Electric water pumps

[Bill Shurvinton]

The only time my car has overheated without a sudden 'shock' was driving up to monserrat. Low speed, steep gradient. 2 tonnes of detroit's finest couldn't cope with this. Totally inconclusive evidence, but it happened.

 

Back to empirical: I checked with Geoff, who checked his notes. On the racetrack (which is not constant load, but is pretty severe), the deltaT is 2 degrees. That is with a peak of 230HPish, which is way above the design spec of the engine.

 

I believe Geoff has some measurements from a sustained WOT run on the dyno at powerstation when he was testing the cooling system efficiency with the EWP. IIRC this was a 10 minute run to check that the EWP really did cool properly. Have your hat ready.

 

I need to check my textbooks, as I was fairly sure that the pressure in the cooling system was mainly caused by the expansion of the water as it warmed. One lives and learns. Of course this is dangerously close to getting to the arguament about pressurised vs non pressurised system and whether it is better to stick with water or go to an NPG system where you have a lower specific heat capacity in your fluid, but a much higher boiling point so do not need to pressurise it to prevent localised nucleic boiling. But that is another discussion for another time :-)

[beaufort]

This seems to solve the pump rates -

Options for pump control:

1. With EWP 'smart' Controller (Part No. 8010), Recommended method. Use the Davies, Craig EWP controller for optimum temperature control. The EWP Controller has a microprocessor, which will supply the pump with the voltage, that will run it at exactly the right flow rate to maintain the set engine temperature. You set the temperature you want on the 'smart' Controller for maximum power and fuel efficiency!

 

With the ignition on, the EWP will run on after a hot engine shut down, eliminating heat soak.

 

This option requires the removal of the thermostat and either the mechanical pump impeller from the pump shaft, or the bypass of the water pump pulley from the belt set-up, using a shorter belt.

 

2.

[Bob Corb]

I still agree with you entirely Bill, we're just coming from two different sides. The overall pressure is developed by the expansion of liquid and compression of the air in the header tank. The cap pressure (~1.5Bar) is maintained in the tank. The bottom hose from the tank feeds to a point close to the pump entry to keep the minimum system pressure up, this helps prevent cavitation. Normally the pump generates a pressure rise of several Bar so that the engine entry is around 4Bar. The boiling point at 4 Bar is very high and this usually prevents boiling. An EWP can only generate about 0.3Bar which will change the boiling point significantly.

 

If you take an engine that produces 40 kW of coolant heat at max power and then change the heat to coolant to 30kW by reducing water flow then the 10kW of heat has to go somewhere else. It usually appears as heat to oil instead or increased heat directly off the engine external walls possibly but the heat will be made and needs to be removed to stop the cylinder head melting. I'm working on several concepts right now and I've got metal temperature measurements here which show how an uprated engine with a slightly larger water pump just manages to maintain acceptable metal temp whereas the stock version (20% less power) is reaching worrying metal temps. Even these engines are using low flow pumps but they are designed with cooling jackets with jets around the exhaust valve/port areas. These engines are the latest cutting edge designs from an OEM who I believe have a very good grasp on cooling jacket design.

 

At the end of the day all I'll say is that you will probably cool your engine fine but dont be surprised if you find cracks or distorted heads at your next rebuild.

 

BC

 

Edited by - Bob Corb on 9 Jan 2004 12:27:36

[old captain slow]

If you do change the pump, what do you do with the mechanical pump and if you remove the impeler, what timing belt do you fit?

 

C7 CDW

[Bob Corb]

I wouldn't touch the timing belt, you may get into serious belt flap issues. I'd leave the water pump there as an idler.

 

A serious electric water pump is in developement and can be seen here:-

 

http://www.kolbenschmidt.de/pdfdoc/elektrische_kuehlmtelpumpen_e.pdf

 

These are apparently going into production in '05.

 

Notice on page two of the .pdf that the anticipated size range for passenger cars is 200 to 750 watts. This is a little larger than Davies Craigs 100W unit...

