N54 Cooling - Temp Control Logic and what are your temps during extended track use?

[gmx]

While cold, the PFC08 will still have more initial bite and co-efficient of friction than OEM pads so the bias is if anything more forwards. Just have to ease into and be more careful patient heating up the rotor/pad to optimal temps but also prevent them potentially glazing on the rear.
Didn't have time to do the rear pads whilst finalising, installing the oil cooling solution. I brought the rear pads down with me but had no space for a trolley jack + stands to switch them over. I'm sure I could've borrowed tools at the track to do it but I honestly didn't see the need, car felt quite good. It started to rain just after lunch and I just had fun, practiced some wet lines then headed back home.
I have another one coming up in July where I'll put the rear pads on as well, however this track has a much smoother surface in the braking zones. It's also high speed, vmax around 220km/h on the straight about 25km/h higher average speed overall. Will be dumping the oil and start fresh since I did a lot of laps over the weekend. For me, anyway.

 

[Asbjorn]

My friend send this: "When you have a lot of airflow concerns":

Meanwhile I had this done. Might work, might not, but now the option is there.

I ordered AiM sensors to log temps, but it turns out my AiM Solo 2 DL is not compatible with data expansion and only logs OBD. Trying to get a refund and then figure out another direction.

 

[Davidwarren]

My friend send this: "When you have a lot of airflow concerns":

View attachment 28586

Meanwhile I had this done. Might work, might not, but now the option is there.

View attachment 28587

I ordered AiM sensors to log temps, but it turns out my AiM Solo 2 DL is not compatible with data expansion and only logs OBD. Trying to get a refund and then figure out another direction.

If you have the space, I had great results with brake ducts tied to a NACA duct. I can’t imagine that much air finds its way into a hose without an inlet.

Making Brake Ducts

Making Brake Ducts - Page 2

www.1addicts.com

 

[Asbjorn]

If you have the space, I had great results with brake ducts tied to a NACA duct. I can’t imagine that much air finds its way into a hose without an inlet.

Making Brake Ducts

Making Brake Ducts - Page 2

www.1addicts.com

Couldn't find any naca ducts locally, so we just made some square ducts and called it a day.

Waiting for new flexible ducts to run from the cover to the rotors.

 

[Asbjorn]

Finally got a chance to measure some S55 coolers that had been removed from an M4 after an accident.

Main radiator
490 x 570 x 30mm

Aux radiator
170 x 240 x 30mm

Oil cooler (still mounted on car)
640 x 100 x 20mm (my guestimates)

Conclusions

Total radiator surface area
3,201cm2 for S55
2,175cm2 for N54 (approximately)
2,959cm2 for mine (plus another upto 2,200cm2 from the CSF behind aux radiator, depending on how much the dct core uses)

Total oil cooler surface area
640cm2 for S55
370cm2 for N54 (guesstimate)
522cm2 for mine (plus another 522cm2 now covered behind aux radiator as part of semi-dry oil sump)

So it seems I have enough coolers as my power is not much more than M4. I just need either more air flow or more coolant flow. As soon as I am able to log the radiator outlet coolant temperature I will know which one it is.

 

[zcchen]

Looking forward to @Asbjorn giving us the actual data on radiator performance! Thanks for testing.

Now, I just started to think adding an extra electric water pump (EWP) into the system. It could be a small one in parallel pumping flow for the aux radiator circuit, or a bigger one in series after/before the stock EWP. Can it be done? It seems like Asbjorn already have one added in parallel. So I am really curious whether adding one in series would work in series? Can we stack EWPs in series? If so, is it possible to run only one of the two in the system? will it hurt?

 

[Asbjorn]

Yeah, the radiator outlet temp sensor is being installed tomorrow.

Unfortunately I will have to wait until the weekend before I can make the first log. Im putting in an oil pressure sensor at the same time to get some logs with slicks running the semi-dry oil sump system.

