Power Copies vs. Document Templates

I've been wondering if there is an advantage to using document templates over power copies if the model doesn't have any solids. I think I may have my answer.

Currently my model has 150 floors that are created using fills and extrusions, so everything is a surface -- no solids means I can use a power copy to make each floor. Trouble is, after 150 copies of the floor (well, technically, 149), the model runs REALLY SLOW. Each parameter adjustment takes about 30-45 seconds to run. Not much in the grand scheme of things, but I'm used to less than 5 seconds for processing changes...

Now I wonder: if I had set the floor up as one part, and the framework as another, and merged them using a document template, would it run faster? I also wonder if using 1 solid shape for each floor instead of 2 fills and a profile extrusion would help speed things along...

...next time, next time

Awwwwww Freak Out!

Methinks DP (or possibly my laptop) doesn't like to change too rapidly. Case in point:

I have a parameter that adjusts the depth (thickness) of each horizontal plate. This is a very forgiving parameter, and can be adjusted from almost nothing to a whooole lot (it has worked at 1' and at 75', and many points in between).

As can be seen here, it works nicely stepping from 60' down to 15':


The freak out happens when it tries to jump from 60' directly to 15':

Note that the floor_to_floor parameter in this is the same as it was in the previous screenshot... if it's not too small to read, that is...

Why would it freak out when jumping from 60' to 15' when it's perfectly fine jumping from 60' to 30' to 15'??

Two Mistakes

#1: An odd quirk

My floorplate geometry is set up such that the centers of each arc lie on the same circle, and each arc is tangent to the same circle (so all the arcs are the same size). Also, the centers of the arcs are equadistant from each other, so the straight lines connecting the endpoints are all the same length. That's all well and good, but it's not the point.

I have a parameter to set the length of the straight line segments, and the radii/positioning of the arcs adjusts to the line segment constraints. Since there is no upper limit to where I can set the line segment parameter, I can push the line segments until they cross eachother, which creates a bit of a crisis for the arcs that are supposed to be in between the segments. I pushed the parameter beyond this limit and Mickey Mouse popped out.

The trick is, once Mickey appears, he won't go away, even if I decrease the line segment length parameter back to within appropriate bounds. It's like DP has created new geometry to adjust for the segments when they are too long, and then that new geometry stays put even after the line segments go back to their original length. I wonder if there's a way to get around this....


an animation of pushing the parameter limits


#2: Sticking parameters in power copies

I'm still getting used to what it means to include parameters and relations in power copies, both on the input and copy creation side of the bar. My floorplate geometry is supposed to respond to 3 parameters: the length of the straight line connecting the arcs, a factor that adjusts the amount of rotation, and the distance between floors. In this example I had spent so long worrying about the adjustments I made in plan, that I completely forgot that I had set parameters for adjustment in elevation. Note the missing 'floor_to_floor' parameter:



The worst part is that I forgot to test the floor_to_floor parameter until after I had painstakingly added 18 power copies that lacked that height adjustment ability. The planes that made up the supports for each copy could be readily adjusted, but the extrusion height of each copy was supposed to follow suit. Ahh, I can't wait until we learn how to script...


On the plus side, the straight line segment adjustment and the rotation adjustment worked just fine.

C'mon Baby, Let's do the Twist

Finally got power copies to work correctly!! Well, mostly correctly. In the next iteration I will remember to turn off all extraneous planes and lines before making the initial copy, so then won't have to go back into each instance to hide all that stuff I don't want to see (ex: screenshot #1).

Some things I like: The rotation of each floor is tied to the distance that floor is from the base. So the higher the floor is, the more it is rotated. Also, the overall size of each floor is tied to a profile curve.

Some things I will apply to the next iteration:

• Set extrusion height for floor slab depth to a parameter. Currently I can change the floor-to-floor height, but the slab depth is not controlled by anything.


• Set the "bite" diameter to be controlled by some other parameter. The diameter can be adjusted within the sketch that is within each powercopy instance, but I didn't build in any outside references that guide that diameter.


