 |
924Board.org
Discussion Forum of 924.org
|
| View previous topic :: View next topic |
| Author |
Message |
ideola
Joined: 01 Oct 2004
Posts: 15548
Location: Spring Lake MI
|
 Posted: Mon Feb 18, 2008 9:56 pm Post subject: Intercooling for Ideola's Ultra Wide Body 931 |
 |
|
With the hybrid turbo out for assembly, and head work queued up and ready to go, it's time to get serious about finalizing the intercooling design. I currently have a Holbert-style IC, a 951 FMIC, and a Ford T-bird TMIC. At present, I am leaning heavily toward the 951 unit, given several conversations I've had with Bass GT, and looking at his configuration and results.
However, I still have one major reservation with the 951 unit that I can't get past. Looking at this diagram in Maximum Boost:
Considering that the 951 is configure like the top core depicted above, it sure seems to me that a superior design could be accomplished by building a custom intercooler. For example, from Bell Intercoolers, I could order a custom FMIC with the following dimensions: core=3" deep x 10.5" across x 6 " high, which is capable of flowing over 600 CFM, which should be more than enough for my target of 350 BHP. I would choose end tanks configured as in the bottom core depicted above (only with the inlets on the oppposite sides). It's also worth noting that I could increase the frontal area fairly dramatically and still stay within the space constraints in the front of the car.
Some additional points to consider is that the 951 unit I have will need to modified anyway because the stock end tanks are highly restrictive. For about the same price that Lindsey Racing charges to modify a stock 951 unit, I can get a custom IC with the same or better efficiency, the same or better flow rate, and less pressure loss. Also bear in mind, the front of my project car is already pretty much butchered, so regardless of which route I go, I will have the same effort in terms of bracketry & mounting.
From a bang-for-the-buck perspective, the custom approach looks very appealing. Is there anything I'm overlooking?
_________________
erstwhile owner of just about every 924 variant ever made
Last edited by ideola on Sat May 09, 2009 4:40 am; edited 1 time in total |
|
| Back to top |
|
 |
tuurbo
Joined: 08 Aug 2007
Posts: 1446
Location: East Windsor, New Jersey
|
| Posted: Mon Feb 18, 2008 10:57 pm Post subject: |
|
|
Cost? 
The Lindsey units are tried and true - I think it boils down to your calculations. If you're right about the numbers, your home built route is clearly best. |
|
| Back to top |
|
|
Raceboy
Joined: 01 Mar 2004
Posts: 2326
Location: Estonia, Europe
|
| Posted: Mon Feb 18, 2008 11:50 pm Post subject: |
|
|
Home built is almost always better.
But, for most applications and ease of use stock 951 IC is up to the task. Some flow tests revealed that LR end tanks did NOT increase flow capacity and did not decrease pressure drop compared to stock IC.
_________________
'83 924 2.6 16v Turbo, 470hp
'67 911 2.4S hotrod
'90 944 S2 Cabriolet
'78 924 Carrera GT replica
'84 928 S, sold
'91 944 S2, sold
'82 924S/931 "Gulf", sold 
'84 924, turbocharged, sold.
http://www.facebook.com/vemsporsche |
|
| Back to top |
|
|
tuurbo
Joined: 08 Aug 2007
Posts: 1446
Location: East Windsor, New Jersey
|
| Posted: Tue Feb 19, 2008 12:04 am Post subject: |
|
|
| Code: | | Some flow tests revealed that LR end tanks did NOT increase flow capacity and did not decrease pressure drop compared to stock IC. |
I did not know that!  I was propogating the hype and didn't realize it. Once you mentioned it, I remembered that the origin of that belief in my brain came about from word of mouth reports and scuttlebutt, not actual flow tests. Thanks for correcting me. 
_________________
1980 924 turbo, MSD, Meth. Inj, otherwise stock. |
|
| Back to top |
|
|
ideola
Joined: 01 Oct 2004
Posts: 15548
Location: Spring Lake MI
|
| Posted: Tue Feb 19, 2008 1:00 am Post subject: |
|
|
Yeah, I'm a little troubled by the lack of quantifiable figures for the 951 setup, modified or not. At least with a custom approach, the manufacturer is compelled to produce some numbers. I know that the 951 unit is being used on some pretty impressive setups, both 944s and Bass GT's car. But the issue of pressure drop still concerns me.
