I don't know exactly what the K-Tuned part's ID is, but even if it's only 1", I suspect that it still flows more than enough to meet the cooling needs of even a turbocharged engine. Sure, more is better, but a single 1" restriction in the cooling system shouldn't affect cooling for most of us.
Appears to be a -16 AN fitting, so .88in ID. It does look pretty small, but was likely chosen because that's the fitting size the Meziere water pump that they built the system around has.

Couldn't find any real specs on the pump other than a flow rate at a "who knows what" pressure differential.

Screenshot_20220706-142122.png
 
Last edited:
Appears to be a -16 AN fitting, so .88in ID. It does look pretty small, but was likely chosen because that's the fitting size the Meziere water pump that they built the system around has.

Couldn't find any real specs on the pump other than a flow rate at a "who knows what" pressure differential.

Those style pumps are used for everything under the sun - W/M injection, Vaccum pumps, etc., I have a couple of higher output ones I played with for those purposes, however I just don't trust them, there is such a wide range of vague specs for them. I'd feel much happier using a Pierburg.
 
Not too much progress, I've been down the past few days as I had a wisdom tooth removed (so much fun).

Anyway, last night I got the controller to power up for a moment, just to check that the relay setup works as intended. Thankfully all seems well so far. Progress has been slow as my "summer "break" " has gone to crap, but bleh. It'll get done by the end.

The 1 1/2 inch pipe I need to finish up the plumbing is only sold in 10ft increments around here and is very expensive, so thats fun :) I was hoping an exhaust shop around here would have stainless but none of them stock it or have the ability to buy small amounts.
 
Not too much progress, I've been down the past few days as I had a wisdom tooth removed (so much fun).

Anyway, last night I got the controller to power up for a moment, just to check that the relay setup works as intended. Thankfully all seems well so far. Progress has been slow as my "summer "break" " has gone to crap, but bleh. It'll get done by the end.

The 1 1/2 inch pipe I need to finish up the plumbing is only sold in 10ft increments around here and is very expensive, so thats fun :) I was hoping an exhaust shop around here would have stainless but none of them stock it or have the ability to buy small amounts.

The price on SS tubing is through the roof - 1.5" OD SS 304 1/8" wall is almost $100 a foot(!) form online metals.com, where I typically buy my metals. 1/16" wall drops to about 1/2 that.

I went to the local metal recyclers - you can buy SS for scrap value, that's where I got my 1.5" SS tubing for the water pipe, mandrel bend & all, paid about $10 for the piece with a U bend

IMG-6550.jpg
 
Back to the choice of pumps - the Lotus Elise K2o swap - he said this about the 200 vs. 400 series pumps, when I suggested that the 200 might be enough for mid engine/K24, referencing the BMW application;

View attachment 63584

Hmm.....

Well, from a coolant flow standpoint the N52 (CWA200 OE application) and N54 (CWA400 application) are almost identical. Extremely similar open deck block, same coolant outlet housing, shared hoses, similar radiator, same thermostat, and so on, except it has two turbos to feed coolant to as well.

I'm not sure how valid that logic is either. Many, many cars cannot run on a dyno continuously and keep themselves cool without absolutely ridiculous fans. At APR/Dinans main facility, for example, they have to have massive centrifugal blowers with intakes 5 feet in diameter in order to move enough air to keep even stock cars cool on the dyno for any period of time. It sounds like the poster above was simply using the radiator cooling fans, and while the CWA400 may be the difference between being able to get away with that and not, that does not mean it's necessary or optimal.

Imagine using fourty amps (500w) just for your water pump before factoring in the fans, on a measly (lets be honest, in the scope of things) 240hp engine. No wonder it was mentioned as "not being an efficiency/power increase"- that's an issue with the implementation, not the water pump itself.

If you're having to blast coolant through the block and radiator at 100mph in order to keep the engine cool, you have an undersized radiator and are wasting electricity (power) as a coverup. The minimum flow determination should come from little more than making sure your temperature across the engine is not excessive- for maximum efficiency, once that is met (plus a little bit more to help out your radiator), there is no need for more coolant flow. Also, when fitting such a ridiculous pump to a system, you also have to be aware of restrictions on the intake side. With a pump so powerful, it would not be hard to pull enough of a vacuum on the intake side to create significant cavitation- something the CWA will not hold up to.

For reference: the largest pump Davies Craig makes, the EWP150, which is regularly used on Big Block V8s and track cars, flows 60lpm at 6psi pressure differential. Under the same conditions, the CWA200 flows over 120lpm. At flow rates approach 20lpm, the favor begins to shift towards the DC pump, but at flow rates you're more likely to see in the real world, even the little CWA200 blows it out of the water. This isn't to say the EWP150 is bad- it's absolutely not, but just for comparison. It uses 2/3 the power of the CWA200 as well.

