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.
Intake design/shape debate, end-feed vs center feed
Would like to hash this out via theory and experience.
Looking at the two available types of intake manifold for a 944 turbo (and really, for most stuff) - "end feed" like the stock 951 intake, Revline 16v intake, 944 LeMans, SFR. The other option is "center feed" like the LR intake or Shawn's custom piece. See pictures at bottom of thread, hopefully Shawn doesn't mind me borrowing his excellent photos representing both options.
Since this is an intake thread and I know it'll be mentioned, please set aside any consideration of resonance tuning and just focus on air flow distribution.
With the center feed, LR and Shawn say that they have not seen any evidence of the center cylinders being favored for air over the outer cylinders.
It is possible to angle the "entrance" to be somewhat perpendicular to the axis of the runners so the air has to change direction rather than shooting straight across to 2/3, and this is how lots of "other marque" manifolds are made, see the Macan 2.0T intake for example.
With the end feed, supposedly the rear cylinders (at the far end of the plenum) are favored because "air has mass and wants to keep moving."
However, in searching the internet for info there seems to be just as many people saying that the FRONT cylinders get overfed and the rears get starved...
The most glaring thing I see when someone invariably posts a CFD graph is that 9 times out of 10, all the runners are shown as flowing at the same time, which is just not the case in a poppet-valve piston engine with a fixed firing order. This "mistake" DOES seem to support that the rear cylinders would be favored with the air molecules taking basically a ballistic path (in a straight line but curving gradually towards another force, in this case low pressure outlet instead of gravity).
In a 4-cylinder you have at most 2 cylinders' intake valves open at the same time, and one of those is almost closed (pushing compression back out the port) when the other one is starting to open.
(On this point I also think there's not a whole lot of point in making the plenum super-huge when operating a non-restricted engine in a normal sort of RPM range, ie south of 7000 rpm)
So...if anyone has thoughts or experience about one type of manifold FLOWING MORE EQUALLY versus another, please chime in, along with thoughts how to help a given design flow more equally.
The action of valves opening and closing have such a huge effect on the pressure waves generated inside any manifold, that if you dismiss resonance tuning, or it’s effect on air distribution, imo the exercise is pointless.
The action of valves opening and closing have such a huge effect on the pressure waves generated inside any manifold, that if you dismiss resonance tuning, or it’s effect on air distribution, imo the exercise is pointless.
The point of throwing it out was to avoid discussion of runner length for boosting torque, and the usual bitching by certain vendors and posters about how they dislike another vendor's intake.
How about something different spencer. Like a twin intake manifold with two thottle bodies to match a twin scroll exhaust setup? Would that work? Let’s all think about it.
Sorry Spencer, I think this is an over-the-top exercise and there is very little to be gained by theorizing/optimizing/testing the intake manifold flow. However, I am with you on any improvements that can be made reasonably.
I think the best example of the argument between end-fed and center-fed is the GM LS series engine intake manifold. It's end fed with a plenum, and flows massive amounts of air. No one seems to have complaints or substantial redesigns (except those high rise tunnel ram, dual quad intake racers). GM put millions into the research and design of the LS series engines and basically got almost everything right, no, perfect!
The end-fed works very well and I'm sure that a CFD analysis would bear that out. End-fed is mostly a substantially linear flow approach with flow bending edges at each cylinder's intake port. End-fed have four bend edges total (one at each intake port riser), whereas a center-fed would have a bend flow at each intake port plus two bends at the manifold inlet and one additional bend outside the manifold, for a total of seven edge bends. Further, valve reversion pulses in the end-fed are muted by plenum flow and size, and provide very little disruption of the overall flow, since little or no port runner tuning exists. The only way to improve an end-fed (and center-fed) is to use intake bellmouths inside the plenum for flow smoothing.
While the arguments between manifold types might be analyzed based on natural aspirated design, a turbo charged engine is a pressurized system flowing into a vacuum (piston downward in cylinder) which almost makes the issue moot.
I realize that there is much, much more about this topic, but I am stopping here.
Go with the end-fed manifold and start enjoying the car!
While the arguments between manifold types might be analyzed based on natural aspirated design, a turbo charged engine is a pressurized system flowing into a vacuum (piston downward in cylinder) which almost makes the issue moot.
Agree 99% with the one caveat, the turbocharged engine is an NA engine at idle and off boost, that's where the long runners keep velocity to atomize fuel and have more volume for inertia to aid in smooth idle.
I can understand from one point of view how end-feed might over-feed the rear cylinders due to momentum (high velocity). But bearing in mind the pulsing dynamics inside the manifold of a 1342 (or better imagined, 2134...front front back back), there is going to be a depression at one end or the other depending which cylinder is on intake stroke so I think it would sort of even out....maybe.
Lots of the high-end stuff without budget or packaging constraints (944 LeMans, 944 GTR, 917/30, Ferrari F40, etc) has still used end-feed so it might not be so bad.
IMO center-feed is more interesting looking, and the "outer pair" of cylinders only has another bend and a few more inches of plenum space to navigate vs the runners.
Yet, again from velocity/momentum point of view, it makes sense that 2/3 could get more air than 1/4 but if there's pressure in the tank and incoming air is just replenishing pressure that was drawn down by a previous intake event it might not matter...
Unless I can squeeze a center-feed inlet pipe in my cramped bay my intake will look very much like the one Jay shared, minus the ITBs...but everything else about the engine will be different
Not sure why this is getting over complicated. Are you building some over engineered motor that will be pushing out over 500hp and trying to squeeze every last 0.1hp out of it?
calculate your runner length
calculate your plenum volume.
THEN work out where you want the inlet from. This will all depending on your room in your engine bay. There are many high end cars with both.
If you want a perfect design, then design one in solidworks and do some CFD on it. keeping in mind you or you'll be paying someone to make the final product.
Sometimes I wonder about "factory" (Audi intake) techs.
I was watching a video a couple days ago about revival of the famous BMW Parmalat vintage F1 car.
Factory techs assembled the engine, video breaks to on track running in, then back to engine room where they disassembled, inspected engine and reassembled it.
Factory guys, suited up in BMW clothing, etc.
Piston/rod combo inserted into cylinder with no rod stud cover with what seemed to be no care at all at nicking a cylinder wall or crank journal.
One of the guys smashed the cylinder head into a head stud while swinging it into position.
They just wiped oil off the block deck and the rod journals with a rag that wasn't even lint-less.
Amazing.
Even on my own Porsche Club racing meager engines, all my components are squeaky clean/dry.
I even use a tack cloth to capture dust and/or particles from my block deck before I lay a head gasket.