Engine Cooling: Water Pump Flow Data

[NewbVetteGuy]

Is there a standard testing methodology used to test engine water pump flow rates or hp lost for a given flow rate?

I keep reading that electric pumps are tested with essentially zero pressure, but how are mechanical pumps tested?

If the manufacturers each pick and choose how to test their water pumps, is there any good apples-to-apples comparison data out there on water pump flow or hp loss per flow data?


If the typical testing methodologies don't involve actually running the pumps on a running engine, then where can you get flow numbers that are realistic?
(If an engine is tested on an engine dyno with a mechanical water pump, won't the dyno's built-in flow meter show ACTUAL coolant flow rates, or is the dyno still controlling the flow rate in that situation somehow?)




Adam

[Walter R. Malik]

Is there a standard testing methodology used to test engine water pump flow rates or hp lost for a given flow rate?

I keep reading that electric pumps are tested with essentially zero pressure, but how are mechanical pumps tested?

If the manufacturers each pick and choose how to test their water pumps, is there any good apples-to-apples comparison data out there on water pump flow or hp loss per flow data?

If the typical testing methodologies don't involve actually running the pumps on a running engine, then where can you get flow numbers that are realistic?
(If an engine is tested on an engine dyno with a mechanical water pump, won't the dyno's built-in flow meter show ACTUAL coolant flow rates, or is the dyno still controlling the flow rate in that situation somehow?)
Adam

Water pump flow rates mean little if impeller RPM is not given along with what pressure head is restricting that water flow.
Some pumps, (impeller designs), flow best at lower impeller RPM and some flow better at higher RPM; (considering the same pressure head).
Actual pressure within the water jackets will also have a large influence on the cooling.

[fabr]

The charts in this Stewart link may shed some light on how it should be done.

https://www.stewartcomponents.com/index ... tion_id=15

[Tom68]

Is there a standard testing methodology used to test engine water pump flow rates or hp lost for a given flow rate?

I keep reading that electric pumps are tested with essentially zero pressure, but how are mechanical pumps tested?

If the manufacturers each pick and choose how to test their water pumps, is there any good apples-to-apples comparison data out there on water pump flow or hp loss per flow data?


If the typical testing methodologies don't involve actually running the pumps on a running engine, then where can you get flow numbers that are realistic?
(If an engine is tested on an engine dyno with a mechanical water pump, won't the dyno's built-in flow meter show ACTUAL coolant flow rates, or is the dyno still controlling the flow rate in that situation somehow?)




Adam

Flow rate won't be the HP hog, pressure differential will be the greedy one. I like mechanical pumps pushing against a restriction after the heads, reduces steam formation around the chambers, probably no big deal with aluminium heads though.

[Rick!]

The pump curves below ]represent most typical centrifugal pumps.
The curved lines are the flow at an RPM and a specific discharge outlet diameter and then a simple flow restriction (ball valve or high zoot motorized valve) device is used to control flow from no flow to max flow.
The one important line is the net positive suction head required (NPSHR) - it's always positive.
A centrifugal pump always makes its design head and the system governs suction head, assuming no other pressure inputs into the system (combustion leaks, etc.)
I used to do this on HO two strokes at my former employer to add a little science and engineering to designing cooling systems. We used "2 Bar" impellers to generate between 20-30gpm depending on cooling system and head configurations for 180hp/L engines. 2 Bar converts to roughly 69 feet of head.

The Stewart data can be made into interesting surface plots but without knowing head, they are just pretty pictures. One thing you can glean from their data is the power required to create a specific flow volume. Then, knowing how large an electric motor needs to be to achieve the same flow rates, you can see even the best Meziere pumps can only achieve their numbers at very low system restriction values.

[NewbVetteGuy]

Is there a standard testing methodology used to test engine water pump flow rates or hp lost for a given flow rate?

I keep reading that electric pumps are tested with essentially zero pressure, but how are mechanical pumps tested?

If the manufacturers each pick and choose how to test their water pumps, is there any good apples-to-apples comparison data out there on water pump flow or hp loss per flow data?

If the typical testing methodologies don't involve actually running the pumps on a running engine, then where can you get flow numbers that are realistic?
(If an engine is tested on an engine dyno with a mechanical water pump, won't the dyno's built-in flow meter show ACTUAL coolant flow rates, or is the dyno still controlling the flow rate in that situation somehow?)
Adam

Water pump flow rates mean little if impeller RPM is not given along with what pressure head is restricting that water flow.
Some pumps, (impeller designs), flow best at lower impeller RPM and some flow better at higher RPM; (considering the same pressure head).
Actual pressure within the water jackets will also have a large influence on the cooling.

