Deceptive Simplicity

November 7, 2011 / no comments


I’ve been thinking about the relationship between simplicity and complexity in design. Why do some design problems initially appear simple but then upon investigation, turn out to be very complex? Why does the solution to a complex problem often, after lengthy analysis, turn out to be the most simple answer? Or why does it sometimes take a very complex technical solution to produce an elegantly simple visible end result?

During the design and construction of a recently completed project, I asked myself some of these questions. Although I don’t have all the answers, this project provides some examples to demonstrate what I’m talking about. The project is the United States Institute of Peace in Washington, D.C., designed by Safdie Architects.

Here is one of the design challenges presented to us: make it look like the architectural model – make that translucent roof glow at night. And, oh by the way, you have to light the space underneath at the same time. It seems simple, right? But it’s really complicated.


We did lots and lots of computer modeling to test out different ideas. Now, you’re probably wondering, why is there a picture of the Lincoln Memorial? Well it turns out, the roof couldn’t be any brighter than any of the surrounding monuments and memorials, so we had to do a complete luminance study and a presentation at the National Capital Planning Commission to show that our roof wouldn’t be any brighter than the memorials.


Then we had to estimate the transmittance and reflectance of the roof, and it turned out the roof system was going to be in two layers: an outer layer of translucent glass and an inner layer of fabric membrane. So, estimating this was actually quite complicated because of the inter-reflections. We started with back-of-the-napkin sketches and then moved on to tabletop mockups with some of the possible materials for the roof.


Then we moved on to full-scale mockups, and these had to be done in Germany because that’s where the roof was being built (at Seele, outside of Munich). First we looked at the mockup in daytime to see how the different combinations of possible inner and outer materials would perform.


Then we tested the different material options outdoors at night with the proposed lighting solutions. And we did visual evaluations of how it looked outside and inside, took all kinds of meter readings, and of course when we were done, since we were in Munich, we had to have a beer!


So after that very complicated design process here is the solution – really simple: for the interior portions of the roof, linear fluorescent forward-throw cove fixtures. And for the exterior overhangs, in-ground metal halide adjustable accent lights.



And here’s the visible end result: very simple and elegant. Architectural forms are revealed and the space is well illuminated. I knew we were successful when a visitor said to me, “So, you’re making the roof glow, but I don’t see any light fixtures.”



Moshe Safdie’s vision was realized, and we’re a good neighbor to the surrounding memorials and monuments. So was all this complicated design process really necessary to reach this beautiful end result? That’s what I’m thinking about. I think it was. All the modeling and mockups and testing and head-scratching gave us much, much more confidence that our very simple solution would work. Without it, I think we would have been inclined to make the solution much more complicated, and in the end, that could have given us a final result that was cluttered and incoherent.

Photo Credits: Safdie Architects (1), Glenn Heinmiller/Lam Partners (2-10)

Measure Twice, Cut Once

April 11, 2011 / no comments


The old adage of measuring twice and cutting once applies just as well to design as to construction, and especially to the design of our lighted environment. Lighting can certainly be judged quantitatively, and it often needs to be, but it is also always qualitative and very subjective. The perception of brightness, the balance of light and dark, shade and shadow, and the appropriate contrast to enhance a sense of dimensionality without visual clutter, all play a role in the quality of lighting.

Sometimes numbers aren’t enough, or they don’t show the whole picture; sometimes seeing really is believing. In the case of a recent high-rise tower project, the penthouse screen wall was backlighted to create a glowing crown within the nighttime skyline. This started as a computer model to study the relative output needed, along with the general patterning of the screen wall structure. Then, during the glass selection process with the architect, a series of full-scale mockups were done on the rooftop of the existing building to evaluate the appropriate brightness.


Numerous glass samples were explored and tested, during the day and backlighted at night. Various frit patterns and interlayer films behaved differently at night, lighted from one side, than they did during the day when daylight illuminated both sides.

Care was taken to accurately simulate the structure supporting the screen wall, in order to represent realistic shadowing; however, the primary purpose of the full-scale mockups was to determine with the client the preferred brightness of the backlighted glass. There was no right answer in this case.


If the process had stopped there though, we still would have missed the mark, because brightness was only one piece of the composition. The next step went back to small-scale modeling. A 1/4”-scale physical model was constructed of one entire façade, with all of the structure and fixtures accurately represented. This allowed the overall pattern of the wall, with all of its potential shadows, to be scrutinized and explored.


The interplay of shadows from vertical and horizontal structural elements, in relation to standard fixture lengths, created a depth to the lighted wall that, at this grand scale and from normal street-level vantage points, became a positive quality that the entire design team, along with the client, preferred over a relatively uniform and flat appearance. The distinction between texture and visual clutter can be a very fine line, and neither the computer model nor the full-scale mockup really told the full story.

Sometimes the answer is quick and right in front of you, and other times, especially when perception and subjective qualities are in question, multiple methods need to be explored to make sure that the right measurement was made. A single measurement would have cut this design too short.


Photo Credits: D Sharon Pruitt (1), Lam Partners (2-5)