Electric water pump and power saving?

[turbofiat]

One day I was driving my Spider and when I kicked in the turbo, the car felt like it had more power than usual. About 1/4 mile down the road I noticed my battery light fully lit and the temperature guage about halfway between normal and the red zone. So instantly I new I had lost the fan belt.

Turned out the belt had stretched, turned itself inside out and was still on the pullies! But it had stretched about an inch so I couldn't tighten it up. Luckily I was within walking distance of an autoparts store. I keep a spare in my tool box now.

We all know that dragsters like to run electric water pumps and those tiny alternators/generators to minimize drag on the crankshaft.

The question is how much drag does a waterpump and an alternator create on a Fiat twin cam? Anybody ever dynoed this?

Maybe it was just in my head but it really did feel like a significant increase in power.

I wonder if it would be practical to mate an electrically driven clutch type pulley, either on the crankshaft or waterpump that will disengage the V-belt only under boost? And control it using a pressure switch and a toggle switch so you could override it if nessessary. Like if you were driving across a mountain where the turbo was kicked in most of the time.

This may not be such a good idea on race cars where the turbo is kicked in most of the time but on a street car where the engine only sees boost for a brief period of time then disengaging the waterpump and alternator should not cause any issues. Like with overheating or draining the battery.

The only downside is at night it may cause the headlights to go dim going from 14 to 12 volts but otherwise.

Any of you guys ever thought about this?

[Guy Croft]

So totally unreliable running an electric water pump, forget about it.

If you want more power build it into the engine, this is a race engine website after all. That's what we do here.

GC

[Acki]

So totally unreliable running an electric water pump, forget about it.

Sure? BMW makes it in his new engines. My Prof tells much about this, he's a fan of this!

[Guy Croft]

Well, maybe but that's a top-of-the-range road car and this is a race engine site!

Show me a race proven aftermarket one - that top drivers trust as much as toothed belt/micro-V driven systems and that they don't have to change after every event for fear it packs up halfway thru a rally stage.

And as for more power? Nothing is free where energy use is concerned.

GC

[Mats]

Davies Craig? Prodrive seemed to think they were reliable enough to use them on their F 550 FIA-GT racer.

[Julian]

We ran a Davies-Craig in my Dallara replica and it was quite superb, two seasons without any problems although it did take quite some time to get it to do the job (something to do with my 'race mechanic' fitting it wrong so that it just pumped the water around the engine instead of through the radiator).

What it does offer is optimal cooling at all times rather than when you are sat at optimal revs (when does that happen in a race engine?) and proper engine cooling after the engine is shut down. Add to this the various problems we've had with the original water pump which had all sorts of bad habits and it adds up to a big improvement.

The downside is bigger power draw through the electrical system and another circuit to wire in.

There is always that nagging feeling that it won't last and it is another electrical system to be reliant on when there is more than enough already.

[Guy Croft]

The topic was centred on whether it will give a net power gain and my view is no.
The energy to drive the pump has to come from somewhere. It makes no difference whether you take that power from the engine direct or via the battery. Unless you run the battery flat that energy has to be recovered.

In a GT car I imagine that the cramped and complex architecture of the Ferrari engine installation made an electric pump a desirable option, no more than that.
On full load the pump is going to be running at max power, at part load it will require lower speed - much as it behaves when connected to the crank with a belt.

GC

(sorry, rather badly worded in original)

[Julian]

In terms of gaining power you are absolutely right - there is no gain in peak power to be had except when the engine is cold and the pump is running at minimum (and you shouldn't be loading the engine up under those circumstances anyway).

Under high load conditions it will drain more power than a mechanical pump so one might argue that you actually lose power by using an electrical pump but in terms of cooling efficiency I believe that the electrical system is more effective and as such can do a better job of keeping the engine at optimum temperature. In theory you could balance that increased effectiveness off against a smaller capacity cooling system but personally I would rather go with some redundancy.

If the racing comes down to ultimate power then it has to be a mechanical system but if it is a reliability or budgetary concern I would still go with the electrical pump. Yes they need replacing regularly but the cost of a water pump is rather small compared to the cost of a full rebuild.

[Guy Croft]

Some valid points as usual Julian, and yes, naturally one can make the electric pump load sensitive, but, again, more complexity and more to go wrong, whereas one might argue that that belt driven pumps have served us well for 100 years or so. And whilst they are not load sensitive - the cooling system does have a thermostat and the mean loss of thermal efficiency at part load is very modest compared with other losses in the engine system, most notably the exhaust.

I'm right in the middle of examining an important proposal (for an innovation firm) regarding vehicle radiators (and potential enhancements to them) so forgive me if I am circumspect about changing things that work fantastically well just as they stand.

