Turbocharger Pulse Tuning & Other Mysteries

[Dynoroom]

Well it's the end of the racing season and we didn't even get to test out our hot rods at Bonneville this year. That means we have time to tinker with the little things to make our rides faster & to BENCH RACE!
We can start off with this subject and then move into other areas of interest as we go.

So, the question was asked on another thread: Is pulse tuning in a turbocharged engine using a divided turbine housing worth the effort?

Explain your answer, meaning small vs large engines, engine speed range etc.

I'll will start off by saying that at Bonneville with a large racing engine (300+ CI) that pulse tuning is a waste of your work hours & dollars. Pulse tuning was originally developed for relatively slow speed (<3,000 rpm) diesel engines where individual exhaust pulses will help a large turbo spool up to maximum speed. As diesel engines in the past did not use a wastegate the turbo was sized for it's maximum engine speed and was quite large. This led to poor low speed acceleration. In a high speed racing engine where the rpm never drops below peak torque during a shift (or at least shouldn't) pulse tuning has no real value. Do the math and see how many pulses you have from a V8 engine turning 8,000 rpm or 7,000, or even 6,000. There really is no pulsing at this engine speed that can offer any extra spooling effect to the turbine.  

[redhotracing]

Mike-
Well written. Having "only" run mile and 1.5 mile
events, I can't speak to WOT for 5+, but in our
364ci motor (88mm turbo, T6 flange, 1.32 AR)
we had no spoiling issues once pushed off. I can
definitely see your point about low RPM diesels
needing the extra help, but ours pulls like a SOB
once we're past the initial 1st gear hit.

[wobblywalrus]

Does this apply to turbos on single and twin cylinder engines?

[Stainless1]

like he said, count the pulses... flow turns the wheel... the more it looks continuous the better it works.

[kiwi belly tank]

Turbocharging is so misunderstood by most people, this thread should help dispel some of that mystery for them.
  Sid.

[Crackerman]

You covered larger v8 engines with a mindset towards lsr.
What about smallish (1.5-2.5liter) 4 culinders running a large turbo. (65+ lbs/min 60mm inducer or much larger, up and over 75mm inducer) could they benefit from pulse tuning and divided exhaust housings? Especially on boost recovery after a missed shift or a really slow shift.

[Dynoroom]

You covered larger v8 engines with a mindset towards lsr.
What about smallish (1.5-2.5liter) 4 culinders running a large turbo. (65+ lbs/min 60mm inducer or much larger, up and over 75mm inducer) could they benefit from pulse tuning and divided exhaust housings? Especially on boost recovery after a missed shift or a really slow shift.

Fair question. In my own personal experience I have never tested any engines smaller than 122 CI (2.0 liters). So, I'm not qualified to answer a question you my already have an answer to. But I will chime in with my thoughts.
A turbocharger that pumps 65 lbs/min of air can make roughly 600+ hp. The mass flow through the engine doesn't care if it's 1.5 liters or 3 all it knows is air density going through it will make X amount of torque.
Worst case scenario, a missed shift with a loss of boost on a 90" 4 cylinder engine. If the engine makes peak power @ 10,000+ rpm I would say pulse tuning would have a minimal effect on boost recovery and if you're running a "mile" or El Mirage type event if you miss a gear or have boost fall off you might as well turn out and wait for your next run because you are not likely to set a record even if pulse tuning helped with boost recovery as you still will have lost to much time. At Bonneville, I don't know. Is it worth it to get up to speed? I doubt it. When I watch bikes or small inch engines run to the "5" they are all hung out, picking up 10, 20, 50 rpm in a mile. Pulse tuning has nothing to do with that.
So educate me, what really happens?   

[redhotracing]

Mike Reichen ran a Bullseye s488 (T6 flange) turbo
on his 2.0L '95 Evo to the tune of 237mph @ Maxton
but missed shifts weren't an issue

[vwpsycho]

We run a Precision Turbo 6768 (67mm inducer) on a 2008cc gasoline 4 cylinder. It uses a conventional, undivided T4 exhaust housing. We use an equal-length fabricated manifold with 2"x12" (approximate) runners. Our Production Supercharged Volkswagen has made 15 or more 5 mile runs, reaching roughly 200 mph at the 2-1/4. I just took a peek at a few logs from last July and September. On each shift (1-6) it takes roughly two seconds from throttle lift at 35psi intake pressure @~8500 RPM, until the same boost level is regained in the next gear, at 6000-6500 RPM.
Our technical partners have been testing with the new Borg Warner EFR turbos on similar 2-liter fours. The split, twin-scroll design does allow for a lower RPM spool up. Our Passat setup described above doesnt produce measurable positive intake pressure until over 5000 RPM.   

