valve and get round the problem in the short term. Long term I thick we need to find a more perment soluation. The bit ticket concern I have is the coolant not going through the radiator and therefor reduce effiency. We need to pressure feed the radiator to get the flow up.

Danny thats good to hear, Do you have any photos of the top and bottom pipe also were the cap is in relation to the outlet.

Tony

[SilverSpear]

Originally Posted by Dessertrunner

SH_T "hell must be freezing over" Trevor agrees with Harvey I need to get that posting framed. Great to think you guys CAN agree.

In all seriousness I agree that we can put in a bypass valve and get round the problem in the short term. Long term I thick we need to find a more perment soluation. The bit ticket concern I have is the coolant not going through the radiator and therefor reduce effiency. We need to pressure feed the radiator to get the flow up.

Danny thats good to hear, Do you have any photos of the top and bottom pipe also were the cap is in relation to the outlet.

Tony

Actually this is the second time it happens , should have saved the other post...

Tony, you are in Luck (you could have asked for a million dollars ). I was checking my files and I do have some pictures of that Legacy rad fitted before I switched to PWR... It was a tight fit especially with the legacy overflow tank. Also in the pictures you can see "Legacy" "97" (pointed to it in red, 97 written in arabic). The upper and lower metal frames are just frames to hold the rad and fans, the tanks are on the right and left.

[svxfiles]

Originally Posted by Dessertrunner

Was looking at a Ford radiator with side tanks and even it has the cap on the suction side.
Tony

A Ford radiator in an SVX...http://www.subaru-svx.net/photos/use...svxfiles|37228
Side tanks, what six years ago?
Why is Trevor being,,,
like he is?

[Trevor]

Originally Posted by Dessertrunner

SH_T "hell must be freezing over" Trevor agrees with Harvey I need to get that posting framed. Great to think you guys CAN agree.

Tony,

Therefore you should be guided accordingly and take very special and serious notice of what we have to say.

In all seriousness I agree that we can put in a bypass valve and get round the problem in the short term. Long term I thick we need to find a more perment soluation. The bit ticket concern I have is the coolant not going through the radiator and therefor reduce effiency. We need to pressure feed the radiator to get the flow up. Tony

The radiator flow should not be increased above the designed limits. The radiator has been designed to dissipate heat at a certain level of flow, i.e. the optimum level of flow. Once this is exceeded, the cooling will become worse, not better. I have explained all this in great detail, but unfortunately you chose to spit your dummy.

If the flow rate is increase a larger radiator will be required, but then you will be up against the problem of a flow rate which is too high, passing through an engine designed for much less. The problem involves two very distinct factors and both must be taken into account.

Both Harvey and I are not making suggestions on the basis of a temporary or short term fix. We are are talking about probable permanent and final solutions.

Please note my P.S. now added to my previous post, regarding a way to make a suitable valve to suite what Harvey has in mind.

[Trevor]

Originally Posted by svxfiles

A Ford radiator in an SVX...
Side tanks, what six years ago?
Why is Trevor being,,,
like he is?

Tony for goodness sake, I certainly am what I am.

You posted:-

“Was looking at a Ford radiator with side tanks and even it has the cap on the suction side. With out trying to sound to insulting to Subaru you got to wonder what the hell they were thinking when they did the cooling on these cars. The problem seems to be they assumed it would work.”

You stated, “according to Subaru”, and then by association clearly referred to Subaru engineers.

I was again trying to press home the logical and valid point that the SVX radiator was designed to operate at an optimum level. It was/is not badly designed.

[Dessertrunner]

I am amazed the ford radiator looks exactly like I would like to use. The outlet size is large and also up a bit from the bottom to get better flow. Would love to test one of them in the bench.

Danny I think the problem with the one you had was the tank were very small and my have killed the pump, the PWR are better then the standard SVX radiators.

There are a couple of issues with having a bypass valve back to the pump,
As I understand it the cap lets go at 14psi. Currently on a standard PWR the top tank pressure at 6k rpm is 7psi What ever we set the bypass at it has to come off the cap release pressure. We can't simply put a higher pressure cap on because when the revs drop the blow off Temp will change.

The second and my biggest concern is getting rid of heat load and to do that I need every litre of coolant I can get. If you remember all attempts with electric pumps have not had enough coolent flow to cool the engine so I not sure how the bypass will get around this problem.

What I am aiming for is to cool a race engine that develops higher power then standard and to that end I need everything in the system to work to maxium.

Have a great day.
Tony