Anyone running the GDT cylinder 4 cooling kit

[motorbreath]

So I cannot argue with this mod being stock on the JDM, but I assume that the hot coolant is not routed directly to the suction side of the water pump, bypassing the radiator. Does anyone have images or an understanding of the JDM/OEM version of this mod?

 

Regarding lower coolant temperatures being observed after the mod, if the mod is redirecting hot coolant directly into the water pump, bypassing the crossover tube and radiator, are those lower temperatures you observe a result of the thermostat being heated using hot coolant versus cooled coolant exiting the radiator?

 

This would explain the lower temperature observed at the coolant temperature sensor in the crossover tube. If that is true, an alternative option would be to install an aftermarket thermostat.

 

The other question is whether or not cylinder 4 is an issue with our engines since they have EL headers. I'll do some forum digging, but this I am still not sold on this mod, apart from it being found in JDM models. If so, I would want to exactly replicate the JDM mod...

[silverton]

the jdm parts are essentially the same as dom's, except the hoses are formed to go around the steering shaft.

 

cylinder 4 is certainly a problem, ask any NA->T guy, hi that's me, I broke so many ring lands. 95% of them cylinder 4, the other 5% cylinder 2, and cylinder 2 only broke WITH cylinder 4. never messed up 1 or 3. And that was with the parallel fuel mod

[Fabsx]

So I cannot argue with this mod being stock on the JDM, but I assume that the hot coolant is not routed directly to the suction side of the water pump, bypassing the radiator. Does anyone have images or an understanding of the JDM/OEM version of this mod?

 

Regarding lower coolant temperatures being observed after the mod, if the mod is redirecting hot coolant directly into the water pump, bypassing the crossover tube and radiator, are those lower temperatures you observe a result of the thermostat being heated using hot coolant versus cooled coolant exiting the radiator?

 

This would explain the lower temperature observed at the coolant temperature sensor in the crossover tube. If that is true, an alternative option would be to install an aftermarket thermostat.

 

The other question is whether or not cylinder 4 is an issue with our engines since they have EL headers. I'll do some forum digging, but this I am still not sold on this mod, apart from it being found in JDM models. If so, I would want to exactly replicate the JDM mod...

 

 

Forgot to mentioned that my “observations” are from my external coolant gauge. Not using or looking at the factory oem one. Therefore I’m sure that my readings are indeed from the mod itself.

 

Also, I do already have a oem lower thermostat that has been on for years before doing the cyl4 mod.

 

 

Sent from my iPhone using Tapatalk

Edited March 29, 2022 by Fabsx

[motorbreath]

Ok, thanks for taking the time to answer all of my questions! I guess the only issue I have now is that I have no more excuses to not do this mod...

[Clubchico15]

I see people keep on mentioning the "JDM OEM Coolant mod fitting" in this thread. I looked at a JDM motor at a engine retailer that had a Auto tranny mated to it, that fitting was feeding a tranny cooler that is attached on the right side of the motor. Same line then T's off to the heater core. Is it a 4th cyl cooling mod, or just a coolant line that goes to JDM spec'd Auto tranny vehicles... just my .02 cents on this discussion.

[rhino6303]

I see people keep on mentioning the "JDM OEM Coolant mod fitting" in this thread. I looked at a JDM motor at a engine retailer that had a Auto tranny mated to it, that fitting was feeding a tranny cooler that is attached on the right side of the motor. Same line then T's off to the heater core. Is it a 4th cyl cooling mod, or just a coolant line that goes to JDM spec'd Auto tranny vehicles... just my .02 cents on this discussion.

It is just a line to the transmission cooler. It was done this way because it's a direct line to it. It just so happens that a happy result is that the cylinders run cooler with the additional flow through both heads (cylinder 3 line and cylinder 4 line).

[silverton]

long way to travel from the left side of the engine to the right side for a rubber hose..... the hoses that were on the motor I got definitely went to the heater core, no other routing. and it's not a mod if it's oem. I wish I could find the pictures I took of it as delivered but I can not find them.

[Clubchico15]

long way to travel from the left side of the engine to the right side for a rubber hose..... the hoses that were on the motor I got definitely went to the heater core, no other routing. and it's not a mod if it's oem. I wish I could find the pictures I took of it as delivered but I can not find them.

