SVX Engine cooling "Again & Again"

[Trevor]

Originally Posted by Boxersix

Possibly. A test certainly worth trying. But as it's also been said numerous time...modify the cast impeller too....

Not knowing much about impeller CFD, I'm curious as to why the two impeller designs, but more importantly why they are both uses between a mix of cars. One could look at the WRX impeller and come to a conclusion that it may flow less than the cast unit.

I'm looking to get my hands on an EZ30 pump at the moment too(even though I know the EZ coolant configuration is nothing like any other subaru motor i want to see the impeller)

Boxersix,

Looking at your photo, it would appear that the cast impeller is shaped such that it would be more efficient than the pressed design. The difference can only be assumed to be based on economics or the desire to alter the designed performance. The latter would appear to be more logical.

There are two ways of reducing the impeller's efficiency used in days gone by. Those without machine tools simply broke away a portion of every second bucket. If the Subaru casting will not break, the outside corner/edge can be ground off.

You are a relatively late arrival within these forums and I have observed that thankfully you have injected a level of logic previously often absent from many technical discussions. If you will please take the time to backtrack within this tedious rigmarole, you will discover the reasoning behind my proposal which must warrant trial. This crude approach has worked in the past.

[Trevor]

Originally Posted by JONES OFF ROAD

If someone wants to post me a steel impellor, I would be happy to try it for next race in about 3 weeks, I was going to drill hole in the current impellor anyway!
Steve

Steve,

An old mate of mine now dead, always with grease under his fingernails, would have put it this way. ---

“Bugger it mate this is not rocket science, all you have to do is stuff up the pump so that it does not work so good. Drill it if you can, but cast iron usually breaks, so belt it. If you can’t break it, grind it or file it and one sort of way or another stuff up the pumping bit. Don’t worry if it's a bit out off balance, water is damp and will dampen. Don't be so bloody fussy.”

The same George Smith, won many, many races having given up on the fancy modifications and ways of keeping a gearbox from jumping out of second gear. His way, a length of welding wire attached to the dash and wound around the lever as and when required. It never failed him.

Well worth trying an impeller grind and best of luck.

Trevor.

[Boxersix]

Has anyone tested the flow capacity of the stock thermostat?

Just another # I'm looking for here for something I'm making that's all

[Boxersix]

Originally Posted by Trevor

Steve,

An old mate of mine now dead, always with grease under his fingernails, would have put it this way. ---

“Bugger it mate this is not rocket science, all you have to do is stuff up the pump so that it does not work so good. Drill it if you can, but cast iron usually breaks, so belt it. If you can’t break it, grind it or file it and one sort of way or another stuff up the pumping bit. Don’t worry if it's a bit out off balance, water is damp and will dampen. Don't be so bloody fussy.”



Trevor.

This sounds like something you'd see engraved on a plaque in a museum somewhere Logic though is sound.

There are a couple things I'm going to try on a junk case here for S&G's, just have to pop the rotating assembly out of it and gut the valve train from the heads. Also going to hook a complete case/head assembly to the dyno's coolant pump(~500GPM max) and flow the case to it's limits to test balance.

However first, since the front end of my car is torn down to the timing gear I'll drill the impeller a bit and see what it does on the mustang dyno in a couple weeks. I have yet to experience this massive overheating issue

[Trevor]

Originally Posted by Boxersix

However first, since the front end of my car is torn down to the timing gear I'll drill the impeller a bit and see what it does on the mustang dyno in a couple weeks. I have yet to experience this massive overheating issue

A worthwhile experiment, but unless the overheating condition is replicated with a standard pump installed and then the conditions repeated with a modified pump, nothing will be proven. It is impossible to exactly reproduce the conditions relative to the problems reported with an engine set up on a dyno. However this could be useful provided that the overheating condition which occurs at constant high RPM, is first duplicated using a standard pump.

This thread has illustrated a series of unscientific tests and has gone no where for that very reason. In order to obtain a properly definitive result, the modification must be carried out by a user who is experiencing the reported overheating, while repeating exactly the same operating conditions. The exactly same SVX running under the exactly same conditions, driven in exactly the same way and most importantly with no other simultaneous structural alteration.

[Boxersix]

Originally Posted by Trevor

A worthwhile experiment, but unless the overheating condition is replicated with a standard pump installed and then the conditions repeated with a modified pump, nothing will be proven. It is impossible to exactly reproduce the conditions relative to the problems reported with an engine set up on a dyno. However this could be useful provided that the overheating condition which occurs at constant high RPM, is first duplicated using a standard pump.

This thread has illustrated a series of unscientific tests and has gone no where for that very reason. In order to obtain a properly definitive result, the modification must be carried out by a user who is experiencing the reported overheating, while repeating exactly the same operating conditions. The exactly same SVX running under the exactly same conditions, driven in exactly the same way and most importantly with no other simultaneous structural alteration.

Agreed. I simply what to test what a modified pump will do, and possibly obtain some flow data to work with for future tests. You know, "X amount of impeller drilling yields Y amount flow at Z pump RPM"

[Trevor]

Originally Posted by Boxersix

Agreed. I simply what to test what a modified pump will do, and possibly obtain some flow data to work with for future tests. You know, "X amount of impeller drilling yields Y amount flow at Z pump RPM"

Special thanks for again demonstrating that we now have someone posting on the basis of worthwhile, valid, logical common sense, as well as sound technical know how. You are a breath of fresh air.

