Spalato

iTrader: (33)

Thank you all who contributed to the thread, I hope we can continue the discussion in a more civilized manner.

KNONFS

iTrader: (14)

From what you saw on the dyno sheets, plus the engine port setups, was there more power down low even tho the turbo was spooling slower? Or was the adiitional power only after max boost was reached?

mannykiller

iTrader: (93)

I know A lot of people would disagree but I too have been doing my own research on spool and response....and I've taken a different approach as well. I am a Huge fan of Howard and I take EVERYTHING he says into consideration? why? because he's not only my engine builder...but he's always been helpful to this forum and whether I agree with him or not...he has more experience than me and that says something. Simply talking about turbo exhaust manifolds I take the side of the Long Runner, slimmer ID. Even though our ports are 2" The new piping I ordered for my New long runner design measures 2.00"...that is the OD (outside Diameter). I chose schedule 11 or ".120" wall. 120+120=.240" .......So 2.00"OD -.240 Wall" = 1.76" Inside Diameter. With the Pipe being 120 wall it should help with strength...Although I'll still be using a supporting brace to help support the turbo...this 11 gauge 304 SS will help keep the manifold reliable in that it wont crack as easily from fatigue (hopefully). Also, with the heavy Gauge wall .120... it keeps the ID at 1.76".....so this IN THEORY (the one I am following at least) will keep Exhaust Velocity up. The length of the runners is very important because the longer the runner (to an extent) The more time you'll have to take advantage of Exhaust Gas Expansion. With this theory...The shorter the manifold.. the gas has less time to expand out of the motor...and thus you won't be able to fully take advantage of the explosion. For instance the air fuel mixture hasn't had time to mature straight out of the motor... It hits the turbo and is still expanding in the downpipe...far after it's fully reached it's potential. With longer runners....it gives the exhaust time to fully expand or Mature...so that when it hits the exhaust turbine wheel.... it is at it's maximum energy potential. Maybe it's too late to take advantage of the exhaust pulse...but what you lost with the pulse... you gained in an extra 20-35% of Expansion. So if you take the pulse effect and put a number on it you'd subtract that number of that exact energy potential from the expansion/heat loss (If there is any) and you'd get an exact number of how much more exhaust gas energy you are taking advantage of. So potentially you should be able to keep response up by taking advantage of this expansion, make more power, use less fuel, and be overall more efficient. I know I'm not an engineer, but there is a reason F1 turbo manifolds are long.... everything in F1 is done for maximization. Again, this is just my Opinion and this is the THEORY I have chosen to follow when designing my manifold.
In regards to wastegate placement... Many F1 turbo manifolds place the gate directly inline with the flow of the exhaust.. I think Howard is right, and I agree with him that you should find the least flow interrupting spot possible...but it is a balancing act. too much and you interrupt flow..too little and you risk not being able to control boost pressure. IF my legit rocket lab welder can do it i'll be integrating my wastegate runners into my exhaust wheel Housing on my new EFR. Their iwg housing looks like it uses 4 ports on the inside...which look damn similar to eachother give or take some surface area. 2 go to the exhaust wheel, and 2 go to the internal gate. I'll be doing a bit more research but I can't wait to get my manifold done! This thread could become a VERY good info source!

mannykiller

iTrader: (93)

p.s. I tried a bigger intake plenum..with longer runners. It created this weird turbulence that was instantly fixed by switching back to the OEM intake set-up.

Speed of light

iTrader: (8)

The turbines in the turbos we use are known as impact turbines, which primarily use the kinetic energy from the gas velocity to excite the turbine; the secondary stage of expantion across the turbine also contributes energy recovered once its going--but doesn't do as much for initial spool. It does, however, have a substantial effect on turbine efficiency.

Given the foregoing, I want the turbine to receive the most dense gases possible AND I want the expantion to take place across the turbine (not before). For best response, I would bolt the turbo right to the exhaust port--if I could.

