The Quest for 600 bhp - Part 3 - "Cooling System"
Discussion
My first tasks in relation to this project were to try and cover all aspects of the car that would change due to the installation of a supercharger. So I will be trying to cover all these issues BEFORE I fit the blower, not try and cure problems found afterwards. Issues include cooling (engine, oil and charge), fuelling and control, mounting and drive, lubrication, intercooling/chargecooling, induction and exhaust systems.
THE ENGINE COOLING SYSTEM
Following some research into advanced cooling systems in Motorsport applications and further research with a BMW engineer, a vehicle cooling systems competition parts adviser and others, I have decided that to improve both the cooling CAPACITY and the EFFICIENCY of my cooling system, I will be going all electric.
I have decided on an electric water pump delivering 110 litres/min that will be installed in place of the mechanical, belt driven pump, in the lower radiator hose. The stock thermostat is to be removed completely. The stock waterpump housing will be refitted but without the rotor. I will use a shorter drivebelt as the waterpump pulley will also be deleted.
To drive the new pump I will be using a new microprocessor controller which will monitor engine temperature and adjust water pump voltage, and hence flow accordingly to alter the required flow rate. The desired temperature can be set manually in 5 deg C increments from 75C to 95C, higher settings for fuel economy, lower for power. The controller monitors engine temperature from a sensor installed in the thermostat housing, and below 75C does not run the pump to allow rapid warm-up of the engine. From 75C up to the set desired temperature, the controller starts to pulse the pump with 5 volts, ramping it up to continuous battery voltage as the desired temperature is achieved. The controller also has its own self-diagnostics system to trace and warn of any system faults.
Should the engine coolant exceed the preset temperature by 5C,( such as when stationary), the controller will then switch a new electric radiator cooling fan/s on to reduce temp back down to the preset level. I am enquiring whether or not it is possible to anodize my aluminium radiator black to increase efficiency, as it has plastic side tanks and is only a couple of years old, and will get this done when I get the intercooler/chargecooler black anodized.
The reason I am doing it this way is 3-fold. Firstly, I want to increase the capability of my cooling system to handle the additional load I will be putting on it with a supercharger, secondly, although minimal, the deletion of an engine driven pump should release up to 5bhp, not much but every little helps and that is 5 less I have to find elsewhere, and last but not least, I wish to route the pipework from the supercharger to the inlet manifolds from under the radiator, up behind the nosecone, through the chargecoolers then back into the throttle bodies, as I do not want to run the hoses through the front panel cover and over the radiator as in the Dinan turbo cars.
I would dearly love to do a before and after dyno run to measure the gains with the electric pump but I hope you will understand that dyno time is not cheap and this test alone would cost almost as much as the mod so I will have to rely on others dyno results. (BMW themselves saw a 1.5% improvement on an E46 3-Series in power and fuel economy) so my basic estimate is a 1.5% increase on a 380bhp CSi = 5.7bhp.
The theory behind this is that the stock pump delivers sufficient coolant at full load on maximum rpm’s. For the most part, the engine is operating at part throttle where the pump is too big for the task and is consuming unnecessary power and fuel. Additionally, the pump stops running the instant a very hot engine is switched off, whereas the controller runs the pump for up to 2 mins after shut-down to prevent heat soak or until coolant temp drops below 75C. With all new BMW engines now running electric pumps, I am hoping that my theories will prove correct in practice.
Thomas Bruner, Head of Department Simulation and Testing Engine Mechanics and Heat Management at BMW, and the innovator of the production BMW electric water pump states:
“Up to now, a cars water pump has been supplied with constant power by the engine. As the pump needed to maintain sufficient coolant circulation at all times, it would often soak up as much as 2000 Watts of energy. We are now using an electric water pump which works according to need. It requires a maximum of 200 Watts, just 10% of its predecessors energy claim”.
Now whilst BMW’s concern is for fuel economy and efficiency, my goal is to reduce the parasitic losses on the engine, to free up more useable horsepower, and at the same time, reduce the loading on the main drive pulley which will now have the additional task of driving the supercharger.
I have modified the stock waterpump, removing the impellor and blanking off the drive to use a shorter belt to drive the AC compressor only.
There are other modifications done to allow this to the tensioner arrangement but I will detail this in the “supercharger drive” chapter.
8Tech.
THE ENGINE COOLING SYSTEM
Following some research into advanced cooling systems in Motorsport applications and further research with a BMW engineer, a vehicle cooling systems competition parts adviser and others, I have decided that to improve both the cooling CAPACITY and the EFFICIENCY of my cooling system, I will be going all electric.
