911 Cooling Shrouds - Improving Internal Air Distribution
 

Hi all,

I have been playing around with this shroud for a few weeks now and along with my datalogger, have what I believe to be an accurate baseline to compare any modifications to my FAT 911 cooling setup. I'm using stock T4 under cylinder tins.

I have tuned my engine the best I can and it seems to be running well. I have quite a few graphs produced by my logger and have noticed some interesting points:

1. With the unmodified shroud and stock 11 blade fan/ alternator ring, as the engine warms up, #1 and #2 warm up slower than #3 and #4. The slope of the graph for these temps over time indicate that #4 warms up twice as fast as #1. (Is #1 receiving twice as much cooling air as #4 ? :shrug: )

2. After a hard run up to 140 kph from a dead stop, and then coming to a dead stop and idling (900 rpm) for 5 minutes ... #1 cools down twice as fast as #4. The cylinders seem to cool down with #1 cooling the fastest, then #2, #3 and then #4. #1 cools at the rate of -0.35F / Sec and #4 at -0.18F / Sec.

3. By taking the fanbelt off and starting the engine and letting it idle, all 4 cylinders warm up at the relatively the same rate ... making sure its not a tuning issue causing the temp imbalance.

4. Driving the car in 5th gear at 115kph (3000 rpm) vs 4th gear at 88kph (3000 rpm) creates temps 25 degF hotter (engine load im thinking)

I'll post some pics of how the 911 alternator ring has 4 built in diverters that seem to direct airflow in this shroud. The airflow of a 911 axial fan spins clockwise and I have a cool picture of an axial fan showing the airflow pattern and air distribution in free air.

My plan of attack is to place diverters on the back of the alternator (alternator cap ) similar to a stock 911-6 setup to start, and working with the alternator ring diverters

So ... does anyone have any suggestions on how to equalize the airflow (swapping to a DTM does not count :D) ? I have my own ideas but it would be great to place all of this information in one thread, where everyone could benefit from this information. I'm not looking to market this product, just love the look and sound of this setup and want it to cool more effectively.

Sandeep

Hey Sandeep,

Cool idea, lets give Jake a head start for s/th he plans on doing in the future as well :D

Funny, I experienced roughly about the same findings as you did. They are also consistent with the data that Jake had on his site (and this forum) AFAIK some time ago, just after his tests.

I say roughly thesame, because my temp differences between 1 and 2 AND 3 and 4 were very close, but that may partly be due to the fact that I used (porsche 964) Nikasil cylinders.

I also installed the stock diverter (it just fits a BAS shroud with only very minor mods) and even enlarged it. Utimately, it made nr. 4 cylinder run cooler than nr. 3 : Oeps :o

Since my oil temperatures kept a stunning 85 degrees Celcius (185F) at all revs (!) and since I believe that the cylinders are a major factor in cooling the oil (apart from the actual oil cooler of course) and because my top cylinder fins are very small, which creates a huge barrier for air to flow over the first cylinderhead fin, I went as far as to put an almost horizontal plate over the cylinders to increase flow towards and over the heads. I will have pics in a few days I hope.
Anyway, that gave a 50F lower head temp! with the same oil temps. 1 and 2 are still lower in temp overall than 3 and 4, but that will be the goal for step 3 ;)

The difficulty is that there are a lot of factors into play here. For instance, I have used the 11-blade fan with the 245mm diameter blades and a 90mm pulley on the fan and a 140/145mm pulley on the crank.
I have also tried the 12-blade fan which has a 100mm fan-pulley and 255-ish fan diameter. The 12-blade also has a totally different housing air-outlet; more straight than the 4 curved ones you spoke of. I have a picture of the latter one if you want it posted (I have before). I have done too little runs with the same conditions to be able to say s/th conclusive about the differences on that one yet.

- end of report - (for now)

Regards,
Walter

Wally,

Its great to hear you are having success with your 911 setup. I too believe that the stock 911-6 diverter on the back of the alternator is key. The 964 fan diverter is indeed a work of art, perhaps an improvement over the SC 11 blade setup.

I have ordered an SC diverter, hope to have it in a few weeks. I will install it with mods if need be and see where my temps are going to end up. Up till now, my max temp difference between #1 and #4 are averaging 70 degF while 1-2 side and 3-4 side are up to 60 degF apart. 1-3 and 2-4 are 24 degF apart because 3-4 are higher than 1-2.

Please post pics of that horizontal plate you are talking about .. would be interesting to see.

Sandeep

Sandeep, the first step you should take would be to increase the drive ratio like we spoke about when I met you. This is easily done with the later model Porsche top pulley.

