Printable View
https://rumble.com/v35jsi0-20230804-fatt-session-3.html
Quote:
After lunch - still dealing with heat issues; oil and water temps are good, but the intake charge air is high. Started playing with line through 3 and into 10 while continuing to short shift.
Unfortunately, the session got red flagged because a Porsche lost its brakes into T1 and found the tire barrier.
https://rumble.com/v35noyy-20230804-fatt-session-4.html
818 is home in the garage, intact. I cleaned the gravel and grass out of it and other than the missing roof scoop it seems intact. Still need to check fluids, pads and wheel bearings. I did notice a tiny glistening at the passenger transmission output after the first session but didn't see any actual leaking fluid.Quote:
Last session of the day, I was feeling pretty tired and didn't want to push the car too hard. Ran some laps, called it a day about 1/2 way through the session.
Printing a new version of the roof scoop, will fully duct the radiator inlet (the hood was rising a bit) and am looking at tunneling the turbo and exhaust to better isolate that heat from the engine compartment.
And of course, looking for my next track session!
Nice work! I can’t wait to be able to get mine out on track. I am not sure if you have done this but wrapping your headers, exhaust, and adding a turbo blanket will help try to lower some engine bay temps. This has worked well on other Subaru engines I have been around and something I am working on myself. Plus like you mentioned ducting your radiator and any other heat exchangers you have will help them work better. I have to think with the coupe you might also look into ducting your intake since the engine bay gets so hot. You might even consider adding some other engine vents to pull air through the engine bay to exhaust the hot air.
Already done, except the turbo blanket.Quote:
Originally Posted by blomb11
Yup! Intake is fed directly from the passenger side panel vent, which was enlarged. Once upon a time I had an aluminum airbox constructed that captured most of the incoming air and isolated the filter, I'm considering reimplementing that with an additional small outlet to a tunnel that houses the turbo and exhaust. Need to do some CAD work (cardboard, not computer) to see if it's practical.Quote:
Originally Posted by blomb11
Additional venting in the engine bay would probably be a good thing; I've looked at replacing the rear glass with a Lexan panel. If I really wanted to burn some money, I'd contract for a vacuum formed one with venting; but maybe I'll just go cheap and cut out some holes as a first attempt. But the rear hatch is right under the wing, so not sure how good it will be to disrupt that airflow...
I like how Gator's rear bumper cover came out after his mishap... might take inspiration from that also.
Been driving mine around for a bit now and yes, the engine compartment does get hot. Hotter than I like. I added a small fan to one of the intake ducts from the roof, and cut away part of the trunk to give the exhaust more venting area. But now realize that stuff like that doesn't help much without a place for the hot air to get out. I'm looking at some sort of heat shield that helps direct heat from the exhaust out the rear grill. But I don't have much grill area directly behind most of the exhaust, so hot air really has no place to go. An additional grill beside the stock one is probably needed.
I made rear hatch louvers and am not using the hatch glass. In addition I made a mold for some acrylic side quarter window scoops.
Would love to see pictures of that! If the roof scoop thing doesn't work out replacing the quarter windows with scoops was my next step.Quote:
Originally Posted by lance corsi
Yep they look wicked cool. They are actually 2 piece scoops. One piece is the transition from the forward edge of the window opening and it extends just past the roll bar because there is a little bit of usable window that I didn’t want to block off, then the scoop starts where the first piece drops off and covers the entire quarter window opening. I cut off the flange area that the glass was supposed to glue onto, just leaving the rest in place in order to give enough lip to attach my small clips that bolt thru that area. I countersunk flat head screws into the acrylic scoop and this is what holds everything onto the roof. I used some .030” thick stainless to fabricate the clips. Stainless screws hold everything together. I have since decided to make some more scoops from cf in case/when the acrylic breaks, although I do like the old school look of the clear acrylic as it allows a view into the jewelry box! Pm me if you want pics because I don’t have any to put on this forum with the proper protocol. Maybe I could get Bob-n-Cincy to post them so everybody can have a peek.
Obviously I had to make a mold for each side. And in the process, the inconsistencies of the roof were revealed, so I’m not sure if my sizes would fit all roofs. Each scoop had to be carefully fitted to the window opening.
I've been working on 3D printing some scoops for the quarter windows as well, would love to see pics. I'll DM you so you can email them and I'll post on here.
Thanks Aj
Here are the pictures of Lance's body mods:
Attachment 188551
Attachment 188550
Attachment 188549
Attachment 188548
Air scoop function is complicated. Airflow is stratified and has varying energy. Laminar airflow is desired, but it does not exist on surfaces, especially deviant surfaces.
