Suggestions to run cooler ??

[K3rm1tth3fr0g]

Originally Posted by KillerBMotorsport

^ this right here is your problem. You have less thermal capacity than stock with the mishy.



It's not about thermal mass, it's about thermal efficiency. A fat $200 radiator will perform nowhere near a $2,000 radiator that's half its size. Thermal mass has benefits, but not if you can't actually shed the heat.



Going back to stock is going to a far more thermally efficient core, even though it's smaller. I'd start there.

My temps dropped several degrees on average in all conditions with just the mishi rad to start.

Not sure I buy the stock rad having higher cooling capacity than a massive aluminum rad, given its designed properly. And regardless of anecdotal failures, mishi hasn't been demonstrated to have a core design that doesn't work.

Anyone cut one open yet?

* Registered users of the site do not see these ads.

[m5gorrf]

Originally Posted by KillerBMotorsport

^ this right here is your problem. You have less thermal capacity than stock with the mishy.

It's not about thermal mass, it's about thermal efficiency. A fat $200 radiator will perform nowhere near a $2,000 radiator that's half its size. Thermal mass has benefits, but not if you can't actually shed the heat.

Going back to stock is going to a far more thermally efficient core, even though it's smaller. I'd start there.

I respect that vendors "usually" have better testing after working in that vertical for a number of years also SCCA tech inspection for more years.

Where I truly question the OEM over a good aftermarket is that when the car was still stock I had this issue with less then 30K miles on it. After upgrading all the components to the level of Cobb stage 2 & loading that map as well as changing the radiator. "Assumable" with the added power came added heat & the aftermarket rad kept it cooler in the same temp & climb. Correlation is not causation however in this case is a good indicator.

With the Stage 2 components & map & with the OEM radiator I could only "guess" that I would have been in a worse case as to coolant temp under max stress of outside air temp of 104 & climb over a mountain pass. More so with vehicle operators (they are NOT drivers) that break for every corner.. & act like two legged cattle when behind the wheel.

I do truly get thermal dynamics of different materials, surface area & air flow.

[KillerBMotorsport]

The testing was both with both new radiators? Fresh coolant of the same mix? Same humidity, barometric, and temperature conditions? Same mods and power output? Controlled loading with the same airspeed over the car (dyno)?

Opinions are good, and there are many here, both ways, but data will always best opinions. Post what you've got and settle the debate.

Anyone that has the parts are more than welcome to bring them here to test. We've got temp and pressure sensors for both supply and return, and a controlled environment for loading and airspeed. If my opinion is wrong, dyno, materials, and tech time, is on the house.

[KillerBMotorsport]

Another marginal improvement is to make sure the sides of the radiator are sealed. Ambient airflow like to take the path of least resistance, around. OEM uses adhesive foam, which is OK, better than nothing.

[K3rm1tth3fr0g]

Originally Posted by KillerBMotorsport

The testing was both with both new radiators? Fresh coolant of the same mix? Same humidity, barometric, and temperature conditions? Same mods and power output? Controlled loading with the same airspeed over the car (dyno)?

Opinions are good, and there are many here, both ways, but data will always best opinions. Post what you've got and settle the debate.

Fresh coolant, same conditions, and mods, yes. Brand new oem rad no - had 59k miles when pulled.

[D-Rodman]

Do you have an egt gauge installed?


Sent from my iPhone using NASIOC

[benflynn]

Correct rad cap and coolant tank cap? They have to be diff

[m5gorrf]

Originally Posted by D-Rodman

Do you have an egt gauge installed?

EGT yes in the Cobb DP.

OEM caps (no change from aftermarket)

Yes to checking that the airflow is sealed

The only data I have so far is what I started with, that of changing back to the OEM Coolant dropping my top temp by 5 to 7 degrees (f) Cobb AccesPort.

I am still wrestling with the OEM Radiator claim... So I will escrow the cost of a new radiator from Subaru, you ship it to me. I will pay for the OEM coolant. The test, on a hot day will run it with a data log. Make the change and re run the same route within 90 minutes and re data log.. you are right I will release the escrow.. wrong, you pay to ship the OEM rad to you & keep it. Proving yourself correct is its own reward!

Re Greddy.. great to know others have found a way to use the space behind the fog lights to keep the heat out of the engine compartment.

Thanks for all the on going thoughts.

[rtv900]

Originally Posted by blurred

I dont need to "get back to you" as I have real world experience in seeing them fail to cool cars.

Seems like they win people over with "pretty welds" (maybe you are, but are you a certified welder or welding quality inspector?

You are forgetting that Kermit doesn't consider real world experience,

Originally Posted by K3rm1tth3fr0g

You need to because the thermal mass is extremely different between the two

Obviously failures are going to occur... What matters is the response to the failures and the frequency.

No, he doesn't need to get back to you
Thermal mass has less than nothing to do with how heat exchangers work, it's a mass to surface area ratio and the ability to flow inside and outside.
All the mass in the world won't do squat once it's up to steady state.

The underlined part is hilariously contradictory to your stance on oil pickups btw, just saying.

Originally Posted by KillerBMotorsport

^ this right here is your problem. You have less thermal capacity than stock with the mishy.

It's not about thermal mass, it's about thermal efficiency.

^yep

Originally Posted by K3rm1tth3fr0g

My temps dropped several degrees on average in all conditions with just the mishi rad to start.

Your temperature sending unit doesn't have the resolution to even make a claim about a few degrees. Rule of thumb for measuring devices is that you need about 10x's the resolution of what you are measuring.
Your temp sender would need to be capable of reading in about 0.2 degree increments to make any legitimate claim about "several degrees".

Dial calipers read down to .001 or .0005, but you can't use calipers to measure something within a tolerance of .001 just because they read that number, it is horribly inconsistent and inaccurate.
You need a mic that goes to .0001 or .00005, basically a full decimal place past what you want to measure accurately.

[benflynn]

Originally Posted by benflynn

Correct rad cap and coolant tank cap? They have to be diff

For real, do you have the one way cap on the rad and the two way cap on the reservoir? If you have a cap on the rad. This will cause cooling issues under load

[K3rm1tth3fr0g]

OP make sure both caps are in the right spot. I've seen MANY people have them installed wrong - it's extremely easy to do.

You are forgetting that Kermit doesn't consider real world experience,

Right - anecdotal claims cannot be relied upon solely to discredit products and actual evidence needs to be provided. Another reason I had asked if anyone cut one open earlier, to look at the core that is being criticized up close.

My experience was never presented as hard proof of anything, it's just one piece of the puzzle.

Thermal mass has less than nothing to do with how heat exchangers work, it's a mass to surface area ratio and the ability to flow inside and outside.
All the mass in the world won't do squat once it's up to steady state.

Cars do not always operate in steady state anyway You know like when doing a pull...

The underlined part is hilariously contradictory to your stance on oil pickups

^^Nuance is hard, I get it. ^^

[BlackFighter]

I'm 99.5% sure Killerb cut one in half and made sure their design is worth $2500. Atleast I freaking hope so.

[KillerBMotorsport]

Originally Posted by BlackFighter

I'm 99.5% sure Killerb cut one in half and made sure their design is worth $2500. Atleast I freaking hope so.

LOL, we actually very rarely recommend it. It's wicked expensive and not necessary. For most higher power guys a C&R or Ron Davis is more than enough, they perform very well. Ours is the BFH when no other tool will work. A great example is a Middle East customer that runs open class T/A and Hill Climb events. The hill climb being the biggest strain and the load is more continuous. They make +700whp, with a very big FMIC, at high altitude (lower air density), in desert temps, and because of that, A/C is a must. They came to us after trying several high dollar options that did not work. No one else makes or uses a core of this capacity, because really it's overkill for 99.998% of the people out there. The ONLY reason it exists is because we got interest to make a stock fitment unit after posting results with one of our high HP shop mules, using the 3/4 size rad, on our low mount setup.

We get this performance from a single row core that is much thinner and lighter (when filled) than the 'big' aftermarket radiators. It's all in the design, materials, and advanced manufacturing. Any heat exchanger core's efficiency, whether it's for a radiator, intercooler, oil cooler, heater, A/C, etc., is going to be proportional to the price tag.

Most rads you don't need to cut apart, just peek into the filler cap, inlet or outlet ports, to see the internal core design

[rtv900]

Originally Posted by K3rm1tth3fr0g

1) Right - anecdotal claims cannot be relied upon solely to discredit products and actual evidence needs to be provided. Another reason I had asked if anyone cut one open earlier, to look at the core that is being criticized up close.

My experience was never presented as hard proof of anything, it's just one piece of the puzzle.

2) Cars do not always operate in steady state anyway You know like when doing a pull...
3) ^^Nuance is hard, I get it. ^^

1) Says the guy who in another thread was pontificating about how you should absolutely change the factory oil pickup despite not a shred of evidence or data existing to back up that claim, all while basing his statements on anecdotal claims and nothing more.
God bless the internet. But I am glad you are coming around to the whole evidence and data thing.

2) This displays less than no understanding about how a heat exchanger works. Unless you pull over before that pull and ice down your intercooler for 10 minutes that is totally baseless. Heat exchangers don't rely on masses of aluminum to pull heat from intake air. . . . . .because of course once it's up to temp those masses would do less than nothing anyhow.
They rely on material as thin as possible to exchange heat with ambient air.

Or are you talking about the radiator before a pull? Because that makes even less sense.

3) Facts and actual real world understanding of these fundamentals is hard, I get it.

Anyhow, nice to know you modded a bunch of cooling parts for the sake of modding to say you modded, and now you run "a few degrees cooler" as measured by a device that can't measure a few degrees anyhow, so to summarize your results objectively:
You are either running a few degrees cooler, a few degrees hotter, or no meaningful change in any way after modding because racecar.

[SchoobieDoobie]

there's probably an equation somewhere relating air flow , surface area of the fins the air contacts, and coolant flow rate, and the surface area coolant the flows across, the heat capacity of the specific coolant, air temp and coolant temps

weld quality or how shiny the radiator is probably does not appear in the equation.

the coolant flow would only be affected by the cross-section area of the tubes in the radiator, as i don't think it's very practical to try and improve the water pump mechanically. you would to hire an engineer to change pully size, timing belt, etc.

you possibly could increase the coolant flow by going electric.

to increase air flow you would need to increase the fan speeds. i heard somewhere one fan is 5 blade the other is 7 but my car is not here right now or i'd go look under the hood. someone said you could switch to 2 7-blade fans. i cannot verify this.

you could also increase size of the air coming in. not sure if you want to take a sawzall to your grill and front bumper. not practical

[SchoobieDoobie]

double post bug

[K3rm1tth3fr0g]

Originally Posted by KillerBMotorsport

LOL, we actually very rarely recommend it. It's wicked expensive and not necessary. For most higher power guys a C&R or Ron Davis is more than enough, they perform very well. Ours is the BFH when no other tool will work. A great example is a Middle East customer that runs open class T/A and Hill Climb events. The hill climb being the biggest strain and the load is more continuous. They make +700whp, with a very big FMIC, at high altitude (lower air density), in desert temps, and because of that, A/C is a must. They came to us after trying several high dollar options that did not work. No one else makes or uses a core of this capacity, because really it's overkill for 99.998% of the people out there. The ONLY reason it exists is because we got interest to make a stock fitment unit after posting results with one of our high HP shop mules, using the 3/4 size rad, on our low mount setup.

We get this performance from a single row core that is much thinner and lighter (when filled) than the 'big' aftermarket radiators. It's all in the design, materials, and advanced manufacturing. Any heat exchanger core's efficiency, whether it's for a radiator, intercooler, oil cooler, heater, A/C, etc., is going to be proportional to the price tag.