• Build the profile curve that controls the overall floor diameter as one piece. Currently it's three pieces (each ~1/3 the overall height). As such, when I instantiated a copy, and gave it the profile line as an input, it only locked onto one portion of that line. Then, if I tried to adjust the floor-to-floor height such that the top floor went above that first profile portion, I got an error. Bummer.

Operation Intersection - FAILED

Ran into a bit of an issue trying to intersect a helix with a series of stacked planes. Here's the setup: in 3D space I built a helix that made one rotation about a vertical axis while rising 2000' (the current height of my tower). Along this axis I set a series of 20 horizontal planes, each offset 100' from the previous one. I then started the Intersection tool and picked the helix a the first element and one of the planes as the second element. This went along fine and dandy until I hit plane #11, which is at the vertical midpoint of the helix. That's when I got an error message claiming there was no solution for that operation. Any thoughts?

I have feeling the error has something to do with the little orange point that can just be seen above the "Helix.1/Twist Setup/Part2" label

Here's the error message.

An animation of the steps I went through to make the intersection (I did these exact same steps for the other 19 planes, and they all worked).

Learning Curve

Long day fighting with old Rhino habits. I successfully created a framework that allows me to adjust the major and minor "bites" that are taken out of the overall circle of the floor plan. The trouble now is lifting a copy vertically in such a way that I can then specify a rotation. I don't think I set it up in the correct way to use powercopies. The image on the right is its own sketch, but it is built from 14 arcs that reference a whoooole bunch of points in the image on the left. So when I tried to define a powercopy, there were a bajillion defining points. Tomorrow I'll take another stab at it, maybe setting some constraints based off of formulas rather than just geometry. Also, I'm still not sure how to set it up to allow changes in the number of "bite" circles (getting it set up so that a set number were all equal radius and so their centers were equadistant from one another was a pain in the butt).

(LEFT) the sketch that forms the controlling framework. (RIGHT) an animation of what happens when controlling framework is tweaked.

Preliminary Sketches

Didn't go into the studio today, so can't use DP (I wonder if it'll be worth it to purchase a student license so I can work from home...) So instead of starting to model, I started sketching out ideas for that model.

First up: thinking about the radii of the indents in the perimeter of the plan. The indentation circles are all tangent to one another and all have the same radius. Their centers for another circle, and the radius of this circle could be used to drive the width/depth of the indentations.


Next up: As the floor area increases/decreases, I see a couple of things that could happen. Either the radii of the indentation circles could increase to maintain their tangency to one another with the same number of circles, or the number of circles could step up/down once a critical radius is reached. This second option would be interesting to test, since the radius of each floor decreases as the floors acend.

Last up: Thinking about the twist of the tower. Again, two possibilities I'm looking at. As overall tower height changes, does it maintain a full 360-degree rotation, or does it maintain the angle of offset between each floor?

Floor plans!

Some pertinent information that I've dug up:

• Proposed height - 2,000'


• 7 levels of below ground parking


• 56' high glass-enclosed atrium


• 150 floors


• Each floor rotated relative to it's vertical neighbor, such that the entire tower twists through 360-degrees

This is a photo of a drawing, so there is some distortion, but it shows how the floor plan is essentially circular, but with 7 large "bites" taken out of the perimeter, and 7 smaller "bites" taken out of the arcs created between the large "bites" (yeah, that's a bit of a convoluted description, I know):

I believe the current design of the base is slightly different than this, eliminating the inner support legs and with the atrium expanded out so that the outer support columns die into the glass. Some have likened this rendering to a stiletto heel:



Item of note here: the central core around which the floors are built:

Hopefully I can use this drawing to extrapolate the diameter of the floors and the amount of variation as the tower tapers from base to tip (It's a shame it show any of the elevator core):

Something to aSPIRE to.

First post...


...Daunting, to say the least.

I'll keep it brief. To begin, I'll be examining Calatrava's Chicago Spire: structure, rules governing the development of form, core-envelope relationship, patterns and deviations, etc. This exploration will unfold over the next couple of weeks, so I'll begin this beginning with some basic math:

+=

More to follow later.