Honda / Acura K20a K24a Engine Forum banner
1 - 3 of 3 Posts

13 Posts
Discussion Starter · #1 ·
I see a lot of sidewinder type manifolds for the K with unequal length runners. I also see some people pairing the wrong runners together ( why even bother twinscroll at this point?) and getting equal length and equal number of bends on various motors, Ie pairing 1+2 and 3+4 instead of the correct 1+4 and 2+3.
I do not want to call any manufacturer out because I understand engine bays complicate matters and compromises must be made, but I would like to know if anyone has some empirical input on what should be maximized:
- the length of the runners to be equal length
- the number of bends in each runner to be minimized ( but different between cylinders)
- the number of bends to be equal in each runner

I am building a twinscroll for a MR-S to no off the shelf solutions exist for me. I had been going into the design process with the mind to optimize for equal length, but not really care if some runners had more bends than others. I saw a fleeting comment that suggested to minimize the number of bends and to hell with equal length which got me thinking and google image searching, and seeing some terrible stuff out there, and then seeing some of the best and very $$ name parts out there with issues such as runner 1 and 4 being half the size of each other, when those are supposed to be equal timed apart for their own turbine volute.. I guess without knowing otherwise I would expect keeping 1+4 merged and close to the same length as each other, and 2+3 merged and equal length to each other would be crucial to actually getting the benefits of a twinscroll system where the exhaust pulses all hit the turbine wheel equally spaced out from each other and completely separated by adjacent cyl firings.. ( which again is completely lost if you pair 1+2 and 3+4 like a top mount I saw with an EFR mounted to it)

optimize for fewest bends, equal number of bends, or equal length for a twinscroll manifold?
1 - 3 of 3 Posts