Request pros and cons high volume water pump

[Doug Flaten]

Re: I tried to just walk away.......

But I have time on my hands, I agree that time is a factor and your arguments are valid. However, waiting until the two fluids have reached equilibrium will take a long time because the temp differential is continually shrinking. As you point out, the temp at the radiator is likely to increase as the circulation rate increases bringing more heat from the block. The temp differential is analogous to the voltage which provides the driving force. If the radiator temp is hotter and there is sufficient air to remove the heat, then it follows that more heat could be dissipated at a faster rate. Increasing the air flow increases the amount of mass brought into the equation to absorb the heat at a larger temp differential since you are not trying to establish equilibrium across the radiator.

[Everett "Heat Transfer" Ogilvie]

Re: I tried to just walk away.......

Agreed. The "radical" example was just to point out that time is a factor with any parameter that has units of time, ie; a rate. The point is that older, less efficient scaly radiators CAN start to operate in a regime where the flow can contribute to heating problems (we are not talking about scale to the degree where whole rows are clogged, which is a completely different problem). Like everyone here has said, some of these designs were near the limit with brand new components, and when the radiator efficiency is compromised, flow rate can become a factor when the conduction coefficient changes due to the insulative properties of scale.

[Iron Duke NCRS #22045]

Heat exchanger design parameters

If we take the "faster is better" argument to its extreme, coolant moves though the radiator at infinite speed, spending an infinitesimal time in the radiator and transfers an infinitesimal amount of heat. The fact of the matter is that all heat exchangers (the proper name for the misnamed automobile radiator) are designed to operate over a range of temperatures and flow rates. The amount of heat transfered will vary depending on the operating conditions. Standard heat exchanger catalogs are usually accompanied by tables that provide the designer with heat transfer characteristics for varying temperatures and flow rates of both media. Without access to all this data I would be reluctant to redesign the GM engineers' handiwork. I'm not saying that it can't be improved on, but to say that a higher flow coolant pump will improve cooling system performance is ludicrous without a thorough engineering review of all the data, little of which is available to us.

On Special High Performance engines, the coolant pump drive ratios were reduced to slow down the water pump. This compensated for the fact that these engines idled faster and operated at higher peak revs. Slowing the pumps down kept them out of cavitation at high revs, but still provided adequate coolant flow for engine cooling and cockpit heating because of the higher idle speed.

Duke

[Everett "copper big block radiator" Ogil]

Re: Heat exchanger design parameters

Well said. My point all along, since last summer in fact. If we had an open system with unlimited air flow and infinite supply of liquid coolant, more coolant flow would make sense because your input coolant (to the block) would always be ambient temp. In this open system you would not be returning the "non-cooled" fluid back to the motor. Maybe I will hook the garden hose to my water pump inlet for a continuous supply of fresh water, and just go on short drives.

[G B.]

Using numbers

I believe the peak flow of a stock small block water pump is in the 30 to 40 gpm range at 6,000 rpm. No, I do not have a pump curve in front of me. I am remembering figures from past reading. Please just listen to the logic I'm using. If the actual peak turns out to be 25 gpm or 45 gpm or whatever else, you can beat me later.

Since the pump is centrifugal, not positive displacement, I'm guessing the output at idle is probably arount 10 gpm. A high volume pump water pump might increase these figures by 30%. Now , does that mean that 13 gpm or even 50 gpm is "moving too fast to have time to cool"? I don't think so. The rear end ratio of your car can vary the engine speed by 30% or more. I have gone from a 3.08 to a 4.11 without creating overheating or cavitation problems.

In fact, I don't know anybody who has experienced high rpm cavitation problems on a small block Chey engine. In the interest of accuracy I want to stress that I don't know any Bonneville racers or dyno operators. Last night I called the only stock car racer I know well enough to grill about race engine cooling. He runs 15-20 minute heats with his engine speed varying from 3,000 to 7,500. This is harder than I usually run my Corvette on the street, and probably more severe duty than some drag racing engines see for a few seconds over a quarter mile. He uses a high volume pump "because it cools better". His water temperature varies from 180 at the start to around 200 during the race.