Dynamic air on the surface flows slower than air above the surface.
OTOH dynamic air does not want to turn easily and has force. In Mike's case the air coming off the windshield forced the air scoop up and off. Airflow does not stay attached over 7 degrees surface angle.
Highly developed aircraft (P51) and race cars (Mclaren) position air scoops off the surface for clean laminar flow, and no lift.
jim
Attachment 188560
Attachment 188561
Just so long as no one ever mistakes my 818 as a highly developed anything... but good to know. Since I'm reprinting the scoop, I'll try including a separation between the windshield and the inlet.
Lance asked me to post these two pictures
Attachment 188580 Attachment 188581
Mike, copying is the most sincere form of flattery. Besides, it’s nearly impossible to come up with a truly original thing nowadays. We are all subliminally tainted in some way by the things we have seen in our lives. Very few original idea’s floating around. Besides, when a design has been proven to work, why change it? I like problem solving but I don’t regret using other peoples solutions.Quote:
Originally Posted by mikeb75
You guys got me all motivate now. 107 degree heat.... damn the torpedos full speed ahead, I'm doing fiberglass work this Friday.
While working on creating a heat shroud for the turbo and exhaust I dropped a bolt on to the belly-pan/diffuser and into a puddle of transmission fluid. :(. Looks like the passenger side axle seal is weeping. So while I have that corner unbuttoned, guess I'll be replacing that... and maybe backing the diff preload a bit.
The plan:
Attachment 189127
Attachment 189129
Attachment 189128
Aluminum chimney with some Thermo Tec heat barrier to try to isolate the turbo and exhaust a bit more and lower the heat soak on the intercooler. I printed an inlet or the passenger side scoop to duct some air to the turbo side, ideally blowing through the chimney and out the back.
The problem:
Attachment 189130
Noticed a tiny weep after one of the track sessions... guess it was a bit more than that.
Replaced both transmission seals and after discussion with Andrewtech, tightened the sundials (passenger more than driver) to take some of the play out of the differential/bearings. It will have some affect on the pinion/ring gear spacing but should be safe enough. This adjustment firmed up the output stub axels so the seals actually work. After a few test loops it seems to be working, don't see any further leak from the transmission, so that's good.
Put a not-awesome updated roof scoop on the car, it's a bit ugly. Tried my hand at vacuum bagging the CF layup but there is at least one wrinkle, so no need to sand perfectly smooth and polish. If it works well enough I'll pull another mold and make a nicer part.
Finished the exhaust chimney, insulated it and finished the scoop in the passenger side vent. Definitely noticed less heat in the engine bay after a drive and the pre/post temp gauge looks better; won't know until I can really stress the car.
Looking for a late season track session opening so I can get more data...
Late last year we got some bad news on the medical front, for which I cancelled any scheduled or planned track events for the first half of the year. We've cleared that hurdle, and I can start looking for/planning events again.
While treatment was ongoing, I took on a few additional/unplanned car projects while I was needing to be available to do things, ferry kids, whatever:
- Remade and replumbed intercooler plenums
- Replaced and fixed transmission output seals
- Built a heat shield for the exhaust from turbo to wing mount
- Added roll hoop and roll bars
- Added water (not meth yet) injection between the intercooler and throttle body
Short clip of the last shakedown drive where I was specifically monitoring intercooler temps (approx. 80* ambient today) and the manifold pressure/injection system during some 3rd and 4th gear WOT runs:
https://rumble.com/v5bgpn9-818-scr-t...024-08-18.html
Take aways:
- Intercooler is still heat soaking a bit as engine compartment temps rise
- Seeing approx. 20* drop over IC when on boost
- Boost looks good with expected maxes on the map (no obvious leaks, just loud)
- Water flow rate might be too high (6 gal/H), little stumble at WOT in 3rd
Track event is scheduled in 2 weeks, will be nut & bolting the entire car. Since I'm not running methanol and do have an intercooler, maybe I'll change out the water injection nozzle to the 4gal/H unit I already have or try to get a 5 gal/H one here in time (according to the chart, between 5 and 6 is the sweet spot for power and boost levels).
Mike: Glad to see you back at it. Have a great track day. I can't wait to see the results.
Short lame update: weather was rather hot - car didn't like it; was still pulling timing, boost and fuel cut; exploring lowering boost target while I figure out next steps.
Track was greasy and crowded; I spent more time pointing cars by than driving clean laps, which was frustrating - but I ran solo all day and didn't wreck the car, so that's a win.
The pros:
Roof scoop didn't fly off the car this time.