What we seem to know so far is that increasing radiator surface area to the size of an M4 is not enough. So either the coolant flow needs to increase, or air flow needs to increase. Since the M2C has no cooling problems, while the M2 has, I still hope it is "just" the water pump that is the bottleneck. But the logs will offer all the proof needed in one weeks time.

Since both my car, and the M2 has aux water pumps, and both still overheat, I believe we have somewhat confirmed that this is not the final solution. Aux radiator pumps were abandoned entirely on the S55 and S58 which both runs only one mechanical pump instead (ignoring the turbo cooling etc). But what is great about aux pumps is that they can run safely even if the main water pump decides to turn off. The coolant will just circulate in a radiator loop. And it doesnt matter if the aux pump is the same size as the main pump in this setup.

I haven't given much thought to the idea of running two main pumps in series yet. Maybe others have? Anyhow, I want the data to prove if that is the direction actually needed first.
But the first idea that comes to mind is to simply install another OEM 9000l/h pump on the engine outlet side, and then connect it to the same control as the stock location one, while ignoring the self diagnosing feedback connection. In this way the pumps would match each other in speed.
But I have no idea if this possible. It would double the power needed and might trigger something. The benefit of such a setup would be lower coolant pressure differential leading to higher overall flow from the same pumps. The disadvantage would be that the second pump gets very hot coolant, sits in a high location, and has no self-diagnosis for when it stops working.

 

[Asbjorn]

Finally got the radiator outlet temperature sensor working and made a test in +100F ambient. It looks like the water pump is not the bottleneck of my cooling setup as it is now with three auxiliary radiators and booster pump.

This was not the result I was hoping for...

 

[MDORPHN]

Can't back it up with data, but I suspect the problem is inadequate air flow. Accordingly, I'm planning to replace the middle section of the crash beam with tubing and vent the front bumper cover so they'll be far more direct airflow through both the intercooler and radiator.

Neil

 

[Asbjorn]

Can't back it up with data, but I suspect the problem is inadequate air flow. Accordingly, I'm planning to replace the middle section of the crash beam with tubing and vent the front bumper cover so they'll be far more direct airflow through both the intercooler and radiator.

Neil

Since it is not water pump at my power level, it can only be airflow, radiator size or ambient temperature and air density. The last two wont change. Radiator size is ruled out for my setup. It must be a combination of lack of incoming airflow and radiators being too thick (8cm/+3in) everywhere.

Im at a point were I just want to start over and straight up copy the M4 cooling setup. Perhaps running a large slim fmic instead of the bottom config.

 

[Jeffman]

Finally got the radiator outlet temperature sensor working and made a test in +100F ambient. It looks like the water pump is not the bottleneck of my cooling setup as it is now with three auxiliary radiators and booster pump.

This was not the result I was hoping for...

Just so that I understand (without re-reading the entire thread), this was not the result you were hoping for because the temp of the coolant leaving the radiator increased from 93C to 100C in less than two minutes?

If so, please remind me of your (steady state) radiator coolant temp goal and for what time period (15 minutes? 1 hour?) you’d like this temp to be stable.

Thanks.

 

[Asbjorn]

Just so that I understand (without re-reading the entire thread), this was not the result you were hoping for because the temp of the coolant leaving the radiator increased from 93C to 100C in less than two minutes?

If so, please remind me of your (steady state) radiator coolant temp goal and for what time period (15 minutes? 1 hour?) you’d like this temp to be stable.

Thanks.

I was hoping that the radiator delta would increase when pushing the car, ie I was hoping that the N54 water pump was under dimensioned, and that increasing coolant flow would solve the problem of keeping ect below 100C.

 

[Asbjorn]

Since it is not water pump at my power level, it can only be airflow, radiator size or ambient temperature and air density. The last two wont change. Radiator size is ruled out for my setup. It must be a combination of lack of incoming airflow and radiators being too thick (8cm/+3in) everywhere.

Im at a point were I just want to start over and straight up copy the M4 cooling setup. Perhaps running a large slim fmic instead of the bottom config.

