SL-C electric water pump questions

[John]

I'm considering adding a supplementary electric water pump at the radiator to improve flow.

At this point I intend to leave the mechanical pump on the engine (LS3) and wire the electric pump to come on when the fans come on at 180.

Any thoughts or suggestions are welcome.

I have the upgraded fans and shroud but the car overheats when idling a while after running it hard.

 

[Stephan E.]

Hello John,

Take a look a Davies Craig EWP water pump systems. They go up to 40 Gallons per minute and have a programmable controller. I am just done installing mine on an LS 7 based engine removing the stock pump entirely so I am no sure about supplemental applications. In my case, no more mechanical thermostat and I can determine when the fans and pump to kick in, after run time etc. It also gives you a ton of room behind the bulk head and a very reduced accessory drive. See my posts on that subject.

Good luck

 

[Julian]

Davies Craig make both booster pumps to compliment your mechanical and full electric. They are a pass through design (some of the Meziere offerings are not) so coolant flows via mechanical only when the pump is off.

 

[Howard Jones]

I would recommend deleting the mechanical pump all together. Mount the electric pump back at the motor on the inlet (cold) side, and buy the most pump GPM you can get. Both Davies Craig and Meziere make suitable pumps. The DC system uses a controller and I like that feature.

Be aware that in my case I needed to upgrade my standard radiator. I had one made by C&R.

I am using a DC EWP150, the controller that comes with it, no thermostat, and a C&R radiator. I have paid careful attention to ducting the radiator and confine airflow into the cars nose to going through the radiator and out the top of the car. Otherwise its pretty well sealed up. This has worked very well for me with the original fans that came with my SLC from RCR.

Mount the pump in the location in this video or as close as you can to it.

 

[John]

Good info, thanks guys.

Howard, has removing the thermostat materially affected how long it takes your engine to get up to operating temperature?

I have also ducted the radiator to help with the airflow.

 

[Howard Jones]

No. The electric pump will not cycle on until the controller sees water temp at the target temp + 10% then it turns on the pump and runs it at the speed required to keep it target temp.

It is important to place the sensor for water temp that the controller is referencing in the same place as the thermostat was located, on the top of the engine where the water temp is highest. Drill and tap the thermostat housing and screw the sensor into it if you don't have a threaded bug already in a suitable location.

I played with using a restrictor in place of the thermostat, 3/4 inch diameter hole in a round stainless steel sheet, but I didn't see any difference in how the engine controlled temp. Just remove the thermostat and let the controller do the work.

If you use a Meziere or another electric pump such as Stewart without a controller then you would want to control coolant temp with a thermostat. Call either of them. Stewart was especially helpful as well as DC when I talked to them. A Meziere was used buy the 25 hour blue car with good results. So they all will work well I would think. I just saw DC as more suitable for a dual use car.

Remember this: the controller will only control water temp if it can sense the water volume that it needs to control and that is the engine water temp in the heads, the top of the motor or where the water leaves the engine at its hottest. If you use a thermostat then you will be waiting for the thermostat to open before the controller can begin to regulate the coolant flow if you mount the sensor down stream from the thermostat. Do not position the sensor in any other location in the system such as up near the radiator. The engine will overheat long before it can heat the entire coolant volume to above the target temp.

If you want to talk about this PM me with a Ph # and I will call you. Don't post Ph# here.

 

[Randy V]

I think I would make sure there was coolant flow at all times - even when stone cold.
My concern would be that the coolant surrounding the exhaust valve / port would quickly overheat before any appreciable temperature reading could trigger the flow of the pump. Once you boil the coolant in one area, it is all over as you now introduce bubbling into the system.
I designed the cooling system for my car (now Mitch’s) to use a thermostat with a full flow bypass. Yes, it will take longer to come up to temperature, but it will ensure proper flow through the block and heads at all times.

 

[John]

Randy, agree totally. How did you do your full flow bypass?

Howard, thanks. I appreciate the detail.

 

[Stephan E.]

