bics56
January 3, 2007, 5:49am
71
I spend about 3-4 days doing a street/strip type port job on 3400 heads. If I “flow match” them on my flowbench, it was taking me 3 weeks to do this when I first started. Last time took me 4 more days. Now I don’t work 8 hours non stop as my wrists won’t take it. I also do more than port work. My flowmatch is 2% variance max between 6 ports. The flowbench I use has 1% repeatability, proven in magazines and my own testing so that leaves 1% for my error. I try to go as close as possible without chasing my tail past where I want the shape to be but I have got them as close as 1% max variance (.5% from the mean value). That is average for all values from .050 to .500 in .050 increments. While I do my best at each measurement, the casting difference between heads/cylinders makes it difficult to do that. Its interesting to note that if I turn the screw a little, I can usually get the values desired. To do a true graph of matching, you need to measure in tighter increments…say .010.
I was told cnc can do some crazy % variance, like .005 or something. I don’t know so I won’t comment on that. I am not trying to say a human can beat a well programmed cnc machine because of all the variables. Popular engines for modding will get that sort of aftermarket attention because there is money to be made. CNC is faster for repeatability and from that stand point is far superior. You can’t really compare that with a custom hand porting though.
If you are a 2% Variance and the machine is a 1%…wouldn’t that make a possible variance of UP TO 3%? Not that it makes a difference, just wanted to point that out (I am somewhat of a #'s nerd)