---------------------- multipart/alternative attachment
At 1:54 AM +0100 12/2/05, Richard Brekne wrote:
>
>Udo Steingr=E6ber and I got into a conversation=20
>about this in Helsinki a couple years back on=20
>the only opportunity I've had to sit and chat=20
>with him. It was part of a discussion where he=20
>was explaining why he felt compression=20
>soundboards were the way to go.
You've mentioned this before on the list, but the=20
question I asked at the time of your original=20
post, "has Udo done any modelling experiments=20
with both types of construction?" remains=20
unanswered. Would you ask Udo this question,=20
since you have already established a relationship=20
with him?
I have done modelling with both types of=20
construction. I have built full size models and=20
tested them on the bench. For those who might be=20
interested to see an image of our experimental=20
setup, go to our second web page on my I-rib=20
development.
http://overspianos.com.au/iribbd2.html
Regardless of the sound board construction method=20
chosen, the maximum stress on the sound board=20
assembly will be directly under the bridges, if=20
the ribs are of a uniform cross-sectional area.=20
When RC construction is employed, the depth of=20
the ribs under the bridges can be increased to=20
reduce the stress point under the bridge. Our=20
I-ribs are flat on the bottom flange and crowned=20
on the top flange, so they will be relatively=20
stronger in the middle where they are deeper=20
(standard solid RC ribs can be profiled in the=20
same way). Rib feathering allows for all boards=20
to be appropriately weakened at the edges where=20
there is less stress, to allow for greater sound=20
board activity. Our I-ribs are tapered also in=20
the width of the bottom flange (we call this=20
secondary feathering), which allows us to further=20
vary the strength of the rib along its length,=20
creating a sound board which deflects more=20
uniformly under the downbearing load. The same=20
result can be achieved with a solid RC rib, by=20
varying depth along its length. Terry Farrell has=20
shown some excellent images of RC ribs which had=20
significantly greater depth in their mid=20
sections. Would you like to post a couple of=20
links, Terry, to show the troops what you and=20
others are doing? Varying the strength of the=20
ribs along their length is, I believe, the way=20
forward in sound board design. I can't imagine=20
how similar rib strength variation could be=20
achieved with a CC board, unless the builder is=20
prepared to forgo crowning in the proximity of=20
the bridges, since a deeper CC rib will prevent=20
the re-hydrated panel from forcing the board into=20
a crown.
>He mentioned something about the so called K=20
>point (I think this is a Klaus Fenner coined=20
>term), which as I understood it was that point=20
>on the soundboard where crown from both along=20
>the grain and cross the grain intersect and is=20
>at its highest point. This area is exactly (and=20
>always) in the area typically refered to by=20
>american techs as the <<killer octave>> area,=20
>and of course varies somewhat from piano to=20
>piano.
The highest point of the sound board, from the=20
instruments I have measured, is somewhat further=20
down the long bridge towards the middle of the=20
board.
The killer octave, for most 'conventional'=20
designs, is the grossly overloaded part of the=20
sound board. They also are often overloaded=20
directly under the bridges. Try placing a 150 mm=20
steel rule under the bridge, almost anywhere, on=20
the underside of the panel of any CC boarded=20
piano. The rule almost always will rock, which=20
indicates that the crown has reversed directly=20
under the bridges. I have observed this many=20
times, even with pianos which were under one year=20
old. Quite a number of the new grand pianos=20
exhibited at last year's Adelaide piano=20
technicians convention had reverse crown directly=20
under the bridges. And this observation was not=20
limited just to the lower priced pianos either=20
(the checking with rule was done late in the=20
night to avoid upsetting various exhibitors -=20
some examples already had 'nicely developed'=20
killer octave zones - Ron N was party to these=20
observations also).
> It is then also the exact point then that is most vunerable to failures.
How do you arrive at this conclusion? The only=20
part of any sound board which is prone to failure=20
would seem to be those areas that are=20
under-engineered and overloaded.
>Udo was of the position that along the grain=20
>crowning was every bit as important as cross=20
>grain crown.
I would have expected this point to be a=20
'position', rather than a 'conclusion'. Crowning=20
along the grain is something that will occur=20
naturally in any sound board once it is fitted to=20
the case, whether CC or RC crowning is used.=20
However, the resultant crown along the grain=20
won't really help a sound board to better=20
withstand the downbearing forces. While it might=20
appear to be beneficial, the span along the long=20
bridge relative the section sizes of the=20
materials used, will make it impossible for it to=20
be of any serious assistance.
> Something to do with along the grain=20
>compression due to downbearing having a=20
>stablizing effect on cross grain crown and=20
>strength.
I doubt it. Very slight soundboard compression=20
will result along the grain as the downbearing is=20
applied, especially since the bridge is located=20
above the 'mounting-plane' of the inner rim, but=20
it won't help or hinder the net ability of a=20
sound board to resist sinking. Furthermore, it=20
would be very difficult to build a sound board=20
with compression crowning along the long grain,=20
since the bridge would need to be dried before=20
gluing it to the panel, and there would be=20
insufficient shrinkage of the bridge, along its=20
length, to achieve an alleged worthwhile crown=20
along the grain.
