When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Twin turbos, switching from serial to parallel operation
Okay, here is the background.
The racing class I am in requires stock engine short block (only rotating assembly balance allowed) and stock turbos (compressor housing, CHRA & exhaust housing). Intake and exhaust manifolds and all engine and turbo peripherals are free as is boost control as long as you do not add or modify wastegates.
Allowed fuel is 110 unleaded
How would it work to have both stock turbos working in parallel on the exhaust side 100% of the time, but start with the turbo compressors in serial mode (one turbo compressor feeds the other compressor) and then switch to parallel mode.
I am thinking this would allow for maximum low/midrange torque since you are halving the surgeline of a single stock turbo and doubling its pressure ratio (while being limited by the total cfm flow of a single stock turbo).
I think you should be able to get ~35psi safely (for the turbos) out of the turbos in serial mode while peaking right around 200hp (I have seen up to 225 on primary turbo alone) and then switch to parallel mode once the turbo shaft speeds are at the max you want to run (so no further turbo shaft increase/spool necessary once switched into parallel mode).
I think with really open exhaust exits on each turbo housing and a moderate amount of antilag in the exhaust manifold this could net a boost pattern like 35psi by 2,500rpm holding to ~3,000rprm and then suddenly switch to 17psi from 3,000rpm to redline.
Since the motor won't be (relatively) choked on the exhaust side ever (twins 100% in parallel on exhaust), this should be good for ~420ftlbs torque @ 2,500rpm, fading to 350ftlbs by 3,000rpm and then bam switch to parallel mode slowly building back up to 380ftlbs ~5,300rpm and then fading down to keep hp ~400hp till ~7,500rpm (as stock non-sequential turbos do).
Question is- what happens to engine output when you take two serial turbo compressors at X rpm and then suddenly switch them to parallel compressors at the same rpm. Pressure ratio halves and flow suddenly doubles.
When I estimate stock non-sequential twins at 17psi @ 3,000rpm it sure looks a lot like the same 350ftlbs torque you should get from serial twins at 35psi @ 3,000rpm...
Thoughts?
Googling this led to the Garret website where they are working on somewhat similar configurations for production vehicles.
don't get me started, i ordered my sxe7670 last week along with two 60mm wastegates... the other sxe7670 will be made of wood for the mach up.
also i think that the primary stage turbo would have to work less hard then the 2nd stage and the exhaust would fight the 2nd turbo. if using two stock turbos..
however what if you paired a stock HT12 with say the sxe7670... the ht 12 feeds the borg 10 psi, the borg(being a high pressure turbo) efficiently turns that into 30 psi... then later on they switch to parallel..
i'm driving my twin seq'd FC to work for this next month or so.. hoping to get it on the dyno and get some data.
You're talking compound. You will be limited by flow of the primary turbo, but the sky is the limit with the amount of boost you can generate. Lets say each turbo can generate 1.5 bar, that's a pressure ratio of 2.5. 2.5x2.5=6.25 = 70-80 psi. But the first turbo will fall on its face for flow.
Also i forsee sharing the exhaust energy eithe in serial or parralel might hurt spool?
Will be interested to see how you get the transition from serial to parallel.
In any case i love this kind of thinking outside the cube. I got me popcorn for this one!
OK, I think I have figured out what would happen at serial to parallel compressor switch-over.
Turbo speed would stay constant, engine flow would stay constant, so we would follow the turbo speed line on the compressor map back to half of the total flow.
So, if I was running 29psi serial at 3,000rpm (or wherever flow maxed) at 140,000rpm turbo shaft speed, 1.5PR per turbo, 0.195kg/s flow per turbo when I switched to parallel flow you would just follow the 140,000rpm shaft speed line to 0.0975kg/s flow per turbo and that shows you you would now be at 2.22PR or 18psi boost.
You're talking compound. You will be limited by flow of the primary turbo, but the sky is the limit with the amount of boost you can generate.
Technically, compound means the turbo exhaust wheel is geared way down and feeds power directly back into the engine crankshaft. This has been used on production vehicles.
But, yes, serial turbo charging is what diesel guys mean when they say compounded turbos.
Yes, in serial mode you are limited by the total CFM flow of 1 turbo (unless you bypass) or about 200-225rwhp on an FD from the dynos I have seen. You would want to switch out of serial mode as soon as you hit that peak HP.
Also i forsee sharing the exhaust energy eithe in serial or parralel might hurt spool?
Moot point. The class allows any ignition and fuel setting/component and allows modification or removal of emissions components. Antilag is legal.
You take off the air pump and run a boost line from the turbo directly into the ACV and signal the ECU to pump the boost into the exhaust sleeves (port air). On my FC I have already shown that with good ignition I can make best power at 9-10AFRs, so there is the fuel for the air you are sending to the exhaust manifolds.
This strategy will require turbo shaft speed sensors on both turbos for the ACV control and exhaust shutters in the downpipes to prevent overboost as well (could use aftermarket diesel "jake brakes").
Will be interested to see how you get the transition from serial to parallel.
What I first sketched up was a rotating sleeve between the low pressure and high pressure turbo with passages 180 degrees apart.
So, it appears it is really difficult to get turbo shaft speed without modifying the turbo (ie, drilling a hole in compressor cover).
You can go with an optical pickup and fiber optic lines, but without modifying the compressor cover it becomes very difficult to get a vibration free mount that does not impede airflow excessively.