I have decided on an electric water pump delivering 110 litres/min that will be installed in place of the mechanical, belt driven pump, in the lower radiator hose. The stock thermostat is to be removed completely. The stock waterpump housing will be refitted but without the rotor. I will use a shorter drivebelt as the waterpump pulley will also be deleted.
To drive the new pump I will be using a new microprocessor controller which will monitor engine temperature and adjust water pump voltage, and hence flow accordingly to alter the required flow rate. The desired temperature can be set manually in 5 deg C increments from 75C to 95C, higher settings for fuel economy, lower for power. The controller monitors engine temperature from a sensor installed in the thermostat housing, and below 75C does not run the pump to allow rapid warm-up of the engine. From 75C up to the set desired temperature, the controller starts to pulse the pump with 5 volts, ramping it up to continuous battery voltage as the desired temperature is achieved. The controller also has its own self-diagnostics system to trace and warn of any system faults.
Should the engine coolant exceed the preset temperature by 5C,( such as when stationary), the controller will then switch a new electric radiator cooling fan/s on to reduce temp back down to the preset level. I am enquiring whether or not it is possible to anodize my aluminium radiator black to increase efficiency, as it has plastic side tanks and is only a couple of years old, and will get this done when I get the intercooler/chargecooler black anodized.
The reason I am doing it this way is 3-fold. Firstly, I want to increase the capability of my cooling system to handle the additional load I will be putting on it with a supercharger, secondly, although minimal, the deletion of an engine driven pump should release up to 5bhp, not much but every little helps and that is 5 less I have to find elsewhere, and last but not least, I wish to route the pipework from the supercharger to the inlet manifolds from under the radiator, up behind the nosecone, through the chargecoolers then back into the throttle bodies, as I do not want to run the hoses through the front panel cover and over the radiator as in the Dinan turbo cars.
I would dearly love to do a before and after dyno run to measure the gains with the electric pump but I hope you will understand that dyno time is not cheap and this test alone would cost almost as much as the mod so I will have to rely on others dyno results. (BMW themselves saw a 1.5% improvement on an E46 3-Series in power and fuel economy) so my basic estimate is a 1.5% increase on a 380bhp CSi = 5.7bhp.
The theory behind this is that the stock pump delivers sufficient coolant at full load on maximum rpm’s. For the most part, the engine is operating at part throttle where the pump is too big for the task and is consuming unnecessary power and fuel. Additionally, the pump stops running the instant a very hot engine is switched off, whereas the controller runs the pump for up to 2 mins after shut-down to prevent heat soak or until coolant temp drops below 75C. With all new BMW engines now running electric pumps, I am hoping that my theories will prove correct in practice.
Thomas Bruner, Head of Department Simulation and Testing Engine Mechanics and Heat Management at BMW, and the innovator of the production BMW electric water pump states:
“Up to now, a cars water pump has been supplied with constant power by the engine. As the pump needed to maintain sufficient coolant circulation at all times, it would often soak up as much as 2000 Watts of energy. We are now using an electric water pump which works according to need. It requires a maximum of 200 Watts, just 10% of its predecessors energy claim”.
Now whilst BMW’s concern is for fuel economy and efficiency, my goal is to reduce the parasitic losses on the engine, to free up more useable horsepower, and at the same time, reduce the loading on the main drive pulley which will now have the additional task of driving the supercharger.
I have modified the stock waterpump, removing the impellor and blanking off the drive to use a shorter belt to drive the AC compressor only.
There are other modifications done to allow this to the tensioner arrangement but I will detail this in the “supercharger drive” chapter.
8Tech.
If you're thinking that a black radiator will give better heat transfer, bear in mind that the colour only affects radiated heat, and radiators don't radiate (despite the name!). What you're after is a surface with the best heat transfer efficient. Anodising would prevent corrosion and might be worth doing purely from that point of view (maintaining 'as new' heat transfer for longer) but anodising a finned shape like a radiator matrix would be nigh on impossible; you'd have to be pretty certain you need that extra 0.001% and absolute throwing money at the problem for this to be worth while.
When you're considering water flow rate requirements for mechanical/electric pumps, it's relatively easy to control the average engine temperature but far more important to prevent hot spots forming in the heads. I can't tell you whether that's an issue on that engine, but I think you'd be advised to find out before you do anything that drops the water flow rate.
When you're considering water flow rate requirements for mechanical/electric pumps, it's relatively easy to control the average engine temperature but far more important to prevent hot spots forming in the heads. I can't tell you whether that's an issue on that engine, but I think you'd be advised to find out before you do anything that drops the water flow rate.
What about transission cooling?
I also agree with the 'black rad' comment too, not really relevant if you have sufficient airflow and surface area for heat transfer.
