Anyone running the GDT cylinder 4 cooling kit

Curious if anyone is running this kit:

http://shop.getadomtune.com/shop.getadomtune.com/cylinder-4-chamber-cooling-system-for-2010-lgt/

saw some banter on it over at NASIOC in this thread

https://forums.nasioc.com/forums/showthread.php?t=2882734

relatively inexpensive... not sure if it's needed or not, though.

Post 1

Did he try the kit for the WRX or the one for the 10+ LGT? The pictures are generic on the GDT website, but the cost is different between the two kits, suggesting the parts are different. I wonder if the LGT kit has something to help deal with the lack of space?

Did he try the kit for the WRX or the one for the 10+ LGT? The pictures are generic on the GDT website, but the cost is different between the two kits, suggesting the parts are different. I wonder if the LGT kit has something to help deal with the lack of space?

This was the fix that the getadomtune sent me. They told me to cut the T off the original hose, attach it to the open end of the new one, and connect the 90 degree coupler to the old hose. I'm gonna wait till I get everything i need to upgrade the radiator since I'd have to drain some coolant to do this anyways.

The GDT cooling mod is just snake oil, imo. Subaru knew cylinder 4 would be the weakest cylinder and that's where they put the knock sensor. The whole idea is that if any cylinder is to see knock, it should be #4. This mod raises the temps of cylinders 1-3 so that they're more prone to knock now, but you won't be able to see it because the ECU can't monitor that.

Who do you think knows the cooling system better, some aftermarket tuner, or the company that spent millions of dollars designing the engine?

Unless you're adding knock sensors to every cylinder, just leave it alone.

The GDT cooling mod is just snake oil, imo. Subaru knew cylinder 4 would be the weakest cylinder and that's where they put the knock sensor. The whole idea is that if any cylinder is to see knock, it should be #4. This mod raises the temps of cylinders 1-3 so that they're more prone to knock now, but you won't be able to see it because the ECU can't monitor that.

 

Who do you think knows the cooling system better, some aftermarket tuner, or the company that spent millions of dollars designing the engine?

 

Unless you're adding knock sensors to every cylinder, just leave it alone.

Amen. This seems like the most questionable non-aero/body mod I've ever read about on legacygt or nasioc.

to be fair, a large number of the folks who have had issues have had them with cylinder 4, and overheating is often blamed as the issue. Perhaps Subaru intentionally reduced cooling to cylinder 4 to make it a canary of sorts for knock, but weakening one cylinder doesn't seem to be a great strategy...

the EGT results seem pretty definitive in that in makes things a whole lot more uniform. Cylinder 4 is still hottest in their data, but the delta between hottest and coolest is only 20 degrees instead of 75 degrees. If you are tuning the engine to make more power, that seems to be a difference which will enable you to hit bigger numbers with less risk of blowing cylinder 4...

The tuner himself on the topic:

[ame=http://www.youtube.com/watch?v=X82Atqn5pX8]Reverse Cooling and how it applies to Subarus - YouTube[/ame]

From the comments:

Q: Hey Dom, if you bring down the temp on cylinder 4's head, decreasing its knock tendency with this mod, doesnt that mean we would lose cylinder 4 as a "reliable" way of sensing knock? I mean cylinder 4 will no longer be as prone to knock as it was without this mod, doesn't that mean we could potentially miss knock events that are more likely to happen in cylinder 1, 2, and 3?

 

A: A completely valid concern and one of the reasons why I didn't make the mod flow the maximum amount of available coolant (I strategically sized the orifice in the fitting). I have some EGT and temp data that shows 4 is still hotter, but the delta is reduced dramatically. It's still the harshest cylinder, it's just [sic] as harsh. https://cdn8.bigcommerce.com/s-gfm7bog6j4/images/stencil/1280x1280/products/112/388/CoolingModEGTs__30473.1533290766.png?c=2&imbypass=on

I assume he meant to write "it's just not as harsh."

