<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD> <META http-equiv=Content-Type content="text/html; = charset=iso-8859-1"> <META content="MSHTML 6.00.2800.1226" name=GENERATOR> <STYLE></STYLE> </HEAD> <BODY> <DIV><FONT face="Comic Sans MS" color=#000080></FONT> </DIV> <DIV><FONT face="Comic Sans MS" color=#000080>We have just = seen a bunch of numbers carried out to several decimal points of accuracy being tossed = about to prove one thing or another.</FONT></DIV> <DIV><FONT face="Comic Sans MS" color=#000080></FONT> </DIV> <DIV><FONT face="Comic Sans MS" color=#000080>May I remind the = gentle reader that any numbers given for such parameters as <STRONG><EM>compression = strength perpendicular to grain</EM></STRONG> and <STRONG><EM>shrinkage or = expansion relative to changes in moisture content</EM></STRONG> are averages. They = are derived through the careful testing of a variety of wood samples within = a given species but taken from a variety of locations and growing conditions. = Some samples will have physical characteristics considerably higher and some considerably lower than these averages. These numbers are not, = sadly, The Gospel Truth!</FONT></DIV> <DIV><FONT face="Comic Sans MS" color=#000080></FONT> </DIV> <DIV><FONT face="Comic Sans MS" color=#000080>To complicate things = even more, they are based on tests taking a relatively short period of time -- = minutes rather than hours or days. Hoadley, in his book Understanding Wood, = suggests that we must degrade these figures by approximately 40% if the = <STRONG><EM>time under load</EM></STRONG> is going to be appreciably longer than the time = taken for the specific test in question. This number also appears to be an = average as I've had other wood technologists quote degrade figures of 30% to 50% = for time under load periods of five to ten years. It depends, apparently, on the particular characteristic and the particular species of wood under consideration. For a characteristic such as tensile strength parallel to = grain the degrade figure can be considerably lower. For a characteristic such = as compression strength perpendicular to grain it will be some = higher.</FONT></DIV> <DIV><FONT face="Comic Sans MS" color=#000080></FONT> </DIV> <DIV><FONT face="Comic Sans MS" color=#000080>To use the published test figures in an attempt to show that a soundboard made to a = given set of specifications can safely be taken right up to the point of short = term damage without sustaining either short-term or long-term degradation = overlooks the cold, hard reality of the natural variableness of wood in = several ways. </FONT></DIV> <DIV><FONT face="Comic Sans MS" color=#000080></FONT> </DIV> <DIV><FONT face="Comic Sans MS" color=#000080>First, it overlooks = the problem that, out of any given batch of wood, some samples are going to measure = high and some are going to measure low. If we could depend on every single piece = of wood that goes into a given soundboard panel testing on the high side, or at = least on its being average, we'd be in great shape. But in the real world we = can't count on this. Inevitably, some samples are going to test low. In fact, about = half of them are going to test on the low side of average.</FONT></DIV> <DIV><FONT face="Comic Sans MS" color=#000080></FONT> </DIV> <DIV><FONT face="Comic Sans MS" color=#000080>Second, it overlooks = the problem that no lumber grader, regardless of training and experience, can = accurately determine any of these physical characteristics without getting into = some form of destructive testing which, needless to say, is not being done by = any piano manufacturer. Nor has it ever been that I am aware of. At least = not in production. We can't get by with random testing, you see, we would = have to test every single piece of wood.</FONT></DIV> <DIV><FONT face="Comic Sans MS" color=#000080></FONT> </DIV> <DIV><FONT face="Comic Sans MS"><FONT color=#000080><FONT>And, = finally, it overlooks the degrade factor that must be applied if the structure is = going to be expected to support a load over some extended period of time. Even if = we accept Hoadley's figure of 40% -- rather than the more probable 50% for = a parameter such as <EM>compression strength perpendicular to grain</EM> = for a soft wood like spruce (it's all that soft earlywood that does it) -- = this still means that we have to work from a number closer to 350 lbs/in2 = </FONT><FONT>than to the 580 lbs/in2 quoted in the USDA's Wood Handbook. And all the time = we have to keep in mind that about half of the lumber we use is going to test = lower even than that. Just how much lower, of course, we have no way of = knowing because there is no practical way of either determining or = predicting the specific and exact mechanical characteristics of any given sample of a = natural material like wood.</FONT></FONT></FONT></DIV> <DIV><FONT face="Comic Sans MS" color=#000080></FONT> </DIV> <DIV><FONT face="Comic Sans MS" = color=#000080>Del</FONT></DIV></BODY></HTML>