I thought I read once about a rotary valve setup, sorta like a ball valve -- I can see there'd be problems there but sure seems like you could have a bigger opening w/o interference and virtually no loss of compression from a big cavity.

http://upload.wikimedia.org/wikipedi..._desmo_8x6.jpg

Mercedes used those for a bit in the 1930's as well.

If you want to build a 10k+ rpm valve train, get rid of the spring. Don't let it jump time though...

I think the more preformance you force out of an engine, the closer the valves are to the pistons.

You forgot about piston design. most have a cutout for valve clearance

My 2000 Toyota Tundra (4.7L V8) was a VVT-i engine and was an interference engine. On the other hand, the same year the 3.4L V6 they put in the Tundras as the base engine was not.

In comparison, my 1991 Lexus LS400 (4.0L V8) has a non-interference engine. Strange that Toyotas FIRST V8 engine was a non-interference one, but that changed in 1998, when the Lexus switched to a VVT-i engine and it became an interference one. Still mixed thoughts on to whether the 95-97 LS400 is interference or not.

The Ducati setup look to be what's refered to as a babbit ear setup. As mentioned Mercedes used it in the 30's, I think. Also OSCA, formed by the Maserati brothers, used it in the late 50's early 60's. I think the other poster is refering to a new setup, thats still being patend.
There are also cam less engines. Ford research these in the 60's. Renault used a camless engine in F1, during the late 80's early 90's. It was pneumatic powered. They were 2000 RPM above Honda and Ferrari's redline at the time. CAT and Cummins and some developement engines that were camless and I think BMW too. I am thinking with the cost of Piezo electronics, this might be a possiblity. Also, any friction or drag in an engine is becoming a concern. It effects MPG.
Briggs & Stratton had some rotary valve engines on there were some for model airplane engines, HP and Austrian company I think.
I've had two cars break their timing belts. One was a 78 Honda Civic. Non-interference. I just put on a new belt, that was fairly easy. The other was a '96 Ford Contour. The first belt went for 126K, when it was changed. The belt that broke went about 45K. The 2.0L Zetec is listed as an interference motor. I was doing about 20 to 30 mph and slowing down to turn, when it broke. I replaced the belt and the motor had a slight tick. I did check the valves by turn the motor over to see if any stuck. They seemed OK. The motor is running OK at a bit over 200K.
Tom

Several years ago, I got to change the timing-belts, including the balance-shaft belt, on a friend's '90 Accord. The ordeal wasn't helped by having to rig a special fixture to hold the crank-pulley, as the Honda pulley bolt is installed with about 250 foot-pounds!

Having owned a few timing-belt vehicles myself, a timing-chain helped move a '98 Nissan Altima to the top of my shoppng list a few years ago, when it replaced my rusty, tired '84 Honda Accord. Plus the fact that used '90s Nissans are reliable, but tend to sell for much less than comparable Toyotas and Hondas. While my Altima isn't as refined, I was willing to trade some refinement for not having to perform an immediate timing-belt job on a used Camry or Accord.

Happy Motoring, Mark