The shroud you are working with has a few physical challenges that will limit your work that can be done without actually modifying the shroud's skin.

Your primary objective should be to effectively divert more air flow to the #2 and #4 cylinders as they are the hottest due to being almost directly under the fan and air generally does not like making a 90 degree bend... At least without being directed efficiently.....

I am glad to see that you are making this development- You will undoubtedly see just how difficult air can be to manipulate, I never would have dreamed that it was so difficult to work with.

The differences in temps arenot as great as I saw. This could be directly related to two things:
One is your climate, its cooler where you are on ambient temps and also that your gearing may be shorter and not placing as much load on the engine thus not showing all the data to its max extreme. In some of my tests I installed 28" tall tires on my car to effectively gear the vehicle up to place more load on the engine and create more heat with less fan speed- thats when the differences really came out in the car just like they did on the dyno, and even more in some instances. Doing this allowed me to develop the TI shroud to even a higher level than most people would ever be able to push it to..

Also the smallest engine I tested was 153 BHP-

Kep it up, you'll learn a ton!

Sandeep,
This is the 12-blade backside:

http://www.keversite.com/fotoalbum/f...2004231006.jpg

Note that the original 964 cone is missing here (still need to get that one :o )

I really like the way the air is directed out in a straigth way. Also that that is done by a multitude of air diverters, which are also longer overall than the 4 on the old fan houses. Theoretically, it would make it easier to divert the air to 3 and 4, if half the work is already done inside the fan housing!

It would only be logical that porsche would upgrade its parts during the production of its aircooled boxers :)

Greets,
Walter

As promised, here's a picture of the original cone air deflector of a porsche 11-blade 245mm fan/alternator combo:

http://www.keversite.nl/fotoalbum/fo...2005194727.jpg

And here's a preliminary air deflector I made to direct more air to the heads and to overcome the barrier of the first cooling fin. This is because the fins on the nikasil 964 cylinders I use are very small (therefore low, so the heads fins stick out a lot). It gave a 50F lower head temp :D
The 964 cylinders have M6 thread on top of the cylinders, so I could mount to that. Originally on the porsche 964 those holes are for mounting the knock sensor bridge...

http://www.keversite.nl/fotoalbum/fo...2005195149.jpg

The original air deflector/cone did gave slightly lower head temps for 3 and 4, but not as much as I thought they would. Enlarging the area of the cone air deflector did work, but still not enough to my liking compared to the temps of 1 and 2.

Work in progress ;)
Walter

911 fan shroud
 

Thank's alot wally that's the 911 fan that i have but one of my deflectors is missing.

Thanks for the comments Jake.

I reprogrammed my logger to flash an LED when the hottest temp from any cylinder reaches 375F. On a full acceleration run from a dead stop, gas pedal to the floor in each gear, 5500 rpm shift, #4 hits 375F at 4th gear, 3800 rpm consistently.


Thanks for the pics Wally.

In the diverter picture, it looks like you may only be getting 3/4 of the airflow from that quadrant as it looks like some air may getting past that last diverter to the left in the pic, if you compare it to the alternator guide diverter. I know you are using the 964 setup now though.

My idea is to use 3 diverters, one each directly behind the stock alternator ring diverters and one in between to make 2 channels to capture all of the airflow of that quadrant of the alternator ring. This will mean changing the positions of the factory alternator cap diverters.

The factory alternator ring diverters look to be of an airfoil design and I will try to incorporate the airfoil shape into my diverters placed on the alternator cap. I will keep placing diverters around the alternator cap, taking more air from each neighboring quadrant on the alternator ring until I can equalize the temps. That is the plan anyways ! :D Who knows if it will work until you try it. :eek:

Sandeep

Yes, that should be 'relatively' easy to do. However....it may result in that you take air away from your cooler running 1/2 side to a level where both sides run equel temps, but generally too high!
I think you need to direct the air also to flow from top to bottom over the heads. That is the direction of the fins after all!
To accomplish this, I think I also need more room in my shroud, at least more above the heads. My simple 'steel plates' over the cylinders won't cut it, but it is a start in that direction. I also think that dumping the air from above is the key succes element of the type 1 cooling way: its in the direction of the fin area (duh..!)
The more advanced shrouds (which I consider the Klaus, Remmele and RMB ones, maybe even Gerd Weiser's shroud) do seem to have more room above the heads in the shroud and even more above the 2 and 4 cylinder than over the 1 and 3 cylinder. Not by chance I think. Those shrouds are also 'flatter' right behind the fan. Probably to even out the left/right side flow better. Ever noticed that?

