Several possibilities; all treated differently.
The cylinder head sealing contact surface itself could require lapping. Best way
to affect this is removal and bench lapping with 240g paper glued onto the
exhaust
header with contact cement; some kerosene as lubricant and rotate header
back n forth about 50deg for a bit, then lift and rotate 30-70deg and repeat
several times Not much pressure required. Easiest to clamp header pointing up
and rotate head upon it; makes cleanup easier... can use valve grinding past
as well... soot or bearing blue at the start and end can be used to check
sealing
surface for gaps. Sealing surface only requires 1mm cross section.
The 914 has part of the header extending into exhaust port; it is a very firm
fit
when new; long term slop flogs out this area in the head casting and requires
sleeving insert to prevent leakage.
Copper compress rings can be used as mentioned and should be set centered and
tensioned
again when hot. This is a bandaid approach to rotax factory sealing method
and may work totally fine. Have found copper rings on some engines.
Heat expansion of the header pipe may be causing movement at the sealing point
and returning to normal once cooled down. Stainless steel pipes grow a fair bit
when heated up. Some stainless steel exhaust systems on cars can grow up to
30mm... Try loosening off header at cylinder head and collector; give the header
a few light taps with hide hammer and retighten just under spec and again once
cooled down. Observe for outcome.
Check the clamping collar when tensioned; there needs to be a gap between the
collar
and the head; the collars bend quite easily when over tensioned.
Read this topic online here:
http://forums.matronics.com/viewtopic.php?p=513027#513027
|