Radiator capacity

[aeronaut]

Anyone know the fluid capacities of the OEM E36 M3 radiator, and the commonly used OEM Z3M radiator?

[pbonsalb]

No, but I don’t think the difference is significant. It’s probably about cup more, maybe 2 at the most. The core is a little thicker but not as tall and I don’t think the side tanks have any more capacity.

[aeronaut]

That sounds about right. I'd love to have actual numbers.
I have a CSF E36 radiator (the cheapo $90 one), and a Nissen Z3M radiator, and they are the same capacity (within a few ounces anyway). Which kinda surprised me.
I haven't run the Nissen yet, but the CSF definitely runs hotter than my previous Behr Z3M radiator.

So, now, with Behr's Z3M NLA, I'm wondering what to do.
And knowing actual capacities would help.

[aeronaut]

Damn I wish I had my old Behr Z3M to measure.
For the record, measured volumes:

CSF 2526, 2.65L
Nissens Z3M, 2.65L (https://www.fcpeuro.com/products/bmw...ns-17112227281)

[pbonsalb]

I have been running a Mishimoto Z3 S54 M radiator for 7-8 years and it is definitely bigger than the Behr Z3 S54 M unit that it replaced. But Mishi gets mixed reviews. I recall another company, maybe Zionsville, that is expensive but gets great reviews, and there are probably a few other expensive alternatives.

[scoobiedoo2029]

i have a oem e36 m3 radiator i just replaced with a mishimoto because i have an ac tig setup to fix it if it leaks. just went with the expansion tank too. and ill say the product is great for its price point. but is not aviation grade parts. zionsville might be but ive never seen their product in person. the mishimoto welds and engineering are exactly what should be expected in the price bracket. its decent but you might do some quality control for them, but hey lifetime warranty;


just layed the radiator flat on the ground, hoses up, plugged the temp switch hole and it took a little over half of a gallon milk jug full of water until it was pouring out. this was without hoses.. the hoses look to be nearly another half gallon. that 2.65L is probably spot on.

[aeronaut]

I had a Mishimoto years ago, for about 5 years. Fitment was messy. But no leaks or other issues.
Zionsville and some others are certainly higher quality, but radiators on track cars are a every-2-year-consumable. And, FCPEuro carries Mishi and CSF, so a replacement would hopefully only take a few days.

But, crap, seriously, 2.6L in the Mishi?
If your Mishi is 2.6L, I'd be disappointed, their web site says 3.13L

[scoobiedoo2029]

I had a Mishimoto years ago, for about 5 years. Fitment was messy. But no leaks or other issues.
Zionsville and some others are certainly higher quality, but radiators on track cars are a every-2-year-consumable. And, FCPEuro carries Mishi and CSF, so a replacement would hopefully only take a few days.

But, crap, seriously, 2.6L in the Mishi?
If your Mishi is 2.6L, I'd be disappointed, their web site says 3.13L

i measured my oem radiator. a .5 liter increase sounds completely correct on the mishimoto given the square sides and thin wall of the aluminum side tanks over the rounded 3/16 thick plastic part. it is only a minor increase in capacity.


if im recalling correctly the z3m radiator was slightly shorter to clear an oil cooler? maybe its capacity isnt much of a change.

[pbonsalb]

Which model Mishi does scoobie have?

[scoobiedoo2029]

BMW E36 M3 S52 3.2L Engine Cooling - MMRADE3692 - Performance Aluminum Radiator (ecstuning.com)

seems fine so far. was miserable to bleed with the mishi expansion tank and dear god the dip stick is booby trapped, i fired it into the ceiling and sprayed coolant everywhere.

[pbonsalb]

Ok. Normal capacity makes sense. I have the x-line, but it did not cost $573 when I bought. https://www.mishimoto.com/bmw-e30-e3...tor-92-99.html

[aeronaut]

i measured my oem radiator. a .5 liter increase sounds completely correct on the mishimoto given the square sides and thin wall of the aluminum side tanks over the rounded 3/16 thick plastic part. it is only a minor increase in capacity.


if im recalling correctly the z3m radiator was slightly shorter to clear an oil cooler? maybe its capacity isnt much of a change.

Yea, the Z3M is shorter.

Here's all I know, my current radiator is a CSF OEM replacement (plastic end caps), that I measured at 2.6L. For whatever reason (volume, design, etc) it's not enough capacity.
0.5L increase is 20%, and that should be enough to stabilize my temperatures.

If the Mishimoto listed volume of 3.13L is wrong and the real volume is close to 2.6L, then hmmmm.

[propcar]

That sounds about right. I'd love to have actual numbers.
I have a CSF E36 radiator (the cheapo $90 one), and a Nissen Z3M radiator, and they are the same capacity (within a few ounces anyway). Which kinda surprised me.
I haven't run the Nissen yet, but the CSF definitely runs hotter than my previous Behr Z3M radiator.

So, now, with Behr's Z3M NLA, I'm wondering what to do.
And knowing actual capacities would help.

