Skip to content

Finding What’s New Under the Sun

September 25, 2013
by

Dr. Christoph Reinhart spoke on NPR last week about Raphael Vinoly’s Walkie Talkie making the news  for melting a nearby Jaguar automobile (See my previous post on the subject here.)  Dr Reinhart, who leads the Sustainable Design Lab at MIT is responsible for the DIVA plugin for Rhino/Grasshopper which interfaces Greg Ward‘s Radiance ray tracing engine to Rhino models.  He gave a good interview, mentioning several recent examples of the solar light pollution problem including the Museum Tower in Dallas and the Vdara in Las Vegas.

Unfortunately the interviewer gave the impression that ray tracing was a new technique of Dr Reinhart’s, a solution now being applied to buildings  that is somehow sufficient to solve the ongoing problem of reflected light pollution. Of course ray tracing has  been around for decades, and Radiance for over 15 years.  DIVA, along with a number of other interfaces such as OpenStudio and LadyBug, provides access to the much more complex underlying software engine that they all have in common.

The key feature of Radiance is that it is one of a few physically accurate rendering engines, where most architectural rendering tools shortcut reality in order to quickly generate beautifully lighted scenes.  There are, however, two fundamental flaws in the ray-tracing approach used by Radiance.  First, the power, accuracy and detail of the model lends itself to analysis rather than design.  Its one thing to model a building after it has been designed and built and then figure out how much window film treatment is required to reduce the resulting to tolerable levels.  Its another to provide a fleet footed capability to help design the building right in the first place.  We need tools that help designers as the form is being created, not after.

And that leads to the second flaw, Radiance analyzes form but doesn’t require actively thinking about it.  The problems in these new buildings are all about the geometry and designers need to understand the effect of the geometry directly.  Radiance generates random vectors from a source or observation point and follows them to see where they happen to end up.  Simple and easy to generate when you know nothing about your environment, this approach makes no attempt to use underlying geometry to guide the analysis process.  As a result it has to work harder and provides little information to the designer to increase his or her understanding of problems in the underlying geometry.

Returning to the example of the University Of Maryland Ellipse I recently studied for HDR Architects in Maryland, I did much of this analysis using the solar vector generator in Heliotrope guided by my understanding of the geometry I was examining.  Essentially I used my knowledge of the building geometry to construct a custom ray-tracing of the specific problem occurring there.  I later ran a Radiance study of the same space and compared the two results over an annual period.  Notice the green dots speckled through the Radiance output?  Those dots represent points at which the random ray generation algorithm concluded an insignificant amount of energy was falling on that location over the annual period. The difference between a green dot and an neighboring red one is purely the result of the random number generation and an insufficient number of test iterations run to smooth out the results?   How many runs would be sufficient?   That is unknowable ahead of time, just more than the number I ran.

Annual irradiance study using Radiance and the DIVA interface.  The green dots mixed in with red in the center of the image are random anomalies.

Annual irradiance study using Radiance and the DIVA interface. The green dots mixed in with red in the center of the image are random anomalies.

The directed ray-tracing approach generated smoother and more consistent results.  Both studies say the space below is hot.  Radiance gave energy numbers as well, however the quality of those numbers is quite suspect since I had no real materials data to base the window reflectivity upon.  Furthermore, the directed ray-tracing approach gave a much more precise understanding of where and the hot zone would occur and how it would travel over the space during the day.

Annual reflection study using custom raytracing provides a smoother and clearer assessment of the hot zone coverage without the anomalies.

Annual reflection study using custom ray-tracing provides a smoother and clearer assessment of the hot zone coverage without the anomalies.

We can do even better, however, by projecting the actual geometry along the ray paths to understand how it obscures or reflects the sun.  In the physical world a Heliodon does exactly this, taking the surrounding features and projecting them onto a spherical dome surface that provides an immediate understanding of when an observer will see the sun throughout the year.  It eliminates the need for the hour by hour animations typically used and described in Reinhart’s interview.  I developed a virtual Heliodon component in the new version of Heliotrope which provides the same capability.   That component incorporates a proprietary spherical geometry engine which makes it blazingly efficient.  So efficient that it is possible to project urban scale geometry and update it in real time.  Below is an image of solar access at a point in one of Portland’s downtown open pedestrian plazas, with the surrounding building geometry projected onto it.

This virtual Heliodon placed in Director's Park in downtown Portland, illustrates how urban geometry can be projected onto a spherical element giving a bird's eye view of annual shade coverage at a point in the square, much in the same way that hand-drawn Pilkington sun calculators were used in the past.

This virtual Heliodon placed in Director’s Park in downtown Portland, illustrates how urban geometry can be projected onto a spherical element giving a bird’s eye view of annual shade coverage at a point in the square.  If flattened to two-dimensions, this projection is effectively the same as the hand-drawn Pilkington sun calculators used in the past.

Dr Reinhart brought up an additional example of a problem they had recently studied using traditional ray-tracing that might have benefited from a geometric projection approach, a recent analysis they performed for a airport where a solar pv panel was causing reflected glare issues into the control tower.  Using geometric projections we can answer the question of when the sun will be reflected into the observer’s field of view instead of using animation studies and provide a quick, clear and accurate understanding of the overall design situation.  Please return next week and I will work the details up for that example in the next post.

Advertisements
No comments yet

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: