dholdgreve wrote:Balsa Man wrote:I misunderstood your first post, thinking you were working with a 3-leg tower configuration- sorry ‘bout that. Comments on getting the angle right for the critical joint where upper and lower leg segments meet still apply.
Using a 4-leg, square configuration, as Raleway says, if you have the legs really symmetrically lined up, the faces of any adjacent leg pair will be parallel, in the same plane, and bracing pieces will go on outer leg faces straight, in full contact with the leg faces; no gaps.
While the elongated rhomboid cross section you show in your drawing is correct if you’re looking down perpendicularly onto the cut end of an angled leg, the four faces of that leg stick remain at 90 degrees to each other. See my recent post in the “Base” thread for photos of a tower on a jig – this is a tower from last year, but legs are at an angle. The fact that outer leg faces where the braces go, for any pair of adjacent legs are/will be in the same plane is….supported by the fact that in building this the way we confirmed that the ‘rotational’ alignment of the legs was correct was to tape leg pieces onto the jig and lay a flat plate up against each tower side. That plate would come into full/flat contact with the legs. We actually did this with the little, white plastic angle iron pieces (leg holding strips) in between the leg edges and edges of the tower jig plates. Then, we put some glue spots along the angle iron/jig plate edge joint, and put a second plate against the adjacent tower face, forcing/holding the angle iron in correct “rotational” alignment.
If you are actually seeing, with legs on the tower jig, the gapping/mis-alignment you describe, you have alignment/symmetry problems with the jig. It doesn’t take much rotational mis-alignment (or jig plates being not identical, or not arranged at true 90 degree angle to each other to see the leg face mis-alignment and gapping you’re talking about.
Hope this helps
I'm not sure sure of that, Len... I believe the issue is that, as the lower section legs "lean in" toward the "waist" of the tower, if you were to take a horizontal cut through that plane at anywhere in the lower section, you would indeed arrive at the drawing shown, due to the lean. The radial dimension from the center of tower of the columns would indeed be greater than the opposite dimension of the column due to the lean, creating the diamond appearance shown.
Frankly, I just now realized this myself, but it has proven to be a problem in our early testing with braces and tension bands breaking adjacently to the column, proving the issue.
Well, Dan, after putting some pieces together and playing with them on a jig, I have to say, you and Jinhu are right and I was just plain wrong. This is a real issue. What can I say, except …always learning. I’ve always tried very hard to be sure before I say… something is right, ‘this is the way this/this works’, or something is wrong, ‘no, that’s not how that works. Blew it big time on this one.
I think because of the lesser angle to the legs in last year’s towers, the effect was less- enough to not see it in putting plate up against a side, combined with what we knew to be less than perfect rotational alignment of the angle iron we were using for leg holding.
So, to check things out, I put together a (1/8”) a leg segment, then glued 3 pieces onto one face of it- 1 perpendicular to the long axis, one at about 10-12 degree angle to the long axis, and one at about 24-25 degrees. Then glued ‘corresponding’ pieces on adjacent face of the leg stick – parallel to that face, so looking down the line of the leg piece, the glued on pieces were perpendicular to/at 90 degrees to each other. Then, with leg pieces taped onto two opposite jig plates, took the leg piece with the pieces glued onto it, aligned it parallel with the edge of the jig plate in between them, and moved it in toward the jig plate edge. Before it got into contact, the pieces glued onto it perpendicular to it contacted the adjacent legs.
Continued pushing it in toward the jig plate edge (with the perpendicular pieces bending), and next the pieces at 10-12 degree angle came into contact, kept pushing until the pieces at 24-25 degrees came into contact (which was about when the leg segment was fully up against the jig plate edge. So, while I recognized early-on the issue of square deforming into elongated rhomboid as cut angle got away from a 90 degree/perpendicular cut (and need to cut both lower and upper leg ends at the angle bisecting the angle between upper and lower leg segments), I missed/blew off how the 3-d geometry of leg faces was working- that adjacent faces weren’t in the same plane, and how the extent to which this was true varied with the angle of bracing pieces connecting two adjacent legs.
So, recognizing and understanding what’s going on, what does that mean, and how can it be dealt with?
Problem is most severe in a C tower base, essentially non-existent in C chimney, and almost non-existent in a B chimney.
You definitely want brace pieces to be straight, not put on with significant bowing in them. (as discussed before, even with 1/32” thick X braces, there is very slight bowing- half of 1/32”-1/64”; no way around that, but you don’t want any more than you have to.
As previously discussed, filing/sanding of leg faces where braces go definitely compromises leg strength, by reducing cross section (hence buckling strength) in the filed down/sanded down zone. However, the presence of the glued on brace pieces increases cross section in that area. Last year (thinking we were dealing with just rotational mis-alignment, we did do some careful filing along the contact lines, and legs worked – had the calculated buckling strength; held under full load. The other thing we did was careful trimming of X brace piece ends so they contacted each other, and put a tiny spot of glue on at that contact. That may help the tear-away problem you’re encountering.
The greater the angle (from horizontal) of brace pieces, the less the problem, which says wider bracing intervals may turn out to be ultimately more efficient (if calculations show tower weights to be close between two bracing intervals. Also, in situations where the angle/gap is small, using medium/thick CA as gap filler is viable- just a bit of extra glue weight.
