turbo manifold design
#61
This is the design for the one I built for my car.
Each runner is 16.7” long, equal length.
Merge collector is smooth transistion straight into turbo.
Another thing I see on a lot of other custom turbo manifolds is I think the runner diameter is too big. If the cross sectional area of the runner is much bigger than the exaust port cross sectional area at the flange between the head and the manifold then you are just loosing energy by slowing down the exhust flow where it has to speed up again to fit through the turbo. There is more loss in doing that than there is running down smaller diameter pipes to the turbo. Velocity is your friend, energy is a fuction of the velocity squared. you want to maintain all the peak velocity you can from the exhaus pulses till they hit the turbo.
FWIW I also made one just kind of a short smooth runner non equal length. The difference in spool, throttle response, and power is dramatic.
Bob
Each runner is 16.7” long, equal length.
Merge collector is smooth transistion straight into turbo.
Another thing I see on a lot of other custom turbo manifolds is I think the runner diameter is too big. If the cross sectional area of the runner is much bigger than the exaust port cross sectional area at the flange between the head and the manifold then you are just loosing energy by slowing down the exhust flow where it has to speed up again to fit through the turbo. There is more loss in doing that than there is running down smaller diameter pipes to the turbo. Velocity is your friend, energy is a fuction of the velocity squared. you want to maintain all the peak velocity you can from the exhaus pulses till they hit the turbo.
FWIW I also made one just kind of a short smooth runner non equal length. The difference in spool, throttle response, and power is dramatic.
Bob
#64
Schedual 40 mild steel has lasted through serious abuse on the track for a couple years now since a few mods and do over of my first attempt.
Also Went to Inconel studs to hold the turbo on. I did notice ATP has made a V-band compressor housing for the GT3071R-WG looks like the just take the T2 flanged stock internal WG housing and machine a V-band connection on it. My next version probably will be that but for now this is working.
Bob
#65
Some pictures. And of course the turbo flange studs stretched and the nuts fell off as is typical for turbo miatas driven stoopid fast. Inconel studs and Resbond in now.
My first one cracked at the collector to turbo flange joint. The welded in gussets that the turbo flange studs thread into seems to have cured that problem. I also learned to compleat all the welds on each runner completely before putting them together in the manufacturing process and don’t do relief cuts in the head flange.
Bob
My first one cracked at the collector to turbo flange joint. The welded in gussets that the turbo flange studs thread into seems to have cured that problem. I also learned to compleat all the welds on each runner completely before putting them together in the manufacturing process and don’t do relief cuts in the head flange.
Bob
#67
fast forward to 2:20.... this is the type of person that does a remote turbo install:
http://www.youtube.com/watch?v=m67ZMn0tzg8
yes, he has a remote turbo...or had. it "blew up" on the dyno, twice.
http://www.youtube.com/watch?v=m67ZMn0tzg8
yes, he has a remote turbo...or had. it "blew up" on the dyno, twice.
#68
Cpt. Slow
iTrader: (25)
Join Date: Oct 2005
Location: Oregon City, OR
Posts: 14,429
Total Cats: 1,207
...The welded in gussets that the turbo flange studs thread...
Gorgeous manifold bbundy. Eventually I need to upgrade to a AF design piece for a simpler dp, better/bigger turbo, and the like. I'm years behind with my greddy piece.
#70
It's got a little diving board support underneth the turbo to the block very similar to the MSM factory one. The support cracked made out of cheap weld steel from the hardware store. Made it out of highr grade 4130 plate and it hasn't cracked.
Bob
#74
Elite Member
iTrader: (9)
Join Date: Jun 2006
Location: Chesterfield, NJ
Posts: 6,930
Total Cats: 404
Yeah I wanna know too. I hope to make a true equal length ramhorn style manifold for someone here (ha!) to find out for sure myself, but my only sorta equal length ramhorn manifold was much more laggy than a shorter manifold, but when it did spool it was nuts.
#77
Don't want to hijack the thread. more manifold and build pics in my thread over in the build section. title is "f20c engine conversion"
#78
What we need is someone to build three different manifolds on the same jig: semi-log (like ETD), Absurdflow and ram's horn. Keep everything else in the system the same, except for the tuning. Best to do it on the same car.
Spend some time arguing about what supporting mods to start with. I'd suggest having all other elements optimized approaching overkill so the only limits are the turbo manifold itself. Three-inch exhaust and a better intake manifold. '99 head. Etc.
Dyno them all under the same conditions using strict procedure discipline.
Then have a bunch of people say the results are not relevant because of the supporting mods.
Any volunteers?
Spend some time arguing about what supporting mods to start with. I'd suggest having all other elements optimized approaching overkill so the only limits are the turbo manifold itself. Three-inch exhaust and a better intake manifold. '99 head. Etc.
Dyno them all under the same conditions using strict procedure discipline.
Then have a bunch of people say the results are not relevant because of the supporting mods.
Any volunteers?
#79
Elite Member
iTrader: (9)
Join Date: Jun 2006
Location: Chesterfield, NJ
Posts: 6,930
Total Cats: 404
My ramhorn's going to be a drop in replacement for sav w/results if everything goes to plan. No plans for a log style, especially one with that low/bottom mount location & especially especially not with his sick setup.
#80
Cpt. Slow
iTrader: (25)
Join Date: Oct 2005
Location: Oregon City, OR
Posts: 14,429
Total Cats: 1,207
ITT: Sav installs greddy mani/turbo. Lol.
This is both a joke and a serious note, but I wouldn't want any manifold test bed to have anything but a stock manifold. You might not get the best test results since the intake manifold may stop the turbo manifold from performing at it's best, but 99% of us have a stock intake manifold.
Then have a bunch of people say the results are not relevant because of the supporting mods.