Been thinking for a while about my Ducati's bad habit of getting bloody loud if rolling slow... waiting in queues, traffic lights, city riding and so on. It's worse on hot days. I hate it, when the effect takes off it can get ear-splitting.
Long story short: I think it's engine heat conducting to the inlet stub pipes. I think they're pre-heating the mix (before induction and compression), so I'm getting super hot combustion, hyperexpansion and high pressure exhaust... hence the appalling noise. Park the bike up, let it cool down, come back and it's quiet again (well, for a carbed 900SS). It's quiet for about five minutes, then it's back to the racket. Only quietens down again once I get some airflow going.
So I got thinking about how to insulate the stub pipes. Not easy: the usual way is to use a phenolic resin spacer and extra-length engine studs. On this bike if you do that then you screw up where the carburettors mount up, unless you can go to split singles, or get really fancy and make your own inlet pipes. I did some thinking about that, then a question occurred: am I sure about how this heat is conducting? Maybe a thick paper washer is enough. Then: How is that heat being conducted in?
So, there are only three ways, really:
1) Bulk conduction from cylinder head, through paper washer, to stub pipe flange
2) Bulk conduction from cylinder head, through stud bolts and nuts, to stub pipe flange
3) Swirling fluid conduction from turbulent inlet charge being pulsed back and forth while the engine breathes.
Can't do anything about 3. Can use a thicker paper washer, or a proper phenolic resin spacer, to sort out 1. But what about 2? It's metal to metal contact the whole way, no paper or anything... is 2 actually significant?
Thermal conductivity paper: 0.05 W / mK. Path length: 0.4 mm. Surface area: approximately 1545 mm2
Thermal conductivity M8 metal studs: 45 W / mK. Path length: 10 mm. Surface area: approx 100 mm2, for two studs.
The basic equation for thermal conduction is:
thermal energy per time = conductivity x area x temperature difference / path length.
Assuming that the cylinder head and inlet stub are both a uniform temperature (yeah, I know, they aren't), that swirling fluid conduction isn't significant (cross fingers) and that we're not interested in absolutes like what the temperature difference actually is, then we can say:
Thermal energy transfer is proportional to conductivity x area / path length, and...
Total = conduction across gasket + conduction through stud
Gasket: 0.05 x 1545 / 0.40 = 193
Studs: 45 x 100 / 10 = 450
Total: 643.
So thermal conduction through the stud bolts is roughly 70% of the transfer.
From that, the best way for me to cool those inlet stub pipes down is to go to a 1mm paper washer, and get some kind of top hat insulating washer onto the studs. Don't let the stud bolts make direct metal to metal contact through their flange nuts. That's it. You could halve or even quarter the heat transfer with relatively thin insulators and little need for special parts. Certainly it's easy enough to test, you just need a fibre washer that'll take the heat placed under the inlet flange nuts.
I won't be in a position to test this myself for some time (see the winter layup thread I started), but maybe this'll be useful to someone else out there?
Reply With Quote
Originally Posted by AllanB
Bookmarks