In the end, you very likely to end up redevelop/design your own shroud. Jake is probably right about that :( :D

Regards,
Walter

Great points Wally. By equalizing the temps in the way I mentioned, you are right ... they won't nessesarily be cooler (hotter I would think) but they may be even :)

The airflow must be directed from top to bottom as you have mentioned. Check out Remmele's website under his pictures section and look at the pic of the inside of his shroud ... it is quite flat right behind the fan and it looks like this dumps the air vertically between the cylinders.

I'm sure this shroud uses the 964 setup and with the vanes you have shown on the 964 fan, you are much further ahead than I, I would think, down the road to improving cooling.

Perhaps a 964 fan setup is in my future ? :laugh:

Sandeep

Thanks Sandeep :o
The 'new' curved fan blade houses and fans/generator combo's you can buy at bug stores 'everywhere' do NOT use the housing of the original 964!! So, AFAIK, you still get the old 4-vane style fan housing. That fan housing is a new mould, where a Valeo generator fits in nicely. Also, the original 12-blade fan is turned down to a 225 size if I am not mistaken. Anyway, a considerally amount of original fan blade is removed from the tips. I don't know how that effects cooling air.
Probably 'they' thought that the amount of air was more than enough, maybe even to much (as Jake also has said sometimes), so it would be no problem. I don't know...Most 'old-style' german tuners still prefer the old 11-blade I heard once...
So don't be put down by the 11-blade: its damn good; it just needs to learn to behave in the presence of our type 4's :D

Regards,
Walter

I feel left out. I don't have any data to share but i do think if done right the 11 blade will work. Samcat had a 11 blade setup that di work well on a stock 2.0l type4.

I have finally received my 11 blade fan factory diverter in the mail ... time to start playing and logging the data :D

Its a good thing the 911 fan/alternator setup can be removed from the shroud and unwired in about 3 minutes :agree:

Sandeep

Cool!
Curieus what your accurate measuring device spits out ;)

That datalogger of mine is using I/C's accurate to 1.0C, converting to F, accurate to 1.8F.

My VDO oil temp gauge/sender is off by about 12F at 180F oil temp measured in the DS tank. My logger reads 180.8F, and an infrared thermometer measures in at 179.8F :cool:

Sandeep

Coincidence ? 911 Alternator Cap W/ FAT Cooling shroud
 

Wierd .... I just installed the factory 911 alternator cap with 2 diverters into the FAT shroud .... it fits with NO modifications and the back and top of the diverters hug the back of the shroud .... almost like the shroud was made to fit this like a glove. :cool:

Now for a road test/logging session to see if it amounts to anything positive.

Sandeep

:D

Looks like that diverter does make a difference after all ... 30degF cooler on average for 3&4 and 15degF cooler for 1, but 2 stays the same on average :shrug:

I think that 3&4 are getting more air now and 1 is getting less (but might have been getting too much in the beginning and the shroud may not have been able to get rid of the air around #1 :confused:

Out for some more testing and more diverters ! :laugh:

Sandeep

Yes, the same with the BAS shroud: I thought about it as well and I think they probably made/modelled the shrouds of an original 911 shroud for the part thats directly behind the fan housing.

Interesting data!

Greets,
Walter

Wally,

I've made quite a few changes to the diverters on the alternator cap but I can't seem to get the temps on 3-4 to drop anymore, 3-4 are still about 40 degF higher than 1, and 2 is about 25 degF higher than 1.

I am going to try the T1 under cylinder tins now, but still wonder why #1 runs so cool with the T4 tins ... can't get it above 325 degF no matter how hard I try.

#1 has a nice airflow path from the alternator ring ... almost a direct shot of air to the head. Perhaps the diverters are directing the air to the cylinders and not the head on the 3-4 side :confused:

I will try the T1 tins first and then try your method of the horizontal plate over the cylinder bank as the airflow may be having a hard time getting past the first fin on the head towards the case. It is much taller than the last cylinder fin towoards the head and perhaps this is an airflow barrier.

More testing ! :D

Sandeep

Yeah, my thoughts also and with the scarsely finned (at the top anyway) 964 nikasils, that may be even more true.
The shroud should be a bit higher at the heads therefore imho, at least at the front cylinders (2 and 4) to have room for better air guidance and to direct the airfow 90 degrees there, so the air comes from the top down eventually...

My cylinder plates may work with nikasil cylinders, but you might get in trouble with steel cylinders because of the blocking of air flow essentially, so be carefull!
When you block off the cylinder air too much, you should see an increased oil temp very soon,...I think.

Take care,
Walter

BTW I have decided to have the heads 'done' after all and even increase cylinder capacity quite a bit (its never enough) while we're at it, so my testing has halted momentarily..