So taking inspiration from the S55/S58, and knowing I have either an airflow or cooler size problem, I have been doing some thinking...
People I have spoken to generally agree that increasing water/oil cooler thickness beyond 40mm yields diminishing returns. I assume the same concept is true for FMICs, in the sense that doubling the depth gains less than doubling the frontal area. It is clear when studying the S55 that none of the coolers are very thick, but the frontal area is as large as possible without using angled coolers. Finally low temperature circuits are placed in front of higher temperature ones.

It is difficult to find good FMICs locally in China, but my buddy @CNM135i happens to have an unused N55 Evolution Raceworks Competition FMIC that might very well be up for sale locally

Competition Series Front Mount Intercooler (M135i/M235i/M2) |

With some cutting and welding I should be able to fit this FMIC to my car and rearrange the coolers like this.
Present setup (left) vs proposed setup (right)

With this setup I would throw away the custom non-stepped 7.5HD VRSF FMIC, as well as the large 40mm 1-pass auxiliary radiator, and convert to a thinner, perhaps triple pass radiator, with inlet at the bottom and outlet at the top.

As an alternative, this could also be done with a smaller thicker single pass aux radiator.

Again the new aux radiator would be driven by my 3600l/h booster pump in series with the other auxiliary radiators running AN10 lines. The CSF radiator is still getting the main coolant flow of upto 9000-3600 = 5400l/h through 38mm tubes.

The whole idea behind the above changes would be to reduce the total cooler thickness while increasing airflow and the effective frontal area of both the FMIC and radiators.

Any thoughts?

 

[Jeffman]

I like it, as long as you maintain high air flow with the radiator fan when speeds are low - do you have an upgraded higher power fan as well?

 

[Rob09msport]

Pure stage 2 and ADE650 IC

I have been telling people my iat read lower than my ambient sensor on sunny days . I know it's not possible and is simply due to the sun shining on panel next to temp sensor but the point is it knock vrsf out of park and is bar and plate. I never have heat soak except for first 15 to 30 sec after sitting in traffic or idling in lanes etc. I would rec this fmic over everything else for daily and for track only still would say toss up with wagner.
Anyway I wanted to post referencing the m2 and m2c I would really think the a2w ic has to be helping cool the head by acting like a big heat sink I could be wrong but would seem to be big thing to me and could def explain why it cant keep iat down on track when modified but it's still cool enough to pull heat from block.

 

[Asbjorn]

I like it, as long as you maintain high air flow with the radiator fan when speeds are low - do you have an upgraded higher power fan as well?

No. People have recommended I remove the fan together with the AC to improve airflow on track... I guess Im just not ready for that yet... also I might occasionally do a bit of drifting. Upgraded fans are mainly for street and drag racing. I don't even know how WOT feels in 1st gear, because I just don't have the heart and courage to do so haha

 

[Bnks334]

Since it is not water pump at my power level, it can only be airflow, radiator size or ambient temperature and air density. The last two wont change. Radiator size is ruled out for my setup. It must be a combination of lack of incoming airflow and radiators being too thick (8cm/+3in) everywhere.

Im at a point were I just want to start over and straight up copy the M4 cooling setup. Perhaps running a large slim fmic instead of the bottom config.

Since you have two radiators stacked one in front of the other, which radiator outlet are you measuring from? the the "behind" (CSF) or the one in front? (AUX)"

 

[Asbjorn]

Since you have two radiators stacked one in front of the other, which radiator outlet are you measuring from? the the "behind" (CSF) or the one in front? (AUX)"

Yeah so I mean the csf is around 4cm thick and the aux radiator is 4cm thick bringing the total thickness up to 8cm.

Same for the stacked aux oil and radiator on the driver's side... total of 8cm.

Oh hold on, you mean the temperature measurement. It is taken right after the second T coupler. So Im measuring the temperature of the coolant that has been through all four radiators and before entering the main water pump. So for instance if I am cruising with the aux water pump turned off, and I turn it on, the temperature will drop drastically because all the cold slow moving coolant from the aux radiators suddenly starts circulating faster.

