Terra Tettris

Introduction

TerraTetris was developed as a participatory design game to enable the dwelling organisation and configuration in the Zaatari camp in Jordan for the Syrian refugees generatively. It was realised by looking into the existing housing situation, identifying the problems and focusing on the opportunities that would give way to propose a more systematic and modular approach as part of a universal system.

The proposal is carried out in various interrelated scales in a feedback-loop logic, connecting the design phases from urban configuration to the construction details under one manifold system logic.

Game Instructions:

The intention of playing the game is to involve the camp dwellers to participate in the decision-making process involved in planning of their Earthy houses which is a more permanent alternative to their current container/tent dwelling.The participants playing the game are the chosen representatives of the individual family who will try to put forward the interests of the family involved in the game. Depending on the type of dwelling the number of participants may vary. For example, in a family dwelling where there are 3 families of 5 members each the number of participants playing will be three. similarly, in a communal dwelling there may be many more participants since it comprised of a lot of smaller families and individuals.

The participant must fill in a list of his /her preference in terms of importance of the various spaces in the house and the priority for the direction of the spaces if any.

If the participant has no preference regarding space planning, then they can rely on the Architects decisions which are the default values in the game. The Architects decisions are based on the REL-Charts and the Bubble diagrams made for the various dwellings based on the level of privacy and the ideal connectivity between the spaces.

The process is a simplified framework for generating courtyard type of houses in a participatory manner.

Urban aggregation logic (Scale03):

Four main objectives were proposed at an urban scale level as a combined strategy to address some of the main design problems identified during the research process.

1.Maintain the current grouping of families 2.Interaction with surrounding context 3.Introduction of communal functions 4.Rearrangement Strategy

To avoid the constant relocation of the users, the dwelling arrangements (Family and Communal) will be implemented one by one, where (if possible) only the families that will be accommodated on each step will be moved on the process. This rearrangement (as the Tetris game) will be directional, following the attractors (the existing or proposed functions) attraction/repulsion effect which ideally will be different on each urban block to create more variety. On this manner, the attractors will determine the start and the order of the rearrangement process. The dwellings that are in direct contact with the attractors, depending on the function, will create the hybrid configurations as a combination of commercial and housing.

The size of people on one block, based on the program of requirements, will determine the amount of square meters of green areas and communal areas that need to be implemented, converted in grid units. As a result, the total number of modules of each function (housing, communal areas and green areas) will be known and the order of implementation will be established. The process of implementation of the modules will be divided in a certain amount of repetitions to achieve an interspersed arrangement of functions.

Dwelling Configuration Logic (Scale02):

This stage creates a courtyard type of house since it had contextual benefits in terms of climatic comfort. A typical courtyard house is configured in a way that all the spaces in the house connect to the courtyard and the courtyard becomes the primary source of access to all the spaces in the house.

This provided a starting point for our breakdown of the design process. As the first step we generate the courtyard and the connecting passage around it. Then based on the size of the connecting passage we calculate the available sides for connection with the other spaces in the house.

The size of the courtyard is dependant on the number of people in the house and a per person area consideration. It can be argued that the courtyard can also be calculated based on the number of sides needed by the various spaces in the space program so that the size will always be enough for the spaces to fit in . The main problem with this approach is that the courtyard becomes so large that it no longer functions as a courtyard but behaves more like an open space both spatially and climatically.

The second step in the process is to start arranging the spaces along the courtyard till all the available connecting spaces along the passage are filled up. The remaining spaces then can either go on the first floor or can connect to an extra passage originating from the main courtyard passage. A combination of both is also possible . The order of arranging spaces along the courtyard is based on the decision tree as generated by the player. The decision tree records the preferences aof all the players and guides the block placement process The Diagrams shown below and on the folowing page describes the whole process diagramatically for communal house within the camp. The construction of the modules happens in a modular way where the attemp is made to generate compression only blocks for cost-efficient and monomaterial construction in adobe bricks .

Unit Development Logic (Stage 01):

The same systematic approach is followed during the shaping process and the unit development. The main component that connects all the focus points and scales of the process is the tartan grid and the measurement units used, which dictate also the shaping and dimensions. The tartan grid is used in the end as a reference guideline for the urban aggregation, module development and construction. This grid is 1500*1500mm. All the rest of the units used are subdivisions and multiples of each other and correspond at the same time to the ergonomics, the construction and workability requirements, like the brick size and opening sizes. The goal for the form-finding is to create modular and easily repeatable units that can be related to each other to create various adjacencies and connections as part of a flexible and open-ended system. The units developed are of two types the first one increment in size with an increase in the number of perople and the second one increases in number with an increase in number of people.

Structuring (Scale 00):

The structural analysis was carried out in parallel with the design of the room types, their form finding and constructability. The goal was to reach a desired solution within the material requirements and following the same modular systematic approach in having easily repeatable parts. In this approach, certain compromises needed to be made through trial and error and the results from each side were used as feedback to enable decision making and possible shaping or structural improvements. First of all, certain parameters were set as constraints in respect to the systematic modular logic of the whole design. These were, for example, the general plan dimensions, the width of the openings and the height of the arch above, the dimensions of the walls and columns, the spring height of the roof and the maximum height of the modules having a second floor. However, some flexibility is allowed as long as it respects the unit modularity and if it can be accomodated in this logic, by creating multiples or subdivisions of units and measurements. This relates to the roof heights as well, but is influenced also by the constructability and workability at the same time to find a close to optimal trade-off.

Once the parameters were set, a first approach for the shape of the arches, vaults and domes was realised through the use of the Kangaroo plugin for Grasshopper, as explained in the previous section, in order to support their selfweight. However, although this provides with an estimation of the shape in a dynamically relaxed form, it still remains an approximation and it cannot be directly used for a more accurate structural analysis. This is also because the constraints required for the different heights cannot be fixed precisely and the roof structure is not as close as it would have been with the actual construction method. Because of that, a second stage of form-finding and meshing was carried out to optimize the shapes and to use it as an input for the structural analysis, whichwas realised through the Karamba plugin for Grasshopper.

In this process, the final roof shape that was studied for its constructability was used for the analysis. Apart from that, different meshing and mesh subdivisions were tested, in order to find the one that gives more accurate results, since the software interprets the meshing in a specific way and that influences its internal calculation process. In this continuous process to optimize the shapes, the loading conditions were set for each case, the final geometry was structurally analyzed and the results were evaluated according to the limitations of the material taken into consideration.