RENESISFD

iTrader: (13)

agreed completely.

If he does not understand that statement he should not be in this thread, it is over his head.

mannykiller

iTrader: (93)

But more expansion means more volume..nah?. I know too long and you lose energy..but I think it's more about finding just the right balance so that you get maximum exhaust gas expansion coupled with the the right length. I believe that length to be right about 16-18 inches.

Howard Coleman

iTrader: (6)

"Given the foregoing, I want the turbine to receive the most dense gases possible AND I want the expansion to take place across the turbine (not before). For best response, I would bolt the turbo right to the exhaust port--if I could."

that got me laughing as it is my line.

hey, we are all pursuing our theories here and crazy things sometimes work. crazy things happen, Packers/Cowboys yesterday anyone? i have seen long manifolds on super fast drag cars, i have seen some F1 cars w long manifolds.

i have experimented w intake runner lengths on my twin Weber 45 DCOE manifolds on my 4 cylinder race engines and found major output variations caused by one inch of runner length diff. after you let your super computer cool down from running the numbers you will find that nothing beats just doing it and looking at the results for guidance.

my money is on short and straight w 50 mm ID exiting the motor and making the turn. so far i am not displeased w my initial results. 403 SAE ponies at 5000 rpm and over 160 mph in 3/8 of a mile. of course there is still a lot of work to do and some of my ideas may not be optimum but i both enjoy working them as well as sharing. there are loads of people on this board that have more experience and they are always welcomed to share.

one of the areas that interests me (and was brought in the initial post) is our throttle bodysizing.

the OE TB has 8.546 sq inches of butterfly area.

our LIM runners are 5.7 sq inches

just as an FYI a 90 mm is 9.857

given our runners are of modest area would a 90 mm TB be a plus over OE? the other consideration is the fact that the single butterfly doesn't line up well w the UIM. Luke has always been wanting to fab a UIM for me. he thinks it would broaden and move the torque curve right. the stock UIM is probably well engineered for its purpose. given OE is about 216 rwhp we are operating in another zipcode.

howard

mannykiller

iTrader: (93)

^I was going to say ....I've heard you say that hundreds of times howard..."I would bolt the turbo right to the exhaust ports if I could."

bumpstart

i think much contention comes from quantifying the energy in the exhaust , either as heat,, as pressure, as velocity, or as flow
all of us not quite grasping that not one sole variable is a complete quantification of that energy


ie.. simply heat and pressure need to be observed together,, to grasp flow..

.. and so none of us really have a full count of what is going on in terms of work being done by the exhaust

i think one point we are all trying to grasp is that if entering into blow down earlier,, or if timing combustion later,, so there is more energy in the exhaust \

is that there may be,, at least for certain rpms ,, ideal turbo manifold length and runner size equations that allow the afterburn principle to happen to most effect , yet not cool or lose energy before hitting the turbine wheel

clearly it wont be the same on all setups, given differences in porting and timing , and quite possibly the best way of going about finding it is suck see,, alongside maybe getting a thermocouple to measure exhaust temp and pressure pre,, and post turbo,, aiming for the biggest differential

Vicoor

temp differential across the turbine would be a really interesting data point and may yield some truly valuable information when compared to spool.

However I'd be skeptical of pressure differential as a indicator because at the extreme velocities that must be present, static pressure differential would not reflect the potential energy due to Bernoulli's principle.


It's great that this kind of informal theory is brought forward in this forum. It may be over most of or heads to fully understand the dynamics of a turbocharged rotary, but being able to go through this brings us all a little closer.