The ability of a sound board to support=20
downbearing is related primarily to the ability=20
of each rib/board segment's ability to support=20
the load to which it is subjected.
This, I think, ties in with Ron N's reply to Terry Farrell's original post;
At 6:41 AM -0600 12/2/05, Ron Nossaman wrote:
>
>>At 5:22 PM -0500 11/2/05, Terry wrote:
>>When I raise the pitch of a piano, typically I=20
>>find that I need to pull an area of the treble,=20
>>commonly around the sixth octave or so, a bit=20
>>extra so that area does not end up flat when=20
>>the pitch raise is completed. I use typical=20
>>pitch raise overpulls - 20% in bass, 25% in=20
>>tenor and 33% or so in treble - but that one=20
>>octave or so in the treble needs to go a little=20
>>further - maybe 35 or 38%. I find this to be=20
>>true on most pianos.
>>
>>My understanding is that one factor that may=20
>>conspire to produce a killer octave (low volume=20
>>and/or short sustain) in a piano is the fact=20
>>that the killer octave area is also=20
>>the area the long bridge is curved most -=20
>>rather than having the downbearing supported in=20
>>part by a straight (or nearly so) bridge (like=20
>>in the tenor), the curved part of the long=20
>>bridge in the killer octave area is more prone=20
>>to rolling - I know, not rolling - actually=20
>>soundboard deformation - but I'm trying to=20
>>point out that it can rotate in this area more=20
>>easily than other areas.
>>
>>My question is - might these two phenomena be=20
>>related? Is the killer octave area more prone=20
>>to going flat because the bridge is rotating (I=20
>>suppose in part due to soundboard not having=20
>>enough support in that area)?
>>
>>Thanks for any thoughts.
>>
>>Terry Farrell
>
>I think so. The tenor isn't heavily loaded, and=20
>the soundboard (usually) supports it, though it=20
>sinks some. The upper treble section is heavily=20
>loaded, and barely sinks at all because the=20
>bridge is very close to the belly at the top=20
>end. It's sitting on a brick, essentially. With=20
>the top end of the bridge as a fulcrum, any load=20
>put on the treble levers down to the curve in=20
>the bridge, where it loses beam support because=20
>of the curve. So the killer octave has to=20
>support not only it's own heavy bearing load,=20
>but gets additional load from both the tenor and=20
>treble by virtue of being at the end of two=20
>third class levers courtesy of that curve.
Exactly Ron N, and the situation in many cases is=20
made even worse due to the lack of an appropriate=20
sound board cut-off, in this area of the board,=20
which causes the ribs to be way too long and way=20
too weak to withstand the excessive forces under=20
which they placed.
>The soundboard deflects proportionally more at=20
>the curve, and the bridge rotates with it's=20
>center of rotation being a line through=20
>somewhere at the top end, and somewhere in the=20
>mid tenor. The part of the bridge furthest away=20
>from this center of rotation is the curve, where=20
>the killer octave is. This is also why it's=20
>possible to have both negative crown and=20
>negative overall bearing in the killer octave.
Well put Ron N.
And now, back to Mr Steingr=E6ber's claims.
> He also meant that both these were needed to=20
>produce the kind of acoustical results he wanted=20
>from a soundboard.
Mmm? For those of you who haven't seen a sound=20
board being glued into a case, the 'free board'=20
doesn't even nearly conform to the profile of the=20
inner rim. The sound board will always have to be=20
clamped to the inner rim (which places it under=20
some initial stress - which will occur regardless=20
of the construction methods used).
>I'm under the impression, correct me if I am=20
>wrong, that most of the rib crowned methods=20
>extoled on this list do not employ any=20
>purposefull along the grain crowning as part of=20
>how the panels are attached to the rim, and=20
>therefor do not have this vunerablity. It would=20
>stand to reason that they too are not able to=20
>create the same acoustical results.... (whether=20
>that is preferable or not is an entirely=20
>different subject).
I cannot see how it 'stand[s] to reason' at all.=20
You seem to be implying that RC sound boards are=20
not able to 'create the same acoustical results',=20
but I don't see how Udo arrived at this=20
conclusion. You seem to be implying that the RC=20
school membership don't build crown into the=20
board along the grain. It can't be avoided once=20
the sound board is installed, regardless of which=20
construction school you happen to be sitting. But=20
you seem to go further by implying the crown=20
along the grain is somehow a critical component=20
of tone building. I am not convinced that it is,=20
but even if it were, its an unavoidable result=20
with both methods once the panel is fitted to the=20
case.
I believe the important issue is that the=20
majority of boards are grossly overload at the=20
second top string section.
Ron O.
--
OVERS PIANOS - SYDNEY
Grand Piano Manufacturers
_______________________
Web http://overspianos.com.au
mailto:ron@overspianos.com.au
_______________________
---------------------- multipart/alternative attachment
An HTML attachment was scrubbed...
URL: https://www.moypiano.com/ptg/pianotech.php/attachments/64/6f/95/80/attachment.htm
---------------------- multipart/alternative attachment--
This PTG archive page provided courtesy of Moy Piano Service, LLC