Computer-Aided Design (CAD) software is now so ubiquitous in architecture schools and practices that we rarely question its presence—indeed, many cannot even imagine a time when it was not part of our design “toolkit.” As an umbrella term, CAD covers a vast array of programs that produce different results. Some only create two-dimensional drawings, while others are capable of highly sophisticated three-dimensional renders and animations. The initial benefits of CAD principally related to its suitability for repetitive work, as it facilitated the use of “copy,” “cut,” and “paste” functions. These generic tasks enable a designer to very quickly add, reposition, or remove elements of the design by selecting the appropriate part(s) and then carrying out the required function, typically through an on-screen menu. These functions raise an interesting issue that will have greater implications further on in this section, that of independence. The simple fact that design elements can be so easily manipulated is due to their independent nature, i.e. even though they may appear visually connected to other elements within the design they are not, so if one element is moved then those surrounding it remain unaffected. This functionality of CAD software affords rapid construction and transformation of designs through simple stages. In order to position elements relative to each other, CAD users set up grids or guidelines as a background layer to the drawing to enable them to connect or “snap” the elements together effectively. However, this behavior also highlights one of the disadvantages of the independent nature of this technique, since making significant changes to the drawing may result in many other parts also having to be copied, repositioned, or pasted. Clearly, the greater the degree of complexity in the CAD drawing, the more manual work involved in revising it. As such, CAD programs based on this type of functionality are very useful once a design is well developed, but may limit early exploration and constrain the design process.
Revit screenshot from ONL’s drawing package for the Al Nasser Group headquarters, Abu Dhabi. The tower’s gently curving form renders each floor unique, which makes the use of digital design tools very pragmatic.
Photorealistic computer-generated images (CGIs) are frequently used to sell the potential of a design prior to development, and may be advanced and detailed enough for it to be difficult to know exactly what is real and what is virtual— as shown in this image for Rogers Stirk Harbour + Partners’ design for the Leadenhall Building, London.
Accurate geometrical description and curvilinear deformation, illustrated in this detailed panelization screenshot by Zaha Hadid for the Haydar Aliyev Cultural Center, Baku.
One of the primary benefits of CAD data is its transferability into other software platforms and formats. The majority of computer-generated images (CGI) readily used in architecture and in the visual encounters of our everyday lives—such as advertising, television, movies, animations—are not simply the product of one single software program but are hybridized images that were developed using numerous packages to optimize their visual impact. However, once the final images are produced there remains the fundamental process of showing how the objects can be constructed within a CAD environment. A key issue in using any CAD software prior to fabrication is scale. It is important for any designer to consider the limits of the machine they intend to use since there will be constraints in terms of the size and number of elements it can make at any one time. Herein, we discern the need for geometry to create, manipulate, and refine our design ideas. In contrast to physical modelmaking—through which we manually alter the shape, form, and geometry of a material depending on its properties—in a CAD setting we are unable (yet!) to literally reach into the screen and directly change a surface or form. We therefore need suitable mechanisms by which we may control appearance and other formal characteristics. This is where the geometry of CAD software comes to the fore, as it affords the designer a level of command over creative ideas. There are two basic ways of making three-dimensional forms digitally: Non-Uniform Rational B-Splines (NURBS) and meshes. The key difference between them is that NURBS facilitate smooth surfaces and curves, whereas meshes approximate these formal elements via polygons and subdivisions. Projects may typically require designs to be defined in both formats, depending on the stage of the design process alongside software applications.
The contemporary city is increasingly viewed as a series of informational and material flows and exchanges, as investigated in the dynamic “infrastructuralism” of the West Side Convergence project, New York, by Reiser + Umemoto.
The seminal “datascapes” used to illustrate MVRDV’s research opened new avenues for discourse on design ideas and the nature of landscape and urbanism, as shown by their image for Waste Sector from Metacity Datatown (left). Alongside such research, the office continues to develop striking projects such as the Sky Village proposal for Rødovre (above).
The organic geometry of the bays for FOA’s International Airport in Shenzhen was inspired by woven bamboo, sea ripples, and the floating ribbons in traditional Chinese dance, coupled with highly pragmatic responses to the effect of the sun on the building. The resulting asymmetrical roof system is set in relation to sunpath and shading parameters, while also providing skylights.
This geometry study for the Taipei Performing Arts Center by EMERGENT/Tom Wiscombe was developed in CAD to enable circulation, structure, and program to be evolved in an integrative manner.
In these CAD images for Neri Oxman’s Raycounting project, the designer explored a novel method for developing form through light-ray orientation and intensity. By assigning light parameters to flat planes, 3-D double-curvature surfaces are produced that are subsequently translated using digital fabrication techniques into physical objects.
STEP BY STEP ADVANCED COMPUTATIONAL DESIGN
The design development of many contemporary projects requires designers to engage with a variety of digital tools to explore the implications of their ideas and generate appropriate data for the manufacture and assembly of the building. In this example, a range of analytical and design software is used to allow the architect to thoroughly test and understand their creative ideas from concept to fabrication.
The House in Dublin project by Amanda Levete Architects (AL_A) uses daylight to dramatically animate a deep plan. This is achieved through the careful articulation of a curved roof-shell surface, perforated with glazed elements. The plan is penetrated by a central courtyard, providing high levels of daylight and natural ventilation to all internal spaces, and a discreet external private space that also receives sunlight.
1 Volumetric studies use a series of physical models to examine the design’s curvilinear geometry. Given the complexity of the form, these are a product of digital fabrication themselves—using 3-D printing and contouring for accuracy so as to explore more refined aesthetic decisions.
2 Surface-stress analysis is carried out on the proposed curved roof-shell structure to check for stability and areas of high stress. The continuous skin functions as a funnel for the collection of rainwater, so additional “live” loads are added to test their implications.
3 Once the curvilinear form is sufficiently developed, patterning parameters are applied to the skin, as the composite roof panels will be manufactured off site and installed in large sections.
4 This then enables individual panels to be determined, a number of which are fabricated as digital, then 1:1 physical, prototypes. This allows the design team to conduct further tests on detail components, such as structural and thermal integrity, as well as to explore more refined aesthetic decisions.
A photorealistic digital render using data from analyses and design development to give an accurate visualization of the proposed project.