GC

[Mats]

How does one find out the best RPM to run a mech pump at?

Example, a standard engine has a red line of 6300 but is rarely driven above 4000 (and I bet the engineers had that in mind). The race version of that engine has a red line of 8500 and is driven to that on every gear. Surely the pump (a radial one) will not produce a linear flow depending on RPM and I'm quite sure it will cavitate badly over a certain RPM.

So what to do? Use the speed at max std RPM and match a pulley so it will turn at the same speed at the new redline?

To use an electrical pump here would be so much easier. But there are some other issues to think about such as reliability.
The Davies Craig unit is specified to draw a maximum of 7.5Amps, 14.4 x 7.5 = 108w. Factor in losses in the alternator and it will maybe come to what, 150-160w? Surely a mechanical pump draw more then that? But then again it maybe need to cool more then the electrical unit can handle? Then again the mech pump is maybe overspinning and therefore wasting power...

A friend has seen a 5Hp increase on the crank with an EWP over the std setup mech pump, that sounds very interesting!

[Julian]

Mats - you've made some good points there and both are quite right, to an extent.

Normally you want more waterflow at higher revs as this is where you generate the most thermal waste but when racing the waterflow is less than optimal. More importantly you may exceed the optimal flow rate at those higher revs, this can cause cavitation at the impeller which damages the pump and causes overheating. You can change the size of the pulley to correct the gearing - or you can swap to a system that is not dependent on engine revs to maintain optimal flow.

The point about power draw is significant - the pumps do draw considerable power and in terms of efficiency are not as good as a mechanical pump due to losses converting from electrical energy to mechanical energy but it is nowhere near as bad as you suggest.

Peak power draw occurs only when the pump is really struggling, normally due to a blockage in the system (such as an air lock). Once the water is flowing and at normal pressure the draw is much lower but the overall power loss is going to be worse than just using the mechanical pump. Your friend may have recorded a 5bhp increase at the crank but that will be simply the engine not having to mechanically turn the water pump and the electric replacement not drawing much power during a dyno run. Under race conditions the pump works very hard indeed and that wonderful gain is neutralised.

[Mats]

Thank you for the reply.

I just wonder, if there was a repeted 5hp loss with a mech pump and the EWP according to you draw less then the max load of 160watts as I suggested, where does the extra ~3.5kw go that "neutralise the gain" as you put it?

Logically I naturally understand that the power to drive the pump isn't free just beacuse it's electrical but the above suggests that either the EWP isn't up to the task or the mechanical one is doing a lot more work then needed. Then engine temperature gauge seem to think that the EWP is doing fine so I'm kind of leaning towards the other alternative.

There are other questions too, some people say that the mech pump build pressure against the thermostat and this helps against micro boiling. I think this sounds a bit strange since I imagine that the thermo flows quite well when fully open. Myth?

I love the possibility to run the EWP after the engine is shut off but the mech pump is just so easy and reliable...

[Julian]

I could be very wrong on the mechanical efficiency of the water pump compared to the EWP, in which case the EWP is a far better option.

As for pressure differentials - I don't have an explanation for the exact why but on a mechanical system you do need the restriction of the thermostat plate. The instructions for the EWP recommends against leaving the thermostat in though as it just makes extra work for the pump (I'm not sure if that is the instructions for the controller or not - it has been a while since I looked at them).

Given the experience of my friends who have played around with this sort of thing a lot I would go with a restrictor plate instead as it allows the engine cooling to work more as the manufacturer intended but still allows the EWP to pump around the full circuit.

Much of the "pressure" you are talking about comes from the system being sealed (ignoring the bypass pressure rating on the cap) and the fact it gets hot. The thermostat on most modern cars won't open in a cup of boiling water, you need to get the water under pressure so the effective boiling point is higher. If you really, really want to find out more about this the man you want to talk to is Dr Meir Shillor at Oakland University, Michigan (sorry I don't have any contact details to hand) - he developed a whole set of new "models" for simulating the cooling system as part of his research.

[Mats]

Well yes, naturally the pressure comes from the heat but some also claim that the pressure inside the engine (or rather the pressure between the pum and the thermostat) is greater because the waterpump works against a restriction (the thermostat). I myself think this sounds a bit strange and that's why I asked.

I realize I have taken the subject very off course and I'm sorry about that, just seemed that there was a good discussion going and it was heading my direction anyway. Thanks for all the replys.

My conclusion is that I'll probably make a smaller pulley for the crank and therefore gear down both the alternator and water pump. Robustness is king!

[Colin Haggett]

Would just like to add that one advantage of an electric water pump over the mechanical unit, is that once the engine has been turned off the electric water pump will continue to pump coolant around the engine.