Also, to be clear, Lindsey has two units available: the improved stock version, and a custom unit that looks similar to stock, but is actually a completely custom-built IC. It is configured similarly to the stock unit but has a larger frontal area and requires their Stage V header panel. It still doesn't overcome what I perceive to be an inherent design flaw (the orientation of the runners).
One little, minor clarification is that I am considering a custom-built alternative, which is not the same as home-built.  Bell has a configurator that lets you specify the core that you want, along with the end tanks, and then they assemble it for you. The core I'm looking at is ~$220. I'm still awaiting a quote for the complete assembly.
Incidentally, a similarly configured, off-the-shelf option is available from PWR.
Its dimensions are 6" x 18" x 2.87", rated at 900 CFM , at a cost of $500. It's probably a bit too large for my target BHP; since I don't want to induce any unnecessary lag (due to the larger volume), I would lean more toward the Bell custom-built unit at this point.
_________________
erstwhile owner of just about every 924 variant ever made |
|
| Back to top |
|
|
Rasta Monsta
Joined: 12 Jul 2006
Posts: 11724
Location: PacNW
|
| Posted: Tue Feb 19, 2008 2:00 am Post subject: |
|
|
This is the next step for Shaggy. . .man, I can't wait. The heat exchanger and pump are installed, once the battery is moved it will be pretty easy to wrap up.
Not a hijack, just excited.
rasta
PS: Dan, are you grafting in the 951 rad subframe to hold the IC? If so, pics, etc, would be bueno (no offense to Tigger's phone cam shots!)
_________________
Toofah King Bad
- WeiBe (1987 924S 2.5t) - 931 S3
|
|
| Back to top |
|
|
flosho
Joined: 01 Jul 2004
Posts: 3155
Location: Eau Claire, Wisconsin
|
| Posted: Tue Feb 19, 2008 2:10 am Post subject: |
|
|
Dan, I'm sure you've talked to guys over in the 951 world? I believe a lot of those guys don't even upgrade the stock 951 unit around 300-350whp if I'm not mistaken.
I guess it also comes down to intercooler placement. Your custom intercooler design would be a front mount, while the 951 would probably be in the nose.
Just my opinion, but I'm not a big fan of the endtanks on the top/bottom, but thats purely a cosmetic preference that I don't think looks as good as the standard side end tanks.
_________________
[This Space For Rent] |
|
| Back to top |
|
|
ideola
Joined: 01 Oct 2004
Posts: 15548
Location: Spring Lake MI
|
| Posted: Tue Feb 19, 2008 3:20 am Post subject: |
|
|
| Rasta Monsta wrote: | | are you grafting in the 951 rad subframe to hold the IC? If so, pics, etc, would be bueno (no offense to Tigger's phone cam shots!) |
No, I'm not planning to do anything that extreme. The two options on the table are to go with a 951 FMIC and mount it pretty much like Bass GT has done. The other alternative is to play around with the positioning of the rad, and to put a custom, conventional FMIC somewhere in the front of the car. I have the holbert location available due to my ultra wide fenders, and lots of other space up front due to the front valance design. In any event, I will probably just fab up my own bracketry, rather than going to all the trouble to graft in a 951 subframe...just seems like overkill to me. When I get to that point, I'll take lots of pix as always
| flosho wrote: | | Dan, I'm sure you've talked to guys over in the 951 world? I believe a lot of those guys don't even upgrade the stock 951 unit around 300-350whp if I'm not mistaken. |
Well, I don't doubt that the 951 intercooler will work. I know that it will. But my point is this: the 951 unit will require mods, no question about it, and those mods will cost me $$$. If I'm going to spend money, why dump it into an inferior design to begin with? I'd rather put the money into an optimal design.
From my research, I just feel like the downflow design (as its referred to) provides superior pressure drop characteristics, and also provides greater frontal area, both of which will contribute to a better intake temp-to-boost level ratio. Lindsey's Stage V unit reduces pressure drop from 135 to 120 (when compared to their modified stock unit). Based on Maximum Boost and other sources, it just seems like a downflow design would be better. In my case, I'm not constrained by sticking with the original 951 dimensions, mounting points, or plumbing (which were ALL design contraints factored in by Lindsey). Since those aren't constraints for me, for the same amount of money (or significantly less...the Stage V costs $1200) I think I can come up with a superior FMIC.
