Learning from Japanese Structural Design: Reflections on the Symposium

MoMA’s exhibit on Japanese architecture (through July 31, 2016) examines the “constellation” of influence in the country’s early-21st-century architecture and design community, but perhaps not so explicit in the exhibit are 1) the structural engineers’ parallel relationships of influence and 2) the structural engineer’s role in collaborating with architects to produce these works. In an effort to explore these characteristics of structural engineering influence in Japan, Prof. Guy Nordenson (of Princeton University and Guy Nordenson and Associates) and Prof. John Ochsendorf (of MIT) organized a symposium, titled “Structured Lineages: Learning from Japanese Structural Design,” which brought together some of the top structural designers from both Europe and the US for discussion.

Most of the lectures presented by the guests focused on the works and experiences of specific Japanese structural designers and educators such as Yoshikatsu Tsuboi, Mamoru Kawaguchi, Masao Saitoh, Gengo Matsui, Toshihiko Kimura, and Mutsuro Sasaki. Each half of the symposium brought the speakers together for a vibrant panel discussion moderated by our Prof. Sigrid Adriaenssens and MIT’s Prof. Caitlin Mueller. The final panel discussion welcomed Prof. Sasaki himself to the mix.

First panel discussion moderated by Prof. Adriaenssens. Left to right: Seng Kuan, Marc Mimram, Sigrid Adriaenssens, Mike Schlaich, Laurent Ney.
Second panel discussion moderated by Prof. Mueller. Left to right: Guy Nordenson, Chikara Inamura (acting as Prof. Sasaki’s interpreter), Mutsuro Sasaki, Caitlin Mueller, Jane Wernick, Bill Baker.

Several fruitful discussions and themes arose from the independently-constructed lectures. Reflecting the literal implications of “lineages,” Prof. Seng Kuan referenced the traditional lineage model in which Japanese arts and crafts get passed down for seven or more generations. As Prof. Ochsendorf demonstrated in his lecture with the help of Chikara Inamura, such a “lineage” is visible in 19th-20th century Japanese structural engineering:

Continue reading “Learning from Japanese Structural Design: Reflections on the Symposium”

A Physical Costa Surface 2/3: Form Finding Process

How can we algorithmically approximate the form of the mathematically defined Costa surface? This question is at the center of this second blog post of the “physical Costa surface”series.

The form finding approach introduces a physical dimension to the equation generating the minimal surface. Finding the shape can be done in several ways. However, whether it is physical form finding or numerical form finding, the process implies that the initial geometry of the material is relaxed in the final state under a given set of loads (such as gravity, or pre-stressing). In our case, an initial mesh is pre-stressed isotropically and then relaxed to find the final shape (see figure below).

Form finding process of the prestressed membrane for the Costa surface 

The initial mesh is created by defining a four hole costa surface geometry. The image below shows how this intricate surface appears when cut along a vertical plane of symmetry. In this image two holes are cut and one is hidden. In order to create the initial mesh, the surface is divided once more along the vertical symmetry plane perpendicular to the first cutting plane. The observation of a quarter of the Costa surface geometry helps us create an initial mesh made of low curvature surfaces. this mesh is then adjusted for the six hole geometry. The initial mesh for the analysis appears in the image above.

A mathematically defined four-hole Costa surface slit along a symmetry plane (credit: math.hmc.edu)

The actual form finding takes place once the initial mesh has been determined. Three numerical methods were used in the form finding process: (1) Direct resolution of the system by Newton’s Method (2) Natural Force Density Method (NFDM) and (3) Dynamic Relaxation Method (DRM). The initial mesh is prestressed isotropically and allowed to relax into the final geometry with the three fixed ring boundaries seen the figures. The surface area of the initial mesh is 9.083 m2.

numerical analysis
Comparison of three form finding techniques (left) – Displacements obtained from Newton’s Method after 18 iterations (right)

The three methods used converge to final mesh areas between 8.25 m2 and 8.29 m2. The three numerical methods used effectively reduced the surface area of the initial mesh under the given boundary conditions, and created suitable final configurations that are close to the actual minimal surface. Among the three methods, dynamic relaxation produced a final configuration with a smallest surface area of 8.248 m2, but needed the longest computation time to converge. NFDM and Newton’s method both achieved convergence with satisfactory results and much shorter computation time

It is not possible with a material to achieve the final form without taking into account the construction phase.  The construction of the textile surface will be detailed in the next and final post of the series. Stay tuned!