Hello guys,

I had such concerns about head temperature before I purchased my kit from Davies Craig. In fact the pump is not steady mode during engine warm up and the controller algorithm relative to target temperature is working as follows.

More than 20 degrees (Celsius) under target the pumps cycles every 30 seconds for 10 seconds at half speed.
More than 5 degrees it cycles every 10 seconds for 10 seconds at half speed and within 5 degrees of target it goes into ramp up to operating speed to maintain set temperature.

 

[Randy V]

Randy, agree totally. How did you do your full flow bypass?

http://www.gt-forty.com/GT-Forty/Cooling_Sys.html
The link above takes you to my build site. Mitch has made great progress here and may be able to post up some pictures to support the implementation...

 

[Howard Jones]

Stephan's explanation of the controllers process is more detailed and if you are interested in exactly how it works then I believe there is a explanation on their website. My 10% is what it seams is going on when I drive the car . My target temp is 181F if I remember correctly and if I forget to turn the fans on as I leave the track the water temp goes up to about 200-210 ish as I craw back to my paddock spot. I can hear the pump running at what seams like full on.

However from the moment I start the car when cold the controller very rapidly allows the coolant temp to rise to 181F maybe something like 5 mins from overnight cold. Then thereafter on track the temp never changes during the session even with air temps in the high 90sF. If I turn the fans on when I leave the track coolant temp remains at target until I turn the car off.

Another thing. I have a reason that I think the pump needs to pump directly into the engine inlet. This allows the full pressure that the pump produces to be applied to the engine where the heat is being generated. Mid engine cars with their long and multiple turn coolant pipe runs down the side pods are just about the most effective way to design a system to be the least efficient. If I remember correctly every 90 degree turn is the equivalent of 3-6 feet in straight tubing length depending on tube diameter.

When designing a system try and not make 90 degree turns and instead do two 45s a few inches apart and if a 90 is necessary, make the radius as long as possible depending on how much space is available. If your are working with stainless steel then the welds MUST be purged otherwise every weld will introduce turbulence.

When I was doing the research on coolant systems I talked to the Stewart water pumps tech department and one guy in particular that had worked on a ALMS mid engine prototype. He spent a hour on the phone with me and recommended that I duplicate that system in which they used TWO pumps, one at the engine inlet and one at the radiator inlet, effectively dividing the complete system into two.

He must have said it ten times that the two most important things was placing the pump inlet at a location that had as much of the water volume above it as possible and reducing turbulence in the system, especially in turns once your had the total flow volume/time variable identified. One area that they had made a big improvement was the actual fitting that turned into and out of the engine block.

The other very important thing to be concerned with is the steam vent and air purge system. Both my cars have the back of the heads vented, pipes run back to the thermostat housing and then onto the expansion tank. The radiator also has two vent lines, one at the top of each end tank that also after being combined run back to the expansion tank as well. This system works so well that "burping" the system isn't even a consideration. Just fill it up and run it. Place the expansion tank has high in the car as possible to aid air in the coolant system to rise through the bleed system and end up back in the air pocket in the expansion tank.

 

[Howard Jones]

Stephan has got me to think about it and the pump does comes on at high speed when it is first turned on (wired to run when ignition is on) then cycles on and off until target temp is reached. Once reached, the temp never changes during the session.

 

[Mesa]

Hind sight 20/20 and all that. I would never build my car again with a mechanical pump in place. Electric for sure, along with other attributes, you can replace the pump without pulling the engine (which would/will be required in my case).

 

[Julian]

Howard, why stay with manually switched fans on an automated system? The ideal would be to have a thermostatic switch in the outlet of the radiator for the fans, thus if the radiator becomes overwhelmed and return is still hot due to lack of air flow the fans will switch on.

 

[Howard Jones]

I like complete control over systems in a track car. It aids in troubleshooting and it adds the ability to turn them off if I want them off. Such as low battery, hot engine, few cranks left, need low current draw.

