Independant Intercooler Test
Registered User
Joined: Feb 2011
Posts: 357
From: DFW
Rep Power: 40
Interesting results. I wonder why 9 m/s is the fastest airflow measured? Barely 20mph.. aren't the ICs seeing considerably more than that? As in, at 100+mph (where the heat becomes an issue), they're seeing 100+mph airflow through the core?
I'm converting knowledge of radiators for water cooled PCs here, but generally at lower airflow speed, lower density cores (ie .2s) way outperform higher density cores (proto/tpc, champion, AMS). Yet when you get good airflow with higher static pressure (as would be caused by the ducting vs having these open air as in this test), the higher density cores come alive and completely dominate the low density ones.
So while this is a fantastic data point, to me it seems somewhat pointless b/c the air was so slow across the cores. Nobody has issues with cooling at 20mph right? Or am I missing something (entirely possible)?
I'm converting knowledge of radiators for water cooled PCs here, but generally at lower airflow speed, lower density cores (ie .2s) way outperform higher density cores (proto/tpc, champion, AMS). Yet when you get good airflow with higher static pressure (as would be caused by the ducting vs having these open air as in this test), the higher density cores come alive and completely dominate the low density ones.
So while this is a fantastic data point, to me it seems somewhat pointless b/c the air was so slow across the cores. Nobody has issues with cooling at 20mph right? Or am I missing something (entirely possible)?
Thread Starter
|
Registered User
Joined: Apr 2006
Posts: 1,677
From: London (UK)
Rep Power: 156
Thanks Nick for doing this. A few thoughts:
1) You tested a group of really good intercoolers, one should not think of the fvd as a low performer by any means. I really wish you could have gotten a 3.5" bell core in the mix as it is probably the most common aftermarket core and would have given an aftermarket baseline, disappointing few would play. Same with a .1 core.
2) The gt2 rs cooler flow myth is debunked yet again. Thats still an acceptable delta p at over 700 hp worth of airflow.
3) Do you have any pics of the test setup? The 9 m/s if cooling airflow may not be a simple as it seems. If they were able to maintain that across the core, it may be more realistic that we think. Anyone know the air velocity across the core installed in the car running at various speeds? With the backpressure from the outer core itself, it may be much lower than the car's speed. Just curious if they know something more here as I'm impressed the efficiency is as high as it is at .3 kg/s of charge flow!
4) I've never had good luck with those inlet/outlet temp gauges on automotive apps. They are good in airplane apps where you're at steady state for hours but imo, way too slow to capture useful data over the span of 10 to 20 seconds where most of these in car intercooler tests take place. Maybe theyre much faster now though...
1) You tested a group of really good intercoolers, one should not think of the fvd as a low performer by any means. I really wish you could have gotten a 3.5" bell core in the mix as it is probably the most common aftermarket core and would have given an aftermarket baseline, disappointing few would play. Same with a .1 core.
2) The gt2 rs cooler flow myth is debunked yet again. Thats still an acceptable delta p at over 700 hp worth of airflow.
3) Do you have any pics of the test setup? The 9 m/s if cooling airflow may not be a simple as it seems. If they were able to maintain that across the core, it may be more realistic that we think. Anyone know the air velocity across the core installed in the car running at various speeds? With the backpressure from the outer core itself, it may be much lower than the car's speed. Just curious if they know something more here as I'm impressed the efficiency is as high as it is at .3 kg/s of charge flow!
4) I've never had good luck with those inlet/outlet temp gauges on automotive apps. They are good in airplane apps where you're at steady state for hours but imo, way too slow to capture useful data over the span of 10 to 20 seconds where most of these in car intercooler tests take place. Maybe theyre much faster now though...
Thanks Earl.
1)It is a shame more cores were not added to this test but that's not to say it can still be replicated if people still have a change of heart and still want to test their cores as it can be repeated over and over again.
2)yes the Gt2 core for the money and its capabilities is a no brainier but still will hit a brick wall at some point.
3)Yes they were able to maintain that speed across the core and no I don't have any pics but will ask if they are willing to take some for me.
4)This is what I have found out myself is that these monitoring systems will not read the airflow fast enough for the testing I want to do and dura metrics would be far more superior as the car ECU reads so much faster.
Thread Starter
|
Registered User
Joined: Apr 2006
Posts: 1,677
From: London (UK)
Rep Power: 156
Interesting results. I wonder why 9 m/s is the fastest airflow measured? Barely 20mph.. aren't the ICs seeing considerably more than that? As in, at 100+mph (where the heat becomes an issue), they're seeing 100+mph airflow through the core?
