Probably the hardest part of all of this, aside from
deciding what shape you want to model, is how to lay out
triangles on it so that your structure is rigid and won't
collapse under its own weight!
Unfortunately, this is something I don't understand
very well -- I have no mathematical or engineering basis
to analyze a structure and determine if it would be
rigid. However, I have some ideas, and I used these
to guide my own design of the Desert Nose. Since I haven't
actually built the nose yet, I don't know if these
concepts are correct or not.
- Convex three-dimensional curves do not need extra
bracing.
- Concave three dimensional curves require extra bracing
in the form of extra triangles to form a truss which
is rigid and extends from one edge of the concavity
to the other.
- Avoid long struts unless they are always under
tension (not compression). Long struts under
compression are more likely to bend. "Long" for 3/4"
conduit begins at about 4'.
- There can only be triangles in the structure. Any
other shape will create a weakness. The nostrils
in the Desert Nose are such weaknesses. That was one
reason for the strong column between the nostrils.
Still, vertex 8 is liable to be forced toward
the central column when weight is applied down
on vertex 14. It might be applied down on vertex
14 as a result of weight being distributed through
the entire nose tip structure.
- The more struts connecting at the same vertex, the
harder it will be to make that vertex actually be
solid. The Desert Nose has one vertex, dubbed the
"Booger Joint" (vertex 23) at which 14 struts meet.
Will a carriage bolt be adequate for this joint?
- Struts must always be straight, but you can build your
model with some curves in the struts for your own
convenience. When you measure your model, you're
going to ignore everything except the positions of
the verticies.
- Forces will transfer through your structure and
focus on the area(s) where there is a weakness. To
repair a weakness, add more struts. These must form
new triangles, and you must add enough to form pyramids
to reinforce the weakness.
I recommend building a model and applying some
pressure to various places to see how the model flexes.
Better to break the model than build something full-size
which is going to break.
I built my model of the Desert Nose by soldering copper
ground wire together. Another approach would be to
use 1/8" dowels and 1/8" inside diameter vinyl tubing.
The latter material is better suited to showing the effects
of stress. The copper model is more durable.
A model could start as a real-world object, like a shoe.
(There was an old woman who lived in a shoe...".) You
could draw some dots on the shoe with a grease pencil or
some safe kind of marker (is there such a thing for shoes?).
You'd just place the dot's in a way that you can draw
triangles with the dots as the verticies.
Then you could take an initial set of measurements from
the dots. You'd have to secure some graph paper under
the shoe and measure the X, Y and Z coordinates of each
dot. You'd want to label each dot (vertex) and each
line connecting them (each strut).
You could build a model from just that, but you might
want to apply the
above concepts first and anticipate where you'll need
some trusswork to stupport a convex curve, such as the
one over the top of the foot. You could use a new color
pen or grease pencil to represent vertexes "inside" the
shoe as opposed to on its skin, and draw the struts which
connect to these interior trusswork vertexes in the same
or other colors. Then you can estimate coordinates for
these interior vertices, since you can't actually measure
them.
One of the important considerations when laying out the
dots (verticies) for your model is how big the final
structure will be. You want to lay things out so that
the final struts are between 1 and 4 feet long, with
the majority in the 2 to 3 foot range. You can use longer
struts but your project will be weaker and more pointy
and flat -- less curvy. Shorter struts make a project
more curvy, but it increases the weight and effort
needed to build the structure and its parts. It takes
a lot more shorter struts to build a structure
than longer struts.
So, if you were scaling a boot 1' tall to be a 10' tall
boot structure, then a 1" strut drawn on the would be
1/12 * 10 = 0.8 feet (9.6") -- too short. A 2" strut
would be 1' 7.2" -- that would be OK. But take a look
at how the structure might look with straight lines
between the vertecies. Again, shorter struts make the
structure look more curved, longer struts make it look
more angular and flat.
If you need a door you can plan for three struts which
are at least 3' in length to meet in a triangle. You
can get in through a hole that large. A larger hole
will be easier to get in through. If you want a non-triangular
door then you'll need some trusswork behind it to support
wherever there aren't exclusively triangles.
You might want to position some of the vetecies strategically.
For example, the Desert Nose has an interior deck. I
adjusted all the vertexes so they were all at the same
height (Z value) so the deck would be flat.
If you wanted to build a shelf you could position two neighbor
vertexes at the same Z value and use them as anchor
points for the table.
If you needed mounting points for speakers or lights
or to hold a sign, etc. and you can anticpate where those
locations are, rearrange your struts so there are vertexes
there. Then use longer carriage bolts at those vertexes.
Let the longer threaded end point inside if you need an
inside connection, or outside if you need an outside
connection.
The idea is that within reason, you can move struts
around the surface of your model and it won't drastically
affect the shape of the model or the strength of the
final structure, as long as struts don't get too long or
short and as long as there is adequate trusswork for
convex curves. |