[beaufort]

I admit I am getting rather bamboozled with a lot of this talk, I checked the site -

 

http://www.daviescraig.com.au/

 

Click on Tech Trade Info button, lots of interesting graphs and figures, still got me convinced that EWP is the way to go.

[Bob Corb]

Yeah it sounds great doesn't it. Even his own measurements show that 80l/min are about right for a 1.3l shopping trolley.

 

He has done no measurements to show that the engine internally is reaching higher temperatures. Who knows, perhaps on that particular low rated Aussie 6 cylinder engine its fine. I'll go and bolt one of his pumps to a Bentley Conti GT. It will solve all the problems overnight. Internal combustion engines will become 30% more efficient instantly and we'll all save £500 a year in fuel costs.

[normalbloke.29]

On the YB in my Cat, I have removed the stat,pulley, and have removed all bearings,shaft,impellor from the waterpump assembly, and am having the holes welded.I maybe pedantic, but didn't like the thought of the old waterpump assembly sat there as a potential leak point.

I will try the pump this summer, and if it proves unsuccessful it will be relatively straighforward to fit a conventional pump and stat back into it.

Bob, once again, many thanks for your input on this, it has been extremely eye opening.

p.s, know anyone with a decent YB 4x4 bare head for sale?

 

[Bill Shurvinton]

But are we talking apples and apples?

 

(Cut to conclusion: If you are not willing to look into this carefully then don't do it)

 

I don't have data on mechanical pumps flow vs pressure vs speed, but am concerned that we are looking at one facet rather than the complete system. In the case of a focus this is fine, as the only thing you are looking at changing is the pump. In the case of a se7en you are looking at a car where the owner can carefully optimise the performance of each part to end up with the compromises that suit their requirements (see conclusion).

 

Take for example the radiator. I know for a fact the geoff tried several radiators until he found one that worked. An awful lot of other se7ens just sling in a Polo unit. Now Polos were never renowned for high output, but the rad has done sterling service with much higher output cars. The point being that you don't always design a part exactly for the envelope.

 

Then you have the other aspects that you have to take into account which can and do affect CHT (which most of us can't even measure). Inlet air temp is one. A se7en with cutouts for throttle bodies will generally see much lower charge temperatures than a regular car. Then you have the effect of over fuelling on Qdot. Now this leads to an interesting potential area of difference between SVAd cars and OEM cars. I have no idea whether modern OEM cats need excess fuelling at WOT, but assuming that they don't and that the emissions guidelines are getting tighter and tighter then this fuelling will be minimised adding to the heat load.

 

Not forgetting that a Focus has to be designed to handle a far wider range of climates than a se7en, whilst driving the aircon, stereo, power steering etc and packaged in a much tighter engine bay.

 

Sometimes I almost wish I had a cylinder head to measure the temperature of :-) Certainly an interesting area to study

 

 

[henk]

Normal bloke , re 4x4 head, i may be able to help, please ring 07931 772 444 regards

[beaufort]

Not too sure about the electric pumps, had a look at the DCr 80lpm, looked a bit like it was out of a Lada or something of simular build quality, I may just wait a while and see what else comes out.

[Bob Corb]

Bill

 

We're still running on the same lines and you have made many interesting points which I dont usually consider in my analyses.

 

Radiators are funny old things, I'm only just getting to grips with them so I wont comment just now.

 

The air temp issue is an interesting one. If you have colder air into the engine then you may increase HT from the inlet port walls but you will have a denser charge so you will make more heat in the combustion chamber (more air, therefore more fuel...) which will be conducted back up the port wall which means you will probably have higher metal temps. Also the exhaust ports will be hotter due to the increased mass flow of gas at ~1000°C.

 

Modern cars need as much if not more overfuelling than ever. At WOT you need to keep the catalyst below 900°C (or a turbo below a 1000°C). At max power I'd expect most NA engines to have over 5%CO in the exhaust. Some turbo units have over 10%! Modern engines are only "clean" at light loads. In general I relate the heat output to the brake power (just as a rule of thumb). If you compare to fuel input then the overfuelling issue can muddy the waters.

