Electric water pump effect on warm up times

[serialk11r]

Originally Posted by Frank Lee

No way that dinky plastic pulley on my Ford is meant to transfer 10 hp! I can't even imagine a 5 hp engine working hard to spin the pump at up to 7k, tops.

You sure? At 7k, 5hp comes from only 5 Nm of torque. Not a lot of torque, but I don't know the load capacity of belts.

mja1, I'm pretty sure people who race/track will tell you that 10hp is a lot. That aside, the whole point is that an engine doesn't need the amount of coolant flow that the stock pump provides, because the stock pump is supposed to provide adequate flow at idle. Load determines in cylinder temperature, not rpm, so increasing rpm of the engine does not mean the cooling requirements drastically increase. The point is that the pump only needs to turn say 1500rpm while you're rolling down the highway.

You do make a good point about the reliability, but we're talking about efficiency here, and the fact is that the pump on your typical car is spinning way too fast, and that an electric pump provides savings even when driven off the alternator. For racers this matters much more since they run the engine at very high rpm where the pump is cavitating and consuming a sizeable amount of power, but if your engine is to not overheat at 1500rpm, then there isn't any reason for your pump to need to turn much faster than that.

[Tesla]

Just went out to the shed, have an old spa pump motor there, it probably pumps about the same or more than most auto pumps, it is rated at 500w, so about 2/3 of a HP. i think this the ballpark area you would be looking at, there may certainly be transmission losses in the belt, but not to the tune of 5-10HP by my reckoning.
It is a fact that there will be an optimum speed for any pump on any system and you may be able to do get that all the time with an electric pump and save a little bit of energy, but heresay will not cut it, you need to know what the desired flowrate is through the engine to maintain thermal integrity. Coolant temp is fine as an indicator for fan requirements, but not for pump requirements, how do you know if you are circulating the coolant fast enough?

[serialk11r]

Tesla, it's 5-10hp only at high rpm, when you're crusing it's obviously much much lower.

I don't think the problem of determining appropriate flow rate is that hard. Clearly these racing pumps on full power are going to be adequate for maximum load, so if we run a similarly rated electric pump at max speed it should do the job. For part load, we could set maybe 2 speeds that could be exhaust temperature controlled or something like that.

Some of the pumps I've seen are just electric motors that bolt onto the shaft where the pulley used to be, these would be the easiest to use since you can just match the equivalent crank speed to get the correct level of pumping.

I guess a good way to think about it could be, how much is the stock pump flowing at idle, how fast does the engine need to turn to start outflowing the electric pump on full blast. You could do this by just watching the pressure in the system with a pressure gauge attached to a T right after the pump outlet. Then you could have an "idle speed" on the pump that gives you equivalent pressure and therefore flow to the stock pump turning 600 or 800 or 1000rpm whatever your car idles at, and this should be safe for most low load conditions because most cars can cruise with their engines below 2000rpm anyways. Say if electric pump flow matches the stock pump flow at 3000rpm, then you could have a 2000rpm equivalent setting for medium acceleration conditions.

[Tesla]

I agree there probably are some gains to be made, possibly quite large, but that needs to be determined. I suppose the point I'm trying to make is the cost/benefit/risk analysis needs serious attention. Is the benefit large enough to warrant the risk, with the risk being a blown motor if the system is not engineered appropriately.
A lot of 4x4 guys often have an electric pump in addition to the mechanical ones to cover more extreme situations, and through a lot of my research with performance mods, once more power is brought on then upgrades in the cooling system are required, which leads me to believe the cooling system may not be as oversized as many think.
So I suppose the question is: How much is the pump oversized for the job required?

[Frank Lee]

An old school Chevy V8 at 6000 rpm might be losing about 10HP to the water pump. That same engine loses less than 3HP to the pump at 3000 rpm. Where are we usually at- 800-2000 rpm? Educated guess says <2HP for the big Chevys most of the time and then there is my little 4 banger which has a pump that looks about half the size.

I slowed mine down 33% for $20 and INCREASED system simplicity, cost, and reliability over stock. At probably 1HP or less draw stock, I may have saved .25HP? But I slowed the alternator down too, deleted 4 belt driven items, and threw out a nice amount of front end weight.

