4.2.1 Choosing the optimum simulation environment
As it is referred above this step of the experiments try to find the optimum simulation environment that exhibits as much accurate as it is possible the wind funnel effect. The parameters that taking into account in the tests are
1. wind direction (parallel or perpendicular to the passage)
2. wind speed (three different speeds are tested according to extended Beauford Scale
3. buildings height (experiments with two different types are taken place)
4. passage width (narrow or wide according to buildings height)
The diagram below explains clearly all the experiments and the possible combinations (24) among those parameters that were implemented.
According to the literature review and the experiments that took place it is clear that final simulation environment should have same height buildings with a narrow passage between them as well as the wind direction should be perpendicular to this passage with speed 2.
4.2.2 Definition of the problematic area
Also in this part of the experiments the problematic area that defines the space where tensegrity units can be grown was explored. The approach followed here was to try and block the wind penetration and trapping inside the passage, so after thorough experimentation and according to the background of the windbreaks, it was perfectly clear that the best areas to act as windbreaks were the corners of the leeward building and the façade of it. But the maximum available depth of each side should be tested. Three different experiments with different sizes were implemented per area in order to find the final area that GA process would be allowed to operate on. All experiments were carried out under the scope of giving room to the GA algorithm to operate but at the same time assist towards defining an area that would push towards a solution with minimum amount of blockages, hence the windbreak with the maximum permeability which was the actual optimization goal of this step. All experiments with depth bigger than 3 showed that the problem was easily solvable by the majority of the solutions, leaving no room to the GA to optimize stuff and experiment with depth smaller than 3 showed a lack of actually finding a viable solution. Hence it was decided to use the problematic areas explained in methodology step above. Below diagram shows experimental results.
As it is referred above this step of the experiments try to find the optimum simulation environment that exhibits as much accurate as it is possible the wind funnel effect. The parameters that taking into account in the tests are
1. wind direction (parallel or perpendicular to the passage)
2. wind speed (three different speeds are tested according to extended Beauford Scale
3. buildings height (experiments with two different types are taken place)
4. passage width (narrow or wide according to buildings height)
The diagram below explains clearly all the experiments and the possible combinations (24) among those parameters that were implemented.
According to the literature review and the experiments that took place it is clear that final simulation environment should have same height buildings with a narrow passage between them as well as the wind direction should be perpendicular to this passage with speed 2.
4.2.2 Definition of the problematic area
Also in this part of the experiments the problematic area that defines the space where tensegrity units can be grown was explored. The approach followed here was to try and block the wind penetration and trapping inside the passage, so after thorough experimentation and according to the background of the windbreaks, it was perfectly clear that the best areas to act as windbreaks were the corners of the leeward building and the façade of it. But the maximum available depth of each side should be tested. Three different experiments with different sizes were implemented per area in order to find the final area that GA process would be allowed to operate on. All experiments were carried out under the scope of giving room to the GA algorithm to operate but at the same time assist towards defining an area that would push towards a solution with minimum amount of blockages, hence the windbreak with the maximum permeability which was the actual optimization goal of this step. All experiments with depth bigger than 3 showed that the problem was easily solvable by the majority of the solutions, leaving no room to the GA to optimize stuff and experiment with depth smaller than 3 showed a lack of actually finding a viable solution. Hence it was decided to use the problematic areas explained in methodology step above. Below diagram shows experimental results.