SVX Engine cooling "Again & Again"

[Trevor]

Originally Posted by Dessertrunner

Did this in two as I lost the first attempt,
Trevour I need you to explain to me if the suction is not the cause of the boiling then why would we bother to drill the impeller as its not a problem. The next issue is if you are correct you must be able to explain why the engine is over heating, you can't just throws rocks at everyones ideas unless you can back it up with clear theories that make sense.
It maybe that we need to drill the impeller but it is never done on commerical pumps as it causes airration of the fluid which is what we don't won't.
We need a sound proper fix one that Subaru would be proud of. I am willing to listen, so explain to me were your supporting infomation that the suction isn't boiling.
To everyone else this thing has me stuffed for the minute. Tony

Tony,

The correct line of approach does not have me stuffed.

Within this thread I have gone over all aspects of what I suggest, and have detailed clear theories which make sense. I am not throwing rocks at anyone. There are some who understand what is involved. No one has stated that reducing the pump efficiency is not a worthwhile approach and is not worthy of trial. In fact quite the opposite is on record and is increasing.

You comment, "It maybe that we need to drill the impeller but it is never done on commerical pumps as it causes airration of the fluid which is what we don't won't." Holes will not necessarily course aeration. This is surely a simplistic view of things. This is a fix for a completely non commercial problem.

Please will you, or someone else, explain to me with included facts, that the 'suction' is boiling i.e. the pump is being called upon to move vapour.

Some lateral thought is sorely needed here.

Cheers, Trevor.

[oab_au]

First of all Tony, you have done a fantastic job in doing this work for all us would bees. The placement of the gauges could not be better. There is a lot to be learned from these test that you have done.

Looking at the first graph, without the thermostat the first interesting point is at 2500/3000, when the pressure at the top hose and the inlet to the pump start to separate. The thermostat is between these two, so it would seem that is a restriction as the difference between the two pressures has grown at 5000, to about 10 psi.

The other point to see is the way the pressures between the top hose and the left side of the engine block, are separating at 2000 becoming larger to 3500, then increasing pressure up to the same, as the left side output. By rights these two pressures should stay the same, unless there is a restriction between them. The only bit there is between them is the junction of the right side block outlet.

Other than that the inlet pressure becomes more negative while the outlet pressure becomes more positive, as the revs go up, showing that there is a restriction in the circuit. As Dan said the pressure around the system should be about the same.

The second graph, thermostat in and out, hits us right in the eye that the thermostat is restricting the inlet to the pump. At 4000 the top hose pressures show the difference between the two increases as the rpm goes up. With the thermostat out the top pressure mimics the pump inlet, minus the restriction of the radiator. With it in the pressure in the top hose is rising, as the thermostat becomes a restriction between them. We can see that reflected in the pump output that is rising as the pump can pull more water in, to pump a higher pressure out, without the thermostat.

There is more to be deduced from this, as we look longer.

Harvey.

[Trevor]

Originally Posted by oab_au

First of all Tony, you have done a fantastic job in doing this work for all us would bees. The placement of the gauges could not be better. There is a lot to be learned from these test that you have done.

Praise is but a ploy, when suspect. If your pet theory is to be properly tested, pressure should be tapped at the left head outlet, to be compared with that at the radiator inlet.

Looking at the first graph, without the thermostat the first interesting point is at 2500/3000, when the pressure at the top hose and the inlet to the pump start to separate. The thermostat is between these two, so it would seem that is a restriction as the difference between the two pressures has grown at 5000, to about 10 psi.

This in fact indicates that the restriction of a thermostat, presumably open, is an advantage. The difference is approx. 5 PIs, not 10, as claimed. Once again my theory of an abnormal pump outlet at high RPM is upheld.

The other point to see is the way the pressures between the top hose and the left side of the engine block, are separating at 2000 becoming larger to 3500, then increasing pressure up to the same, as the left side output. By rights these two pressures should stay the same, unless there is a restriction between them. The only bit there is between them is the junction of the right side block outlet.

Nonsense. The top of the radiator and the pump outlet can not be related left right, in structuring a biased conclusion in support of one's pet theory. There can be no differntial at these points.

The passages within the block, and in fact the complete circuit, result in the difference between the outlet of the pump and the inlet to the radiator. As must be expected this variant extends with increasing pressure/flow/volume.

Other than that the inlet pressure becomes more negative while the outlet pressure becomes more positive, as the revs go up, showing that there is a restriction in the circuit. As Dan said the pressure around the system should be about the same.

Of course there is very real restriction within the circuit, What is news? If this was not so, a good old thermo syphon set up would be adequate. Those with wit, will note that the restriction has an increasing effect as engine RPM rise i.e. the very reason why pump output should be reduced at high RPM.

The second graph, thermostat in and out, hits us right in the eye that the thermostat is restricting the inlet to the pump. At 4000 the top hose pressures show the difference between the two increases as the rpm goes up. With the thermostat out the top pressure mimics the pump inlet, minus the restriction of the radiator. With it in the pressure in the top hose is rising, as the thermostat becomes a restriction between them. We can see that reflected in the pump output that is rising as the pump can pull more water in, to pump a higher pressure out, without the thermostat.

The difference as indicated is very slight, i.e. say less than 1.75 PSI. Hardly hit in the eye stuff and an allowable contingency which should be ignored.

There is more to be deduced from this, as we look longer. Harvey.

The lecture floor is yours, but please remain in touch with reality.