 

Motor below shows the routing exactly like what i saw at the local engine importer.

 

https://www.ebay.com/itm/274540798769?_trkparms=amclksrc%3DITM%26aid%3D111001%26algo%3DREC.SEED%26ao%3D1%26asc%3D20180816085401%26meid%3Ddecaf558c3454591be963c29cea74307%26pid%3D100970%26rk%3D4%26rkt%3D15%26sd%3D194730449045%26itm%3D274540798769%26pmt%3D0%26noa%3D1%26pg%3D2380057%26brand%3DSubaru&_trksid=p2380057.c100970.m5481&_trkparms=pageci%3A1716fbec-b0ee-11ec-8c35-3a62354c3424%7Cparentrq%3Adff2fbba17f0a9bd517b2ca0fffd8c50%7Ciid%3A1

[motorbreath]

After viewing those images I am even more confused! Where is the return line for the transmission cooler? It looks like two different fittings are present on the cylinder 4 side, and both of them seem to lead to the heater core, with one of them teeing to continue on to transmission cooler. That seems to support that that the two fittings on the cylinder 4 side are both the supply and return?!

[Clubchico15]

After viewing those images I am even more confused! Where is the return line for the transmission cooler? It looks like two different fittings are present on the cylinder 4 side, and both of them seem to lead to the heater core, with one of them teeing to continue on to transmission cooler. That seems to support that that the two fittings on the cylinder 4 side are both the supply and return?!

 

I'm assuming that behind cly 3 there is another fitting for return. probably a small L shaped hose gong from cooler to cylinder head.

[silverton]

The hose coming off the water pumps "bypass pipe", is going direct to heater core. the cylinder 4 coolant hose is t'd to the heater core and the cooler. from the cooler, we're likely going back in to the water crossover.

[dgoodhue]

Ok, thanks for taking the time to answer all of my questions! I guess the only issue I have now is that I have no more excuses to not do this mod...

 

IMO I think it is waste of time. Subaru has it set up as #4 being the hot cylinder with the knock sensor on that cylinder. If you cool #4 cylinder better, now the hotter cylinder is going to be another cylinder, the knock sensor and programming is no longer going to be on the hot cylinder. It will just going to stretch (burn) another valve instead of #4.

[Enlight]

IMO I think it is waste of time. Subaru has it set up as #4 being the hot cylinder with the knock sensor on that cylinder. If you cool #4 cylinder better, now the hotter cylinder is going to be another cylinder, the knock sensor and programming is no longer going to be on the hot cylinder. It will just going to stretch (burn) another valve instead of #4.

 

Dom addresses this point in one of his videos. Post-mod, cylinder 4 is still the hottest cylinder so the knock sensor's function is preserved.

[silverton]

With all this talk about not wanting to lower coolant temperatures I'm surprised they make big radiators and fans and such.

 

we have race car goals, I don't know why this is so controversial.

[SAABaruu]

Because cooling systems aren't so simple. Too much flow can impede cooling. People trust that the Subaru engineers didn't overlook this mod, especially since they already produce all the parts to make it. Likely answer seems to be increased knock control. Maybe has some emissions benefit as well.

 

This will always be up for debate and is anyone's guess. I chose to do this mod personally.

 

Id be curious to see what WRC cars did.

Edited April 11, 2022 by SAABaruu

[silverton]

Don't be so curious, ask google! (hint..it's really really big radiators and air)

 

You've not upgraded, or downgraded, the water pump so flow should remain the same regardless of circuits. Mostly because there's only 2 gallons of coolant, and water pumps flow much much MUCH more than that per minute. You've added no discernible capacity, and only gave it another optional route to take. If you look at a cooling circuit for a subaru, you'd see that this mod bypasses SOME, cause it's only a slight optional detour, of the coolant going from the cylinder head to the cylinder block, directly to the heater core which returns behind the thermostat to the water pump to be pumped back in to the cylinder jacket

[SAABaruu]

Don't be so curious, ask google! (hint..it's really really big radiators and air)

 

