Ridiculous and overthought cooling system modifications (M50 specific)

[Dj Buttchug]

Who here knows more than I know? Probably most of you. I have a stupid idea to potentially solve a problem particularly with turbo charged/ high output/m5x engines. If this topic needs to move to the m5x subforum I shall act accordingly.

The goal of this thread is to discuss options for an ADDITIONAL REAR COOLANT TAKE OFF at the back of 400 casting cylinder heads. All criticism, hatred, and constructive ideas are appreciated.

It has become clear to me that cylinders 5 and 6 often run significantly hotter during pulls/ spirited driving. Due to the temp sensor location (between cyl 1-2) I have found that rear engine cylinders are considerably hotter than the measured temp from this location. IR temp laser confirmed. (A 2nd sensor has not been added to the back of then engine but this is a possibility)

This issued has caused head gasket failures in more extreme conditions. I personally have had 3 head gaskets fail on cylinder 6 throughout my ownership of m50 engines. Others as well have documented similar issues. THERE IS NO DOCUMENTATION I CAN FIND ABOUT AN ADDITIONAL COOLANT PORT MOD TO CORRECT THE HIGHER TEMPS AT THE REAR OF THE ENGINE

Makes sense obviously as we know the cooling system flows from the low side radiator coolant hose up into the block. The block is filled up and then coolant flows into the cylinder head and makes its way back to front and flows into the upper radiator hose. As we all know, there is a rear coolant jacket that goes from the head to the heater core which returns into the spider hose under the intake manifold back into the cooling system jacket fitting on the back of the timing case.

This is the most accurate diagram of the cooling flow I have found. Sensor locations are different along with some other small details but I feel this is the correct depiction of coolant flow direction through these engine variants.


I want to create a rear cylinder head port by removing the freeze plug at back of the cylinder head and pipe it back into the thermostat housing on the HOT side port toward the upper radiator flow path. Here is a quick and dirty sketch of what I am wanting to do.


The plan is to fabricate a fitting to utilize the freeze plug port and run a -6AN line under the intake manifold to a welded bung on the thermostat housing. This will create an additional parallel path for heated coolant to escape and return to the radiator. IN THEORY this design doesn't bypass the thermostat and could help stabilize the super heated cylinders 5-6 helping to keep this area cooler.

I have confirmed with an extra head I have that the freeze plug does block off the coolant jacket inside the cylinder head. The area behind this is rather large and does seem to be a good location for a take off. Ideally this would be done with no permanent modifications to the cylinder head so it could be reversed if it does not work.

Here is a photo of the freeze plug removed at the back of the cylinder head.



So the question remains. WILL THIS ACTUALLY WORK? I want to build it and test on my own but before I commit to the project I figured it would be a good idea to get some feedback. There are many other inline 6 cylinder engines that have this type of coolant port. RB25-26 Nissan engines. 1JZ Toyota engines as well as BMWs S54 and I believe S38 as well.

Thoughts??

[Northern]

My dumb take is that fluid flow increase (and decrease cycle time that coolant spends in the block, Q=mcdeltaT and whatnot) should help reduce the temp delta, so reducing whatever restrictions are there (like adding return flow from that port), higher flow (or more likely an electric) water pump and better thermostat control (via the electric pump) should help?

[TobyB]

The only constructive thought that I have is to make sure you're not creating a loop such that hot water can bypass the radiator.
And to check to make sure (I dunno how) that flows are somewhat balanced.

More temp gauges?

t

[Dj Buttchug]

My dumb take is that fluid flow increase (and decrease cycle time that coolant spends in the block, Q=mcdeltaT and whatnot) should help reduce the temp delta, so reducing whatever restrictions are there (like adding return flow from that port), higher flow (or more likely an electric) water pump and better thermostat control (via the electric pump) should help?

Agreed. Maybe I have thought about it too much and now I am overthinking it.

The only constructive thought that I have is to make sure you're not creating a loop such that hot water can bypass the radiator.
And to check to make sure (I dunno how) that flows are somewhat balanced.

More temp gauges?

t

Add in a secondary temp and monitor it with link g4x to get an idea of how much the added line helps or hinders I suppose.

[R3vTechnic]

To me it looks like that would just cause the coolant to take the path of least resistance straight out the back of the head and bypass the rest of the head cooling passages making your problem worse, but it is difficult to see the exact flow paths from that diagram.

Is the radiator working properly and able to keep up with demand.
Do you have an underdrive pulley on the water pump.
Maybe try a higher flow water pump or "overdrive" pulley (not sure if theses exist).
How are Oil temps.

[Dj Buttchug]

Radiator is oversized Zoinsville Motorsport working properly. Water pump is factory upgraded to a metal impeller.
I do not know why anyone would want to underdrive a water pump...

Oil temps are sub 205 after a good beating. (oil has its own dedicated cooler)

For example. IF operating temp shows 180 degrees from temp sensor in cylinder head then IR temp at rear of cylinder head will read out at 205-210 or higher. Nearly a 30 degree swing.

[McGyver]

Photos aren't loading for me, so I can't see how you designed it.

That being said, this could be a cool thing to 3D print in metal vs casting or machining.

[Dj Buttchug]

Photos aren't loading for me, so I can't see how you designed it.

That being said, this could be a cool thing to 3D print in metal vs casting or machining.

Can others not see the photos as well? The image hosting I am using is somewhat shit.

[4bangerfury]

I'm all for experiments, especially if it can involve metal 3d printing. Been wanting to try it and apparently aluminum isn't that bad in terms of pricing.

Couple thoughts on coolant routing:

• From the flow diagram, coolant flows from the pump to the coolant passage in the cylinder head, distributes coolant to both block and head.
• Looks like that passage goes all the way to the back of the head, right where your freeze plug is. I was able to confirm it on an M42 head, assuming it's close enough.
• Do you think adding a line from freeze plug would just route relatively cold coolant straight back to the pump before it has a chance to cool anything at the rear of the head?

Thoughts on implementing a rear coolant exit:

• Add a restrictor like an orifice or even a throttling valve to balance things out and not starve the front cylinders.
• Has to be low profile since the firewall is right there, a 3d print would be pretty sweet.
• Is there room in the heater core return line to add a tee there for the coolant return instead of going all the to the front of the engine? Utilize the existing coolant return pipe under the manifold.

Slap some TCs across the head and plot that data before and after? Oh yea, that's the stuff that makes the 'tism tingle.

[TobyB]

Pix are visible to me, here.

BMW knew about this by about the M50, and that's why the e- thermostat ran cooler
when the engine was loaded to avoid 'cavitation' (boiling) in the hot spots in the head.

Which kind of suggests that if there had been an easy way to fix the head casting itself,
BMW couldn't find it. Now, the E36 was the beginning of the end for engineering over
sales, so it's certainly not impossible that they COULD have fixed it with more plumbing, like
you're proposing, but Finance nixxed it...

t