The brakes are very good, even if the Carbotec rear pads knock around in the stock single piston calipers quite a bit, there is good balance and very good stopping power.
Alignment and tracking feel good, rear doesn't wander under braking.
The cons:
It's annoying to be limited to 110ish on the straight and not be able to WOT the car coming out of corners.
Don't know if tire pressures were too low, if they're coming to the end of their performance window, or track was super hot and green, I never noticed it before - turn-in and mid corner were so loose. Need to compare to prior data, felt like I was 5 to 10 mph slower through corners.
4th gear, 5900 RPM = 100 mph. 5th gear, 5000 RPM = 100 mph. 5 MT FTL.
GoPro cameras do a crap job when in autoexposure mode - the videos I have are so washed out you can't see the track. Guess I need to read the manual for that.
Now that I've had a few days to think, here is where I am at:
I'm not abandoning the 818, yet. As I calculate it, the consumables costs are miniscule compared most alternatives I've considered, and the cost of entry is too high for everything else (not going to entertain the Miata route). I need to figure out what's going on in performance delta between "spirited street" driving and track use... I know there's a huge gulf between the two. No amount of street driving generates the loads and temps that even a few track laps do (duhh). So, all off track testing is pretty useless, and autocross around here doesn't generate the loads - the tune I'm running barely makes boost in 1st and 2nd gears.
Heat evacuation and heat soak in the engine compartment are severe. The roof scoop and ducting do not provide adequate cooling over the stock intercooler, even with a larger scoop inlet. The stock intercooler itself remains a variable, as the Subaru OEM location is poor, especially in this application and even the "STI" intercooler is a tube and fin design, which is cheaper and less performant than bar and plate style. I'm not ready to switch over to air to water cooling yet, I have one more avenue to try with replacing the coupe side glass with inlets and seeing if that provides more air over the intercooler. If that shows improvement, then I can explore rerouting the turbo out to a different intercooler core.
The next track day is scheduled for October, as long as I didn't cook the engine; oil and water temps were fine, just the air at intercooler was stupid hot which pulled timing boost and fuel. The 3rd session of the day the car felt even more sluggish so I parked it. Even if the weather is cooler, I'll be recording the turbo outlet vs intercooler outlet temps to verify if A2A cooling is viable in the coupe.
I've been making the AAIC work just to avoid the complexity of an AWIC. Tried different scoops, ducting etc. All improved IAT delta, but then went with fans. SPAL 5.2" or 6" fans push about 300-400 cfm and seem to really help. Air out of the IC isn't "cool" but it's below any threshold for detonation apparently.
And they're relatively cheap.
Attachment 203564
Mike,
In my development projects I look to the "best in class" for inspiration or outright theft of ideas.
The Subaru turbo intercooler over engine appears to be adequate in their application and not in mid-engine or rear engine application.
Porsche rear engine cars once put intercoolers in the rear wing and at the rear sides as convenient plumbing solutions but contemporary Porsches position intercoolers in the front apron corners. That airflow is dedicated to the intercoolers, nothing else.
My Fiat Abarth incorporates the same configuration, (2) 8X10X1.25 cores mounted facing the side. Ram air enters the front and turns 90 degrees through the cores to exhaust the body at the side, ahead of the front wheels. There are three vent slots into the wheel wells. No fans, no air to water.
The Multi-air engine works pretty hard, 1400cc with 18lb boost nets 160hp.
I do not have IAT presently but I have a data display on order.
BTW I have used the same configuration for racing motorcycle oil coolers and it worked great.
jim
Thanks for the confirm, driveslikejehu; I did try fans over the intercooler earlier in the build with a boxed enclosure. Don't remember if I ever tracked that solution; wasn't super happy with my implementation (heavy and ugly).
Given the less than ideal placement of the stock AAIC and the poor ducting available in the rear of the 818's, I don't think it is a viable configuration for the track. Spirited street driving, sure, but the constant WOT and heat loading of track days is beyond what it can handle. AWIC on the other hand is ideal for the mid engine layout. The high heat capacity of water paired with the second radiator up front makes for stable low intake temps over extended track sessions.
I recently added an extra air temp sensor before my AWIC so I could record the temperature delta across the cooler to gauge it's efficiency. I was concerned it wasn't performing as well as I would like. Boy was I shocked at how well it works. Below is a screen shot of lap 6 of 7 during a session at Watkins Glen. At it's hottest, the pre-AWIC temp was 236.6 F and the intake temp at that same point was 124.6 F making for a 112 degree drop!
The AWIC really stabilizes the intake air temps.
Attachment 203570
Thanks Hobby! Your data is very compelling.