[serialk11r]

iono, I'd be concerned about inadequate flow with an underdrive pulley, not something I'd risk. A lot of cars aren't geared long enough to stay under 2500 on the highway, and by 3000 I imagine the water pump is sapping power comparable to the alternator, but that's just a wild guess.

As far as reliability goes, dual water pumps running parallel will basically guarantee that you won't be stranded if a pump goes out, but that adds to the cost of course, and the higher quality pumps are pricey to start with. With a daily driver I guess you could argue it's not worth the effort to save say a hundred watts for a fraction of a second under acceleration, but for a hybrid car that gets it electrical energy more efficiently, or a car with a high revving motor, I think it could be a good idea.

[Frank Lee]

We'll see how that slow water pump worked out after August.

[oil pan 4]

Originally Posted by Tesla

how do you know if you are circulating the coolant fast enough?

Coolant the engine is discharging vs. radiator outlet temperature.

Originally Posted by autoteach

for $200 you can buy a pwm to control the speed. I would not use a resistor.

A resistor is a lot less than $200 and cant burn out.
If you spend $200 plus the cost of the pump which is in the $160 to $500 range the part about the mod being economical is long gone.
I am sticking with resistors.

Originally Posted by mechman600

The thermostat is two valves: rad closed/open and bypass open/closed.

Smaller engines don't have this, its called a bypass blocking thermostat. Smaller vehicles have a bypass that flows all the time and a thermostat that opens when the coolant gets hot. When the thermostat is open the bypass coolant is still going.
Even on vehicles with a bypass blocking thermostat they still have another bypass.

[serialk11r]

Originally Posted by oil pan 4

Coolant the engine is discharging vs. radiator outlet temperature.

A resistor is a lot less than $200 and cant burn out.
If you spend $200 plus the cost of the pump which is in the $160 to $500 range the part about the mod being economical is long gone.
I am sticking with resistors.

A PWM controller doesn't cost 200, if you're paying 200 you're doing something wrong. A resistor on something like a pump can easily be burning 10, 20 watts or so, I'd rather spend the extra 10-20 bucks to build a real PWM controller. Having a hot resistor sitting in the engine bay near plastic/rubber components such as the coolant hoses doesn't seem like a good idea.

Google "simple pwm controller" and you'll see what I mean by 10-20 dollars. It's probably even less than that. Actually, if you don't want to build it from scratch, you could just get a single channel PWM controller for computer fans, pull the circuit out and attach a larger transistor, and then a capacitor for power smoothing so you don't accidentally kill the pump just in case it's not meant for PWM. It would run you maybe 15-20 dollars over just simple resistors, and the 10-20W it saves over the resistor would pay for itself in a reasonable amount of time.

Heck, people don't use rheostats to control computer case fans anymore, nevermind something as powerful as a pump suitable for a car engine, because of heat management and the risk of burning out the resistors/transistors/voltage regulators from overheating. I've personally had a resistive fan controller burn out within minutes.

[oil pan 4]

Originally Posted by serialk11r

. Having a hot resistor sitting in the engine bay near plastic/rubber components such as the coolant hoses doesn't seem like a good idea.

My resistor is mounted on the under side of the hood, away from everything. It was free, I salvaved it from a broken battery charger. Its rated to disapate 100w worth of power, it absorbs about 10w and it barely gets warm. I mostly use it during the winter when its freezing out, that resistor stays cold to the touch in winter. Running a 10w resistor would be stupid.

I have already thought of and considered all of this, years ago.
The pump runs at full speed most of the time so there is no point in involving a PWM controller anywhere, at that point it just becomes more crap that can break and cant be bought off the shelf. I try not to custom build everything.

I may wire the resistor though a 120'F thermal switch on the engine with a relay, in the spring it takes about 2 minutes for that switch to open, summer about 1 minute. The biggest thing the resistor does is slow coolant flow to try and warm the engine up faster in winter, other than that it frees up some amps for battery charging and the intake heaters that are some times used on start up on cold days.

The pump and resistor system has been in use for over a year and has 8,000 miles on it and its proven its self. Its staying, with a few minor alterations.