Author: Victor Charpentier

Thanks to Princeton University graduate students Xi Li, Max Coar, Tracy Huynh and Olek Niewiarowski for their contributions.


What I Am Thinking: Biologist-Turned-Architect Doris Kim Sung Makes Buildings Breathe

During my Art Residency  in Bellagio, Italy, I had the privilege of interviewing USC Architecture Professor and Princeton alumna Doris Kim Sung. In her work, Doris interprets architecture as an extension of the body and explores how buildings can passively adapt to their environment through self-ventilation and shading by using smart materials and design. 

Russell Fortmeyer, Doris Kim Sung and Sigrid Adriaenssens in conversation in Bellagio, Italy.

Sigrid Adriaenssens: What are the research questions that your designs address?

Doris Kim Sung: Can the geometry or the unit design of a smart material such as thermobimetal affect the architectural performance of a larger tessellated surface intended to shade, ventilate, stiffen, or propel? 

What is “unplugged” architecture? Can you exemplify that concept with one of your projects?

This reference from rock or pop music means without electronic amplification or disconnected from the world of gadgets. I have a deep-seated interest in finding solutions that don’t require added electrical energy or computer controls. For this reason, I have been working with smart materials such as thermobimetal, a material that reacts to heat (it curls), and developing for building use (for auto shading and ventilation in “Bloom”) and construction techniques (for one-hand/one-person assembly systems). Because the use of the material does not require energy, it is a “passive” type of system, but the responsive nature of the material to the sun and ambient temperature make it surprisingly active.

What can you tell us about your latest innovative project?

Continue reading “What I Am Thinking: Biologist-Turned-Architect Doris Kim Sung Makes Buildings Breathe”

Bending Active Systems in Antiquity: What was so special about the bow of Odysseus?

The second Saturday of May is National Archery day (May 14 in 2016).  While today archery is mostly practiced for sport, the bow and arrow have been used by humans in hunting for centuries.  The purpose for bows, regardless of the target, is to propel an arrow using stored elastic energy.  They are one of the most effective ways of storing energy of the human muscle.

The oldest bow in one piece dates back to 800 BC, the same era that scholars believe Homer’s Odyssey was written.  The Odyssey tells the journey of Odysseus, a hero in Greek mythology.  As part of the story, his wife, Penelope, patiently waits for 20 years during and after the Trojan War for her husband to return.  Through this time she challenged her suitors to string the bow of Odysseus.  With the exception of Odysseus himself, none of the suitors possessed the strength needed to string the bow.

Ceramic with depiction of Odysseus holding bow and arrow in The Odyssey

While Odysseus was a Greek hero, there is a limit to human strength when stringing a bow.  The human body allows one to draw their arm back about 60 cm and the maximum force a strong man can withstand holding in a string is about 350 N (Gordon).  Therefore, the available muscular energy is 0.6m*350N= 210 Joules.  Assuming an unstrung bow has zero energy, we can linearly plot the force from the archer’s pull against the maximum extension of the string in Figure 1.  The energy stored in the long bow is equal to the area under this plot (triangle ABC): ½*0.6m*350N=105 Joules.  This stored energy due to deformation is known as strain energy.

Sketches of traditional long bow in unstrung, strung and extended conditions
Figure 1: The archer’s maximum pull on the string plotted against the distance of arm extension.  The area under the curve is equal to the total stored energy in Joules.

Continue reading “Bending Active Systems in Antiquity: What was so special about the bow of Odysseus?”