I'm converting knowledge of radiators for water cooled PCs here, but generally at lower airflow speed, lower density cores (ie .2s) way outperform higher density cores (proto/tpc, champion, AMS). Yet when you get good airflow with higher static pressure (as would be caused by the ducting vs having these open air as in this test), the higher density cores come alive and completely dominate the low density ones.
So while this is a fantastic data point, to me it seems somewhat pointless b/c the air was so slow across the cores. Nobody has issues with cooling at 20mph right? Or am I missing something (entirely possible)?
I'm converting knowledge of radiators for water cooled PCs here, but generally at lower airflow speed, lower density cores (ie .2s) way outperform higher density cores (proto/tpc, champion, AMS). Yet when you get good airflow with higher static pressure (as would be caused by the ducting vs having these open air as in this test), the higher density cores come alive and completely dominate the low density ones.
So while this is a fantastic data point, to me it seems somewhat pointless b/c the air was so slow across the cores. Nobody has issues with cooling at 20mph right? Or am I missing something (entirely possible)?
Right the airflow can only be set to 15m/s.
The issue if you call it one was the the motor used to pump the air was struggling to push air through the high density cores so I asked them to pick the max air flow they could achieve with the most dense core to make it a fair test as its no good running one at 10m/s or 12m/s because then the figures would be all over the place. In regards to the higher speeds at 100mph yes you would see a lot more cooling with the denser cores the gap between the 3 coolers tested would just keep growing!
It's not completely pointless no.
You still get a chance to see which is the better cooler out of the bunch tested.
Okay you do not see max capabilities which you get from reality testing but you sure do get a good idea of what is working better than the other.
You know already from your PC knowledge the higher density cores will win as the airflow climbs.
What would make it even more interesting is if we could add a few more to the mix.
Registered User
Joined: Mar 2009
Posts: 886
From: In my garage
Rep Power: 123
Now this is the question. ICs' location is far from optimal, at least compared to those in front of the car.
When building mine and planning the placement of / ducting to ICs (narrowbodied car) I discussed about the issue with factory trained Porsche Center head-of-maintenace. He claimed that most of the flow across ICs is caused by the vacuum at outlets than overpressure at inlets... Which aren't exactly in an overpressure area on the side of the car. Although the rear fender's shape has an effect here.
As an idea of the issue here's example pics of CFD simulation:
Switzerland-based MATECH Competition relied on Exa Corporation’s PowerFLOW CFD software to make its Ford GT1 race car more aerodynamic.
Another picture here: http://www.flickr.com/photos/jalopnik/422044058/lightbox/
AIR PRESSURE COMPARISON FRONT VIEW - Generated by Computational Fluid Dynamics (CFD), this illustration compares the aerodynamic characteristics of the NASCAR Impala SS (foreground) and the Monte Carlo SS (rear). The colors on the body surfaces represent relative pressure, from red (highest) to green (moderate) to blue (lowest). Similarly, the colors of the streamlines around the cars indicate the pressure in the surrounding air from red (high pressure) to blue (low pressure). High pressure on the nose and headlights of the Monte Carlo SS shows that this area generates the majority of the car's front downforce; this area is blue on the Impala SS, indicating low pressure and lift. Dark red on the horizontal surface of the Impala SS's front splitter shows that this aerodynamic device generates significant downforce. NASCAR allows the splitter to be adjusted from four to six inches ahead of the bumper recess to fine tune the aerodynamics. (General Motors Racing Handout) (USA)
CFD Air Pressure Model
The GT2-spec Corvettes were designed, built and tested on a compressed schedule. The program was approved and announced in September 2008, and construction of the first chassis began in early December. The first track test was conducted at Road Atlanta on April 8-9, followed by single-car tests in Elkhart Lake, Wis., and Sebring, Fla. Doug Fehan, Corvette Racing manager said "The Corvette Racing team had to take on several challenges simultaneously to execute this program." He added "We were preparing for our regular race season with the GT1 cars while designing the GT2 version. The cars were being built and tested in the midst of our preparations for Le Mans. The team was multi-tasking to the extreme, operating on a leaner budget and a faster timeline. It was a monumental effort to have these cars ready for the Mid-Ohio race."
Last edited by pete95zhn; Nov 6, 2013 at 12:15 AM.