But I'm stubborn, I want to exhaust every avenue on the A2A approach. I'll probably wind up spending around the same $$$ as the A2W equipment when all is said and done, but I've got 1 more try in me before admitting defeat and joining the other dark side... :)
Mike this is interesting. Have you considered omitting the rear hatch glass?
It is possible to turn an a2a intercooler into an awic. Might be worth a try.
As I noted elsewhere, with a stock setup it seems the point of cooling the IAT is just to avoid knock. The system doesn't measure IAT except at the MAF; am I wrong? So no more boost or fuel with cooler IAT.
With a tune and added controls to adjust boost, etc, while monitoring it, cooler IAT would result in more power.
I've done a few track days with my STI based setup, in the 80's, never noticed any reduced thrust or anything.
I think for now, I'm good with my 30+ degree delta.
But I'm asking... Will I know if my temp is too high? Would an AWIC net me anything with a stock setup? I may be missing something...
I take the point that AWIC may be where we will end up with enough development, and provides plenty of overhead till you get there, so maybe avoid the middle steps.
Actually, did run without the rear hatch entirely while troubleshooting the water injection system, but that probably hoses the airflow over the wing (sure, not really needed, except when on the track I did loop the car in a higher speed corner last year).Quote:
Originally Posted by lance corsi
The point of having an intercooler is to not only reduce the possibility of detonation, but to also increase the density of the air charge entering the cylinders. Cooler = denser + more fuel = more power. If the temperature of the air is measured at the MAF, the temp post intercooler is a guess. The tuner works with whatever timing/boost/injector duty cycle makes the most power without detonating. He's mainly using the dyno's O2 sensor for fueling anyway. I bought a splitter harness from I-wire to move the MAF temp sensor to the post intercooler side of my AWIC. I don't know that it made much difference. You still have to work with the constraints of detonation avoidance when tuning. My tuner wasn't particularly impressed with the intercooler efficiency on the dyno. I suspect some of that is heat soaking between pulls. I have a temp gauge on the output side of the intercooler and it seems to be doing ok most of the time - 10 to 20 degrees above ambient in street driving and canyon roads. I have noticed that after a short shutdown, it takes a while to get back to that. Even sitting in traffic, it tends to rise some and running the fans on the heat exchanger doesn't seem to make a difference. I have yet to track it so I don't really know how it'll work under the more extreme conditions of sustained boost.
I chose the AWIC because I didn't have much faith that I'd be able to get sufficient air to the IC in the mid-engine configuration, even with the S model, and also because I had dropped the rear end of the transmission. That would make it harder to get an A2A to fit under the hump cover. My nephew intends to solve that problem with an F1 type scoop above the roll bar. He'll probably get more "What kind of car is that?" than I do and I get that a lot. It seems that it might be even harder to get the airflow in the coupe version unless a larger scoop were fitted. Forward mount IC would be a nightmare for boost plumbing and would cause a bunch of lag. A major downside of AWIC is the increased complexity and weight.
My tuner scoffed at the frozen boost intercooler. He scoffs at everything that he didn't build, I think. That said, he does have some first place plaques hanging on his walls. I did what I could to make it work as well as I could. I enlarged the water ports on the IC and the radiator to 1" from the 1/2"NPT, which are very restrictive, and ran 1" tube all the way. I fitted a Davies Craig pump for maximum flow.
Ed
I plumbed an a/a ic into the left side scoop. It's a tight squeeze but I was able to fit one with significantly more cross section and depth than stock.
Used some string and tape to verify it's getting flow but haven't got more scientific than that.
I only do autocross but the stock one got hot enough to boil off water I sprayed on it, after my last runs on a hot day. Pulling timing, feedback knock, all that fun stuff. The new setup stays cool enough to touch. The hot side piping is too hot to touch and the cold side you can hold your hand on.
Don't know if that does you anything but I'm happy with it. I think I'm less than like $400 into it too.
The downside is that I hacked the side scoop up a bit. I intend to fab up a shroud for it and do the same on the other side..m eventually.
My "last" attempt at getting the A2A intercooler solution to work in the 818.
Over the winter I replaced the IC I was running, JDM EJ207 v8 measuring 18.5" x 7" with a USDM EJ255 STI IC I had lying around (benefit of years of Subaru ownership, I guess) measuring 20.5" x 7". I needed to print a new plenum as this IC is larger, and I split the plenum into 4 divided chambers.
Here's the assumptions I'm working with:
"The Internet" says a turbo adds 10* F of heat per PSI of boost to the charge air temp (blah blah blah, compressor map, efficiency, and internet caveats apply).
The map I run in the 818 maxes a 20 PSI target at WOT.
