[pianotech] Reversing Crown

[Ron Nossaman]

George F Emerson wrote:
> Ron,
> 
> Your experience in replacing soundboards in the rebuilding process is 
> more extensive than mine.  My soundboard experience is mostly in 
> manufacturing new pianos.  The instance that I sited was at a company 
> with which I have never been associated.  In any company for which I 
> have worked, I have never seen a new piano soundboard "oil can" 
> immediately after being removed from the rim.  Maybe that should not 
> surprise me, but this one was new to me.

Frank,
I've never taken the board out of a new piano, but I've seen 
plenty of new pianos with zero to negative bearing and 
negative crown in the killer octave. When I tear down an older 
piano in this condition, I often find that the unloaded board 
has measurable positive crown, which may or may not reverse 
when the board is taken out. It depends on the condition of 
the panel and it's MC. Removed boards can go back and forth 
positive to negative crown as the RH% of the room changes. In 
any case, the board is dead.


> Regarding rim bevel, Baldwin beveled only the belly rail, not the inner 
> rim. 

Interesting. I'll have to look for that next time. I don't 
typically pay a lot of attention to rim bevel angles, as it 
doesn't seem to make a heck of a lot of difference. if any, in 
the soundboard installation. At least in a rib crowned and 
supported assembly.


>Call it a bevel if you 
> like, or call it a match in curvature of the gluing surface of the inner 
> rim and belly rail to the curvature of the crowned soundboard.  

Yea, I'd call it a bevel.


>In my 
> design work, I use a 3D computer model of the crowned and loaded 
> soundboard to determine the inner rim/belly rail elevations.  I use 
> another computer model of the crowned, UNloaded soundboard to determine 
> the radii to cut the ribs.  The crowns of these soundboard computer 
> models are "football" shaped, not spherical or cylindrical, resulting in 
> shorter radius ribs for the shorter ribs than for the longer ribs.

I do something similar with a spreadsheet in an existing rim, 
only I ignore the rim bevel angle. My rib dimensions and radii 
are determined by bearing loading, spring rate, and residual 
crown under load. The panel is under very little compression, 
with the ribs supporting the load. In the treble, I clamp the 
end of a 4M radius rib to a horizontal surface with no attempt 
to match the approach angle. The spruce conforms with no problems.
Ron N