However, I think we know the limits of the stock turbos/engine well enough to figure out the shaft speeds we don't want to exceed by resulting boost values.
If I keep it under 47psi and 200rhwp in serial mode and 16psi and 400hp in parallel mode it should be in the live-able shaft speeds.
I can do the serial to parallel switch-over with 3 butterfly valves or one (dual passage) rotating sleeve valve.
Oh yeah, I need to factor in engine flow for turbo shaft speed to get max switch-over rpm.
So, if twins can live with 16psi and 400rwhp OK at 7,000rpm in parallel mode; it should be around 47psi max with 200rwhp max and 3,500rpm max in serial mode.
That is good. I want to get out of serial mode around there anyways since a primary turbo in regular old sequential mode can hit 200rwhp by 4,000rpm.
it wont work.
not enough exhaust energy to spool both turbines, the only reason the factory setup can spool the ht12 is because all the exhaust is going to one turbine... think about it. by the time the engine spools them it will be time for change over to parallel.
it wont work.
not enough exhaust energy to spool both turbines, the only reason the factory setup can spool the ht12 is because all the exhaust is going to one turbine... think about it. by the time the engine spools them it will be time for change over to parallel.
Right, but one turbo will feed the other through the compressor and the engine powers both turbines. The pressure from the primary turbo as well as the flow will help it along until the changeover. I'm not sure it will spool as fast as a functional sequential system, but if someone is willing to try, I'm excited to see if the system can function as desired.
Then again, I just found this site after I had already wrote this post.
The turbos might fight each other and exacerbate surge (at least for one turbo. The larger one?). In a compound setup, one compressor feeds the other, boosting the pressure ratio. I'd imagine that a parallel setup would boost maximum flow, but not the pressure ratio.
My gut says that if one turbo is flowing X PSI, and the turbo is not in the choke region, then having two turbos at X PSI is not going to get any extra air into the motor.
This is out there, but maybe you could re-route the piping to run them as compound turbos. Maybe the small turbo would choke flow at top end, however. There is probably a reason mazda switches off the small turbo.
find a set of dead turbos, and have a drilled set of covers with holes for instrumentation
Definitely I could do this, but I was hoping I could use turbo shaft speed input in the control circuit to keep the turbos alive with real anti-lag.
Not having turbo shaft speed means I would probably set it up without anti-lag first and then just add in moderate anti-lag if the response is too soggy (like maybe only on throttle anti-lag instead of throttle ready turbo shaft speeds).
Better get some good oven mitts and put v-band adapters and dry break quick disconnect fittings on the turbos (and the spare set) if I am going to run anti-lag.
lastphaseofthis
it wont work.
not enough exhaust energy to spool both turbines, the only reason the factory setup can spool the ht12 is because all the exhaust is going to one turbine... think about it. by the time the engine spools them it will be time for change over to parallel.
I figure the exhaust energy will be shared between turbos as you say, but there are many ways to increase exhaust energy.
Maximizing the pressure differential will help by putting a huge downpipe on each turbo. Low restriction intakes with velocity stacks help. Sending air and fuel into the exhaust manifold for anti-lag will definitely help.
The stock sequential primary can get 10psi boost around 2,000rpm with just exhaust mods.
If I can get the primary turbo spooled to 10psi by 3,000rpm sharing the exhaust with the other turbo that would be roughly 28psi boost by 3,000rpm once it runs through the other turbo. 14.7psi atmosphere x 1.7pressure ratio x 1.7 pressure ratio - 14.7psi atmosphere= 27.8psi boost gauge pressure.
That is slightly better spool than my EFR 7670 was on the FC had.
As far as getting both turbos finally spooled just before switch-over to parallel, that would probably work ok. The racing I do you spend most of your time in the spooling area anyways instead of at peak boost.
If it is able to bring the parallel turbo switch-over point lower in the rpms that would be really nice (ie get out of the ~200rwhp limit of one turbo before 4,500rpm).
The stock sequential primary can get 10psi boost around 2,000rpm with just exhaust mods.
If I can get the primary turbo spooled to 10psi by 3,000rpm sharing the exhaust with the other turbo that would be roughly 28psi boost by 3,000rpm once it runs through the other turbo. 14.7psi atmosphere x 1.7pressure ratio x 1.7 pressure ratio - 14.7psi atmosphere= 27.8psi boost gauge pressure.
That is slightly better spool than my EFR 7670 was on the FC had.
As far as getting both turbos finally spooled just before switch-over to parallel, that would probably work ok. The racing I do you spend most of your time in the spooling area anyways instead of at peak boost.
If it is able to bring the parallel turbo switch-over point lower in the rpms that would be really nice (ie get out of the ~200rwhp limit of one turbo before 4,500rpm).
After a lady hit me head on in my 3000gt..I'm currently having to daily drive my 10thae with FD twins, i hit 5 psi right off the bat at around 1900, and 10 is after about 2300. the fc block and manifold being the choke..
my sxe 7670 will be here soon, so i can start looking into fabbing that 7670 seq setup. first is to ring out the 7670 sxe to see if / make changes to get to... the same torque you made on the EFR version. your fc efr dyno is my is my goal but on e 85 some and AI.
your fc efr dyno is my is my goal but on e 85 some and AI.
From what I have seen the E helps spool.
With careful manifold design, I can see your set-up being able to push the surgeline of the 7670 compressor and making more low rpm torque than my single EFR 7670 did.
And of course the top end is going to be so much more torque!
06-04-16, 04:30 PM