Most rads you don't need to cut apart, just peek into the filler cap, inlet or outlet ports, to see the internal core design

This is super interesting and good info. Thanks for sharing KB!

Says the guy who in another thread was pontificating about how you should absolutely change the factory oil pickup despite not a shred of evidence or data existing to back up that claim, all while basing his statements on anecdotal claims and nothing more.

See, the difference is KB provided pictures, explanation, and explicit gathered technical details and info regarding the failures that led to the design of their oil pickup.

In fact they based a product around the failure - again lending credibility to it's inherent design further. I trust that a lot more than individuals claiming a product has a high failure rate with nothing more than words.

Anyhow, nice to know you modded a bunch of cooling parts for the sake of modding to say you modded, and now you run "a few degrees cooler"

More power = more heat

Still waiting for evidence of the Mishi actually failing at a higher rate than competitors.

[m5gorrf]

The 7 vs 5 blade make sense if you consider the hot water enters the rad in the center & exits on the right. The left side of the rad will be less effective but might be helped by a 7 blade as well. Will count blades next the hood is up.

Moving the oil cooler heat exchanger outside the engine bay also makes sense as Greddy does & re-enforces my plans to do it that way. Drain & refill will need to be considered also airflow.

Also vent the hood for better air flow will be the last option. Reading about fitment on those hoods or cutting holes in the OEM.

Still not seeing data or links to known good sources of info. I did look at the Greddy install lit (thanks for that). I can relocate my window washer as I have omitted the 2nd air system. However would rather the heat exchanger be on the left / shorter lines.

Will wait a few months for the recession to hit the vendors before making offers on the hardware needed.

I do welcome all the input & info.. thanks

[KillerBMotorsport]

Originally Posted by m5gorrf

I am still wrestling with the OEM Radiator claim... So I will escrow the cost of a new radiator from Subaru, you ship it to me. I will pay for the OEM coolant. The test, on a hot day will run it with a data log. Make the change and re run the same route within 90 minutes and re data log.. you are right I will release the escrow.. wrong, you pay to ship the OEM rad to you & keep it. Proving yourself correct is its own reward!

There are plenty of opinions out here (both ways) based on similar conditions and situations. It's not nearly as scientific as the controlled environment we have here. We control load, hold a specific RPM, airflow, and so on. Plus, we have very precise instrumentation that would measure inlet temp, inlet pressure, outlet temp, and outlet pressure. Along with making sure other parameters like oil temp, IATs, etc. are being properly repeated.

Originally Posted by m5gorrf

The 7 vs 5 blade make sense if you consider the hot water enters the rad in the center & exits on the right. The left side of the rad will be less effective but might be helped by a 7 blade as well. Will count blades next the hood is up.

I would not recommend changing from the OEM fans, unless something is wrong with them. This is another area where the OEM did a really good job. Besting the performance of the OEM fan (despite what some fan manufacturer's claim) is going to cost you +$600 and will require you to add higher current wiring. You would also loose the variable speed function of the OEM fans as well. Although most don't care about fan noise if their setup requires these kinds of modifications.

Generally, once you're at speed, the fans don't come on. If you have an AP, you can watch when they do, but for street guys doing a 1-4 pull, no. When we were taking our 760whp GD mule (the one with the 3/4 rad) to 1/2 mile events it would have the lowest coolant temps (~190°) at the end of the run, where the load and RPMs were highest. The highest thermal loads we ever saw were actually at a small Auto-X event. High humidity, high temps, low airspeed, and my like of brake boosting created the highest temps we ever saw on that setup ~212°. For reference the ECU starts to pull timing at 216°. For anyone that likes to watch the AP, this temp should be your 'do not cross point'. Not that it's horribly bad, it's just not optimum for performance.