So, to those who believe you can move the water through the radiator too fast for it to have time to cool, I say this. Maybe you can by using a tiny pulley and keeping the engine at 9,000 rpm while you drive down the highway. But then again, maybe at the 150 mph you'll be doing, you'll get enough air through your radiator to cool the engine adequately. As for significant cavitation problems (low flow?, impeller erosion?, noise?, pump vibration?), I think the odds of these occurring at a sustained street engine speed with an anti-freeze mix are pretty slim.

I'd love to hear about your driving experiences that illustrate the ludicrousity of my thinking. And if you drive around in a dynomometer, those will do just fine.

[Iron Duke NCRS #22045]

Re: Using numbers

It's ludicrous, Jerry, because you're still just guessing. No one has presented one shread of solid engineering data - just speculation and anecdotes. I wish I had some solid engineering data to present, but I don't have any. Chevrolet Engineering knew the BB cooling system was marginal, especially in the Shark body with early emission controls. If they thought that increasing the water pump speed would aid cooling they surely would have done it as it would have cost virtually nothing.

Duke

[G B.]

I'm curious as well as ludicrous

Tell me some of your anecdotes about Chevy water pump cavitation and overheating due to high water velocity. Sometimes I rely too much on my own experiences with water pumps.

My own '69 427/390 ran cooler during the hot summer months with a Stewart high volume pump. With a stock pump and a new radiator it would hit 210 in stop and go traffic. After installing the Stewart pump, it never went above 190. But this was five years ago, I never flogged the car, it didn't have a.c., and I never timed the idling periods nor noted the ambient temperatures. In fact, I have no records at all for you to study. This is only an anecdote. It proves nothing to anyone except me.

But I digress. What cavitation and high velocity overheating problems did you experience?

[Iron Duke NCRS #22045]

Re: I'm curious as well as ludicrous

I can't answer your question because I have never had any overheating problems on my '63 SHP. If cavitation occurs on a street engine it may go unoticed since street engines typically spend so little time at high revs. Sustained cavitation will cause overheating and erosion inside the cooling system or other obvious damage due to the high pressure gradients.

I've recently been engaged in a dialog with Mike Hall, Cosworth's Chief Development Engineer in the seventies. Specifically we are discussing development of the EA engine, which was based on the Vega aluminum block. According to Mike both the production Vega coolant pump and oil pump cavitated in the 7000 to 9000 RPM rev range of the EA. The coolant pump was modified with a redesigned "low flow" impeller and a larger drive pulley was fitted to slow the rotational speed. The oil pump design was not amenable to easy modification, so it was replaced by an external pump. It's also noteworthy that the production Cosworth Vega pump has a lower specified flow rate compared to the plain vanilla Vega engines and I have verified this with visual inspection of the impellers.

Your anecdotal evidence has more credibility now that you have attached some numbers to the experience, but did you make any other changes after installing the high flow pump? Also, the question remains: If a higher coolant flow rate is the answer to BB cooling problems, why didn't Chevrolet make such a design change? Chevrolet spent millions of dollars a year on sustaining engineering for the Corvette and a redesigned coolant pump would have been a pittance.

Duke

[motorman]

Re: I'm curious as well as ludicrous

i may have posted this before,but i had a overheating problem with a 427 short track engine back in 1966. chevrolet told me to check to see if i had 40+# pressure in the block at the intake outlet to the rad. it did not and that was caused by pump cavatation. removing every other fin from the impeller did the trick. the restrictor is to help build up pressure in the block not cut down on the flow. early alum heads also had a problem of not completely filling with water around the combustion chamber causing detonation in very high compression drag race engines. the cure for this was to tap the water jacket in the valve spring area with 1/8 pipe tap and install fittings with 1/8 dia tubing and run it into the water crossover to let water flow into and out of the water jackets at the highest point.

[Tom B.]

Re: Dang it, Everett....

...or is it "Heat Transfer Ogilvie", or "Big Block Everett", or "Copper Big Block Radiator.....reminds me of the old schmear of "You can call me Ray, or you can call me Jay, or you can call me Ray-Jay, or....What the heck are we supposed to call you, anything but "Mr."?

