<<<<<------------  Copied from my Facebook notes-January 10th of 2010  ------->>>>>   
Roof'n

Tuesday, December 1, 2009 at 10:30pm | Edit Note | Delete
Trigged out the gables for the GH today, but with an 8 inch runout, some adjustment will be required.... The knee walls are not parallel and constrict almost 8" over the 20' length of the greenhouse. Choices: #1- distribute the problem (symmetric errors but keep a constant gable height of 3' above the frame and vary the rafter and perlin length accordingly), #2- isolate the problem (constant roof height, vary the roof pitch and perlin length on only one side of the ridgeline, the non-visible backside). #3 - Constant perlin length, symmetric errors, and allow changes in roof height as well as changes in roof pitch but with the rafter length staying the same on both sides of the ridgeline, simple construction but the warp in the surface would be visible for miles. Plus it is far too Kentucky red neck to suit me. It may even have been intended by a certain inept and disgruntled contractor, losing 24 million dollars might do that to a man, and all because of a bad concrete hotel foundation, who would have guessed? Final choice was #2, this allows the ridgepole to set level and at a constant wall distance and it simplifies the construction, at least on the front side, the side with the larger surface area (yea!). Because of the intersecting roofline, the backside has to be cut special anyway!! Haven't looked to see if the backside runout could be handled with shims up to the intersection point but that would be sweet. Appearance should be OK. So do we set the ridgepole or build the end gables next? In either event I may need a water level or a transit, If the wall level is as badly weird as the rest of the cement job, then setting the ridgepole to level might be a bad mistake or it could even play good/bad into the current non-parallel wall problem, etc. Both tasks are pretty salty, about $500 to $700 each, if I choose to weld the end gables from 2" stock. I may simply choose to mechanically tie the end pieces into the ridgepole with angle brackets and forego the rest of the costs (looks are nice but so is rice)... It might simplify mounting the cross pieces, the perlins, since everything else would be at the same height. And it would definitely lower the number of calculations and the labor cost if something goes wrong. So why am I writing this? My 4th grade teacher, Ms. Enoch, who later became a professor at Manchester College was nearly 7' tall. Hence, you tended to remember what she taught. She taught us that the best way to flesh out an idea and at the same time clear your head so that you can get on with life... write it down.... several (20-40) times, grappling with it. It does have a certain disenchanting effect. And I used this effect to get through my PhD thesis. But she was, I'm guessing, a very good English Professor. Anytime I want to "see the light" and dismiss the wayward call of the sirens, I write it down. I grapple with it and write it down again. Thank you Ms. Enoch wherever you are.
Written about a month ago Comment LikeUnlike Edward A Kimble Comment.....
A month has passed, I sensed a problem and its good to double check. 12' and 3/16" is not 12.1875 ', although the error was well documented in at least 5 places with carry through errors that would have made an angel cry (or an angle). I worked all the math twice, actually many more times than twice, but on paper and in my notebooks once in decimal feet, once in decimal inches and I couldn't see why I had discrepancies on the order of inches, and you really don't know it's an error until either the source is found or the M.C.Escher reality is thrust violently upon you and you are climbing that stairs to nowhere!! So, tomorrow, the sword of truth, actual aluminum cutting, temperatures and angles like 27.24 degrees permitting :) 2 hours ago