Editorial Policy
   Contact Information
   Advertising Info
   Publicist Info
   Privacy Statement
   Contact 3Dgate

FEATURE • August 28, 2000

Click on image for larger version.

Tube House Design Process

by James Mann

Luz Vargas wanted to break the mold of conventional house building and provide a dramatic, yet workable design for a new house. The RIBA's Concept house competition provided her with the opportunity to explore these ideas and would ultimately have resulted in her design being built had it won. This competition is run every year with a brief to design a house that is innovative in terms of its design, construction, technology and conceptual basis. Espousing traditional ground-up forms of construction, she began experimenting with rolls of tape, cotton thread, and card board in order to develop her ideas for an hourglass shaped structure formed from a lattice of woven timber ribs. This early model proved that such a form could be described and allowed Luz to make early assumptions about how the internal spaces might work in conjunction with the envelope. The conceptual grounds for the design were good, but a lot more information and detail would be needed before Luz and the structural engineers at Ove Arup could finalize their proposals.



Despite having no 3D skills, Luz had been exposed to the potential offered by the medium on a number of previous projects. Luz approached me for assistance after hearing of my own 3D work when I worked as an architect for her husband. Her sketch proposals got me quickly excited by the potential to put together a great model and visual so I set about presenting the concept in a format that would enable the structural engineer to visualise the design and advise as to the structure. The idea for the form was strong but it did not suggest itself to an obvious method of construction. I began by modelling the desired envelope, then inserted floor plates and semitransparent volumes to represent the internal spaces. Luz had developed the building's outer form by stretching a sequence of cotton threads between two parallel rolls of tape. This idea needed to be presented in a diagrammatic form and so a seamless skin conforming to the shape was generated in 3D. These early investigations allowed us to asses the extent to which the internal spaces would be affected by the curvature of the envelope and confirm that height clearances would not become uncomfortably low at any point. Based upon this sketch model, Ove Arup were able to propose a structure of interwoven, prefabricated timber ribs.

Now that we had a strong design concept, coupled with a visually dramatic construction, we needed to establish how this should be set out. The geometry of the construction would be crucial to the size and positioning of the windows and openings, which in turn would affect the vertical positioning of the floor plates. The envelope structure would be all encompassing and have an effect on every component of the house, particularly those elements adjacent to it. Resolving the setting out and geometry of the timber rib construction at the earliest possible stage was essential to the success of all subsequent stages of the design and model building processes. The planned workflow was for the model and texture work to be completed in Discreet 3D Studio MAX, then exported to Autodesk Lightscape where lighting would be applied and the scene finally rendered. I have always enjoyed the power and range of tools within MAX but found it wanting when a high level of accuracy is required from the model so I used Robert McNeel & Assoc. Rhinoceros to set out and construct the rib structure.

The method of construction proposed by Ove Arup required strings of individual ribs to span between large structural hoops at equal increments along the hoop's circumference. By altering the spacing between each rib string and the sequence by which points on each hoop were connected, it was possible to proportionally increase or decrease the tightness of the waist to the hourglass form. A sequence of 2D and 3D studies were used to explore suitable arrangements in relation to the external appearance and the interior volumes. Once a suitable relationship between the external form and internal spaces had been developed as a diagram, it was a fairly simple task to create the ribs along the paths that took the place of the cotton thread in the original model. Fill-in panels were inserted into the diamond shaped voids between the intersecting ribs and arrayed around the house's perimeter to form the enclosed hourglass shape.

At this point, Luz reviewed the model against the relative position of the internal floor plates, then marked the agreed positions of window and door openings onto printed elevations of the model. These panels were then deleted and replaced. Rhino proved quick, versatile, and ideally suited to this task, allowing me to experiment with a great deal of freedom yet still produce reliable information.

Click on image for larger version.

Reference material for the proposed structure was extremely limited. Where it did exist, it was relevant to conventional tube forms only. For the design to succeed at the Concept House competition stage, our images would have to demonstrate that the proposed construction was viable, the geometry could be achieved, and what form it would take.

My architectural experience was essential in assembling this stage of the model. Design and model were developing at a close parallel rate with each leading the other at varying times. Typically, Luz would sketch out elements of the building's design giving descriptions of the size and spacing for specific items. Many of the details of the design were resolved in her head, but not in a technically drawn format. As we were both working within the same building, we could meet during lunch breaks. I would present quick renders showing the model in its latest state or aspects that Luz had asked me to explore in 3D detail. These meetings also gave me the chance to indicate what new or additional information I required. Luz would give me her latest sketches and describe how the new and amended elements should work. The time before the competition was short, so for me to be able to take information and understand how it was to work with minimal explanation meant that we then had more time to discuss the actual design, how the building was shaping up, and how we might improve it. More architects should work this way. I have personally worked on many visuals and many building designs, but to work on both, simultaneously, would seem to be something of a Holy Grail.

Most architectural visuals serve as a rough draft of a design in terms of the quality of spaces, forms, and materials. With the Tube House's explicit and exposed construction, every significant component needed to be modelled. Elements that did not work in 3D would not work in reality and would appear incongruous in the model. Briefings with the architect took place on a daily basis when necessary. Information passed both ways, architectural experience kept me moving forward as conversations regarding construction, composition, and spacing involved me making informed decisions about what would or would not work and having a clear understanding of what the designer hoped to achieve. This approach allowed me to not only avoid the hazards of a trial and error approach by instruction, but to make suggestions as to how elements may be achieved within the framework of the design. I did after all have the advantage of working in and around a developed 3D representation of the building.

Continued on page 2. >>