Update
 

An update ...

During this week of testing its been averaging mid 30C with humidity in Toronto so I'm happy with the results so far.

That SC diverter works wonders with this shroud along with T1 under-cylinder tins. I have modified the stock alternator cap diverters and added a 3rd diverter to the cap to direct the airflow more to #4. Initially I had too many diverters and the temps actually got worst the harder I drove it, but looked great when idling / light driving.

So I'm down to 3 diverters and here's the stats:

1. #1 and #2 run on average within 10F of each other.
2. #3 and #4 run on average within 13F of each other.
3. 1-2 bank runs on average 36F cooler than 3-4 bank.
4. 1-3 run on average 24F cooler than 2-4.
5. Hottest temp with current setup is 341F on #4 (Highway running, 5000 RPM in 4th - at this RPM I think there is too much air for the shroud to get rid of, but I rarely drive at 5000 rpm)

On my drive to work (20 mins with idling in traffic, 80kph top speed) highest temps for #1 are 289F, #2 301F, #3 312F and #4 319F

The interesting thing about this whole diverter setup is that all 3 diverters are on the 3-4 side of the alternator cap. 1-2 get a direct shot of air with no mods required.

If 1-2 run so cool (hottest temp for 1-2 with current setup average 323F) with no diverters/deflectors of any sort, I'm confident 3-4 can run just as cool.

Next step is to tweak the diverters a bit by placing tufts of yarn on the diverters and looking through the fan to see which way they blow. The stock diverters blew the air right to #3 cylinder and not the head.

After I get the air poining in the right direction, I will make the pulley mod Jake has recommended. I don't think I need to use Wally's horizontal tin mod as my shroud sits higher than his does, and it does not look like the airflow is being blocked by the first fin on the head towards the case.

Thats it for now !

Sandeep

Excellent work Sandeep. I would like to see how those mods held up in a higher load arrangement, what is your gearing??? Tire size"?

#1 is generally, no matter what is done the coolest cylinder- This holds true on all shrouds except the stock TI doghouse which robs its air for the oil cooler from just above #1 and heats it up 20 degrees in my testing.

#1 and #2 are on the same tangent as the fan so you have increased pressure and volume there, but normally the FAT shroud bakes #2 and freezes #1, but the smallest engine I have used for the testing was a 2270. I would be interested in seiing the plenum pressures associated with the shroud with these mods compared to what I saw here. Do you have an old Boost gauge laying around that will go to 15 PSI??

Your temps are pretty close together now and definately acceptable, as far as the deltas between the cylinders, but your stock engine is still running 25 degrees hotter than Zens 2316.... BUT it's DTM cooled.. LOL

I really like your work on the shroud, it makes me want to speed up my development and get to the 911 shroud faster, but it won't happen till 2007 at this rate.

Like I have said all along, the 911 shroud is not horrible, but off the shelf none of them have touched a DTM, and Sandeep is proving that as well as the fact that they can be made to work if you have the instrumentation and time to make the mods.

I love the fact that he is doing this work. If all goes well we will have a 911 system that's a bolt on. I don't mind cutting the back of my deck lid.

The system that I already have the molds for does not require any cutting of the decklid. Since the outside physical parts of the shroud have already been completed (Joe did this before he died as he was working on a 911 system that actually worked with even air distribution) I only need to take the time to test it the way that Sandeep has and right now I have other priorities that are more important like a heater box exhaust system and my Billet heads.

I don't know if we can ever get a 911 system that will beat a DTM, it is my goal to make them exactly the same but I do believe the DTM will win the challenge in the HP loss from the fan category due to the radial fan primarily. I have an idea for this that would release pressure within the upper plenums of the shroud at high RPM when pressure with the axial fan skyrockets and reduce the HP loss- You'll like it.

Sandeep is having an easier time doing his work than I had imagined. Sandeep you should send me that shroud for a heavy load test when you are all done toi see how the mods impact the cooling on a bigger engine on the dyno. I now have my logger in place and can measure the temps much faster and easier. That would be a fun comparison.

I'm using stock 914 gearing and using a 265/35/18 ZR 18 rear tire on an 18"x10" wheel ... that sucker with tire weighs 52 lbs. I'm also using a 930 rear rotor (309mm x 28mm) so I'm thinking that adds some additional load when compared to a beetle 165/80/15 rear tire and drum brake setup :D .. not sure how much more though or if it is significant.

No I don't but that is a good idea ... I will see what I can find at JJB 2005 on June 26th in the swap meet. How should I hook the hose up as I've never used a boost gauge ?