Agreed. The flow rates should all be at a certain head pressure and show the flow at every 500 RPM of impeller RPM.
-I'm more familiar with magnetic drive pumps used in brewing, which are rated at specific head pressures. The mag drive pumps also make it VERY easy to tell when you're cavitating the pump (they let out a HORRIBLE, loud squealing sound); not sure how to tell when an engine / direct drive water pump starts cavitating...


Now where is a good source of getting this objective flow data for common water pumps?
-The pump manufacturers all say things like "Flows 20% more", "30% increased volume" -vs WHAT? At what pressure and at what RPM?



Adam

[NewbVetteGuy]

Some pumps, (impeller designs), flow best at lower impeller RPM and some flow better at higher RPM; (considering the same pressure head).

This is another subject I'm interested in.

What are the common features of "Good, Better, Best" water pumps?

My current rough understanding:

• OEM replacements: stamped impellers with sloppy internal clearances
• Good: Cast impellers with tighter internal clearances, and impeller designs that resist cavitation for longer
• Better: Cast impellers that are precisely CNC machined, or billet CNC machined impeller designs; tighter internal tolerances; available with designs that flow more at lower RPMs and then cavitate at higher RPMs, or with designs that trade off some low RPM flow for more flow and avoid cavitation for higher RPMs -> More flow and more pressure capability
• Best: !??! Advanced impeller designs designed for a specific purpose; tightest tolerances; closely balanced flow between cylinder banks; reduce HP loss at a given flow output vs. lesser pumps

I've been working pretty hard to find information on water pumps and impellers and I'm finding it shockingly difficult compared to most other subjects.
Find anyone with actual test data for different water pumps to enable a comparison is particularly difficult...

Even finding information as simple as "How much does a stock, stamped GM water pump flow" is difficult with lots of HUGELY different reports.



I've found the FlowCooler and PRW PQx water pumps at least specifically advertise that their impellers are designed for lower RPM improvements, with the PRW "high flow" going so far as to say that at 4,000 RPM it has "minimal cavitation".

The PRW claims to flow "35 GPM at 4,000 RPM with minimal cavitation" which actually sounds like a truly terrible flow rate, so I'm guessing that's while on a running engine. -A basic Stewart Stage1, which I think is just a GMB pump with Stewart's impeller and some machining done to it, claims 90 GPM @ 4,000 RPM. I can't see two water pumps using nearly the same basic cases and just different impellers having a difference of 55 GPM between them.

-So where can a person go for some fairly reasonable comparative data?
Right now my approach is: Cooling matters for circle track guys a lot, so what do they use? (Seems to be Stewart, Edelbrock, and Adams that are deemed to be in the top-tier in Circle Track)- but the lack of comparative data or even in-the-open discussions on the features of a "Good, better, best" class pump is something I'm still trying to find and it's not a very fruitful search so far.



Adam

[NewbVetteGuy]

The charts in this Stewart link may shed some light on how it should be done.

https://www.stewartcomponents.com/index ... tion_id=15

I totally agree. I love those. Now where can I go to see test data for OTHER pumps? Even a stock pump?

How does a person move from picking a pump based upon marketing or upon what someone else uses to more objective data?
-Where's "Consumer Reports" for water pumps?

-Would a well instrumented dyno capture actual flow rates and IF (a big if, I understand) the water pump was noted in a NOTE somewhere, be a good, fairly objective source of flow data on an actual engine for actual water pumps?


Adam

[NewbVetteGuy]

The pump curves below ]represent most typical centrifugal pumps.
The curved lines are the flow at an RPM and a specific discharge outlet diameter and then a simple flow restriction (ball valve or high zoot motorized valve) device is used to control flow from no flow to max flow.
The one important line is the net positive suction head required (NPSHR) - it's always positive.
A centrifugal pump always makes its design head and the system governs suction head, assuming no other pressure inputs into the system (combustion leaks, etc.)
I used to do this on HO two strokes at my former employer to add a little science and engineering to designing cooling systems. We used "2 Bar" impellers to generate between 20-30gpm depending on cooling system and head configurations for 180hp/L engines. 2 Bar converts to roughly 69 feet of head.

The Stewart data can be made into interesting surface plots but without knowing head, they are just pretty pictures. One thing you can glean from their data is the power required to create a specific flow volume. Then, knowing how large an electric motor needs to be to achieve the same flow rates, you can see even the best Meziere pumps can only achieve their numbers at very low system restriction values.

The data that Stewart gives out is just mind-boggling (see attached XLS from Stewart when I asked about flow data for their Stage 2 C3 Corvette short pump). -I'm not sure how to go from that data to what the flow would be on an actual engine, though, either. (What's a reasonable pump outlet pressure for a pump running on an aluminum headed SBC at 6,000 RPM?)

