Electric waterpump project
Turbo Mini Forums

lol thats just fucking shoddy!

That impellor is horrific on the leccy pump, surely there are versions with a little more engineering design for efficiency than that, from the picture it looks like it will be a shiny rotating restriction

I was picturing the high capacity pump impellors, there a really rough casted material which can't be good for efficiency, but whether it makes a notable difference is anyones guess!

could you use the mini impeller on the electric pump?

We use pump impellors not dissimilar to the lecy one pictured. They seem to work well. They work fairly well running backwards too, well, better than a closed face curved vane impellor does

Right, I've found some more heat

Problem is the Sim gives a 60 page report and does not sum all the heat loads that could go to coolant.

The latest sim is for 116kW brake power.

Heat from the cylinders - 33.5 kW
Friction Power - 13.2 kW
Exhaust Ports - 8.6 kW

Total - 55.3 kW

Lets say 35 kW at 100hp.

I'll update the calcs when I've worked out how to derive the rad outlet temp using a Log Mean Temperature Difference method.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Paul, seems like our posts crossed.....thanks fo the heads up on the additional heat found!

The fact that this is a small turbocharged engine does make a significant difference.

I have a sim for a standard 998. It give 45hp (33.5 kW) at 4500 rpm which is reasonably accurate.

The total heat loss allowing for cylinders, friction and the exhaust ports is 25.1 kW, ~ 75% of brake power. Add to that the heat loss to the oil, lost through the gearbox case fins, and we are getting closer to the rule of thumb of 100%. So I have confidence that the sim in predicting correctly.

When you turbocharge a high performance version of the same engine and treble the power output as a result of boost and higher revs, a lot of the factors that determine heat transfer to the coolant do not go up proportionately. It still has the same surface area in the ports and combustion chambers where most the heat transfer takes place. Combustion temperature does increase but not hugely. It's the extra pressure that creates the higher power but that does not increase the heat transfer.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

I do love this tread. The topic has interested me for some time.

However. On a previous occasion my input did course some anger towards me as I had only ideas and questions. Not data or hard facts.
This time I have some data to offer. In an earlier tread, I was told that if I ran an electric water pump, I would have to run it at full speed at all times. Nothing else would work. I did find that odd and I still do. If you take a pump large enough for a 3 liter engine then it should be possible, to run it at less than full speed with a 1 liter engine (998ccm).
I have tried to search for the capacity of the standard pump (belt driven). I did not find anything. There may be factory data out there. With the latest tread I got the idea, to test it. So I did.
I took a 998 ccm a+ mini with standard pump with a cast impeller and large pulley. Removed the thermostat and blocked the hose to the heater. Took the top radiator hose off. Fitted 2 long hoses instead on the thermostat housing and radiator with the open ends into a large 100 liter bucket. Filled the bucket with water, started the engine and ventilated the cooling system for air.
Then I set the engine at 2000 rpm. Took the open end of the top hose out and into an empty bucket for 30 seconds.
The result was 10,5 liters. Now, there must be room for inaccuracy. The rpm gauge is the original mini rpm gauge fitted to the car. I took the hose out of the bucket looking at the watch, but there might be up to a second error. There might be other inaccuracies. But end result is 21 liters per minute +/- 1 liter @ 2000 rpm. @ 1000 rpm it would be less. @ 3000 rpm it would be more. But I cannot say it would be 10,5 liters @ 1000 rpm and 31,5 liters @ 3000 rpm. I have not tested it. I have been told that the water pump starts to cavitate at 3000 – 3500 rpm. I conclude it would then pump less.
Now a lot on here have had cooling issues and many have stated that fitting a front mounted radiator solved all problems. That be metro or Cinquecento radiator. I fitted a Polo radiator on my Clubman 15-20 years ago and unless I got stuck in traffic on a day with 30 degree Celsius I had no problems. I didn´t have a fan fitted. 
This is just an idea, but maybe the problem isn´t the waterpump capacity, but radiator capacity. Say if you fit a 20 liter per minute electric pump and a large front mounted radiator instead of an 80 liter per minute pump and a sidemounted radiator.
I started with a Bosch booster pump with a capacity of approx 15 liters per minute, but decided it was not enough. I have now made a setup of 2 pumps mounted parallel. Hopefully I will have it fitted this winter.
Everyone feel free to voice your opinion on my thoughts. Try to keep it in a positive spirit thou. It is hard to take people serious, if they have a negative attitude.

If i have more toys than you when i die, I WIN

That's very interesting. I've calculated around 1.0 litre/second for 100hp and you measure about half that at 3000rpm. It's all pointing in the right direction.

As for radiators, if you look at the calculation I posted then you will see that the temperature drop is proportional to the rad surface area and the flow inversely proportional. So if you double the surface area of the rad, you only need half the water flow.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Turbotom, love your experiment!!! Thanks for sharing this data....


....I can only imagine the look on my girlfriend's face if I were to perform an experiment like this

interesting although most of it goes over my head so i cant add much..
this type of thread/conversation is what makes the forum so grate

In addition to pump capacity and radiator efficiency, the coolant properties (specific heat index and viscosity) itself can also play an important role in cooling performance. For future reference I have taken the liberty of copying an earlier post on "waterless coolant" into this thread....

....okay, I'm to lazy to retype:


<<I have been looking into waterless coolant recently, my curiousity awakened by recent full force marketing presence in all the magazines and even wheeler dealers. Let me start by saying all is not what it seems! Some remarks:

Evans is really good at:

-preventing corrosion
-preventing overboiling (not the same as overheating!) on pre war non pressurised colling systems

Evans is not so good at:

-performance cooling!

