[pianotech] Curved cast iron plate question

[George F Emerson]

> Chuck Behm wrote:
>> .... what seems unusual is that the plate is bowed in the webbing area 
>> from side to side. I have the plate upside down on sawhorses, and if you 
>> put a straightedge from side to side across the area for the pins, there 
>> is nearly a 1/4 inch gap in the center.

I doubt that this condition has developed over time.  More likely, the plate 
left the foundry with the curve you describe.  1/4" may be a bit excessive, 
but not by much.  What would be more surprising would be to find a grand 
plate that was absolutely flat, unless it were machined flat.  In the 
extreme case of a 1/4" curve, if the manufacturer attempted to machine that 
out, there would be little or nothing left at the extreme treble, as the 
webbing is often only 1/4" thick in the treble.  However, it is not uncommon 
to machine the pinblock matting surface a little.  Typically in modern 
production this surface is machined to produce a reliable reference surface 
for positioning the plate on the work table of a computer-controlled 
machine.  This typically results in a flat surface at the treble end and at 
the bass end, extending about10-12" toward the center.  Flattening out that 
much of this characteristic curve could reduce a 1/4" gap to something 
closer to 1/8".  It's no unusual for a manufacturer to "fit" the pinblock 
face to the plate on upright pianos, but not in a grand.  The pinblock will 
flex enough to conform to the curve of the cast iron when it is an offset of 
around 1/8" across the scan of 4+ feet.

Such a curve does not necessarily translate into errors in string heights. 
In modern manufacturing, the v-bar is precisely machined, and the agraffes 
precisely counterbored, to produce the correct string heights.  While I 
cannot testify to the way it was done 150 years ago, I can say that methods 
were available to produce similarly accurate results.  I can also state that 
a computer-controlled machine can produce some surprisingly inaccurate 
results.  If a piano cast iron plate is not well supported from the bottom, 
in the process of counterboring the agraffes on a CNC machine, the plate can 
flex enough to result in the counterbore being significantly shallower than 
the coordinate to which the tool travelled.  The supposed accuracy of CNC 
technology sometimes gives a false sense of security, and errors such as 
this go undetected.

Frank Emerson