The GDT cooling mod is just snake oil, imo. Subaru knew cylinder 4 would be the weakest cylinder and that's where they put the knock sensor. The whole idea is that if any cylinder is to see knock, it should be #4. This mod raises the temps of cylinders 1-3 so that they're more prone to knock now, but you won't be able to see it because the ECU can't monitor that.

 

Who do you think knows the cooling system better, some aftermarket tuner, or the company that spent millions of dollars designing the engine?

 

Unless you're adding knock sensors to every cylinder, just leave it alone.

are there not knock sensors on the other cylinders?... and if not how does my ap read them?

are there not knock sensors on the other cylinders?... and if not how does my ap read them?

I think the AP reads roughness on all 4 cylinders, not knock. I belive roughness is sensed by the coil packs.

Sent from my SM-G965U using Tapatalk

I think the AP reads roughness on all 4 cylinders, not knock. I belive roughness is sensed by the coil packs.

you can track knock sum for each cylinder. but cyl 4 is scary to watch.

Hey guys,

I work in hydraulics making hose assemblies and such. I've been a little concerned about the sealing method on the metric to JIC (AN) adapter. Adding gasket compound to the metric side might work but I wouldn't rely on it alone.

In my industry we use adapters often to lower the cost of replacing hose assemblies. The metric adapter going into the engine block is being used as a port adapter. A metric adapter and a BSPP (British Special Pipe Parallel) to JIC (Joint Industry Council) are similar in that the threads are on a parallel pipe, it will have a concave tip to seal on a pipe connection, and a recessed space between the threads and the hex base to seal on an o-ring or bonded washer.

I would suggest adding a bonded washer to the metric side and doing or not doing the gasket material to the threads. Some adapters might have a bonded washer or flat o-ring built into the base of the adapter. An ideal adapter would be constructed of Aluminum and include a bonded washer. In my work, we carry adapter constructed of steel. I don't like the galvanic potential in a coolant system with an aluminum engine block and a high carbon steel adapter. I bet corrosion wouldn't take long to form.

Metric port, British port, and O-ring Boss area all very similar in design. In fact a JIC male can be used as an O-ring Boss fitting if there's enough shoulder to support an o-ring or bonded washer because they share the same thread sizes.

A JIC connection doesn't require a sealant as the cup and cones smash to seal. NPT or pipe thread fittings seal by deforming the threads, so does a BSPT (British Special Pipe Taper) connection. Thread tape is a lubricant allowing the tapered threads to mash easier.

Instead of using a 45 degree hose fitting I would suggest trying a 45 degree female JIC to male JIC adapter and a straight fitting hose adapter or hose assembly.

I would also like to see a hose tee with a JIC middle connection so either end could be capped if there is a leak. Hopefully that's also available constructed of aluminum. Looks like a difficult area to work. Crows feet help but they area pain themselves.

This 'mod' is a factory item on the JDM legacy engines and my '08 EJ20X U20 cylinder heads have the fitting on it.

My problem with the GDT kit is why is it sending hot coolant from cyl4 straight to the water pump? It's not hitting a radiator and being reintroduced back into the system.

I haven't been able to figure out where Subaru routed the coolant from cyl4 on their JDM legacy's. I've scoured the JDM parts diagrams, can anyone help?

Pic of factory cyl4 cooling mod on U20 JDM legacy cylinder heads

https://imgur.com/a/aFH7u7h

I have been looking into justifications for this mod and have not found anything convincing, apart from the previous post in which it is claimed that this mod is the stock configuration for JDM legacy engines. From my understanding, the stock head gaskets have small oval cut-outs to allow coolant to pass from the lower interior of blocks into the cylinder heads and then back into the upper interior of the blocks.

