just curious, why not widen the transfers over across towards above the intake????
interesting exhaust duration on the stock 460. I've never seen anything on a saw above even 165. not to be critical, but any chance you measured wrong????? That just seems wild to me.
The transfers have been widened a bit on the back side, did not go crazy with that, go too big and velocity drops and scavenging is then hurt. The saw will rev well but will be hurt torque wise. Also the rear side of the transfers should be aimed so the two streams flow into each other at the back cylinder wall so they then rise up the back wall of the cylinder, aimed too much towards the center or too far apart and the scaving loop is not as effective. Thats the way I read it anyway.
The porting numbers on these new stihls is high, it did not seem right to me, tripple checked and it is indead that high. 660 is higher than the 460 yet at 180 deg. Thats higher than I would port a work saw, more where a stock appearing racer would be.
For determining numbers I use a few pieces of computer software, some I came up with myself, to look at predicted port velocities, cylinder preasures, temperatures, even sonic wave action, then work up changes experimenting in theory with different changes to port and duct areas, heights, angles, timing, compression... It's fairly time consuming, takes 3-4 hours just to go through a saw and measure up everything, then each simulation takes 10-20 min depending on how complex the motor is, more ports, more detailed port and duct profiles and it takes longer, pipes or tuned exhausts also add time to the number crunching..
I'v been attacked before on using computers to simulate or predict what in theory could work, and I conceede it's very true that a simulation or numarical modeling is just that and it can never be 100% accurate on a complex and dynamic real word situation. But from trying maybe hundreds of theoretical designs on dozens of real motors and finding them all to have come out within 10%, and most with in 5% of the expected result gives me a fair degree of confidence that it's a valid method.
Alternatives are to reclcle a lot of aluminium trying different designs or just guess.