[SilverSpear]

Originally Posted by Trevor

... You are a breath of fresh air.

Well said Trev.
Adam, Tony and Matt all are

[Dessertrunner]

Just got back and haven't had time to read all the recent postings. Wasn't able to read postings while I was away but Matt was good enough to let me know about Steve's problems. I then spoke to Steve to get my heard around the problem. As a result of this we have developed a plan.

I think just incase people have forgotten Steve uses a multi pass radiator and I have no understanding of the amount of flow this unit can handle. Next week I will install it on my test bench a see what happens at higher revs on the pump. Given that it has the best possiable cooling ability we may be able to do a mod to the pump that will sort the issue out.

FWIW I feel the difference between the two impeeler types is in the amount of suction that they are capable of generating. The pressed metal type is more of a boaster type pump.

Have a great day/night all.
Tony

[Dessertrunner]

Here is an update on the tests run on Steves radiator and why it overheated.

- I mounted it on my test bench which you have all see photos of.
- Installed a bigger electric drive motor on the water pump (3hp) a bigger drive pully. This enabled me to run the water pump up to a engine rev of 8,500rpm.
- As you may remember Steve radiator is a multi pass meaning coolent enter at the top and flows accross to the other side in the top third, then back again in the middle third and finally back through the bottom third. In short the radiator is super effecent at getting rid of the heat but the flow rate is less then a third of the a nornal radiator of the same size.

What I learned,
From about 3,000rpm up the radiator flow rate failed to keep up with the pump as a result the vacuum on the pump kept dropping to -10psi and the inlet to the top of the raditor increased to a point that the pressure was above cap pressure and coolent was forced out the overflow. At 7,000 the whole system was losing coolent and showing signs of losing prime.

One thing I found.

During the test I installed the standard bypass pipes from the top monifold to the side of the pump. I opened and closed then to measure if they had an effect on the pressure. It was clear that they did they dropped the suction by 2 PSI.

Based on this I suggested to Steve that he feeds his back of the head bypass pipes into the side of the water pump. After sleeping on this idea I think we need to find a addiational improvement as this will work on shorter races but not the real long ones.

Please note that these fixes are may not be suitable for a normal car

1. Okay this is were my head is at on this one I want to try a number of things.
2. Install a tap on the top outlet pipe from the engine to "throttle the pump back to match the maxium flow of the radiator.
3. The water pump has a tag that enables it to increase the vacuum pressure it will suck to so by removing the tag I may be able to reduce the peak pressure the pump curve reaches.
4. Develop a simple way for him to test if he will have a problem one that also works with our cars as well.

Thats it for me for now.

[Trevor]

In short the radiator is super effecent at getting rid of the heat but the flow rate is less then a third of the a nornal radiator of the same size.

1. Okay this is were my head is at on this one I want to try a number of things.
2. Install a tap on the top outlet pipe from the engine to "throttle the pump back to match the maxium flow of the radiator.

In view of the previous disparaging remarks recorded against me within this thread, I will not enter any further debate. However I do again point out my continual reference to a probable excessive rate of coolant flow, at above normal engine speed.

[Dessertrunner]

Hi Trevor our problem is coolant flow is heat transfer and as such in our normal race cars we need all the flow we can get. In Steve's case he can get away with less flow.
Tony

[Trevor]

Originally Posted by Dessertrunner

Hi Trevor our problem is coolant flow is heat transfer and as such in our normal race cars we need all the flow we can get. In Steve's case he can get away with less flow.
Tony

I do not intend to continue with an argument. I simply state, it must be taken into account that if the coolant flow becomes excessive (exceeds what would be normal), due to above normal engine speeds, what would otherwise be normal and acceptable flow paths for the coolant, become abnormal unacceptable obstructions. Heat transfer within the engine becomes subject to pressure pockets, transfer within the radiator is reduced because there is less time available, with the coolant passing through at an abnormal speed.(too quickly)

Increasing speed and rate of flow against a given resistance, results in increased pressure. This is a fact which cannot be denied and careful thought is called for.

P.S. Try driving a herd of cattle through a gate, then set the dogs on them and make them gallop and panic. The result will surely be obvious, i.e the pathway and gate no longer function properly.

[Dessertrunner]

Here is a photo of the current rig with a set up to test the pump flow with out a radiator. The aim was to see if the pump and the block generated any flow problems in the system. With this knowledge I could look at the raditor effect. Also I have attached a photo of the tag on the pump I want to mill off to reduce the pump performace.

As this is a "Tony" type scientific test and I can't measure flow I decided to measure amps at the different revs. Then afterward I will have some base line info.

Next test:-
Standard pump but restricted flow on top pipe repeat power measurments.
Milled pump then repeat with open and restricted top pipe.

Tony

[Dessertrunner]

Okay milled the part off the pump (photo attached) then redid the test. I was suprised at the result. I had expected the maxium pressure (out let of pump) to drop greatly but instead it only dropped by 3 psi to 21 psi at 7,500 rpm the amps dropped from 9.8 to 9. It lowered the maxium pressure curve but in my opinion not enough.


This is were my heads at.
I have done trials restricting the flow out of the motor and it doesn't cause the pump to cavitate. Maybe in Steves case we need to restrict the flow from the motor that goes into the radiator to a maxium allowiable. In that way the pump won't cavitate.

So thats it for me for now.
Tony