I haven't been able to compare capacity... but I installed this one along with a high-performance curved blade spal fan on my M3. I have my old Behr Z3 radiator just in case, but buying this one through FCP does give some piece of mind.

https://www.fcpeuro.com/products/bmw...iator-csf-3054

[aeronaut]

I've got my eyeball on that one. And yea, I plan to buy from FCP Euro which helps limit choices.

That said, there's one Mishi that doesn't come up easily via their search and it's an interesting one.
https://www.mishimoto.com/bmw-e30-e3...tor-92-99.html
The Z3 X-line version has a core thickness of 2.4", and is cheaper than the E36 Xline. (But no advertised volume. Ug)
The E36 X-line version has a core thickness of 2.25"

The Z3 version must be thick enough that it doesn't work on a stock M3. I have no fan, and the shroud could say by by too.

I guess I'll base a decision solely on total core volume, I need to make some more measurements.

[scoobiedoo2029]

id say the trustable numbers of the mishimoto are probably somewhere between the 2.6 an 3.13 numbers, mostly on the high side of that hopefully always over 3L.. probably dependent of what side of the layout lines the fabricator cut the material at.


modifying the side tanks on any aluminum radiator isnt unreasonable im guessing you do not have the throttle body and heater core coolant loop adding volume either. are you running the stewart pump or under/over drive pullys? ive always wondered if the "better flow" isnt moving the coolant through the radiator faster. the positive or negative of such things i dunno.


+1 for reboot engineering radiator stays. %$#! the plastic ones.

[aeronaut]

Ok, made an F-ton of measurements, and here's where I ended up.
Nissens Z3M and CSF M3 OEM are same liquid volume, but Nissens Z3M radiator has 20% more cooling area (fluid tubes that are directly attached to air cooled fins).

I expect the Mishimoto and Nissens performance M3 radiator cores are the same or very similar core design as the Nissens Z3M, but are the height of the OEM M3 radiator, which would give it 9% more cooled volume. Or, 2.9L. If either are actually 3.13L, I'd bet it's in the end tanks.

So at least there is some amount of improved cooling in the Z3M compared to the OEM, so I'm going to install that and run it.

Thanks for all the info and discussion!

- - - Updated - - -

id say the trustable numbers of the mishimoto are probably somewhere between the 2.6 an 3.13 numbers, mostly on the high side of that hopefully always over 3L.. probably dependent of what side of the layout lines the fabricator cut the material at.


modifying the side tanks on any aluminum radiator isnt unreasonable im guessing you do not have the throttle body and heater core coolant loop adding volume either. are you running the stewart pump or under/over drive pullys? ive always wondered if the "better flow" isnt moving the coolant through the radiator faster. the positive or negative of such things i dunno.


+1 for reboot engineering radiator stays. %$#! the plastic ones.

Speed through the heat exchange is secondary to smooth flow, and temperature differential. In other words, at some point, increasing speed won't help cooling, but won't hurt it.

Correct, I have stewart pump, an underdrive pulley, no TB coolant lines, no heater lines.

[RRSperry]

I've had the Mishimoto radiator for 8-10 years with no problems. It fit fine. I'm also running the Stewart pump, lower tstat, and lower fan switch, (no mechanical fan) and supercharged. The car runs at 180F almost always. I've never seen it higher than 205F, and that was sitting in the Dulles parking lot on a 105F day with the AC blasting.

Zionsville is a work of art, but you do have to pay for it.

It's not really a fluid capacity thing, as much as a cooled area thing. 3 rows are (usually) going to be better than 2. I'd just get the cheap Z3M 3 row and call it a day. Especially if you are treating it as a 2 year wear item. (plastic end tanks and necks fail based on age or over tightening the clamps on the neck. At least that's my observation)

[aeronaut]

The cheap Nissens Z3M rad is a 2 row (but row width matters too). But yes, it does have 20% more cooling area than the cheap CSF OEM equivalent.
I'd need detailed measurements on OEM M3 and Z3M radiators to know more.

[pbonsalb]

Would a Zionsville really be good for only 2 years in a track car? What does a racer like ScotchH recommend? Can improving on the factory ducting help?

[RRSperry]

It should last until it doesn’t. But I wouldn’t have a set schedule for replacing it. I’d probably have a old spare just in case. Short of an accident or debris strike, it’s not going to wear out. And you will know if it’s leaking. You’ll hear it,
or see it, or smell it..

[ChuckDizzle]

Would a Zionsville really be good for only 2 years in a track car? What does a racer like ScotchH recommend? Can improving on the factory ducting help?

I think a lot of E36s are missing the factory splash guard which I thought also acted as a duct towards the bottom of the radiator. Without it, there's a gap for air to escape as opposed to flowing through the radiator.

[aeronaut]

Would a Zionsville really be good for only 2 years in a track car? What does a racer like ScotchH recommend? Can improving on the factory ducting help?

Debris is the killer, even with a guard. Unless the guard also blocks a lot of air. l
This is after 3 years. I might have a different approach if not for FCP Euro.

rad.jpg

I found the factory ducting, in good shape, to be pretty decent. Although, I do have a metal under-tray with a radiator "flap".
I've toyed with the idea of blocking off the kidney grills + more custom ducting.