As for the cosmetics, if I go with a custom FMIC setup, my baffling and valance will hide the end tanks anyway...and while cosmetics are important to me, optimal performance will override cosmetics most of the time
_________________
erstwhile owner of just about every 924 variant ever made |
|
| Back to top |
|
|
Raceboy
Joined: 01 Mar 2004
Posts: 2326
Location: Estonia, Europe
|
| Posted: Tue Feb 19, 2008 5:32 am Post subject: |
|
|
Vertical intercooler cores are MUCH better than usual ones and they are indeed the way to go. On our cars they mean shorter intake plumbing also.
As for cores, the best ones are Spearco and Treadstone vertical cores.
_________________
'83 924 2.6 16v Turbo, 470hp
'67 911 2.4S hotrod
'90 944 S2 Cabriolet
'78 924 Carrera GT replica
'84 928 S, sold
'91 944 S2, sold
'82 924S/931 "Gulf", sold
'84 924, turbocharged, sold.
http://www.facebook.com/vemsporsche |
|
| Back to top |
|
|
ideola
Joined: 01 Oct 2004
Posts: 15548
Location: Spring Lake MI
|
| Posted: Tue Feb 19, 2008 11:55 am Post subject: |
|
|
| Raceboy wrote: | | Vertical intercooler cores are MUCH better than usual ones |
And here's why...
OK, guys (& gals), I just had a breakthru in understanding this stuff. Sorry if I'm thick, but after reading & re-reading Max Boost chapter 5, I think I finally figured it out, and I am more convinced than ever that I do NOT want the 951 IC for my race car.
One of the critical aspects of determining proper intercooler sizing is the surface area exposed to the charge air. In other words, the surface of the IC that is hidden under the end tanks. On a stock 951 intercooler, that surface is 5" x 5.5" = 27.5 sq. in. According to Mr. Bell, on average, only 45% of that surface area is actually available for the hot charge air to enter the tubes. That means on the stock 951 unit the "internal flow area" is only 12.375 sq in. If you plot that on Mr. Bell's chart, you can see that that internal flow area is good for just shy of 200 BHP / 300 CFM:
Furthermore, if you plot my BHP goals (~350), you will see that a much greater "charge air face" and "internal flow area" is required:
According to the chart somewhere around 27 sq in internal flow area, which calculates to about 60 sq in of total charge air face. If I use a 3" thick core, the length required would be 20 in long!!! With conventional cores, the ONLY way to go is a downflow design.
However, there is another alternative. I went back and looked at the Holbert-style IC I acquired with the original donor car. This IC is truly huge, especially in light of this new knowledge. It measures 7x7x14. It's charge air face is a whopping 98 sq in, with an estimated internal flow area of 44 sq in. That literally puts it off the chart (Mr. Bell's that is!!!) in terms of the HP / CFM it could accommodate. Sheesh. Maybe those Holbert guys new what they were doing!!!
What this does is pretty much eliminate the 951 IC for my application. That is not to say that the 951 unit wouldn't be suitable for 200-250 BHP ranges...it clearly is. But for 350 BHP, I think it is out of the question. It also puts the Holbert core back on the table, although it is HEAVY, and will require all new end tanks and seals. It has other drawbacks too: being thicker, it will cool less efficiently; it effectively cannot be put in the front of the car; the side location will tend to skew corner weight (did I mention it's HEAVY?).
The upshot is that the downflow design allows for much greater internal flow area of the charge air, which overall introduces much less restriction, which means less introduced heat and less impact to boost level at the intake. The interesting thing is that Max Boost gives no mathematical guidelines to the "frontal area", which is the face exposed to onrushing air when the car is moving. All it says is that everything else being equal, more frontal area = more cooling capacity. So the height of the intercooler, while still important to overall intercooling function, seems to be of much less concern than the width and depth!
The other interesting thing is that Lindsey's Stage V unit has a charge air face of 6.5" x 4.5" = 29.25 sq in, giving it an internal flow area of 13.1 sq in., only marginally better than the stock unit!!! It is no wonder, then, that they were only able to achieve 280 CFM with a core that can accommodate 1000 CFM!!! IMO, it is a poor design to begin with.
_________________
erstwhile owner of just about every 924 variant ever made |
|
| Back to top |
|
|
endwrench
Joined: 07 Dec 2002
Posts: 1631
Location: Victor, Montana
|
| Posted: Tue Feb 19, 2008 2:51 pm Post subject: |
|
|
Corky has a lot of great formulas that get you in the ballpark but most seem to be based in the "lab" and not the "real world". Case in point would be all the theoretically piss poor 951 IC's out there making mega HP. Or the scewed numbers of the PWR IC in your example above. If my math is correct, it only flows about 450 cfm according to Corky's formula and not the 900cfm claimed.
