Task 1:

Here I simply went on to 3D model the "cylinders" by extruding the lines at 5mm intervals. I could then create the bottom polysurface which I copied and placed on the extruded surfaces. I could then trim it down to reach final geometry. The method was slightly tedious, in retrospect I think it would have been easier to extrude all the lines then "cap" them off, creating a polysurface. Additionally I removed some of the interior lines as it turned out less aesthetically pleasing to extrude all of them. The geometry now has a larger flat surface than it otherwise would have, giving a more harbour, or platform, like expression. 

Originally I wanted to try to work on the more complicated fluid lines, however I think the geometry got a litte too chaotic for Rhino, or more likely, I poorly defined the lines such that they turned hard to work- and do commands with.

From there i modeled a simplified version of the rest of the structure, so that I could place the roof in some architectural context. I did this both natively in Rhino3d, and in grashopper. 

All that was left then was to 3d print this structure. Which proved difficult. I tried multiple times, but the structure was to small, and generally too thin. In the end i had to 3d print the roof seperatly, and it had to be both larger and thicker originally intended.

The process of creating it in grasshopper was rather simple. Define a rectangle and a midpint, which would be the basis of the structure. From there I could move both the point and the rectangle to the desired height. Then  by extracting some of the corner points of the newly moved square, I could move these points up a little further, and then create a 4-point surface,  which would create the slanted roof.  From there it was simply to create the desired struss lines, and then piping them.

In the figure above we have the basic structure for the entirety of the first row.  The fist part before the coloured groups simply creates the column line, and slanted roof surface. In the coloured boxes we define the strusses. And finally the last part simpy deals with repeating the structure until we reach the desired lenght of 60m. This script was then copied to create the back row. The height and slanting of the roof can then easily be adjusted by number sliders. 

In conclusion. The structure should be more than good enough to hold the loads. 

In addition to holding the snow load the structure is under, it also serves to drain the water away from the stadium, and into the ground. This is done as the fabric naturally wrappes around into a pipe that goes straight into the ground. This in turn helps giving a clear seperation between the back, and the front of the structure, without gating it of completely. 

From underneath the roof the panels dominate the expression. They are a constant regularity. The columns on the other hand, alternate in a zig-zag pattern, giving sparser looking feel without affecting structural capacity too much. 

We then wend on two calculate the M,V-diagrams based on some simplified models and large assumptions. This was then coupled with a full kamba model which verified our calculations quite well.

The same holds for our Steel beams,  assumption of no rotation works well  as there is roughly equal loading on each side of the columns.

Our assumptions doesnt always hold true. This because there will be unequal loading on each side of a column at certain areas. This holds especially true at the ends. Although the fact that our hand calculations fits so well with the karamba model in most places, gives a sense of trust in the Karamba model.

Finally we got some simple capacity checks!