[benflynn]

It’s the caps

Fans don’t help cool going up hill at speed

[m5gorrf]

Originally Posted by KillerBMotorsport

The highest thermal loads we ever saw were actually at a small Auto-X event. High humidity, high temps, low airspeed, and my like of brake boosting created the highest temps we ever saw on that setup ~212°. For reference the ECU starts to pull timing at 216°. For anyone that likes to watch the AP, this temp should be your 'do not cross point'. Not that it's horribly bad, it's just not optimum for performance.

From river crossing to peak hill climb is about 20 miles & 1700 ft outside temp of 104f. If memory serves the AP temp goes from 190s to (Prestone) 227, with OEM coolant just about 220. The 220 temp (about 2/3 on the OEM gauge) is only the last few miles. I had a passenger reading the AP on that trip while driving the tight twisties. I love where I live & have my own 100 yrd range

As I crest the peak the temp drops back to 190s within a mile at highway speeds. The AC has only kicked off once & I set to heat & opened the windows. Anytime it gets up to 1/2 way up the gauge I baby the engine.

The quick recovery tells me that the thermal load is being dissipated well enough other then the most extreme (regular) situations. Will most likely put a oil cooler behind the left fog light with a fan next. Do a AP read before & after with the same outside temp.. as this holds for about 3 months a year.

Thanks

[m5gorrf]

Originally Posted by benflynn

It’s the caps

Fans don’t help cool going up hill at speed

Brand new OEM caps, but will double check with the dealer

Thanks

[benflynn]

Originally Posted by m5gorrf

Brand new OEM caps, but will double check with the dealer

Thanks

The round one way cap goes on radiator, the two way cap with ‘ears’ goes on tank.

[rtv900]

Originally Posted by SchoobieDoobie

there's probably an equation somewhere relating air flow , surface area of the fins the air contacts, and coolant flow rate, and the surface area coolant the flows across, the heat capacity of the specific coolant, air temp and coolant temps

weld quality or how shiny the radiator is probably does not appear in the equation.

There absolutely are equations that include all of those things. The subject is called Heat and Mass Transfer and the science is well established and extremely objective.
And your right, weld quality and shininess are not included.

One thing to note is that outright mass flow of coolant isn't necessarily the most important thing, in certain cases increasing flow actually reduces the heat transfer, and fluid turbulence is something that increases heat transfer. Small particles are speckled on the fins in many cases to add turbulence to the ambient flow, which increases heat transfer while reducing total mass flow, and turbulence on the liquid side has the same affect although there's generally enough turbulence created naturally.
Bottom line, heat transfer is NOT a case where you want smooth laminar flow.
In the case of an intercooler obviously you have to balance two priorities and get as much heat transfer as possible with a reasonable pressure drop.

Originally Posted by K3rm1tth3fr0g

See, the difference is KB provided pictures, explanation, and explicit gathered technical details and info regarding the failures that led to the design of their oil pickup.

In fact they based a product around the failure - again lending credibility to it's inherent design further. I trust that a lot more than individuals claiming a product has a high failure rate with nothing more than words.

More power = more heat

Still waiting for evidence of the Mishi actually failing at a higher rate than competitors.

So one individual seeing a failure means out of millions of samples there is an unacceptable failure rate?
Nice, contradict yourself much?
(who's the guy who said there will always be failures?)

Probably a nice guy, but sorry he has no legitimate data or evidence to make a claim about failure rates. What he is doing is called surmising to be accurate.

Solid point, yes more power does in fact mean more heat. Didn't know anybody was debating that.

Also, pretty sure I never said I had evidence of mishimoto failure rates, I simply pointed out the gross inaccuracies in your statements and assertions.

To summarize: Your cooling mods have resulted in slightly higher coolant temps, slightly lower, or no change based on your data.
Get a measuring device that can actually support your claims and "get back to us"

[K3rm1tth3fr0g]

Originally Posted by rtv900

So one individual seeing a failure means out of millions of samples there is an unacceptable failure rate?
Nice, contradict yourself much?

A failure?

Yes, the KB pickup was definitely designed after seeing one singular failure.