 

[Bnks334]

Yeah so I mean the csf is around 4cm thick and the aux radiator is 4cm thick bringing the total thickness up to 8cm.

Same for the stacked aux oil and radiator on the driver's side... total of 8cm.

Oh hold on, you mean the temperature measurement. It is taken right after the second T coupler. So Im measuring the temperature of the coolant that has been through all four radiators and before entering the main water pump. So for instance if I am cruising with the aux water pump turned off, and I turn it on, the temperature will drop drastically because all the cold slow moving coolant from the aux radiators suddenly starts circulating faster.

I think part of the issue is that your aux radiator is blocking the CSF radiator making it pretty useless on the track. It might work well on the street but on the heat-soaked track the main CSF radiator is just getting blasted with hot air. Did you really see much drop at all when you installed the aux radiator in the front of the car? You would have to see the outlet temps of both the main aux radiator to know if they are both effective on the track.

Stock radiator was single pass. CSF uses fancy "b" shaped tubes, but, is it single pass or triple pass? I am surprise that the radiator outlet temp is only 12f cooler than the temp taken at the cylinder head... That seems to indicate pretty poor radiator performance and not the turbos or other heat risers causing an issue.

are you still running 100% coolant or did you switch to water + water wetter?

Also, can you possibly provide details on how you retrofitted the outlet sensor. i might throw one into my car for additional data.

 

[Asbjorn]

I think part of the issue is that your aux radiator is blocking the CSF radiator making it pretty useless on the track. It might work well on the street but on the heat-soaked track the main CSF radiator is just getting blasted with hot air.

All the changes I have done have been evaluated on the track only, no street testing whatsoever. The thought of trying to exceed the capabilities of even the stock cooling system on the streets in China terrifies me.

But back to your point. Adding an extra 40mm radiator in this case is essentially the same as doubling radiator thickness. What I am told is that given the sizes here, the thicker radiator wont perform worse by itself, but it certainly wont be twice as good either.

Did you really see much drop at all when you installed the aux radiator in the front of the car? You would have to see the outlet temps of both the main aux radiator to know if they are both effective on the track.

Stock radiator was single pass. CSF uses fancy "b" shaped tubes, but, is it single pass or triple pass? I am surprise that the radiator outlet temp is only 12f cooler than the temp taken at the cylinder head... That seems to indicate pretty poor radiator performance and not the turbos or other heat risers causing an issue.

are you still running 100% coolant or did you switch to water + water wetter?

Also, can you possibly provide details on how you retrofitted the outlet sensor. i might throw one into my car for additional data.

Here's where we put the sensor.

Keep in mind that instead of running the much shorter stock connection from the CSF radiator outlet to the thermostat, it now takes a detour around and below the FMIC. It is not pretty, but again, coolant flow, as a potential problem, has now been ruled out.

I run 50/50 bmw coolant btw. Completely stock. No water wetter. No expensive racing r+ coolant. No 100% water. No 100% coolant. Im done with it all.

And yes, I am 100% sure that adding the first large aux radiator in front of the CSF was a small improvement on track. In essence I just doubled the thickness of the radiator and got some extra coolant volume. The extra coolant volume might have helped just by itself. BTW I also found that the AC would work until coolant reached 105C or so on track. This would tell you something about the air temperature betwen the two radiators. I also found the AC to work really poorly in daily driving when enabling MHD sport cooling, as it would run the water pump at a higher speed, even when just cuising. This problem dissappeared when I added the booster pump for the aux radiators, as long as it is turned off. Anyway, what I am not sure about is whether it was better to run those two original radiators in parallel or series with the oem water pump only. I tried both. It was pretty neck and neck. But both setups were an improvment over only one radiator, as the temperature increase would take longer time on track.

Anyway, with the proposal above, I essentially decrease thickness and increase frontal area of both the radiators and the FMIC. The problem of course being the stacked setup. It is probably great if iat stays low... not so good if iat matches the radiator out temperature as it has done recently on track.

Btw I saw previously that moving from 7.5HD stepped to 7in non-stepped did not improve coolant temps much, but had a huge impact on iat.