I also don't really understand why you're deleting the pulley from the current water pump if you're removing the internals. Not saying it's wrong, but seems to be adding more than a little complication.
I used a Gates poly V drive belt on my s/c installation. What type are you going for and why?
How are you going to manage inlet temp? I'd have thought this was more of an issue that engine cooling in a high boost/high cube motor?
Would love to be present for the before and after dyno runs too!
I also agree with the 'black rad' comment too, not really relevant if you have sufficient airflow and surface area for heat transfer.
I also don't really understand why you're deleting the pulley from the current water pump if you're removing the internals. Not saying it's wrong, but seems to be adding more than a little complication.
I used a Gates poly V drive belt on my s/c installation. What type are you going for and why?
How are you going to manage inlet temp? I'd have thought this was more of an issue that engine cooling in a high boost/high cube motor?
Would love to be present for the before and after dyno runs too!
You need to give a lot of thought to replacing the mechancal pump with an electric pump:
1. The engines with electric pumps are designed to have as low as possible pressure drop across the complete system -heads, blocks, rads are designed to offer the least resistance possible.
2. You still need X litres/min at full power, whether electric or mechanical pump. You still have parasitic losses through the alternator to deliver the required power for the pump, so no net horsepower gain.
You can still incorporate an electric pump to give additional flow, but I would suggest using it in parallel with the existing pump.
1. The engines with electric pumps are designed to have as low as possible pressure drop across the complete system -heads, blocks, rads are designed to offer the least resistance possible.
2. You still need X litres/min at full power, whether electric or mechanical pump. You still have parasitic losses through the alternator to deliver the required power for the pump, so no net horsepower gain.
You can still incorporate an electric pump to give additional flow, but I would suggest using it in parallel with the existing pump.
StuB said:
What about transission cooling?
I also agree with the 'black rad' comment too, not really relevant if you have sufficient airflow and surface area for heat transfer.
I also don't really understand why you're deleting the pulley from the current water pump if you're removing the internals. Not saying it's wrong, but seems to be adding more than a little complication.
I used a Gates poly V drive belt on my s/c installation. What type are you going for and why?
How are you going to manage inlet temp? I'd have thought this was more of an issue that engine cooling in a high boost/high cube motor?
Would love to be present for the before and after dyno runs too!
I don't think that the transmission will require any additional oil cooling.I also agree with the 'black rad' comment too, not really relevant if you have sufficient airflow and surface area for heat transfer.
I also don't really understand why you're deleting the pulley from the current water pump if you're removing the internals. Not saying it's wrong, but seems to be adding more than a little complication.
I used a Gates poly V drive belt on my s/c installation. What type are you going for and why?
How are you going to manage inlet temp? I'd have thought this was more of an issue that engine cooling in a high boost/high cube motor?
Would love to be present for the before and after dyno runs too!
Deleting the water pump pulley gives more room for the installation of the electric fan assembly and improves the routing of the belt for the SC mounting bracket.
I am using a pair of 4 rib poly V belts to transmit the expected power requirement and to give a "limp home" facility if one belt should fail rather than rely on just a single 8 rib belt. Plus 4 rib belts are more readily available in various lengths.
Chargecooling! To be dealt with in a seperate post.
Good queries, thanks,
8Tech.
BOR said:
You need to give a lot of thought to replacing the mechancal pump with an electric pump:
1. The engines with electric pumps are designed to have as low as possible pressure drop across the complete system -heads, blocks, rads are designed to offer the least resistance possible.
2. You still need X litres/min at full power, whether electric or mechanical pump. You still have parasitic losses through the alternator to deliver the required power for the pump, so no net horsepower gain.
You can still incorporate an electric pump to give additional flow, but I would suggest using it in parallel with the existing pump.
Agreed, but the pump is specifically designed for the installation on a competition motor and has an appropriate flow rate.1. The engines with electric pumps are designed to have as low as possible pressure drop across the complete system -heads, blocks, rads are designed to offer the least resistance possible.
2. You still need X litres/min at full power, whether electric or mechanical pump. You still have parasitic losses through the alternator to deliver the required power for the pump, so no net horsepower gain.
You can still incorporate an electric pump to give additional flow, but I would suggest using it in parallel with the existing pump.
There will be a measureable reduction in parasitic losses because the electric pump will not draw all its power requirements from the alternator at full power. A dyno run on a 5.0 V8 showed a measured 6 bhp net increase with the electric pump.
Yes, you can use BOTH pumps in parallel to increase flow rate but this does not help the other reasons I wish to delete the stock pump and you will always have the pump producing the lowest flow rate working against the other pump.
Excellent points though, thanks.
8Tech.
Gassing Station | BMW General | Top of Page | What's New | My Stuff