Also, those pump curves look...interesting.. but I am not familiar with their specific design traits to know if that looks valid or not.

1657340796083.jpeg

1657340808720.png

Even though the safe answer is to just go with the 400 and set the maximum speed to where it begins to cavitate, I seriously do think the 200 would be more than adequate itself. The 400 is just another example of BMWs tendency to go overboard with Factor of Safety on drivetrain components, just like their engines and transmission (selection).
 
Last edited:
But also, at the same time, I hate to question someone's word without knowing anything about them or their experience. It's possible he's absolutely right that the 200 isn't adequate in a properly designed system....but that would be an interesting result, to say the least.
 
But also, at the same time, I hate to question someone's word without knowing anything about them or their experience. It's possible he's absolutely right that the 200 isn't adequate in a properly designed system....but that would be an interesting result, to say the least.

I would say it is due to his specific requirements - the car is for track/hill climb/ slalom - he says in his (setup) testing the 200 didn't provide sufficient volume at pressure requirements, if I'm reading it correctly:

Screen Shot 2022-07-09 at 8.35.47 AM.png
 
So. I'm back... apologies for spontaneously disappearing. I tend to do that when I get busy as I'm sure y'all are aware by now. Tonight I finally finished the final exam for the class that has consumed my entire summer.


I'm in the same place I was before with the car, pretty much need/need to:

coolant pipes of some type

finish the pump connector (got the parts in from BMW to do this)

finalize the CAD for the water pump delete plate

add an additional coolant temperature sensor (M12) thread and shuffle around the sensors until they're optimally setup.

continue ignoring my cracked snail mount

continue ignoring the multitude of other issues

finish expansion tank bracket and route new hoses


And well, I think that's it. At that point I'll have a running car again that will (hopefully) not overheat. Once that happens I can start working on improving the fan control...perhaps I should fix the floppy rear ball joints first though (not looking forward to that).
 
So here's the current state of the system.

Still need a new pump inlet hose as the fitment, despite being sealed, is pretty ugly. Expansion tank mount needs to be made but fitment of it is fantastic, and actually enables the use of the factory lower expansion tank hose.

The FI temp sensor will be relocated to this hose/coolant path somewhere along the way, and the CWA controller sensor will be put in its place.

PXL_20220809_233803451.jpg


PXL_20220809_233754559.jpg


While the implementation is quite lacking, the packaging and fitment of the parts seems pretty well sorted, compact, and efficient.

Now to make it not look like crap...
 
So here's the current state of the system.

Still need a new pump inlet hose as the fitment, despite being sealed, is pretty ugly. Expansion tank mount needs to be made but fitment of it is fantastic, and actually enables the use of the factory lower expansion tank hose.

The FI temp sensor will be relocated to this hose/coolant path somewhere along the way, and the CWA controller sensor will be put in its place.

View attachment 64710

View attachment 64711

While the implementation is quite lacking, the packaging and fitment of the parts seems pretty well sorted, compact, and efficient.

Now to make it not look like crap...
Have you considered adding an aluminum spacer between the head and the thermostat housing to move the thermostat housing outward and mount your head temperature sensor in that and leave the FI temp sensor in the stock position?

For those using the Tipo head a spacer like this is needed due to the number of fasteners it has, perhaps something similar would work here. You would need longer fasteners for the thermostat housing which is a pretty simple matter.

You can find the spacer here:

 
Last edited:
Have you considered adding an aluminum spacer between the head and the thermostat housing to move the thermostat housing outward and mount your head temperature sensor in that and leave the FI temp sensor in the stock position?

For those using the Tipo head a spacer like this is needed due to the number of fasteners it has, perhaps something similar would work here. You would need longer fasteners for the thermostat housing which is a pretty simple matter.

You can find the spacer here:


I like that idea. That'll probably end up being my long term solution. Once I'm back at school (a few weeks) I'll be able to use the resources there to machine one. May as well make a couple of them at that point.
 
I like that idea. That'll probably end up being my long term solution. Once I'm back at school (a few weeks) I'll be able to use the resources there to machine one. May as well make a couple of them at that point.
It would put the new sensor at the hottest point before hitting or any mixing of coolant in the thermostat housing which is what you need to understand the actual engine coolant temp. The other alternative is to move your new sensor to the gauge sensor position which is in the first half of the head (@ above cylinder 2)
 
It would put the new sensor at the hottest point before hitting or any mixing of coolant in the thermostat housing which is what you need to understand the actual engine coolant temp. The other alternative is to move your new sensor to the gauge sensor position which is in the first half of the head (@ above cylinder 2)

Agreed.