You've not upgraded, or downgraded, the water pump so flow should remain the same regardless of circuits. Mostly because there's only 2 gallons of coolant, and water pumps flow much much MUCH more than that per minute. You've added no discernible capacity, and only gave it another optional route to take. If you look at a cooling circuit for a subaru, you'd see that this mod bypasses SOME, cause it's only a slight optional detour, of the coolant going from the cylinder head to the cylinder block, directly to the heater core which returns behind the thermostat to the water pump to be pumped back in to the cylinder jacket

 

 

I have more experience replumbing and changing the designs of EJ cooling systems than I may have led you to believe. Big radiators in my experience (owning, working and collaborating on dozens of EJ time attack cars) only delay how long it takes to overheat. A 'good' time attack EJ sees 210-220F ECT after 5- 10 minutes in 90F temps. Most hit 230F.

 

Anything above 200F reduces engine reliability and risks ringland failure. Above 225F reduces headgasket life. Very common to see pistons overexpanding and scuffing cylinder walls on track EJs with stock pistons. There is a heat problem.

 

With just a few simple plumbing changes and relocating the thermostat, I can keep ECT under 200F on track cars even with stock radiators! Better than any other cooling system I've worked on. Plus it keeps coolant much more stable which is important for piston ring wear. High HP stock engine builds should be running low ECT of 180F. It makes it more reliable. I can cite examples.

 

I haven't been able to find how WRC cars plumb their cooling systems or where they keep the thermostat. Mind sharing if you know anything? Size of the radiator only helps because it's increasing coolant capacity and lessening the effect of the bypassing coolant (which cylinder 4 mod increases this volume thus adding to the issue).

 

Did you know the hottest coolant in the car (exiting the turbo) never reaches the radiator? Or the heater core? Or the oil cooler? Now add the cylinder 4 mod and you'll see where I'm going ... My hypothesis was redirecting this coolant to the radiator would work better (but requires relocating the thermostat).

 

I believe emissions reasons forced Subaru to choose the cooling system design that it did. Runs hotter and improves emissions. The EJ is very interesting considering how old the engine design is yet it somehow managed to survive decades of tougher emissions controls that BMW and VW had to literally cheat to get around. I think this lends to some of the reliability issues. They get too hot and unless you're monitoring ECT separately, you wouldn't know it either (stock gauge is useless).

 

My design I've implemented on two cars was incredible. Coolant never exceeded 200F despite a downsized rad with only one fan. Even in 110F traffic with AC on. OEM oil cooler worked so well it mitigated the need for an oil cooler for a street and weekend track car. Stock oil pan too.

 

Tl;dr cyl4 mod is helpful in balancing ect between cylinders, but it also adds to the underlying issue that EJ cooling systems have too much coolant bypassing the radiator. Relocating bypass lines to the rad and relocating thermostat makes more stable ECT that doesn't climb over 200F without changes to the rad or water pump. Improves reliability and improves OEM oil cooler greatly for less than the cost of a koyo radiator.

Edited April 12, 2022 by SAABaruu

[ilib11]

Hi SAABaruu,

Thank you for the detailed explanation!

Do you have a kit for your changes in the cooling system you mentioned above? Or is there a detailed design specs that we can follow to build the same?

[Scubaboo]

I was curious about the assertion that a lot of coolant does not go directly back to the radiator by design so I searched online for Subaru cooling system diagrams. I was surprised to see it looks to be correct.

 

The diagrams I saw show the waterpump pushing coolant into the engine, and once it exits the engine part of the flow goes to the throttle body and the rest flows towards the heater core.

 

The flow towards the throttle body goes right back to the waterpump, bypassing the radiator.

 

The flow towards the heater hits a fork where half of that flow goes through the heater and then to the waterpump, bypassing the rad. The other branch goes to the radiator.

 

I agree with SAABaruu's theory that the inefficiency was built in on purpose for a hotter running engine for economy and emissions.

 

BMW's also purposely run hot for the above reasons but they go about it very differently. They have very well designed cooling circuits that can overcool coolant to the point that emissions and economy are noticeably affected. Temps are raised purposely by an electro mechanical thermostat controlled by the DME. So the thermostat is not only used to speed warm up time, but to also actively keep coolant temp in a programmed range.

 

I've driven Subarus and BMW's on the same roads equally hard, all the way to top speed, while monitoring coolant temps. I notice BMW's keep their coolant temps in a much narrower and consistent range than Subarus.