Constructing your own Intake Manifold

[vex]

I'm looking to build my own intake manifold. I've done the numbers, and everything but I'm getting snagged up on the "easy" stuff: namely, I'm looking to go to a 90mm throttle body with a single venturi, and I have 2.5" piping. So in other words, I need to go from 4 inches to 2.5 inches. I looked on ATP but didn't see any transition pieces like that. I would like to do this as smooth as possible; so that would mean on single transition... Which lead me to this: Link

Some other throttle bodies I'm thinking of include a 75mm Venturi Sized or a Q45 TB. It really comes down to being able to plumb and run everything, while balancing the effects of the size of the throttle body. So... Suggestions?

[RotorDad]

How long are you doing the intake runners, what material, where are you placing the throttle body & what dimensions were you using for the plenum? You maybe able to make or have someone fab up a pipe to do the transition you want.

[vex]

Originally Posted by RotorDad

How long are you doing the intake runners, what material, where are you placing the throttle body & what dimensions were you using for the plenum? You maybe able to make or have someone fab up a pipe to do the transition you want.

Intake runner length is TBD but is going to be around: .5m
Material: Aluminum
TB placement is perpendicular to runner direction (still working out the finer details)
Plenum is still in the design phase (currently working on ensuring proper distribution of air between all runners)

Consequently after talking with other tuners and fabricators they suggest I use a 75mm TB as 90mm TB is overkill for what I'm planning on.

[TitaniumTT]

Fuck silicone............ sometimes literally

Build it T. Flat AL rolled up and tig'ed and some pipe welded on each end. BAM!

[vex]

ROar! My flow distribution goes to complete crap! Intake inlet velocity is an assumed 47000 mm/s with a total pressure of 24.7psi. (Patm+10psi) This is going to be a lot more trouble some than I thought.


**Flow distribution is important because if you don't have the proper flow distribution you can starve a rotor, while leaning out the other. Which appears to be the case with what I have so far.

[Travis R]

Try putting the TB parallel with the runners.

[infernosg]

Originally Posted by Travis R

Try putting the TB parallel with the runners.

My thoughts exactly. I've not been in the rotary world for a while, but has there been overwhelming success with a setup that has the TB and runner perpendicular? IMO, Mazda designed the system that way for a reason and you'd be better off coming up with something along those lines (TB parallel with runners), but with more volume, longer runners, smoother transitions, etc.

I'm toying with the idea of making a custom manifold for my ITB project, but obviously I won't have a lot of these same problems...

[vex]

Originally Posted by Travis R

Try putting the TB parallel with the runners.

Addressed below.

Originally Posted by infernosg

My thoughts exactly. I've not been in the rotary world for a while, but has there been overwhelming success with a setup that has the TB and runner perpendicular? IMO, Mazda designed the system that way for a reason and you'd be better off coming up with something along those lines (TB parallel with runners), but with more volume, longer runners, smoother transitions, etc.

I'm toying with the idea of making a custom manifold for my ITB project, but obviously I won't have a lot of these same problems...

Honestly if that's what I have to do, that's what I have to do. As it stands I'm trying to avoid doing it so I can have a much more compact set up. It would also eliminate the ninety degree bend that everyone has right in front of the throttle body (can we say pressure drop?). I'm still massaging it a little bit, and I'm confident that I'll be able to get it to flow right with the current thought/process (I have pretty much the entire summer to do it).

[WE3RX7]

Originally Posted by infernosg

My thoughts exactly. I've not been in the rotary world for a while, but has there been overwhelming success with a setup that has the TB and runner perpendicular?

13BMSP - aka Renesis.

[Travis R]

You could also try tapering the manifold to follow the flow shown in the CFD, in your other thread. Makes it a little tougher to construct in real life though.
Good luck

[RETed]

Problem with running TB inlet (technically "outlet") to the runners inline is trying to get the airflow to balance between the runners...
Perpendicular surge tank plenums are popular due to the configuration of the engine - longitudinal - versus location of the IC - up front.
Look at all the popular front-engine, RWD, longitudinal located running perpendicular surge tank plenums:
2JZGTE
1JZGTE
RB26DET
SR20DET
VG30DET
The exceptions:
7MGTE
RB25DET (can this be legit with the RB26 "big brother" listed above?)
VR38DET (current R35 engine)

You can take advantage of turbulent flow when using a perpendicular surge tank plenum, but it's still tricky to get total equal flow between the runners.

This is a daunting task to iron out all the details, but if you do success, it should reward you with big gains.


-Ted

[RotaryProphet]

Your best bet is to taper the plenum along it's length and to lengthen the plenum past the last runner.

One issue is that the tapering is designed to equalize flow between the runners, but our runners are inherently unequal; the primaries in the middle will flow less than the significantly larger secondary ports on the sides. Also, the ports aren't uniformly distributed across the length of the motor. This is why Mazda went with the setup that they did; getting flow balance across this sort of setup is very difficult. A compound curve which narrowed the plenum by a certain volume at or just after the first secondary runner (from the TB's POV), then ran reasonably straight until it reduced again at the primary runners, and then reduced a final time into a rounded cone sometime after the last secondary.

However you do it, you have to get the air to make the turn into runners... and good luck with that! If you figure it out, then you're a better man than I, sir.

[vex]

Haha, thanks for the votes of confidence and suggestions everyone. As it stands I found my error (it was in the form of incorrect initial conditions) Since i've ironed that out my flow pattern is remarkably close to being exactly what I need. I'm working through a single design iteration and I think I know how to get some better flow out of everything. Will post results (if any) in my build thread later on.

[vex]

General update:
tapering of the plenum resulted in poorer flow to the leading runners. Disproportionaly charging the last runners to almost twice as much velocity.