Anyway, dang, the candle almost burned out on this one! I thought I was left for the wolves. Have you been holding back? I know...you just wanted to stir the pot some more. What a friend we've got in Iron. TBarr #24014

[Doug Flaten]

Re: I'm just guessing........but

I think the limiting factor was the radiator size itself. Again, no data, but the cars were built to reduce the frontal area and probably the radiator size was compromised for the sake of styling, which probably never initially considered plugging a big block into the engine bay. I would hope that an aftermarket pump manufacturer, has properly designed the pump to not cavitate at the higher operating speeds. Assuming that it does not exceed the pressure cap or hose ratings, the additional head created by the pump should also help raise the entire system pressure because it is a closed system and additional pressure on the discharge side will result in some head increase on the suction side. Additional system head on the suction side would decrease the tendency to cavitate. The problem is complicated since the block is trying to boil the fluid anyway and vapor may be making its way to the pump. We get back to the fact that if the radiator is marginal, additional coolant flow may not fix the problem. Maybe faster coolant rate could not correct for an already marginal exchanger. Oh well I've started ramblin' now.

[Iron Duke NCRS #22045]

Pretty good theory

I agree that the BB was not anticipated when the C2 was in the design stages, so when market conditions dicatated a BB they redesigned the core support to stuff as big a radiator into the envelope as would fit. Then Mitchell redesigns the car with a more restrictive inlet and wouldn't compromise on the design, so Zora and his engineers got stuck with a problem, and about the same time they got hit with emission controls which just added injury to insult.

On the second part of your theory, there's a limit as to how fast you can circulate the coolant. The discharge head is dissipated in friction as the coolant circulates and sooner or later the inlet head will be so low that the fluid will shear apart i.e. cavitate, which means no coolant flow.

Although I doubt the efficacy of higher coolant flow rates, it would be interesting to hear others experience with it. Particularly on Sharks, I think resetting the ignition system parameters (initial, centrifugal, and vacuum advance) back to pre emission levels will reduce heat rejection to the jacket enough to provide more cooling margin, and peace of mind. Also, I think the AIR system is a culprit, but I need some more data. Looks like TBarr has volunteered his '68 L-89 as a guinea pig for some of my ideas.

Duke

[Doug Flaten]

Re: I guess the point is proper design

Being a centrifugal pump, the pump develops only the amount of head that the system head requires. As the fluid velocity increases, this increased kinematic head is consumed by friction. The friction does not consume the hydrostatic or applied static head. In the case of the car, the static head is the 13 psig developed by heating the fluid and retaining it in the system with the radiator pressure cap. Since it is a closed loop, this 13 psig is applied to both the suction and discharge side of the pump. Increased fluid velocity will not generate excess friction that will erode this applied head. However, the pump has suction head requirements that vary with the pump speed. If you run the pump out of its design parameters, you may experience cavitation because the pump's suction head requirements are greater than that available from the system. If the fluid is already near its boiling point, there is less net positive suction head available. And a heat exchanger of marginal performance would compound the problem.

However, I think the original question was whether he should run an aftermarket pump to increase flow rate. I would hope that the aftermarket pump was designed properly to operate over the anticipated engine speeds such that it would not cavitate given the available NPSH. He was not talking of speeding up the existing pump. Proper pump design can reduce the amount of suction head required in order to reduce cavitation tendancy. So maybe it wasn't just a simple matter of GM speeding up the existing pump, but a problem of redesigning the pump and then still having to deal with a marginal heat exchanger. And then there is the compromise with the bean counters. Most of the automobile water pump impellars that I have seen are crude stampings with large tolerances. A better pump could be built, but at what price.

This thread has been very thought provoking and enjoyable.

[Chuck S.]

Re: I guess the point is proper design

Doug,

Your nice primer on NPSH requirements took me back to the days of having to deal with mechanical seal failures in single stage hot water pumps; cavitation=vibration=seal failure. Boiler feed water pumps are among the toughest duties, but at least they generally have the pressure rise spread over several stages.

Your observation about the bean-counters is also my opinion; I think the decision to use a stamped steel impeller to reduce cost was made FIRST. THEN, I believe that in development, they found even the little cavitation that existed in normal operation would EAT those stamped steel impellers ALIVE, and they added the "hole" in the back of the impeller to reduce the cavitation. In doing so, they saved a little money, but effectively de-rated the output of the pump from the cast iron impeller version.

Chuck Sangerhausen