X-Mazda 3 Racer

Just a curious question, how do you guys figure out that the runners are equal length? I know some CAD programs can do that in design, but the real deal is usually never correct do to imperfections, or are they? Do adjustments have to be made? Do you test each runner for volume? Is there a percentage tolerance that you allow yourself? I know that it may not be that important, but to some getting the exhaust pulse to hit the turbine wheel in exact increments is a big deal. I am just wondering because I remember my dad's buddy used to make turbo manifolds for VW's back in the day and he used sand to calculate volume in each runner ( way before CAD programs ) and because the tubing was mandrel bent, it was all the same ID, the length would in turn be within about 1% between all four runners. At least that's what he said. I am just curious how other people do it like Full-Race Geoff, RPI, Lovefab and others. Educated answers only Please!

agrabau

I have built a couple as well as other exhaust parts for different applications. Your assumptions are all correct. I used CAD to do most of the work but volumetric calculations can be done with something as simple as water from a glass.

If you can get 1% deviation in length you're doing very well. A manifold for turbo is probably one of the most challenging parts to build (equal length aside) because space (runner length) is so limited as the turbo has to remain high, the tubing is thick and difficult to cut and shape. I can go on but i'm getting off subject. Concerning length, you should determine length of the longest spanning runner and then work from there inward to the shortest span (head to collector). CAD is fine and good but when it comes to make something by hand from multiple pieces it's importance diminishes.

ComposiMo

Personally i think there is less importance on making the runners truly equal length on a turbo manifold, and more importance on how well the manifold flows... The 4 runners on our manifold are not equal... they are equal from side to side (1 and 4 are equal, and 2 and 3 are equal), but they are also with a few inches of each other.

I've contemplated making one that actually is equal length just to see if there are any real differences, but the dyno in a recent post by Titan Motorsports on our manifolds shows the improvement that one manifold can have over another based on how it flows... the other manifold wasn't quite equal length either, but the way everything flowed in the manifold was very different.

And to answer the question of 'how to measure', in my cad program, i can measure the length of each runner in the program.

Full-Race Geoff

hint: its not length or volume, its flow rates and compensations made as you turn the bends into straights. the only other person i know who has done this correctly is bisi