The DTM is the King of cooling, you've proven that already. I'll be taking my heads to Mark this winter for a refreshing and the jugs for a light hone ... I'll be looking for a RAT cam/lifters then as well ... maybe by then I can make up that 25F :D

Sandeep

I think the reason I am having an easier time is that I'm only working with 3 diverters .. 2 stock with modifications and 1 that I added, and not trying to engineer a whole system like you did. I can also visually see the results of air distribution by using yarn tufts and looking through the fan while it is running to see where they end up pointing.

I'll send it to you during the winter for testing when I am freshening up my motor.

Sandeep

I have a Race- Technology DL-1 logger with digital SPA instruments for quick reference.

The logger is set up for in car use and Dyno use by simply disconnecting the wire harnasses that are hard wired into the Bug, 914 and the dyno.

The logger even utilizes a GPS to track map the position of the car at the time the data was recorded, so I can literally watch the temps rise going up hills and etc all while measuring TQ and HP and even G forces...

I am having a real learning curve with this logger- It has so many features that its very confusing to get used to, but DAMN is it awesome (and expensive)

Sandeep,

Where did you order your 911 fan diverter from?

Thanks,

Randy

Lucky find on EBay :cool:

Sandeep

Another step forward
 

I had some spare time last night so I thought I would try Wally's horizontal tin mod on 3-4 bank only to see what a difference it would make.

I made a piece of tin 7" long by 4.5" wide and cut it in such a way that one 4.5" side runs parallel with C/L of #4 cylinder and the other 4.5" side runs parallel with #3. The attachment point to the case is the 2 - 6mm holes for factory tinware on top of the case under the shroud.

I then made a U shaped hole in the center of the tin to allow some air to get between #3 and #4, but the majority of the air would go over the cylinders and to the head.

Well ... this decreased the temps of #3 and #4 by about 15F :agree: So now the differance between 1-2 and 3-4 is 1-2 runs about 20F cooler on average than 3-4. I guess the cylinders DO catch the majority of the air flowing across them ... thanks Wally ! :agree:

I am not sure how this will affect oil temps (seemed fine on my testing run) or if this will create hot spots on 3-4 cylinders due to the decreased air to the cylinders.

I have about 40 logs now of different senarios ... 5MB each ... big files.

I think I may be at the end of these mods as the temps seem quite acceptable for the street. Next step is to go to the track ... no track days coming up till September though :( the spaces are all filled.

Sandeep

Here's my datalog from last night ... overall, I'm very happy with the results and I think this is the end of testing for me. Next step is to get out to a Driver's Ed but that will not be until September.



Now to finish collecting the parts for my Turbo project :D

Sandeep

I have considered making a small plenum and using perforated metal plates for the same purpose on each side of a cali setup.

This reduces the velocity, increases the turbulence, forces more air to the heads, and makes the pressure more uniform. (all very good things)

(Think "cool tins" on TOP of the jugs, with lotsa holes)

Cali setup has pretty good side to side, but on each side one jug runs hotter ~30-40 F, Probably the same across the head. (using IR setup)

Any chance we can get some pictures of the third deflector that you added to the 911 divertor? I am running the same set-up.

-Britain

I'll post some pics on the weekend of the setup ... pretty simple and nothing fancy about it.

Sandeep

Cool, thanks.

-Britain

911 shroud
 

Sandeep you can't end the project now i think that 911 shroud will perform as well as the DTM shroud.

:eek: That's right Sanddeep you cannot end now. You have done a wonderfull job and I am sure we are all excited about the finished product. :agree: I am in the Brampton area right next to ya. I have spoken to Jamie about you and I cannot wait to meet you. I am also running a T4
2.6 with Sharp Shroud. When can we meet???? :D I am off Sandy Beach which is off Brisdale one street over from you. U Say When & I am ready.
Thanks............




Big up yardie...........Mandeville area. :cool:

911 shroud
 

Hey Yardie,
Glad you could join in. How does that 2.6ltr run/feel,i am also thinking of running on in the near future. I am in portmore St Catherine. Do you have any problems with head temps and the sharpbuilt shroud. Who built your engine what's the top speed.

I have all the parts but havent built it as yet. This will be my winter project.
But it should be a trip for sure......Are you the guy at Vernam Field with the Burgendy Bug? I was living in the Manchester area and just moved back to the North 2 years ago and started my little project :agree:
Respect.......It's a 56 Oval with T1 Tricked out Tranny and The T4 78x103's

Hey Sandeep you gotta contact me! Hey I am one street over from ya. Just send me a link and will be ready any time......

Great work Sandeep :agree:
The 911 fan shroud needs just that kind of R&D. It is good to know someone can get it to work as good as it looks.