[FriendlyTurkey]

We should not be reducing pump speed if we can help it. It is proven that this pump geometry can do just fine at high rpms in EJ applications. It should be no difference in our EG33 application. We obviously have a large restriction that needs to be eliminated. I think the best method is to attack the major flow restriction in the cooling circuit first. This will reduce system pressure on its own. The other "problems" will go away once we remove the restriction.

[Dessertrunner]

I think the top of block outletis a problem. The image doesn't show it well but there is very little clearance to let the coolent out inspite of the port diamenter. I think I will put a pressure guage into the back of the head and see what number comes up. DOn't have a fifth so it will have to be one of the existing ones.
Tony

[Dessertrunner]

Had to can the last idea, you can't measure up stream pressure if down stream has a higher restriction. Its the old weakest link theory.
Tony

[FriendlyTurkey]

I don't know if I follow. The pressure will decrease through the circuit as you move away from the water pump. If you are able to place the pressure sensor between the restriction and the pump, you should be able to measure the pressure at that point. You could then compare this to the pressure after the restriction, and see the pressure drop.

[oab_au]

Originally Posted by FriendlyTurkey

I don't know if I follow. The pressure will decrease through the circuit as you move away from the water pump. If you are able to place the pressure sensor between the restriction and the pump, you should be able to measure the pressure at that point. You could then compare this to the pressure after the restriction, and see the pressure drop.

Hi Friendly, we could do that if we knew where the restrictions are. That is why we are working back from the pump, to establish what restrictions we can find.
The graphs don't show the real story, as there is not the heat input that a race situation would create. Tony is using his every day car, and can't expose it to too much abuse, it still has to cross the dessert reliably.

With the data that we are collecting, we can then try the modified components on the buggy, that can test them under race conditions.
A slow progress, but as Tony says " Rome wasn't built in a day", but then, he wasn't working on that job.

Harvey.

[Boxersix]

Originally Posted by Dessertrunner

I think the top of block outletis a problem. The image doesn't show it well but there is very little clearance to let the coolent out inspite of the port diamenter. I think I will put a pressure guage into the back of the head and see what number comes up. DOn't have a fifth so it will have to be one of the existing ones.
Tony

I would bet money that an EJ257 crossover tube bolts to the EG33 block. The outlet neck would fall short of the front of the manifold due to the EG's extended length, but I'd say it would physically bolt up to the case. Neck might actually hit the lower runners, but nothing a bandsaw and tig welder cannot correct There is a difference between the junction of the left/right banks on an EJ and the EG crossover tubes. I'll dig through the scrap pile of blow subaru blocks here at the shop tomorrow and see if I have an x-over tube kicking around to show you.



Have you tested pressure post port yet? Meaning, have you tapped the crossover tube just outside the ports both left and right bank to test your pressure at that point.

I know I'm not a "pump" guru but I do have some experience with CFD and loss coefficients with my company building prototype engine induction and exhaust systems. Thermal dynamics are a bit new/rusty to me though. Definitely keep plugging away at the tests that you are doing. I'd be looking at the differences between this system and the EJ system as if there are any that's where you'll likely find an issue, factory corrective manner, or at least a good lead towards the problem if nothing else.

[Dessertrunner]

Were I am at now requires more thought before the next step:-
From what we have seen we need to do the top pipe mod and at the same time I would like to move the thermostate to here and fit a newer type which shuts the heater pipe bypass when the engine is hot so as to get more coolent through the radiator. As part of this PWR have developed a new SVX radiator for me with 45mm top and bottom pipe which should help flow. The second part is related, now I am getting to the point were I can't pull the engine apart and change pipes all the time or grind part of the block on my road car so I really need a test bed.
What I am thinking of doing is build an electric motor driven water pump/engine/radiator power plant. If we are going to conquer the mountain we need to run this sucker at around 8,000 rpm which means the pump will be pumping 266lpm currently its all turn south at around 5k or 166lpm. Using 3 phase power and a VSD we can speed the pump up and measure the improvments we make as they go. Each improvement can be recorded.
The negitive is we don't have heat as Harvey has mentioned. The flip side of this we are not traget boiling which is a result of pressure restriction in the coolent path. In short if were find all areas that cause a restriction we will fix the boiling beacuse we will get the coolent to the pump at a higher pressure.
So what does everyone think, sound like a plan.
Tony

[Dessertrunner]

Boxersix need a tiny favour I got a feeling that the whole in the top of the block were the top pipe bolts on is a problem as well. Seems the left bank is nearly blocked. Would love it if you could get your head around the differences between the engine there.
Tony

[Trevor]

Originally Posted by Dessertrunner

Were I am at now requires more thought before the next step:-

---------------------
So what does everyone think, sound like a plan.
Tony

It sounds autistic.

[Dessertrunner]

New Thermo style.

[Dessertrunner]

Well you may be closer to the truth then you think, I am dyslexia which is similar. Have been paid million for special project that none others could do, so thanks for the compliment.
Tony

[Trevor]

Originally Posted by Dessertrunner

Well you may be closer to the truth then you think, I am dyslexia which is similar. Have been paid million for special project that none others could do, so thanks for the compliment.
Tony

Thank you Tony.

For some time I have been trying to raise the matter in a kindly manner. You confirm what I suspected. I sincerely hope others will understand. I sent up a high one and you caught the ball. A lot has been covered.

No need for explanation, but this is a huge step forward in respect of possible errors in communication. Harvey has only one arm, I only one eye and a hole in my head. Join the gang.