[Sumner]

....Our technical partners have been testing with the new Borg Warner EFR turbos on similar 2-liter fours. The split, twin-scroll design does allow for a lower RPM spool up. Our Passat setup described above doesnt produce measurable positive intake pressure until over 5000 RPM.   

That is going to be my situation, at least for a while (falling below 5000 at least on the first shift).  To put a 7000+ rpm valve train in the motor I'm using is quite expensive, more than I could afford at first.  So if I can take any advantage of the split design I will.  I have to make the headers anyway so it isn't a big deal.

I agree that for larger motors or smaller ones with a smaller turbo on them there is no need.  We haven't and won't do this with the 572 in the Stude.  If you use Borg Warner's Matchbot program for one of their turbos you can get a good idea if considering this might help but it will only work if you are using their combo of turbine/compressor. 

It is easy to get fixated on the compressor side of the turbo but it is the turbine side that provides the power to drive it and make boost.  If the turbine side is sized to help the compressor side make big boost and at the same boost move a lot of air at that boost then it needs to be larger and needs a lot of exhaust to drive it.  Unfortunately the compressor maps are pretty easy to find and look at for your application but it is a lot harder to find good info on the turbine side,

Sumner

Sumner

[skywalker18]

This month's Racecar Engineering magazine had an interesting article regarding this topic based on a paper by Rudolph Van Veen. The paper compares the log style exhaust manifold of the championship winning Mercedes  F1 engine to traditional equal length manifold design. During the simulation, the log style manifold volume was varied from  2 to 6 liters and the equal length manifold had the runner diameter varied from  22.5mm-50mm.  They came to the conclusion that the log style manifold was more suitable for F1 applications due to the capability to maintain the pressure within the system.

The most interesting graphs to me were the plots of pressure drop across the turbine as a function of RPM. The pressure drop with the equal length runners was very dependent on runner diameter.  Whereas the log style manifold had very consistent trends across the rpm range (1k-15k), regardless of plenum volume.  Also, all of the log type manifolds had a temperature drop of less than 200K when the RPMs were above 10.5K. Above this RPM the turbine efficiency also hit a plateu for all plenum sizes.

Very interesting article to say the least...

[SPARKY]

Mike told me several years ago that log manifolds had few drawbacks if any.

[Crackerman]

Dynoroom
I wasn't meaning to sound like i was challenging you. I really dont know what happens, so i figured i would ask from another perspective. I have heard of gaining 200-500 rpm or more faster spool time. Thats as much as i know as i have never had the opportunity to build a manifold and compare apples to apples. my hope was to make more information available.

[Sumner]

..During the simulation, the log style manifold volume was varied from  2 to 6 liters and the equal length manifold had the runner diameter varied from  22.5mm-50mm.  They came to the conclusion that the log style manifold was more suitable for F1 applications due to the capability to maintain the pressure within the system....

Interesting.  Was that with the new 1.6 litter motor or the older 2.4 L???  Did they have any plenum sizing formulas for larger motors?  I'm assuming you can't read the article online?  Did they have any illustrations of the log style they used as there are a lot of options for that type design, any of which could be considered 'log style'?

Thanks,

Sumner

[Dynoroom]

Dynoroom
I wasn't meaning to sound like i was challenging you. I really dont know what happens, so i figured i would ask from another perspective. I have heard of gaining 200-500 rpm or more faster spool time. Thats as much as i know as i have never had the opportunity to build a manifold and compare apples to apples. my hope was to make more information available.

On the contrary, I thought you may of had some real world experience with smaller engines. Sorry if I came off hard nosed. Getting your point across in type is much more difficult than face to face.

Regarding Skywalkers post about F-1, the information is from the 2014 season championship winning engine.
One of the other points they make about log exhaust manifolds in F-1 was it allows smaller more aero efficient body work to be used which also allowed for an speed advantage.