high efficiency intercoolers

[silverbullet]

Its about time for me to ditch my egay FMIC that I've had for half a decade for a real one that isn't over sized either. I was hoping we can all share some knowledge on which cores are quality stuff with overall high heat transfer coeffients. I'm sure many others may benafit from this.

I know used GReddy, HKS and Blitz FMICs pop up on zilvia for fairly cheap and are certainly better than "China Speed" ones but size wise they are just way too large for me (want responsive setup) as they are rated for 600bhp.

Additionally, according to Full Race a back to back dyno test showed a Garrett core FMIC yeiled better results than a GReddy unit throughout the power band see figure 1*.. but are $$! Are Spearco, Precision, Treadstone and Vibrant comparible alternatives?

:eek:


figure 1
http://www.full-race.com/catalog_images/S13_Silvia_FMIC/FRS13S14IntercoolerPower.jpg
* Comparison of a back door setup with Garrett core vs a GReddy off the shelf FMIC. I dont believe Full Race's back door IC mounting position makes any changes to incoming air flow, so i believe this is a result of different cores.

[leetfade]

Interested to hear some input on this as well. I have a Blitz now on the SR and it works great but if I go to the RB I'll likely sell it and get something else.

I know we had a guy seeking out a Forge and we got him a pretty good deal on it but I really haven't had time to do a lot of research on the various brands. It may have just been because he had a Mini and that is the more common one for that car.

Correct me if I'm wrong, but it seems like reduced restriction would play a big part, as well as, cooling ability right? I know you lose about 1-2 psi on average with a standard FMIC but are there worse ones?

[silverbullet]

^ It is hard to do research because all manufactures will claim they have "high efficiency" (including Mishimoto and other china speeds) without giving out actual #s. I forgot which company it was, but during my research, a "reputable" company claimed to have 90% efficiency, which I do not believe to be representative to real conditions.

[Tower240sx]

the good intercooler cores are really mostly honeywell (garrett) cores with per application endtanks welded on...

I use/used a flipped over EVO8 I/C for 6 years on my sr20 cut the pipes off and welded new piping at angles that worked for me, I/C cost $125 (25 i/c 100 welding new piping on)

[KA240SX808]

To my understanding FR's claim for the low end power increase was due to the shorter piping allowing it to build boost and make power slightly faster. Cause the top end power stayed the same. Just the low end seemed to come on a little quicker.

[CodyAce]

While the details at this point escape me, and I'm not sure if his image links still work, but there was an Altima Turbo owner who went from an Early XSTurboEbay intercooler and then went with a Spaerco core, and saw nearly 10 whp gain from the core, nothing else changed.

Now I don't remember if dyno conditions were different (which they could be), or the timeframe exactness (it was over the course of a summer) but it was at the time a significant gain to make you consider upgrading (as better heat dissipation also helps), however the cost per power to some was not worth it.

[Geno]

While the details at this point escape me, and I'm not sure if his image links still work, but there was an Altima Turbo owner who went from an Early XSTurboEbay intercooler and then went with a Spaerco core, and saw nearly 10 whp gain from the core, nothing else changed.

Now I don't remember if dyno conditions were different (which they could be), or the timeframe exactness (it was over the course of a summer) but it was at the time a significant gain to make you consider upgrading (as better heat dissipation also helps), however the cost per power to some was not worth it.

MattBack (Pretty sure you and he know each other) and I were talking about FMIC's a while back and he showed me a dyno where a SR owner switched from chinaspeed to one of those Greddy RSPL fmic's and saw something like 30whp.

I'm curious to see some more information too, I have a stupid large tube fin intercooler from a R34 GT-R on my car now and am looking for something a bit more appropriate than the 4" thick monster I have now.

[silverbullet]

To my understanding FR's claim for the low end power increase was due to the shorter piping allowing it to build boost and make power slightly faster. Cause the top end power stayed the same. Just the low end seemed to come on a little quicker.

Is that really what they said? If so, shame on them. The initial spool up would spool up quicker on the shorter piping system, but after they both hit full boost (<4krpm), it should all be the same. That dyno graph clearly shows the same curve but shifted up in the mid rpm.

the good intercooler cores are really mostly honeywell (garrett) cores with per application endtanks welded on...