And thanks to bumpstart for striving to maintain a board where civil discourse is the rule.

bumpstart

another thing that is overtly hinted at ,even when comparing induction lengths
.. is there is often another or more unseen principle at play

i would think the Helmholtz theory
( which mazda exploit to varying degrees on the induction side with all its engines from about 1983 onwards )

could also apply to the exhaust,, and be complicated by things like pulse frequency from the exhaust
,, and if so,, it stands to reason than MORE than one ideal length/volume equation exist for any one target rpm

ie.. seeing how harmonic theory shows that each frequency will have multiple ideal pipe length/ diameter combinations
.. then it is also very likely, almost certain .. that there is multiple exhaust pipe length/diameter combinations to make best advantage of secondary effects not yet quantifiable ( or easily understandable ) from our viewpoint

im basically saying ,, there is more than meets the eye at play,, more than we can ever really judge , and it is very plausible that two completely opposite paths or POV achieve the same result

and about the only way we can measure it is by power measurement at the wheels at each target rpm

the only way we will draw something conclusive.. ( but not be fully aware of all interactions,, just judging the net effect )

it will be a battle of power graphs ( using the same turbo on differing settups )
displaying area under the curve,, from the rev load point we nominate as the datum point for "early spool "

djseven

iTrader: (56)

https://www.rx7club.com/build-thread...1013942/page2/

Post #38. Looking at them again the only downside was power/trq below 3k rpms. I had not looked at them in awhile and thought there was a larger difference down low but I guess not. This was on same dyno, same tuner, same turbo and I am pretty sure same boost or very close(19-20psi off memory but Im sure its in the thread). I can't speak for boost response I would assume it was very similar, I was meaning to speak on low end torque but obviously it wasn't as exaggerated as I thought.

As you can see with a small turbo and medium/high boost pressure you can create great torque even on our torqueless wonder rotaries. Same concept used on a lot of newer turbo cars and really just like the stock sequential setup. Carter's setup is proven and his driving skills are impressive as well. He also makes some great parts for the FDs for those unaware.


Here is the link for the engine in the car so you can see the port:

https://www.rx7club.com/3rd-generati...usings-976907/

bumpstart

and for howards post.. if we take you average turbo ,, and have a look at the pressure ratio for what most people would consider average boost

...and see that it is usually between 2:1 and 3:1 ( depending how bad your choice was )

every psi drop lost in the inlet tract restriction cost you,, lets say 2.5 psi in exhaust backpressure to recoup the work needed to compensate it

it stands to reason then that if stock throttle body is restrictive above a certain flow point
,, then it needs to be improved,, as removing inlet losses without sacrificing the inlet temperature is the best way to optimise power from what you have,, full stop

i thinks you would need a pressure gauge either side of TB to quantify if you have an issue,, and how much win you get

everything has a cost .. a single throttle body and common mezzanine cost you the lovely Helmholtz and significant low down TQ and drivabilty if you dont have resolution in the ecu and throttle pump to keep up with the changes in airflow during throttle flair

mannykiller

iTrader: (93)

A lot of guys think I'm out to reinvent the wheel with my set up, but really..I'm just out to prove that it'll work and work well. Bumpstart said it well.

arghx

iTrader: (3)

For what it's worth, I've done engine dyno work on piston engines and dealt a lot with simulation results on commercial software. Dealing with static and dynamic pressures either for compressor or turbine side ends up with different numbers but the same story. When you do a dynamic pressure, you have instrumentation issues and other assumptions to deal with. Static vs dynamic (well, static vs static + dynamic) mostly matters for trying to precisely plot on a compressor and turbine map. That in itself isn't fully possible--you can get in the ballpark, but the major turbocharger manufacturers have their own internal calculations and corrections they don't share with the public. All the stuff you read on the internet is only ballpark. Those compressor maps are done in a lab on a gas stand, not on a firing engine. They're often not accurate at the edges of the map, and the manufacturer has proprietary correction factors.

I've had fully instrumented engines on engine dynos with temperature and pressures in manifold runners, before and after the turbo on both the hot and cold side, all through the exhaust, basically everything. The two things that make the biggest effect on spool, besides the geometric parameters of the turbo itself, are the mass flow coming out the exhaust port and the temperature of that exhaust.