The gauge sensor will be relocated at some point as well- once I get the cooling system working as expected and the controller is mounted on the dash (to display coolant temperature) the factory coolant temperature sender is going to be moved to the oil pan or oil galley to measure &display engine oil temperature.


I'm fairly certain the upper section of the thermostat housing contains only coolant at engine temperature. If I understand the flow path correctly (it's been months since I looked, could be wrong) the hot engine coolant floods that housing and gets distributed from there- via either the bypass or the radiator, so that entire area should be engine temperature along with the small expansion tank line.
 
Last edited:
The sensor with the highest priority of being correct (closest to engine outlet for both response and temperature accuracy), is the one for the CWA controller.

The FI one needs to be roughly right but by no means is response or even high accuracy a big deal, so thankfully there's plenty of options. If it reads a bit cold and runs a little bit richer as a result of that...I think I'd be quite happy 😁

They're also the exact same sensor, which is cool I guess.
 
Last edited:
Little update:

WP delete plate complete and installed! It's almost ready for coolant and priming, as the pump setup is electrically complete :)

PXL_20220816_010646235.jpg


Just in time too, I leave for Fall semester of school tomorrow, so I won't get much (any) time to enjoy the new cooling system for quite a while, unfortunately.

After working on the accessory side of the engine I'm again reminded why I decided to ditch the belt and poorly placed component (and frame) mayhem that is the factory configuration. I'll be very glad if I never have to touch it again, though a properly manufactured and thicker blockoff plate (potentially with a sensor bung for inlet temperature sensing) is likely in its distant future.
 
Last edited:
Little update:

WP delete plate complete and installed! It's almost ready for coolant and priming, as the pump setup is electrically complete :)

View attachment 64945

Just in time too, I leave for Fall semester of school tomorrow, so I won't get much (any) time to enjoy the new cooling system for quite a while, unfortunately.

After working on the accessory side of the engine I'm again reminded why I decided to ditch the belt and poorly placed component (and frame) mayhem that is the factory configuration. I'll be very glad if I never have to touch it again, though a properly manufactured and thicker blockoff plate (potentially with a sensor bung for inlet temperature sensing) is likely in its distant future.
Sullied intake water temperature (heater return)

Does the system controller have an input for this? Or is this just for your own interest?
 
Phew. Took long enough, but the car is pretty much done.

The cooling system is sealed and full of coolant. I ended up using an expansion tank from a 2013 Ford Fusion.


It was around $50 locally, quite a bit cheaper than most other options, has a larger capacity than the Abarth tank, has two inlet hose barbs, and fits very well with the shape and position of the pump.

Something quite nice is that the heater return can run to the expansion tank, rather than feeding directly to the pump inlet. This will result in less flow, likely, but it should plenty adequate still, and it greatly simplifies hose routing. If this doesn't work out how I'd like, I can just reroute it again.

It still needs a few things, but for now the car is back on the ground and parked up until I can get it again later this semester. I was able to get the controller to power on, and navigate through the menus with little issue. Unfortunately I was not able to get the pump to come on, but I later thought it's possible the CWA200 is so quiet, it may not have had any coolant make it to it despite being pretty confident it was submerged by that point. I only consider this because when I manually control the water pump on my 330i with the engine off, all I can hear is the water rushing through the expansion tank - the pump itself is not audible.

I won't be able to drive the car to enjoy the system for quite a while but I'm glad the to do list is now almost gone.
 
Last edited:
Phew. Took long enough, but the car is pretty much done.

The cooling system is sealed and full of coolant. I ended up using an expansion tank from a 2013 Ford Fusion.


It was around $50 locally, quite a bit cheaper than most other options, has a larger capacity than the Abarth tank, has two inlet hose barbs, and fits very well with the shape and position of the pump.

Something quite nice is that the heater return can run to the expansion tank, rather than feeding directly to the pump inlet. This will result in less flow, likely, but it should plenty adequate still, and it greatly simplifies hose routing. If this doesn't work out how I'd like, I can just reroute it again.

It still needs a few things, but for now the car is back on the ground and parked up until I can get it again later this semester. I was able to get the controller to power on, and navigate through the menus with little issue. Unfortunately I was not able to get the pump to come on, but I later thought it's possible the CWA200 is so quiet, it may not have had any coolant make it to it despite being pretty confident it was submerged by that point. I only consider this because when I manually control the water pump on my 330i with the engine off, all I can hear is the water rushing through the expansion tank - the pump itself is not audible.

I won't be able to drive the car to enjoy the system for quite a while but I'm glad the to do list is now almost gone.
I've got a 2013 Fusion Hybrid and about the only failure I have had with it was the expansion tank cap. Apparently they had a bad batch that start out OK but don't maintain their 22 psi pressure very long. I'm assuming that you don't want 22 psi for an X so would likely replace it with something closer to stock pressure.
 
Back
Top