I use/used a flipped over EVO8 I/C for 6 years on my sr20 cut the pipes off and welded new piping at angles that worked for me, I/C cost $125 (25 i/c 100 welding new piping on)
Is an evo fmic made by honeywell? Any idea what other cars use honeywell FMICs? Evo, Skyrine, SRT4, trucks?

[Om1kron]

Part shop max made a nice little post about their HMIC intercoolers on *GASP* Zilvia the other day.

http://zilvia.net/f/chat/394043-powered-max.html

[Def]

Honestly, a dyno difference like that could be as simple as referencing the boost pressure somewhere different so you're running more midrange boost.


I've actually never seen a big difference in HP with or without a fan on the IC. It seems like it's mostly down to thermal inertia of the IC to do the cooling since a few shop fans don't really do much cooling for the brief time of a dyno pull. So I don't see how a "more efficient" IC makes a big difference when cooling airflow can be shown to not affect "lower power" dynos that much.

[CodyAce]

MattBack (Pretty sure you and he know each other) and I were talking about FMIC's a while back and he showed me a dyno where a SR owner switched from chinaspeed to one of those Greddy RSPL fmic's and saw something like 30whp.

The big thing in my experience is finding a direct comparison...most of the time when people swap cores, they are changing a ton of other stuff too...which for sure changes results (Yea I know MattV haha)


I'm curious to see some more information too, I have a stupid large tube fin intercooler from a R34 GT-R on my car now and am looking for something a bit more appropriate than the 4" thick monster I have now.

Wow that's a serious core size!

Is that really what they said? If so, shame on them. The initial spool up would spool up quicker on the shorter piping system, but after they both hit full boost (<4krpm), it should all be the same. That dyno graph clearly shows the same curve but shifted up in the mid rpm.

Which has always been my comment with these mid mount setups, is that's most people relaly have no clue how to judge results. Their spool up is mimimal, and power is the same.


Is an evo fmic made by honeywell? Any idea what other cars use honeywell FMICs? Evo, Skyrine, SRT4, trucks?

Truck cores are 90% Garrett. Volvo's have their own cores (who makes them I don't know), same with Mack. (which makes sense as they are now Volvo)

Honestly, a dyno difference like that could be as simple as referencing the boost pressure somewhere different so you're running more midrange boost.

Whichis 100 percent true. Changing reference to the turbo outlet is going to show a ton of more 'response' but it's really just the point where you place the gauge.


I've actually never seen a big difference in HP with or without a fan on the IC. It seems like it's mostly down to thermal inertia of the IC to do the cooling since a few shop fans don't really do much cooling for the brief time of a dyno pull. So I don't see how a "more efficient" IC makes a big difference when cooling airflow can be shown to not affect "lower power" dynos that much.

There is no doubt in my experience (couple hundred cars now) on the dyno that the fan makes such little difference that at our last dyno day, aside from the guys with heat exchangers where it does make a tiny bit of difference (lightning, S/C ecotec cars etc etc).

[CodyAce]

With that said, most of these gains from any upgrade to MMIC or new FMIC is that they come from a better core in general, not from the 'setup'. When going from Chinaspeed to one of those PSM cores, you're going to have better results...but I'd attibute little of the 'shorter pipes' to that reasoning.

[WorkInProgressK]

Can we have a good core vs a bad core example picture wise.

Obviously size will mater, but I would like to know how a good core looks like.

Just bought my MmIC from PBM...and this tread is freaking me.

[Geno]

Can we have a good core vs a bad core example picture wise.

Obviously size will mater, but I would like to know how a good core looks like.

Just bought my MmIC from PBM...and this tread is freaking me.

Bad core, as in way too big.

http://airepoweraviation.com/carprogress/gettinthere.jpg

[Def]

I don't think a big core is going to do anything negative besides weigh a lot.

That core has really nice endtanks. Much better than the sharp 90 you get on a lot of "backdoor" IC.

[Geno]

I don't think a big core is going to do anything negative besides weigh a lot.

That core has really nice endtanks. Much better than the sharp 90 you get on a lot of "backdoor" IC.

It actually weighs 9lbs less than your standard ebay bar plate fmic. Yay for tube/fin?

