>Ron, > >This looks familiar. Isn't this the same as the formula I sent you a few >years back. John, No, this is mine. Yours produced compression levels at half what this one did, as I pointed out when we were comparing methods. >I have lost the file this was on but if I remember I was trying to figure >panel compression using simple moments of force. I wasn't feeling >confident that this approach would work so I have since abandoned it. If >you can get this to work and can explain it to others the more power to you. It's worth a try. Nothing else seems to have shed any light. Panel compression crowning is more like a distributed load than a point load anyway, and that's why I asked for some help from some of the engineering training that reads this list. I have had a distributed load formula for some time, which produces compression figures slightly lower than this formula, but I'm looking for clarification from better trained heads than mine. As to whether it does any good or not, what will come of it will come, or not - as usual. >To be fair you should use a pine rib, this is what Steinway uses on there >panel crowned boards. You also need to scallop the ends of the ribs and >cove the bottom. All of this will effect panel compression. Yes it will, and I'll attempt to try to take that into account one way or another. I suppose I should also thin the panel, as Steinway does, which would further increase compression levels, in the interest of better accuracy. Realistically, there's obviously no way to exactly quantify panel compression levels. This is intended as a demonstration of fundamental principles and approximate stress levels. Actual levels will depend on individual pieces of wood, actual bearing measurements, and MC of the assembly before assembly, and at the time of computation. >I think a direct approach will be more conclusive. Make two cross grain >strips of soundboard panel the same length at the same EMC. dry them both >down to 4.5%. Glue a rib on to one of them using a flat surface. Let both >samples reach EMC with the environment, say 40% RH. Check the level of >crown and compare the length of the strip with the rib to the other strip. >The difference in length will give you the compression. No, it won't. It will give you the difference in length, but that doesn't equate to the compression PSI. Panel compression isn't linear, as that same experiment will show you. Load the crowned assembly and measure deflection with a dial gage. It takes increasingly more load per unit of deflection as the crown is pushed down. It's a variable rate spring, as you have noted yourself. All told, I expect it will take over eight times the load to force that crowned panel flat as it would have to bend that rib (or one just like it, without the panel) to the crown that was formed by the expanding panel. Try it and see what you get. Also, the panel is already compressed by around 1% of it's original width just by shrinking it down, gluing it to the rib, and re-hydrating it, so it's already at or very near it's fiber stress proportional limit when the crown is formed. I suspect that's a lot of why the compression rate beyond that isn't linear, because it's already at it's elastic limit. What's the difference in length between the ribbed strip, and the un-ribbed strip, assuming you dried them down to realistic CC assembly requirements? Is that difference over 1% the length of the unconstrained strip? Again, try it and see. Ron N
This PTG archive page provided courtesy of Moy Piano Service, LLC