Turned out it got that hot that the valve clearances closed up enough to prop open the ex valves. Increased clearances to racing spec.
That would explain why the push rod didn't jump out.
For a Flash in standard tune (I assume it is?) with the valve clearance set correctly this would mean that the carb setup must be on the extremely lean side (or ignition still is too far advanced).
Andy, can you tell us what your carb setup (main jet, needle jet, needle position) and ignition timing is?
My question is, should the clearance test suggested by Brian be applied to either steel or bronze guides. I'm sure the bronze guides have a greated coefficient of expansion. Would that not make them looser as heat builds?
First of all, valve guides are made from cast iron by centrifugal casting. Steel doesn't have the necessary emergency running properties (the carbon in the cast iron acts as a solid lubricant, especially important when used with valve stem seals)
You are right, the bronze guide will expand more than a cast iron one, so if fitted to an alloy head it might become looser, depending on head temp. But not if it is fitted to an iron head. The surrounding material of the head will constrict radial expansion, so due to the difference in thermal expansion the inner diameter might even decrease relatively with higher temperatures, leading to sticking valves.
That's why in iron heads only iron guides are used usually.
In today's modern engines even special cast iron alloys with the same (or very similar) thermal expansion coefficient as aluminium cast alloys are used, so that they can be fitted in alloy heads as well.
General consensus (although some engine experts disagree here) says that cast iron is the superior material for making valve guides due to the excellent fretting and wear properties, which is what I tend to belive, too.
I am also using custom made cast iron guides in the A7 SS alloy head without any problems. You just got to make sure they are a good shrink fit in the head.
EDIT: Ah! but the valve stem probably gets a lot hotter than the guide. Correct?
Yep. Part of the problem is the temperature gradient between valve stem, guide and head.
Cheers, Markus