Some thoughts on the latter; Water is a funny thing, it freezes at exactly 0C, it boils at exactly 100C, and has a specific heat index of exactly 1.0 . As you can see the behaviour of water inspired the scale of many units in physics. Instead of 1.0 for water, the Evans product being 100% glycol (3% water dilution allowed max according to Evans) has a specific heat index of 0.65 . This means it only has 65% the capacity of water to cool a source, in this case our engine.

But that is not all. Water at 90C has a viscosity of 0.31 cP , a 50/50 water-antifreeze mixture already has double that viscosity at 0.70 cP , the Evans product having a higher viscosity still. Although Evans do not specify their viscosity, emperical testing has shown up to 20% lower coolant flow rates using Evans instead of water. The laws of thermodynamics dictate that when mass coolant flow is lowered, a coolants ability to transfer heat gets reduced as well. ( Q= mass flow x coolant specific heat x delta T )

The above two elements lead to exactly what is seen when using Evans in real world applications.....more heat!!! Up to 25C more heat in fact than when running pure water, but hey Evans only boils at 180C right so running at 115C seems perfectly fine, no???? Indeed no!

As we all know an A-series engine performs best at 75C - 80C coolant temp. Running higher temps seriously reduces power, and increases the chances of detonation. Up to a point where ignition advance needs to be drastically reduced, or octance massively increased. Running a mini at 105C requires 3-5 points more octane compared to 75C.

So what are the drawbacks in using water?: corrosion, insufficient pump lubrication, freezes at 0C

Two out of the above three can be easily cured by well known additives such as Redline Water Wetter, Motul Mocool, Amsoil coolant boost etc. What they do not do is provide antifreeze protection, they are however very good at preventing corrosion and have one more benefit:

They reduce the surface tension of water (droplets on a newly waxed car, a mosquito walking on water). Although this doesn't alter water's heat index of 1.0 it has substantial cooling benefits as it makes it easier for nucleate boiling vapour to enter the liquid coolant flow, reducing the chances of gas pockets and reducing coolant temps with about 4-5C compared to the already excellent performance of straight water.

For a summer only mini this is as ideal as it gets, in winter it means your core plugs pop out at best when not stored inside a heated garage

Water, the marketing sucks, it's totally uncool when sitting on a bar stool, but it's pretty damn good for cooling our minis >>

Edited by Evoderby on 1st Nov, 2013.

My wife knows i do silly things. She has stoped wondering. Or maybe i stoped notishing

On 31st Oct, 2013 Evoderby said:
periment!!!

....I can only imagine the look on my girlfriend's face if I were to perform an experiment like this

If i have more toys than you when i die, I WIN

I belive this may turn into one of thoes custard moments. I going to wait to be suprised, good luck

Drives
-Ford S-max Mk2 Ecoboost
-Rover 100 VVC #2 - track project

Searching is all you need on TurboMinis

A 2-core radiator is really struggling with over 30kW of heat.

I've got datalogs from a 100mph cruise on our test track. Engine sat at around 94 Deg C on a cool day in October 2010 on the way back from the RR day.

Plugging that into the calcs and allowing for a top end performance from the rad, then it just about copes with 35kW. It would not be happy on a hot day.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

I've just thought of something.

The thermostat acts as a flow restrictor to prevent overcooling by having the water travel too fast, which is why the blanking sleeve needs to be put in when it's removed (iirc).

Is it possible that the very small orifice in the thermostat is more of problem with regard to heat dissipation than the size of the rad or the coolant flow ?

It would explain why the 'statless electric pump setup appears to work better.

Metric is for people who can't do fractions.

No Alex, my restrictor ring gives (more or less) the same restriction as an open thermostat

I seriously doubt it!

There seems to be a lot of personal experiences in this topic of of this pump works, this pump doesn't, this rad works, this rad doesn't.

From "my" experience, these statements is not applicable to every engine, and not a rule that can be set in stone.
I believe there is a lot of variation, and dowright wierd stuff happening inside the cooling ways on A+ blocks. I personlly had no issues many years ago, then swapped out the block (for a reason I don't recall). After that, I went from an engine that never reached normal, even on the RR, to an engine that would boil over when at idle.
Nothing apart from piston rings, main bearings, head gasket and various gaskets was any different.
Dry decking helped tremendously, but still didn't get me to where I was on the old block.
One day, I'll borrow a borescope from someone and try to figure out just what is different on the "replacement" block - Or I'll just cut it up as it basically seems to be a boat anchor.

Edited by TurboDave16V on 2nd Nov, 2013.

ive herd of casting that didnt form in the moulds and or never had inhibetor witch caused large amounts of nasty stuff that insulated the part from coolant

A month ago I had not thought of using an electric water pump. Now this thread has prompted me to do this:



Electric motor from the scrap project bin. Need to make an endplate and a pulley but have materials and CNC to hand.

Control from the MS3 though a solid state relay of some kind. Mapped duty cycle vs rpm.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

LOL looks a bit heavy duty Paul, you going to drive the alternator off it too?

The major problem with the a series pump design is that it doesn't have a housing around the impeller to help direct the flow of water.

A pump will only ever be as good as the clearances set around the impeller. You only have to look at a turbo impeller and housing to see how important tight clearances are between the impeller and housing.

The electric pump impeller may not be perfect in design, but matched with a decently designed housing it'll always be better then an a series pump.

https://www.facebook.com/pages/Fusion-Fabri..._homepage_panel

www.fusionfabs.co.uk



1/4mile in 13.2sec @ 111 terminal on 15psi

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."