On piston engines, it's easy to manipulate this for spool. Almost all the new turbo direct injection engines have a "spool mode" where the cams phase into a high overlap position as soon as intake pressure exceeds exhaust pressure. Advance the intake cam, retard the exhaust cam, and suddenly you have 100 degrees of overlap for a split second. Typically the AFR is rich in the cylinder but Lambda=1 in the engine-out exhaust due to fresh air passing through unburned during scavenging. Use that scavenged air with retarded spark timing and you get very fast spool.

Look at basically any modern twin turbo or single twin scroll turbo direct injected engine and that's what they do. Since they're not port injected, they don't have to worry about passing unburned fuel through exhaust during overlap either. It's how you see peak torque under 2000rpm on all these new engines using single twin scroll or parallel twin turbo arrangements.

Wastegate pickup for spool, throttlebody size--we can those kind of parameters all day, but they are not nearly as big of contributors as turbo sizing, exhaust mass flow, and exhaust temperature.

Speed of light

iTrader: (8)

....sorry Howard, I wasn't meaning to plagiarize, it's just that Joe Biden told me to say that. Or perhaps great minds think alike.... [rimshot]

But seriously folks, this is sort of an interesting thread. I see this as a problem of how to best harness available energy--as I believe bumpstart was also saying. You have a heated mass, under considerable pressure, and once released, with considerable velocity (and hence mass flow as velocity-pressure). And depending what you want to achieve, there are several different ways to skin the proverbial cat.

Regardless of where you put the turbo, it's all about getting the most energy input to the turbine with the least amount of loss. Ideally, you want to keep the pressure up and density up--when gasses expand, you're losing pressure, resulting in gas that is less dense. In my world, I want this to occur at and across the turbine, where it will do the most useful work. I don't know if a header tube, acting as a [secondary] combustion chamber, can result in additional pressure or density in the tube--this is the realm of the constant pressure cycle engine, e.g., gas turbine. It seems to me for that to work, you would need to introduce addition combustion air directly into the exhaust (as opposed to just lengthening the burn). What you're really doing when you use a two-step, for example, and to a lesser extent when you're tuning for spool, is slowing/delaying combustion so, to some degree, expansion occurs too late to act on the piston/rotor, however, the reaction is still occurring and expanding and its energy can be harnessed to act on the turbine. I don't know that a long tube manifold is necessary or improves on that effect. I will reiterate that most of the work in these turbines occurs as a result of mass flow acting on and expanding through the turbine, only a small fraction of energy recovered is thermal; directly from the heat of the gasses. In fact, these turbines will run just fine on an equivalent mass flow of cold air.

On the subject of long runner manifolds, I agree with those that have observed that they can work well; the caveat is that they introduce a pronounced tuning effect and can work against you (and potentially narrow your powerband) if they're not well sorted. Fortunately, some of this has been done for you by the various vendors. If you're willing to try, test and modify there are probably good gains to be had if you can find the 'sweet spot' you want, for your combination, as was alluded to above.

One decided advantage that many manifolds give you is the ability to run a divided setup, improving response without much downside.

I will relate my own experience about the sensitivity of tuning: I was racing V8 motors a while back and I did an intake manifold change that should have increased power; however, after the change the car would hardly get out of its own way. So where was the power? I'd have to 'find' it.

After some consideration, I decided to lengthen the manifold's runners by 3/4's of an inch using epoxy and reduce the plenum volume somewhat with a wood block (it was a Holley tunnel ram--designed for pro stock use). All the time, the resident bench-racing naysayers were telling me that it would never work and that I needed scrap it and to go out and buy more expensive, trick race car parts from their favorite vendors. Well in the end, it worked like gangbusters....e.t. dropped by about a second (and into the low 9's) over a prior good setup. Indeed, it worked so well that my small block was reclassified to run heads up with big blocks having a liter or more displacement advantage.