Registered User
Joined: Oct 2011
Posts: 429
From: Germany
Rep Power: 61
Very nice flow pictures
Andreas
Registered User
Joined: Nov 2011
Posts: 800
From: Sumter, SC
Rep Power: 130
Interesting results....one thing that caught my eye as particularly interesting is this:
One the "Cooling Side" the outlet temps for both the Aero and GT2RS cores were virtually identical but on the "Charge Side" there was a consistant 3-4 degree advantage to the Aero cores.
So with that information we can deduct that the same temp of heat was being transfered on the "cooling side" but there was a efficiency on the "charge side" coming from somewhere. With an almost identical pressure drop on the charge side, the Aero core seems to get its advantage by the pressure delta on the "cooling side" from a denser core?
One the "Cooling Side" the outlet temps for both the Aero and GT2RS cores were virtually identical but on the "Charge Side" there was a consistant 3-4 degree advantage to the Aero cores.
So with that information we can deduct that the same temp of heat was being transfered on the "cooling side" but there was a efficiency on the "charge side" coming from somewhere. With an almost identical pressure drop on the charge side, the Aero core seems to get its advantage by the pressure delta on the "cooling side" from a denser core?
Registered User
Joined: Nov 2011
Posts: 800
From: Sumter, SC
Rep Power: 130
Now this is the question. ICs' location is far from optimal, at least compared to those in front of the car.
When building mine and planning the placement of / ducting to ICs (narrowbodied car) I discussed about the issue with factory trained Porsche Center head-of-maintenace. He claimed that most of the flow across ICs is caused by the vacuum at outlets than overpressure at inlets... Which aren't exactly in an overpressure area on the side of the car. Although the rear fender's shape has an effect here.
When building mine and planning the placement of / ducting to ICs (narrowbodied car) I discussed about the issue with factory trained Porsche Center head-of-maintenace. He claimed that most of the flow across ICs is caused by the vacuum at outlets than overpressure at inlets... Which aren't exactly in an overpressure area on the side of the car. Although the rear fender's shape has an effect here.
Thread Starter
|
Registered User
Joined: Apr 2006
Posts: 1,677
From: London (UK)
Rep Power: 156
Interesting results....one thing that caught my eye as particularly interesting is this:
One the "Cooling Side" the outlet temps for both the Aero and GT2RS cores were virtually identical but on the "Charge Side" there was a consistant 3-4 degree advantage to the Aero cores.
So with that information we can deduct that the same temp of heat was being transfered on the "cooling side" but there was a efficiency on the "charge side" coming from somewhere. With an almost identical pressure drop on the charge side, the Aero core seems to get its advantage by the pressure delta on the "cooling side" from a denser core?
One the "Cooling Side" the outlet temps for both the Aero and GT2RS cores were virtually identical but on the "Charge Side" there was a consistant 3-4 degree advantage to the Aero cores.
So with that information we can deduct that the same temp of heat was being transfered on the "cooling side" but there was a efficiency on the "charge side" coming from somewhere. With an almost identical pressure drop on the charge side, the Aero core seems to get its advantage by the pressure delta on the "cooling side" from a denser core?
Exactly.
You know your stuff aye. In fact a hell of a lot more than I do as you have studied this kind of stuff.
The reason you can get away with the higher density is down to the special fin design of the aero!
To have this kind of fin density in conventional cores would mean to much heat build up on the cooling side due to the pressure drop?
From what I have learnt or assume.
Last edited by GTRNICK; Nov 6, 2013 at 01:30 AM.
Moderator
Joined: Jul 2003
Posts: 22,235
Rep Power: 1227
GTRNICK .... I can see you are on a IC mission. 
Find yourself a copy of A. Graham Bell's book, Forced Induction and Tuning. There is an extremely good chapter on intercooler science. What works, doesn't work so well and why.
Find yourself a copy of A. Graham Bell's book, Forced Induction and Tuning. There is an extremely good chapter on intercooler science. What works, doesn't work so well and why.
Thread Starter
|
Registered User
Joined: Apr 2006
Posts: 1,677
From: London (UK)
Rep Power: 156
cheers CJ
Holy smokes they are £200
Ill order one
Moderator
Joined: Jul 2003
Posts: 22,235
Rep Power: 1227
About five years ago I paid under $50.00 for mine. I saw a copy in very good condition on the alibris website for about $150.00 plus shipping.
I saw a copy in very good condition on the alibris website for about $150.00 plus shipping.