Street car would be better if they ran automatically.

 

[John]

Howard, I agree with everything you say, except possibly about the turbulence.
It is my understanding that fluid cooling systems work better with turbulence as it helps extract heat from the channel wall.

 

[Howard Jones]

The idea in a mid engine car is that everything is working against you, especially with a mechanical pump located in the traditional belt driven location. Most of the water volume is below the pump inlet, the length of the system is very long compared to the intended design (front engine/ radiator) and the mechanical pump impeller speed is being varied all through its efficiency curve including the high speed portion where it more than likely cavitates.

Add pressure losses in the system due to turns in the pipe, welds protruding into the tubing, and restrictions directly in line with the flow direction like thermostat's and include air pockets in the mix and you can see why our cars overheat.

So why not design a system that addresses as many of these issues as possible. Place the pump inlet in a position that provides as much inlet suction head as possible, run the impeller at its most efficient speed and regulate flow rate with a variable duty cycle, make the effective length of the system as short as possible, and reduce restriction to as little as possible. Once you have done that the radiator must be able to reject whatever amount of heat you efficiently deliver to it.

Why worry about turns in the piping and other restriction? Because they all increase the effective length of the system and in turn reduce system pressure which reduces the rate of flow in the system. Heat does get transferred where the coolant contacts the tubing wall but turbulence in the pipe delivery system just isn't an efficiency way to do it. Besides most of us have insulated the piping so it's not leaving the system there in any case.

Heat rejection in the radiator is a function of, WATER FLOW VOLUME THROUGH IT, total surface area of the radiator, heat transfer efficiency of the radiator materials and construction design, and airflow speed thought it. What we want is as much water and air flow through the radiator in a given length of time as possible. Airflow volume is a very big deal and here is where a carefully sealed and ducted radiator on both the inlet and outlet are important. We want as much of the coolant heat transfer to take place in the radiator as possible and not transfer the heat into the engine bay and cockpit.

I guess I am saying every little bit helps.

 

[Neil]

It also helps to put a small bleed line into the top of the radiator. This allows air to be vented out of the radiator instead of gradually allowing the air to displace the coolant.

 

[Scott]

John,

Here's my understanding of turbulence, you either want it or don't want it depending on where you are in the cooling system. You want it where you're trying to reject heat because delta T (temperature) is the most import factor for efficient thermal transfer. The bigger the delta T the faster the transfer. Consider hot coolant flowing along a wall of the radiator. The molecules in contact with the wall (layer 1) are quickly cooled because they have a large delta T with the cooler air on the other side the wall. However, if the flow is laminar the second layer of molecules must transfer their heat through the cooler layer 1 molecules. The delta T between the two layers is lower than the delta T was between the first layer and the air so heat transfer is slower. In addition, layer one has been cooled so its heat transfer to the wall has slowed as well. Extend that through every layer until to reach the center of the flow and you wind up with rapidly diminishing heat transfer.

When you want eat scalding soup you spoon from the edge for a reason. So, you want high flow to create turbulence in the radiator. However, you don't want turbulence around the impellers, in tubes etc. In those places you want laminar flow.

I agree with what has been said above about bleed lines and expansion tanks. In addition, coolant swirl pots are common practice in race cars. Carroll Smith covers them in at least one of his books and depending on your application you may want to consider one.

I'm in the process of replacing my stock LS7 pump with a Pierburg CWA200. Pierburg was the world's first series-production supplier of an electric coolant pump so they've been doing it longer than anyone. The pumps are manufactured in Germany and are OEM equipment for BMW and others. I know of one proven SL-C race team that uses it and it's my understanding that some LMP teams are using it. They are compact, durable and, to my understanding, can be easily controlled via a PWM signal. Mine just showed up today.

 

[Rich Kruger]

Howard I remember situations in the seventies when removing a thermostat would not allow a car to run cooler, hence the coolant went through the radiator so fast the heat transfer was inefficient.

S2 I don't eat scalding hot soup, I give it time to cool