[Def]

Tube and fin is good, but comparing it to a piece of Chinaspeed boat anchor isn't exactly a fair comparison. :D

[tinto]

I thought someone had posted a shot of my car above.
My HKS IC seems particularly dense compared - is it just a different design?

http://farm5.static.flickr.com/4116/4781784351_a66c74b307.jpg

[silverbullet]

Can we have a good core vs a bad core example picture wise.

Obviously size will mater, but I would like to know how a good core looks like.

Just bought my MmIC from PBM...and this tread is freaking me.

I was hoping the scope of this thread would be comparing results of good cores vs bad cores. I'm almost certain PBM uses an ebay style core, but there are many many air-air heat exchanger manufactures in China so not all of them are the same.

Here are some examples of typical Egay cores:

http://img526.imageshack.us/img526/277/55dx2.jpg
source: http://www.thirdgen.org/techboard/power-adder/368256-cheap-ebay-intercoolers.html

http://www.94supra.com/cooling/cooling_06.jpg
Very similar to (if not same as) Mishimoto
source: http://www.supraforums.com/forum/showthread.php?617183-CXRacing-for-the-win-yet-AGAIN!


Here is a good thread i found on SoCalEvo:
http://www.socalevo.net/index.php?option=com_smf&Itemid=37&topic=38152.0

Quality Core:
http://www.socalevo.net/gallery/albums/userpics/10991/normal_TECH%20ARTICLE%20intercooler%20fin%20engine ering%20and%20design%20%28ins%20and%20outs%20pics% 29%2004.jpg
XS Engineering

[leetfade]

So what is the benefit of the increased fins? It seems like it would be more restrictive and cause more total boost loss.

[V1A]

More fins = more surface area = more heat transfer.

[silverbullet]

doh.

[240sxTTC]

More fins = more surface area = more heat transfer.

Yes but more fins also = more restriction making the turbo work harder for the commanded manifold pressure. This can result in operating pts in a lower efficency pt of the comp map, which cause higher compressor outlet temps. For intercoolers (actually after-coolers) it's a balance between fin design, fins/inch, effectiveness (not efficiency because there is no work involved) and pressure drop.

The general rule is put the largest charged-air-cooler that will fit. Usually you will run out of space before the cooler is too big.

[leetfade]

Ok that helps. I thought there was a too large but you're saying the space available on the car will be exceeded before you hit a diminishing returns on the I/C? That seems surprising too.

So if we stick to the concept of the innards of the core vs. overall dimensions, how do you know the right balance?

For example, if I stick with my SR in my S14 next year and run a GT3076r, what is the best FMIC for the job? Now if I opt to go RB25 does it change? I'm sure it will if I go RB26 due to the TT but it may be the same size, core type, etc. Does the amount of intended boost come into play?

[240sxTTC]

So if we stick to the concept of the innards of the core vs. overall dimensions, how do you know the right balance?

That is what the cooler manufacturer takes care of.


For example, if I stick with my SR in my S14 next year and run a GT3076r, what is the best FMIC for the job?

The biggest one that will fit.


Now if I opt to go RB25 does it change?

No


Does the amount of intended boost come into play?

Yes. The mass flow, which is related to the boost and density, affect the cooler size and hp. Again I would go with the biggest cooler you can fit within reason. I wouldn't start cutting core supports and everything metal in front of the car to fit a GINORMOUS cooler. Below is a link to the cooler info I put together for the catalog and website. This will give you an idea of the core size and the estimated supported hp.

http://www.turbobygarrett.com/turbobygarrett/catelog/Intercoolers/Intercoolers.htm

[silverbullet]

so you work for garrett? Can you spill the beans on what companies use Garrett cores?

[240sxTTC]

so you work for garrett? Can you spill the beans on what companies use Garrett cores?

Unfortunately there is no way to know that. Our distributors sell to their distibutors who in turn sell to shops and end consumers.

[leetfade]

Thanks TTC!

[AutoX_Acura]

the good intercooler cores are really mostly honeywell (garrett) cores with per application endtanks welded on...

I use/used a flipped over EVO8 I/C for 6 years on my sr20 cut the pipes off and welded new piping at angles that worked for me, I/C cost $125 (25 i/c 100 welding new piping on)

How do you like that setup? Can't beat the price even to have a shop do some tig work its still cheap