 

The main problem with modern engines is keeping the coolant below 120°C and maintaining acceptable component temps under bonnet. I've seen lambda sensors turn to dust (on a car I wont name here) after a hot ambient Vmax run. Cooling is usually a problem simply due to small vents and tightly packed engine bays. This isn't a probelm on a 7 obviously.

 

The metal temp inside the engine is mainly related to the HT to coolant and the power of the engine which as long as coolant feed rate and power output dont change means that metal temp wont change either. I'll check to see if we've ever done metal temp measurements on dyno and in vehicle. If I get any answers I'll let you know.

 

Normal

 

I'm still looking for the YB coolant flow.... I thought I may be able to get you a head but alas no I'm afraid.

 

Bob

 

 

[normalbloke.29]

Thanks Bob, all good stuff...!

[beaufort]

I have been looking around for electric water pumps and have come up with one pushing 132L per minute and a mechanical drive simular to dry sump pump pushing 210L per minute.

The previous debate was based on about 90L per minute, does the extra volume change opinions.

p.s. they are both automotive only pumps.

[CHRIS CLARK]

I remember Jason Krebs saying that he had his EWP elec pump running all the time on his R500.

 

As an aside, I have found that the Laminova device (good buy that Bob!) has stabilised water temps as shown by my limited gauge info. Doesn't suffer the high/low fan kicking on/off syndrome when stuck in traffic. Co-incidence? Maybe.

[steve_m]

Is it relevant that a Se7en engine will run at 75C whereas a normal car engine will run at 90C ? Does the 15C difference give enough extra cooling capacity in the water to counter the reduced flow rate ?

 

Also, where does the Hayabusa fit into this ? It produces 180bhp and has such a small waterpump that it won't drive a very large or twin radiators. I guess the pump must work on lower pressure than a normal car pump and probably pushes through a lot less water than a K or Vx ? They don't seem to have a problem with cooling either (even the turbo engines) so whatever Suzuki do must be sufficient.

 

 

[TorAtle]

There's a lot of good stuff in this thread so I thought I'd bring it back. Reason for doing so is that Davis Craig has a new EWP for 2005 which is rated at 110l/min @ 13.5V. This is 37% up from generation 2 which flowed 80l/min.

 

A simple spreadsheat chart comparison with the standard Rover pump which flows 130l/min @ 7000rpm (assuming half that @ half rpm) makes the Davis EWP look very good. The stock pump will only exceed the EWP flow with an average rpm > 6000. Ok so the EWP may not be able to flow that much in reality, but this may be partly cancelled by the stock pump's possible cavitation > 7000rpm.

 

Any comments?

[V7 SLR]

Is there a price on that new pump or somewhere I can look at it?

[elise_s1]

first of all I think there is a lot of missing infomations on this thread.

All this talk about how many litres a pump is flowing is just misleading.

ALL pumps have a typical pressure/flow curve (which means you cannot claim for a particular flow without giving also the pressure) and this is just at a given RPM.

Centrifugal pumps (as used on cooling circuits) in particular, are known for a marked pressure/flow trade-off.

Assumed an electric driven pump could actually flow 110 l/min on an ideal circuit (no pressure), the flow with some pressure difference will be much less than that. To get some pressure (which is a form of energy) you indeed need another form of energy (unlike flow which just needs energy to balance its losses) and this regardless of pump design.

 

Now let's get K specific: as Bob correctly pointed out, the engine driven pump requires up to 1.7 Kw. Even with a much better design giving more efficiency, to have similar pressure/flow specs you are looking for a pump needing at least 1 Kw (which is 74 A at 13.5V). The electric motor required should be bigger than the starter motor.

 

To be honest it’s not clear to me why someone would look for an electric pump instead of the original mechanical one:

If it’s for the power drained, a better design impeller (the one on the K is just pathetic) will provide.

If it’s for the flow when engine off, a small additional pump (as on the A/C Elise/Exige) will do.

If it’s for the worries about HGF, you need the maximum flow inside the engine regardless of engine temp (to kill the gradients) and this is something you can achieve with a PR thermostat

 

Cheers

Carlo

 

Edited by - elise_s1 on 4 Apr 2005 08:02:03