I can't even get the most basic of information for most other pumps, so it still makes comparisons essentially impossible.


Adam

[RCJ]

I've used both the Stewart and prw. They both worked fine,they were on different motors but both motors ran at acceptable Temps. The block and heads are probably setting the flow rate just like a cylinder head port.

[fabr]

I've used both the Stewart and prw. They both worked fine,they were on different motors but both motors ran at acceptable Temps. The block and heads are probably setting the flow rate just like a cylinder head port.

and the head gasket,flow losses/restriction in hoses,radiator flow restriction,etc. all add to total head and that is what determines flow rates SFAIK.

[Walter R. Malik]

Some pumps, (impeller designs), flow best at lower impeller RPM and some flow better at higher RPM; (considering the same pressure head).

This is another subject I'm interested in.

I've found the FlowCooler and PRW PQx water pumps at least specifically advertise that their impellers are designed for lower RPM improvements, with the PRW "high flow" going so far as to say that at 4,000 RPM it has "minimal cavitation".

The PRW claims to flow "35 GPM at 4,000 RPM with minimal cavitation" which actually sounds like a truly terrible flow rate, so I'm guessing that's while on a running engine. -A basic Stewart Stage1, which I think is just a GMB pump with Stewart's impeller and some machining done to it, claims 90 GPM @ 4,000 RPM. I can't see two water pumps using nearly the same basic cases and just different impellers having a difference of 55 GPM between them.

-So where can a person go for some fairly reasonable comparative data?
Right now my approach is: Cooling matters for circle track guys a lot, so what do they use? (Seems to be Stewart, Edelbrock, and Adams that are deemed to be in the top-tier in Circle Track)- but the lack of comparative data or even in-the-open discussions on the features of a "Good, better, best" class pump is something I'm still trying to find and it's not a very fruitful search so far.

Adam

Most drag racing, road racing or oval track people are concerned with higher RPM. (over 4,000), so they probably have tried several pumps throughout the years to get one which works for them.
Now, some people will see this and assume that if they work so well for those cases, those pumps should also work in their street rod, cruiser or street/strip car ... without any understanding of why. They are doomed to be disappointed.

One situation is almost nothing like the other so, getting a pump which works best in YOUR situation is not easy without a lot of investigation or actual testing. This can usually get really expensive without that investigation.

[Tom68]

Most drag racing, road racing or oval track people are concerned with higher RPM. (over 4,000), so they probably have tried several pumps throughout the years to get one which works for them.
Now, some people will see this and assume that if they work so well for those cases, those pumps should also work in their street rod, cruiser or street/strip car ... without any understanding of why. They are doomed to be disappointed.

One situation is almost nothing like the other so, getting a pump which works best in YOUR situation is not easy without a lot of investigation or actual testing. This can usually get really expensive without that investigation.

Cutting every second blade off the old SBC cast iron water pump impellers for high speed use dates back to at least the early 60s, probably not great for traffic.

I've always found SBC's factory water pumps need the impeller pressed on further, just like OEM supplied Delco dizzy's need shimming to achieve factory recommended spec's.

[1980RS]

Well I know one thing about a mechanical water pump, my 30% underdrive lower crank pulley is worth 2 tenths over the stock GM one and my 50% is worth a tenth over my 30% pulley. Have not run an electric water pump yet but I have 2 new CSR's and in used Marzire pump to test.

[blown265]

G'day all.

I have a Davies Craig EWP150 40gal/min electric pump installed in the lower radiator hose, adding flow to the factory mechanical pump, on my personal 1971 street/strip car.

On the subject of cavitation, does this electric pusher pump before the OEM mechanical, raise the rpm cavitation threshold?
For example, if the factory (pressed impeller) mechanical item thrashes the water at 5000rpm, does the added EWP150 electric raise this rpm point appreciably 6000rpm?

The reason I ask, is if the above premise holds true, I was considering raising the mechanical pump speed with a smaller pump pulley for more low rpm flow, and then relying on the electric pump to maintain reasonable system flow at redline. (minimal cavitation) The current cooling is good, but I'd like more safety margin for the occasional worst street situations eg, 110 degf summer days on a slow car cruise, or peak hour traffic jam.

I use the EWP150 thermatically switched on at 185 degf, off at 175, and have an indicator light on the A pillar, showing when the electric pump is running. Interestingly, the majority of the electric pumps' use/need is at low rpm, suggesting the factory mechanical is a poor performer at idle and low rpm.

Thanks
Paulie