Don't get me wrong, I think you're doing the right thing figuring this all out and deciding on the best approach to an efficiant IC. I just wouldn't let it eat your brain or your wallet .
Todd
_________________
'79 924NA. Rebuilt 9.5:1, MSDS header, Mega Squirt Injection, MJLJ-EDIS Ignition, 1.6L Whipple Charger and Intercooler, 10lbs Boost, 944 Trans, Custom HD Clutch.
"simsport" said....superchargers are better than turbos its official!.... |
|
| Back to top |
|
|
-nick
Joined: 16 Nov 2002
Posts: 2699
Location: Cambridge, MA
|
| Posted: Tue Feb 19, 2008 3:55 pm Post subject: |
|
|
| ideola wrote: | | One of the critical aspects of determining proper intercooler sizing is the surface area exposed to the charge air. In other words, the surface of the IC that is hidden under the end tanks. On a stock 951 intercooler, that surface is 5" x 5.5" = 27.5 sq. in. According to Mr. Bell, on average, only 45% of that surface area is actually available for the hot charge air to enter the tubes. That means on the stock 951 unit the "internal flow area" is only 12.375 sq in. |
Not quite true with the 951 cooler. The typical cooler design is a rectangular frontal area and a narrow depth. The 951 cooler is almost square from the side. You can almost think of it as a tube - as air enters it expands evenly in all directions to fill the cooler and also moves evenly through it (relatively).
A narrow depth rectangular cooler is not going to get as much air movement at the top and bottom cooling rows which is probably where Bell gets (part of) the 45% area number from.
That's the hidden advantage of the 951 cooler. The "rear" rows of the cooler aren't getting as much cool air, but the entire volume of air is smoothly passing all the cooling fins.
Both designs have trade-offs, but the 951 cooler isn't as "inferior" as you would guess.
Either way, plan and install the basics. You're not going to get things perfect the first time around. Even Steve's mean machine has had a few iterations. That's what makes this quest so pricey...
Still - lookin' good so far!
_________________
1980 931S
15psi boost, MS-II, EDIS, 951 IC, custom intake, Ford 5.0L throttle body, Forge BOV, WB o2, G31 w/LSD, 964 wheels, 968 rear sway, Bilsteins, 200# Welt. springs. A laptop, and a partridge in a pear tree.
1991 964 C4 Cabriolet |
|
| Back to top |
|
|
bass gt
Joined: 02 Dec 2004
Posts: 971
Location: Johannesburg for now!!
|
| Posted: Tue Feb 19, 2008 9:39 pm Post subject: |
|
|
Time for me to pile in guess.
Ok, re the qustion, "is the 951 unit idea/", well probably not. HOWEVER, it does seem to work. Porsche have always had a history of fitting good quality IC's, hence why the non stock 951's make such big numbers. BUT. Can they make thos numbers in an enviroment other than a dyno? Not so sure. Heat soak is a real issue.
So what are our options?
1) Fit a bigger IC. Easily done, but costly, and may need the radiator relocating to achieve.
2) Improve it's performance. Now this is the interesting one. I found on my racer, boxing in the IC and venting it out the hood via a low pressure area REALLY worked. I mean really!! i was achieving 76% IC efficiency at full throttle on a very large open circuit, where i was WOT for a long time. The charge temps were never an issue.
Then there is the option of IC sprays, if alllowed.
The trouble is with a big IC in the front, is it masks the baseline issue. To make an IC efficient, it neeeds airflow, not just air hitting it's frontal area. If this air goes nowhere, then the IC will start to warm up. I believe the IC in front of the rad is never a good idea. It prewarms the water, limits the through flow ect. And if you look under the front of a '24, the air comes under the front apron, and up into the engine bay. If the pressure differential is not too great, little actual flow will be achieved.
So, in my opinion, an efficient IC is produced by making the max volume of air transit the core, not just hit it.
And i have measured massive drops in the air charge temp following the hood vent mod. IMHO, this makes a moderate IC perform akin to a much larger unit, without the downsides. Give it some thought.