Looking back at video data, I was seeing IC entrance temps (measured at the back of the IC where the Y pipe enters) of up to 200* on main straight of Summit Point Main - it was a pretty hot day, but if we go by the rough numbers of 20 PSI max and 200 charge air temps, the fake math works.
The IC exit temps were reading 150*, so, still painfully high.
The plenum was not optimal, there were 2 chambers with a lot of dead volume and the chambers were not well separated at the bottom, there was possibly cross chamber flow between the bottom of the plenum and the top of the IC core. This plenum was fed by 2x 3" inlets getting airflow from either the roof scoop or side scoops that replaced the quarter windows in the hardtop.Code:4th gear 6000 RPM (105 MPH), 200* IC hot side/160* IC cold side
5th gear 5500 RPM (113 MPH), 185* IC hot side/150* IC cold side
Again, internet math:
A single 3" circular duct will max 215 CFM at 50 MPH (same blah blah blah perfect flow, clean airstream and no restrictions/kinks), which give a maximum of 430 CFM over the IC at 50 MPH. Given the data I have, that's not enough airflow over the IC, and not enough IC core size to deal with the heat being produced by the turbo + heat soak of the engine bay,
New setup:
4x 3" inlets fed from both the roof and side scoops
The plenum is more optimized to flow over the IC core, each chamber is isolated from the others down to the core top and there is considerably less dead volume in the plenum.
After heat soaking the engine bay in the driveway (10 minute warm up, coolant temps 208*F and rad fans running) the starting IC outlet temps read 140 *F. Water Injection was disabled:
This was a pretty low stress drive, it's hard to max out the turbo/WOT on the public B roads.Code:starting exit IC temps of 140, at 15 MPH
30 seconds later at ~30 MPH exit IC temps down to 130
Starting test pull
start exit IC temp 111
pull @ 7 PSI
pull to 5600 RPM
max speed 70 MPH
turbo exit temp max 130
IC exit temp 109 (dropping)
Regardless, the data was promising. The IC temps were dropping the entire drive (on and off boost). The additional airflow over the IC and the revised plenum design seems to have made a difference.
Run 1
Run 2
Attachment 214811
Attachment 214812
Attachment 214813
Will work on getting more data, but here are the compromises made:
- enlarged roof scoop
- replaced quarter windows with NACA ducts and scoops
- 4 chamber plenum
mike, It is optimistic to expect laminar airflow that far back on the side of the car, but smooth air is better than turbulent air to feed a NACA duct.
If you rivit the NACA duct inside the glass, the airflow feeding the duct will not trip over the duct flange edge.
Additionally a sharp leading edge on the duct inlet encourages the air to flow straight aft. A soft edge can encourage the coanda effect to draw the air into the duct.
jim
Yup. Thats why I added the outside scoops to the NACA ducts (hard to see in the 2nd picture).Quote:
Originally Posted by J R Jones
I figured the open window gaps were a bigger issue than the rivets/not smooth transitions. I'm keeping printing up a larger set of scoops in my back pocket, in case I don't see enough airflow.
I did two years racing motorcycles in AMA pro. We had problems with the stock ram air not cooling the engine, or oil and not feeding the induction.
I put a snorkel out in clean air next to the forks to cool the oil and rear cylinder. A modified backwards NACA duct pulled air out of the cooler core.
I put two ports on the front of the faring into the air box and raised power from 125 to 132 at 150MPH. MY race engines ran cooler that the street bikes.
jim
Attachment 214841
I hate to be negative, but I do not think you will be able to get enough air flow for proper cooling. You may have better luck with an active scoop like seen below.
Attachment 214842
I tried a few A2A configurations before giving up and moving to an air to water intercooler setup. I have had great luck with the A2W setup. I get a max of 110 degree F drop in intake temps at the track under full boost!
Best of luck.
P.S. I'm going back to VIR next week to redeem myself and the 818R. Hope to run the entire event with no issues.
Expanding on Hobby's example, I used the Meredith Effect on my motorcycle cooling, inspired by the P51 Mustang:
https://www.supercoolprops.com/home/...h_effect2.html
jim
Attachment 214844
I've stuck with A2A intercooling for simplicity's sake and having a stock engine and ECM. The air just needs to be cool enough to avoid detonation. I'm pretty sure with a stock setup, you won't get any more power with cooler air; the ECM measures the air and plans fuel based on the MAF only.
But with that, I wasn't sure you can get enough air thru the IC with passive air flow, so I added these fans. They're grabbing air from the IC scoops and anywhere, forcing it thru the IC. On the track I see a 50-60deg temp delta.
Attachment 214845