| Elated... Today is that kind of day. A TwinTurbo cylinder head was obtained for the purpose of carving out just a single cylinder section for flowtesting of various exhaust manifolds. The exhaust port was modified to match the ports of the aftermarket exhaust manifolds that will be tested. OEM exhaust valves were installed and positioned at 0.385" opening from the seat and WELDED into place. This is the maximum lift of a Jim Wolf Technology 500-series camshaft. 
The results at this time are limited solely to the driver's side exhaust manifold since we only have a 3D print of that side that is assembled for testing. The passenger side comparison will come shortly. The final results were as follows:
MSP Manifold:
Cyl2:178.6CFM
Cyl4: 150.0CFM
Cyl6: 169.6CFM
AVERAGE: 166.1CFM
Standard Deviation: 14.6CFM ASHSPEC Manifold:
Cyl2: 172.6 CFM
Cyl4: 174.1 CFM
Cyl6: 169.6 CFM
Average: 172.1CFM
Standard Deviation: 2.3CFM From these results it is pretty obvious that the flow capacity of the cylinder heads is the limiting factor given that both of these manifolds are capable of flowing considerably more even in an open-flow condition. However, what came completely unexpectedly from this test is the result we see in the Cyl4 test of the MSP manifold. You can see in the video,
https://www.facebook.com/photo.php?v=796368943725222
that the manometer reading of pressure differential was skewing around quite a bit. There is an audible change in sound of the air that coincides with the fluctuation in the manometer reading. This artifact is a result of an oscillation from laminar flow to turbulent flow within the flowpath. This was an unexpected result that I have never seen before in any flowtest I've personally performed. Peculiar, no doubt, but I thought it best to disregard this anomaly and find a flow control setting that this deviation centered to. The result of this approach revealed a flow value of 150.0CFM, +/-2CFM. This condition was also observed in the central port of the opposing side of the Mike Smith manifold but to a slightly lesser degree. This was not observed in any of the AshSPEC manifold ports. Overall gains in this manifold are around ~3-4% which is a good bit lower than the simulated results, however, the final test conditions have taken the flow study to a configuration that also incorporates the flowpath preceding the exhaust manifold for the best "real-world" comparison. As a result, we also discover that there is a much larger separation in flow consistency in the MSP manifolds than what the previous "open port" tests revealed. Maintaining an even flow across all exhaust ports will ensure that an equal amount of exhaust gases are removed from the cylinder, thus allowing an equal amount of clean air and fuel to enter into the cylinder on the next intake stroke to produce power. The EFI system does not meter airflow into each cylinder independently - it averages the airflow and thus, delivers an average quantity of fuel based on demand to each cylinder. Cylinders that do not evacuate as much exhaust gas will run rich and those that evacuate more exhaust gas will run lean. This imbalance not only reduces the total power output of the engine since wise tuners will always tune for the leaner cylinder bank, but this imbalance also increases the probability of catastrophic engine damage resulting from lean air/fuel mixtures. Maximum deviation percentage between any two runners:
MSP: 19%
ASHSPEC: 2.7% So far for the driver's exhaust manifold our new design shows promise in total flow, average flow, and consistency. I'm considering to take one of the other three sections I cut out and machining it for a larger valve and seat/bowl setup to see how much more we can get by opening up the exhaust bottleneck that is apparent here. Just for comparison's sake: maximum bulk flow of the new manifolds with laminar port entry conditions showed simulated results in the 280CFM regime.....
Enthusiasts soon understand each other. --W. Irving.
Are you an enthusiast? If you are out to describe the truth, leave elegance to the
tailor.
Albert Einstein
|
Straight out of the mad lab.... >> - 