Steve
_________________
Front Wheel Drive is the Devil's work. |
|
| Back to top |
|
|
ideola
Joined: 01 Oct 2004
Posts: 15548
Location: Spring Lake MI
|
| Posted: Tue Feb 19, 2008 11:09 pm Post subject: |
|
|
Guys
This is all fantastic stuff. I appreciate all of the very constructive feedback!
| endwrench wrote: | | Corky has a lot of great formulas that get you in the ballpark but most seem to be based in the "lab" and not the "real world". Case in point would be all the theoretically piss poor 951 IC's out there making mega HP. |
Excellent point Todd! And I think Bass GT's real world numbers are a perfect case-in-point.
| -nick wrote: | | A narrow depth rectangular cooler is not going to get as much air movement at the top and bottom cooling rows which is probably where Bell gets (part of) the 45% area number from. |
Nick, if I'm not mistaken, the flow of air at the top and bottom rows of the IC are a function of the end tank design, something that is widely regarded as a weakness on the 951 unit. In my case, one option I am looking at is creating diffuser-style endtanks (like the diffuser I created for my custom intake manifold) for the Holbert-style core, which would help to evenly distribute the charge air across the core face. Also, take a look at this photo of my Holbert IC with the end tanks removed:
On the left hand side, you can see the "charge air face", and how small the channels actually are for the charge air to pass thru. This is what Corky is referring to...even though the entire "charge air face" is HUGE (in this case, 98 in sq), less than half of that area is actually open for the charge air to pass into (using his 45% rule of thumb, approx. 44 in sq).
As for narrow vs. thick intercoolers, I suspect that there are few advantages for cooling in going with a thicker core. The narrower the core, the better the cooling because the rear tubes in a thick core are exposed to increasingly heated air as it pass thru the core.
The 951 unit is 5" thick. My Holbert unit is 7" thick. A thicker core is a tradeoff that is required in order to produce the necessary internal flow area. In the case of the 951 unit, going thicker was the ONLY option due to the space limitations of where it is mounted. Workable, but less than ideal. Everything else being equal, a 3"x8.5" core will cool more efficiently a 5"x5" unit...obviously, that was not an option given the location of the 951 unit.
| bass gt wrote: | | ... Improve it's performance ... |
Steve, you always have a way of getting me to put that 951 IC back on the drawing board!!!! In fact, while I believe the downflow approach is best, I'm struggling to figure out how to get the "optimal" size installed and not encounter the very issues you brought up regarding flow of air, blocking the rad, etc. etc. I really need to get my hands on the Lindsey Stave V header panel, which I want badly from a cosmetics perspective. This will tell me how much room I have to work with up top. I'm toying around (in my head) with some alternate arrangements, ducting, and orientation for the rad that might allow me to accommodate a large FMIC. Here's a hint:
Again, fantastic feedback everyone...keep it coming!
_________________
erstwhile owner of just about every 924 variant ever made |
|
| Back to top |
|
|
ideola
Joined: 01 Oct 2004
Posts: 15548
Location: Spring Lake MI
|
| Posted: Wed Feb 20, 2008 12:17 am Post subject: |
|
|
Some very interesting supplemental reading.
http://www.bellintercoolers.com/pages/techFAQ.html
One important caveat about Max Boost (in addition to what Todd mentioned regarding lab vs. real world): that book is 10 years old, and it has been suggested to me more than once than an update is in order. While many of the theories may still be practical, it's also likely that some of the content is a bit dated.
[EDIT]
Whoa, check this out...I didn't even make the connection before!!!
| bellintercoolers.com wrote: | Corky Bell and Gerhard Schruf, the founders of Bell Intercoolers, are engineers with a deep appreciation for the performance enhancements available from a well designed intercooler system.
A recognized authority in the field, Corky Bell has served as engineering consultant on many successful racing efforts including the 1986 IMSA GTU class championship Porsche 924 Turbo, the SCCA Escort Series champ in 1994 with the Nissan 300ZX, and a Bonneville record setting Mazda RX-7. Bell is also the author of two publications on the subject, Maximum Boost and Supercharged!. Both providing essential, hands-on, advice for novice forced induction turbo and supercharger system designers.
Gerhard Schruf has spent the vast majority of his engineering career working in the field of forced induction. Having developed turbocharged engines for marine, industrial, street and racing applications, with power outputs ranging from 50 hp to 40,000 HP. Gerhard was also instrumental in the initial development of the unique, pressure-wave-supercharger, the COMPREX. A highlight in his career was working with Gilles Villeneuve and the Ferrari Formula One team, providing forced induction expertise during the teams early ventures into the turbocharged era. As an author and co-author of several SAE-Publications his focus always was on performance, but in conjunction with fuel-economy and exhaust emissions for street applications. |
_________________
erstwhile owner of just about every 924 variant ever made |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
924 FAQ (Technical)