Moral of the story: performance, for any package, can pretty amazing when you get the combo, setup and tuning right.

lOOkatme

iTrader: (19)

My bet is the shorter manifolds with more direct air flow to the turbine and smaller manifold volume will perform better for spooling a turbo than that of a longer one who has more bends and volume to get to the turbo being less efficient.

From the dyno's I see, the smaller turbo's with more direct manifolds have quicker spool times (WHP is shifted to the left), but I don't know many smaller turbo's that have really long manifold lengths like those of the larger turbo's.

I think the best set up for the most efficient highest power rotary is a set up that might include.

12.5 AFR tune in vacuum cells, optimized timing
Optimized turbo that you can fit with a short runner manifold (for whatever WHP the user wants to achieve)
3.5-4" straight through exhaust
V-mount with matched piping of compressor outlet head. less bends the better, smaller ducted intercooler. Smooth and match all internal bends/TB to UIM/ UIM to LIM, etc.
porting, questionable, but probably stock ports for spooling?
Engine with good compression
anything connected to the drivetrain being lower moment of inertia, 15" 8lbs volks, lightweight flywheel, anything to reduce drag on the engine.
Lightweight turbine and compressor wheels
Huge air filter or no filter at all
Optimized placement of wastegate(s)
Optimized placement of BOV


I know the long manifold runner lengths on piston engines are used sometimes for exhaust gas reversion and compression of the engines. reversion can happen causing autoignition. But I am unsure the effects of reversion on rotaries.

ninesixtwo

iTrader: (9)

Resonance tuning on the exhaust side is far trickier than it is on the intake because the underlying force (reflected pressure waves) is heavily dependant on the temperature of the medium through which the waves travel.

gxl90rx7

iTrader: (1)

im no expert on this, but I have tried several different manifold designs, and the results tend to agree with the concepts that mazda seemed to follow in the design of the S5 turbo. I believe the best manifold design would be runners short as possible, with one runner having straight shot to an exhaust port, and match runner area to the housing exhaust port area, which is about 1.3" on stock ports.

also imo external wastegates are bullshit. a well-designed internal wastegate will outperform an external all day.. better location, more reliable, less space, easier plumbing

Grant M

From quick google searches you can find some decent back to back tests on different style turbo manifolds being used on the dyno and chart plots showing boost to rpm.

The one I found was for #cough# honda s2000 #cough# a log style manifold was used, this have earliest spool out of the lot.
Tubular enequal length manifold, this gave slower spool, 300rpms iirc, but much better top end power, 60hp iirc.
A long runner equal length fanimold was used, again slower spool by 200rpms but gave higher too end power due to better fanimold volume allowing the exhaust gases to expand and work better at high rpms.

The turbo used was a gt35 T3.

I have made my own manifold for my rx4, currently at my welder fabricator getting TIG'd, iv have basically just gone with the same design as E&J are using on their rx8 drag car. Making 1,000+ hp 13B.

Iv done it like they have so that I can put the turbo where I want, it moves my red hot turbo away from my LIM, shortens the intake pipe length and in all honesty. Looks cool as hell.

WG placement is at 90*s to the 2 runners, one 60mm WG, so it's neither in the flow of exhaust or against it.

j9fd3s

iTrader: (3)

coming from the NA world, there is something going on, and just for the overall power band there seem to be about 3 lengths that work. for just a specific rpm point though, its hard to say. the traditional system is really long 90-120", short is 20-32", and lately there is a mid length that has caught on ~50"

additionally Mazda published a part of a header length study on the racing 3 rotor engines, and the length of the primary and power aren't linear.

50mm ID is a good choice, a stock manifold is 50mm ID too. the problem is the stock sleeve, the port opening is ~40mm, the sleeve exit is 48mm, and then the turbo inlet is something different again.

the race engines had a 43mm sleeve, and used a 43mm ID header….

DR_Knight

iTrader: (23)

Electronic Boost controller tuning