The nose calculator predicts the lengths of struts assuming that the struts essentially meet at a point. In fact, they don't. The Desert Nose has its "Booger Joint", where 14 struts join together from all kinds of different directions. So, the nose calculator is "wrong" in that you actually can't assemble the struts the way it assumes.

But, in fact, the structure will be rigid if it is designed right, even if the struts meet in a region instead of a point. One way of thinking about this is to envision a sphere about 4" in diameter whose center is the "point" where the struts are supposed to meet. You can imagine that in addition to the center point, there is also a circle on that sphere where that strut can also be placed and still be the same length.

Enter the L-bracket. The L-bracket adds a 90 degree bend at any vertex, and the 90 degree face can be rotated through 360 degrees to be convenient for other struts which need to connect at that vertex.

The L-bracket gives you a lot of flexibility. Some of the verticies in the Desert Nose needed two L-brackets, including the Booger Joint. The L-bracket allows you to join planes. The Desert Nose was based on filling in triangles to support some longer sweeping pieces which defined the overall shape of the nose. So there were some basic planes alread built in. The deck was one plane, but it met at 90 degrees with two planes (one on each side) which supported the tip. The L-brackets made these joints easy to make.

Photo by the Bonitas
click for larger view

Photo by the Bonitas
click for larger view

To make the L-brackets I cut 1.25" x 1/8" angle iron into 2" sections and drilled a hole near the edge of each face centered in the 2" length of the material. I ground the edges and ran them through a wire wheel so they didn't have any sharp edges left.

There was just enough room for a carriage bolt to go through each hole. It was also possible to have the head of one carriage bolt and one nut "inside" the L-bracket.

On the nose, the back was mostly like a dome -- a convex strutcture which pretty much stands up on its own. But the area around the deck and over the deck was a 3-dimensional structure which supported a concave extiorior (where the nostrils meet the tip and the side of the nose).

Wherever the structure becomes "3-d" it becomes more complex and it is harder to get those struts to meet. The L-bracket give you some freedom, but you have to be careful not to create a "problem" by effectively changing the length of a strut.

Here's a metaphor. Imagine you're holding a pool cue and you're going to use it to flip a light switch from five feet away. After you get the tip under the switch, the motion to activate the switch is just up and down -- not in and out. Remember the 4" circle? There is a 4" circle around where one end of a strut meets which doesn't change its length more than a negligable amount. There are also a couple of points in the sphere which change the effective length of the strut by 2". You can avoid problems by avoiding changing the effective length of a strut.

That said, it is safe to adjust the length of some struts -- just not the ones involved in a 3-d structure. So, it was safe to make a few of the top-most struts in the nose a different effective length because all that happened was that some rigid structure had a slightly different shape. But nothing depended on the way that shape looked -- as long as it was rigid. So, whatever fit was fine. These areas seem always to be on the "outside" of the structure and never on the "inside".

Photo by the Bonitas

Struts meet in anything but a perfect manner. There is a trade-off between flattening the end further, making the struts easier to assemble, and the strength of the end. A long flat end works fine of the strut is under tension (being stretched), but doesn't work well if the strut is under compression (being crushed).

If one was building a completely flat sheet of triangles then all the struts would be in the same plane and they could essentially meet at a point too. But since the Desert Nose is made of curves the struts don't meet in the same plane.

When struts don't meet in the same plane, but the planes are fairly close to each other, then the ends of the struts can simply be bent so the strut ends are flat against each other. This works up to about 30 degrees pretty well. The ends get weaker from 30 to 45 degrees. I found that when the joint reached even 30 degrees it was possible to use an L-bracket by bending the strut in the "other" direction.

I didnt' start needing L-brackets until I started building the septum, which had some struts meeting at 90 degrees with the top of the deck and 90 degrees with the front of the septum vertically and horizontally! All three dimensions essentially meet at one vertex on each side of the front of the deck.

So, I'm going to confess here that septum had to be built twice. The first time seven struts were the wrong length due to a bug in the nose calculator. When I built the septum the first time it was essentially like an equliateral trangle dragged upwared -- like the mirrors in a kalaidoscope standing on end. I bent the end of each strut to the "average" of the two planes where each of the "mirror" sections met. So the struts all pointed out from each vertex in one, sharp chunk, which was in its own plane.

click for larger view

This turned out to be a bad thing.

It was very hard to get the septum together because each strut really had to be bent to the right angle or it would fight the assembly process.

I found it was much easier to assemble if I looked at the struts and picked a collection that were more or less in one plane. Then figured out how to attach the rest. IN the entire Desert Nose only a few joints needed two L-brackets, although many more needed one.

So the second time I built the septum the pieces made it wider, so it wasn't an equilateral triangle any more. I decided to favor the vertical plane of the front of the septum. The pieces in the same plane met without being bent. The struts from the other two planes had to be bent to make up the entire angular difference between the planes. But it went together much, much easier because the ends of all the pieces in the same plane didn't need to be bent so they were already perfectly aligned.

click for larger view

I often found myself "unbuilding" a vertex to rearrange the struts so they would sit better. Even among struts in the same plane, some angles fit better than other. If two struts meet at 180 degrees (e.g. they lie on a line and only one of each of their ends touch) then there is a point where the very end of one strut bumps into the "hip" of the other strut -- which is where the flat part transitions into the unflattened pipe.

So, meeting at 180 degrees might mean that you need to flatten then ends a bit more, so the hips are clear of the flattened end of the other strut.

Photo by the Bonitas
click for larger view

Meeting at 90 degrees in the same plane is easy and is perhaps the best case since the hole is near to the side of the flat part of a strut and a longer way from the end of the flat part.

Meeting at 90 degrees in a different plane would require an L-bracket. But depending on where the other struts are coming from you may find that you need to cut off the corner of the flat end of the strut, or cut about 1/4" off the flat end, reducing the material between the edge and the hole, but allowing it to fit on one face of an L-bracket without poking out past the edge of the bracket to interfere with struts attached to the other face.