These oval cut-outs are smaller than the cross-section in the block/cylinder head castings. They are also of uniform of uniform size, which leads me to believe they are purposefully introduced to impose relatively uniform flow from one side to the other. This is the basis for fluid manifold design, where a large reduction in cross-section allows for relatively uniform flow (shower head, sprayers, etc). Thus the placement of the coolant inlets (from the water pump) and outlets (to the crossover pipe) really should not matter much, as long as the cut-outs in the head gaskets are small enough to impose a large pressure drop.

Is there any solid data showing that mod works well at all? I think that we would need to have an idea about the temperatures of the cylinders before installing the mod, and then after. So not just cylinder four, but all of them, since the mod could result in a cooler cylinder four but the other cylinders being hotter.

The other issue is what the cooling performance is at different engine speeds, the cylinder four mod seems like it introduces a significant imbalance between the right and left coolant paths. This imbalance would only get worse at higher pressures.

The other issue I do not see brought-up that could have to do with cylinder four issues is the fact that both cylinders 3 and 4 have more tortuous exhaust paths when exiting the cylinder heads. I would expect more heat transfer from the exhaust existing these cylinders due to this. Combine this with all of the UEL header issues that significant impact cylinder 4, and I am highly doubting that this mod is worth the risk for any EJ engine, especially ours with equal length headers.

Am I over thinking this ?

The GDT cooling mod is just snake oil, imo. Subaru knew cylinder 4 would be the weakest cylinder and that's where they put the knock sensor. The whole idea is that if any cylinder is to see knock, it should be #4. This mod raises the temps of cylinders 1-3 so that they're more prone to knock now, but you won't be able to see it because the ECU can't monitor that.

 

Who do you think knows the cooling system better, some aftermarket tuner, or the company that spent millions of dollars designing the engine?

 

Unless you're adding knock sensors to every cylinder, just leave it alone.

I actually have the parts from an imported motor that look to be an OEM cylinder 4 cooling "mod". so, I don't think this is necessarily snake oil

I have been looking into justifications for this mod and have not found anything convincing, apart from the previous post in which it is claimed that this mod is the stock configuration for JDM legacy engines. From my understanding, the stock head gaskets have small oval cut-outs to allow coolant to pass from the lower interior of blocks into the cylinder heads and then back into the upper interior of the blocks.

 

These oval cut-outs are smaller than the cross-section in the block/cylinder head castings. They are also of uniform of uniform size, which leads me to believe they are purposefully introduced to impose relatively uniform flow from one side to the other. This is the basis for fluid manifold design, where a large reduction in cross-section allows for relatively uniform flow (shower head, sprayers, etc). Thus the placement of the coolant inlets (from the water pump) and outlets (to the crossover pipe) really should not matter much, as long as the cut-outs in the head gaskets are small enough to impose a large pressure drop.

 

Is there any solid data showing that mod works well at all? I think that we would need to have an idea about the temperatures of the cylinders before installing the mod, and then after. So not just cylinder four, but all of them, since the mod could result in a cooler cylinder four but the other cylinders being hotter.

 

The other issue is what the cooling performance is at different engine speeds, the cylinder four mod seems like it introduces a significant imbalance between the right and left coolant paths. This imbalance would only get worse at higher pressures.

 

The other issue I do not see brought-up that could have to do with cylinder four issues is the fact that both cylinders 3 and 4 have more tortuous exhaust paths when exiting the cylinder heads. I would expect more heat transfer from the exhaust existing these cylinders due to this. Combine this with all of the UEL header issues that significant impact cylinder 4, and I am highly doubting that this mod is worth the risk for any EJ engine, especially ours with equal length headers.

 

Am I over thinking this ?

Running it since my last build.

It was much more work then a 4th gen GT. But it has actually brought coolant temps overall. Being with fmic, I saw 10degrees drop cruising and typically in the summer time, i would be around 200+ish. Now seeing 185-194 driving.

Tuner didn’t notice anything different in timing during the tune, but who knows if it helped. More of an insurance being placed then guessing if it does anything.

I’ll spend $100 any day on things that’ll “help” with cly 4.

Sent from my iPhone using Tapatalk