The Abdul-Aziz Center for World Culture (ITHRA)
Mohamed Naeim shares with us how he used computational design tools during the construction of the Abdul-Aziz Center for World Culture (ITHRA) by Snøhetta architects
The Abdul-Aziz Center for World Culture (ITHRA) is one of the most prestigious projects currently being built in the Middle East. It is designed by the international firm Snøhetta from Norway and it’s purpose is mainly to host cultural activities and events. This giant center consists of various buildings containing auditoriums, halls and many other spaces. The project has a free and organic form resembling a cluster of rocks merged together in a homogeneous way. Because of the complex geometry of the project, I developed various computational design tools that were used in different phases of the project. In this blog post I will cover the process of using these tools while working on the ITHRA as an architect.
The building envelope consists mainly out of two layers, the structure and the envelope itself. The structure is made out of typical concrete floors and columns and mass customized steel structures to support the free-form roofs and walls. The envelope consists out of two main elements, an insulated weather-tight layer that is wrapped with sun shading elements, like a veil covering and protecting the inside from the harsh outside climate.
My activities involving this project summarized in five main tasks were; embodiment of the 3D model, BIM data management; fabrication coordination; precision handling and machine automatization. These were challenging tasks, because all of them were new and unknown. I faced new problems which had no previously known precedents. One of those challenges I faced was the digital model we got. This model was a simple CAD model, but because of the complexity of the building envelope, the construction required converting this model to a BIM (Building Information Modeling) model, a model containing information and data embedded into each part of the tectonics. I’m going to walk you through the process I went through doing this project. It consisted of the following steps I’m going to explain more in depth below.
The 3D model had all the elements coded by color and function, but this was not enough information. We needed to be able to recognize the elements and name and classify them in a clear manner to being able to extract more specific data regarding the construction from the pieces. I developed a strategy of tackling this issue, where I used a combination of Grasshopper, a visual algorithmic scripting editor for Rhino 3D and custom Python scripts I wrote myself. These tools were the main generators to get to most of the solutions.
The first challenge I had was to manage fabrication between two different coordinate systems. This is a vital piece of information as orientating elements on a free-form building envelope becomes a sheer impossible task using only 2D drawings. There are over 3000 unique elements and all of them have a different location and orientation. I assigned a self referencing plane to each element so I could alter its orientation between the original site-based plane (world plane) and the workshop-based plane (temporary plane).
Orientating and cataloging the different elements and embedding tolerances for installation purposes
In the second stage I wanted to extract reference points to locate the frames, the mullions, their fixing bolts, the points of the hooking unit to the floor and the point of hanging of the shading elements for the outside. There are hundreds of these points per element, so I automatized the process of finding the reference points, ordering and cataloging them in an efficient and logical manner, so finding these elements and retrieving attributes would be much more easy, fast and efficient.
Detailed cladding system
From this point on-wards our initial 3D model became a fully integrated BIM model, full of useful data that is transferable, reusable and inter-operable. I handed over our model to the surveyors to measure the initial steel frame. We used a laser locator to assign the planes for the elements and I could double check the status of each reference point I organized and cataloged in our 3D model. The machine recorded the data and returned the results to me so I could check the deviation between the ‘real’ reference points and the ones from the 3D model, so adjustments could be made in the workshop if needed. This process was repeated throughout all the other stages of manufacturing.
Curved part of the facade
While fabricating the rest of the elements, some elements required to be produced by a special type of machine, namely folding machines or bending machines. Depending on the element the proper machine was used to produce the desired form. The curving process required a special kind of data, so I developed another Grasshopper algorithm to handle and organize this data. We heavily relied on the use of these machines when we needed to build a mock-up, a one to one scaled structure, representing a part of the building. We needed to produce a number of extremely curved units and install them to the structural portion at the factory. All the curved units were unique with interconnected curved runners, which made it even more challenging.
Specialized folding machines
When dealing with a building envelope consisting of more than 3000 unique elements, orientation and location, manual modeling and organizing these elements becomes a sheer impossible task and 2D drawings become obsolete. This is why I want to stress the importance of computational design tools and strategies. They not only increase accuracy, speed and efficiency but they allow you to understand your design better and most important having complete control over what is happening. Most time gets lost in poor planning, poor communication and the exchange of information between parties. A BIM model is therefore crucial to being able to communicate and exchange information in a clear and efficient manner between teams and externals.
Some things to take into account, for architects designing these complex free-form projects is to give constructors enough margin and tolerances to be able to keep the precision. Especially when the constructors use more primitive methods of building. I found that embedding a sort of intelligence in the elements so they can become adjustable and adaptable for errors that are prone to happen is important to take into account.
Besides the digital computational tools, it is also important to use advance physical tools to assist in the building process, like 3D laser locators and CNC machines. These tools reduce manufacturing time, reduce errors and increase accuracy if used in a correct manner. Therefore it is important to come up with a clear workflow on how to integrate these tools efficiently in the building process for maximum results.