Designs on a grand scale

[John Delacour]

At 4:02 pm +1000 8/10/06, Overs Pianos wrote:

>This would be a dangerously high tension as you say, if the length 
>was actually 197 cm on note A25, which it is not. This concert grand 
>has two extra bass notes...

Silly me!!!  -- and I actually had a good look to see what note you 
started on.  Obviously should have worn different glasses!  I'll take 
a new look at it in the light of this information.

>>As for the covered strings, I have mentioned elsewhere that I never 
>>design bass strings to exceed 70% of the breaking strain, for the 
>>simple reason that sooner or later they will snap, and often sooner.
>
>I agree. In fact I prefer not to go over 60%. Mind you, there does 
>seem to be some variation in our calculations, which could account 
>for some differences in our estimated breaking strain percentages.

Yes, in designing a scale from scratch I too aim for 60%.  Making 
replacement sets for all and sundry means, of course, this is often 
not possible.  With a good grand of about 6'9" and above it is 
sometimes possible to get near to 60% throughout the covered string 
scale, and I like to do that.  So far as I recall, the S&S old B you 
were asking about the other day is an example of a scale where this 
is possible, and almost adhered to by the original designer 
(Theodore?)

As to the calculations, yes they were on the rough side, and 
especially so since I realized after posting that I was using a wrong 
divisor -- not horribly wrong but wrong enough.

What value do you use for the relative density of steel wire?  I keep 
meaning to get access to a precise electronic scientific balance to 
measure this with real wire.  I had a beautiful "chain dial" balance 
but it was stolen.  The value bandied about on the WWW is 7.85, but I 
don't trust it.  I've also never understood why Wolfenden used 7.60. 
As to breaking strains, I use figures my supplier provided over 20 
years ago, and I don't trust them much either, but at least they are 
on the safe side and way below the figures Poehlmann achieved 120 
years ago.

>As a final comment on scaling tensions, I wouldn't trust calculated 
>scale tensions for a new instrument without first testing them on a 
>tension bench for safety margin, prior to building production 
>tooling.

Calculated tensions are fine provided the values are obtained from 
exhaustive tests of real life strings, which I suppose comes to the 
same thing.

JD