I have had an idea to use twin turbos with the exhaust sections fixed in a parallel configuration and the compressor sections initially in serial mode for high boost at low rpms and then switching into parallel mode to extend the compressor flow capabilities.

Background-
This is on a 1993 RX-7 in a racing class that specifies stock engine "shortblock" and stock turbos (compressor cover, chra, exhaust housing) with all engine and turbo accessories and settings free except cannot modify or add wastegates or modify or add pop-off valves. Auxiliary water injection is allowed, 110 unleaded gas is allowed.

The issue is trying to get some more low rpm torque. The stock sequential twins are limited to around 17psi/400rwhp@ 7,000rpm before turbo life becomes very short from excessively high shaft speeds/loading so you get just over 250ftlbs torque on the primary turbo.
Stock sequential dyno-
#ad


So, I was thinking in serial mode I should be able to run much higher boost (up to 47psi?) and up to 200rwhp@3,500rpm while keeping the turbos relatively reliable and then switch over to parallel mode to gain the flow for the usual 400rwhp.

So, what happens at serial to parallel switch-over?
As near as I can tell, the engine flow remains the same (turbo exhaust side always in parallel configuration, ie no change in VE) and turbo shaft speed stays the same. This tells me I should be able to trace the turbo shaft speed rpm line back to half the current total compressor flow and that is my new pressure ratio.

So, if I was near maxing the turbos in serial mode at 140,000rpm and working together to move 182rwhp@3,500rpm worth of air (~26lbs/min on rotary) with 1.7PR each for 28psi gauge boost and switched to parallel mode gauge boost would fall to 18psi? What other changes should I expect?

#ad


I am curious about the turbo response in serial with the exhaust sides in parallel. If its too lazy I can use the exhaust port air injection emissions equipment but hooked to the second turbo instead of airpump and rich AFRs for anti-lag. With a good ignition system I have found the rotary is pretty indifferent to AFRs as you can see below.

So, stupid idea or could serial twins switching to parallel twins work? (I mean, I know Garrett and BW are working on this for production vehicles, but I mean work with the turbos I am stuck with.)

In 2014 I raced my old 1987 RX-7 in the next class up (allows aftermarket turbos) and I used this high boost at low rpm strategy by selecting a little EFR 7670 and running 26psi boost which would fade to 21psi by peak hp. I didn't have enough injector for more low rpm boost. I really liked the low rpm torque, made it easier to drive and fast (roll into throttle instead of whack it open and roll off throttle with boost).
My Dynojet dyno
#ad

Boost and AFRs tuning on DynoDynamics
#ad


#ad


#ad


#ad


#ad

 

Hey RotaryMotoring,

Welcome!

I am always a proponent of trying something, even if it may seem dumb to you. You can observe new things when you do that.


But... If I am understanding your post correctly, then you will have a hard time passing tech. I would imagine that you would need to change inlet piping and add in some valves to covert the stock sequential system into a true series compound compressor layout. Which, I would imagine they would call foul on since they don't allow additional/modded blow off valves.

So, if you were to work that out, then your only other dilemma would be that once you are in compound mode, your maximum mass flow will be that of one of your turbos. Just as you noted above. Now... that may not be such a bad thing, but from the first dyno it seems that car is making ~ 170whp @ 3,500 revs. Would the extra 30hp help you out much? Do you have a track where there are a few corners that you can't come out of hard because of some lag?


If that's the case, then before modifying your hardware, I would try and get proficient at using your left foot to brake and control vehicle speed, mid corner, while giving it a good bit of throttle to keep your turbos spooled. (Brake Boosting)

If I remember correctly, Hans Stuck did this in his Audi days running a turbo, (5 cylinder? 6 cylinder?) keeping up with the V8's when coming out of the corners. Normally there would be a lot of lag.

 

Thank you for the welcome and taking the time to respond to this post.

But... If I am understanding your post correctly, then you will have a hard time passing tech. I would imagine that you would need to change inlet piping and add in some valves to covert the stock sequential system into a true series compound compressor layout. Which, I would imagine they would call foul on since they don't allow additional/modded blow off valves.

The class has "interesting" rules. No changes to wastegate size, number or location. No changes to or adding pop-off valves (over-boost pressure relief valves some cars used to have). All other turbo and engine peripherals and settings are free. I have no idea why they limit the ability to run lower boost, but higher boost is fine.

To control boost creep from very free flowing intake and exhaust I would use exhaust restricting valves (aftermarket "Jake brakes"?).

from the first dyno it seems that car is making ~ 170whp @ 3,500 revs. Would the extra 30hp help you out much? Do you have a track where there are a few corners that you can't come out of hard because of some lag?

Yes, the primary turbo in the stock sequential mode is limited to 200-225rwhp. It might be possible to spool it up to that power lower in the rpms with some anti-lag and then do an earlier than stock 4,500rpm switch-over to parallel. I would still be limited by shaft speeds to around 17psi for reliability.
I desire more boost because though the tiny primary turbo spools to full boost very quickly, torque at 17psi boost isn't as good as torque at higher boost. The racing is mostly 2nd gear in parking lots (auto-x), kart track and hillclimb racing.


If that's the case, then before modifying your hardware, I would try and get proficient at using your left foot to brake and control vehicle speed, mid corner, while giving it a good bit of throttle to keep your turbos spooled. (Brake Boosting)

Better driving is always the best way to go faster!
The local Subaru guys Left foot brake with their laggy cars with lots of acceleration grip, but the rotary spools the tiny primary turbo quickly. Its just that full boost on the stock primary turbo is not as much torque as I want and I don't really want to go back up in class with an aftermarket turbo again since there are so many other mods allowed in that class to spend money on.

One thing I did do is raise the redline to 9,000rpm so I could use 1st gear more often to multiply the torque, but most our parking lot runs are so short (30-60 seconds) that I am usually faster just leaving it in 2nd and concentrating on smooth driving instead of throwing the shifts in there. Obviously, on the kart track and hillclimbs you can't avoid shifting, except it is 2nd, 3rd and 4th.

Here are some diagrams I borrowed from Toyota Supra MKIV site to illustrate what I am proposing if others are having a hard time with my descriptions alone.

Serial turbo compressors with parallel exhaust sides (first mode)
#ad

Parallel turbo compressor with parallel exhaust sides (second mode)
#ad


#ad


#ad