This is my second try to rebuild the parametric wall. It’s a double curved wall partition. it consists of 3 levels of curved edges which give it its free-form appearance. the wall is created by stacking standard-sized bricks along its horizontal contours. the wall bricking is switched simultaneously in terms of laying direction so it gets the famous bond image.
The first challenge was to build the system which can constantly control the stacking protocol. The simplest way was to divide a curve, orient elements and repeat the process vertically. Then I decided to alter stacks so I can give the facade the running effect. It was not easy, I needed to use conditional logics to construct the alterations. Each curve is a plane intersecting with the free-form surface at equal heights. The curve length is measured against the desired brick dimensions. Spacious margins were considered to make sure the bricks doesn’t collide or exceed its extents.
after that, I tried to manipulate the bond patterns. It was the most difficult process,through, It was full of fun. The logic of reproducing the patterns recursively was very dynamic. the process took more than two days to formulate. The looping tool in Anemone helped a lot in driving the production. It’s not an issue to place identical bricks along a curve, but what considered an issue is to reorganise these oriented geometries to take new settings without collisions. the options are endless, but the most familiar one is rotating. texturing is another technique. Moving geometry counter to the normal flow direction is not easy. I can count many other strategies for stacking and bonds making.
The process depends on multiple levels of recursiveness and looping. the first level was concerned with generating randomness of values represent the bricks on its standard world XYZ orientation. Each randomly rotated brick took specific bounds on space while placed at a point. the bounds data is used to build the list of expected sub-curves for the following process. It was also helping in collect another kind of data like rotations lists and curves lists.
The second level of lopping was about orienting the standardised brick element to its corresponding location in the original curve. the loop constantly imports data from the previous loop’s recorded data to feed it to the selected curve level (Row).Because of the multiple levels of recursiveness, I used multiplication of seeds cross loops. That helped to get a variety of results all along the wall facade.
Finally, The collection of bricks patterns were exported up to the first level looping used for wall levelling. The output uniqueness and innovation were magnificent. never saw such possibility ever before. but what really interest me is the potential of such approach on creating different and unique patterns for walls tesselations.
I had few collisions in between few bricks, that happened when I used extreme curvatures for the overall wall form. This why I appreciate my earlier definition(the one I forget) which was operating by rapidly measure distance between each brick and the one lying before it, and then move either toward it or away depending on collision algorithm. This why this is my next task for Parametric Wall 3.0.
In the coming versions, I will try to study other issues I noticed on this one, such as the y – axis deviation and the possibility to optimise the pattern in this direction. the possibility to generate bonds using rotations on other axes is another issue I am looking at at the moment.