ISSUED:                   29 January, 2004 
PHASE I DUE:           17 February, 2004
PHASE II DUE:          24 February, 2004

Exercise 1:

Comparing a physical model to the real world. 

This Proof-of-Concept Exercise has two major objectives.  The first is to model an existing space and have the lighting measurements of the model favorably compare to the lighting measurements of the actual space.  This Phase I comparison is a learning curve experience designed to fine tune your modeling skills (should be a piece of cake, right?).  Phase II will be the design of a major change to improve the daylight character of your existing space.  This improvement can be tested using the daylighting model, giving you a chance to play "what if" with the design. 

1.  The space should have ample contact with daylight.  It also helps if the room is relatively simple.  Complex geometry, furnishings, and fenestration can make your modeling task more difficult.

2.  It is helpful if the space is interesting.  You will spend a fair amount of time looking at, measuring, photographing and modeling this room.  Many of my students end up putting this work in their portfolios.  If you pick a dull space you will grow tired of it.

3.  Access to architectural drawings can be quite helpful though not absolutely necessary.

4.  The space should be accessible for lighting measurements and photographs.  To take these interior lighting measurements you must be able to turn the electric lights off for a brief period (30 minutes).  If the windows have blinds or drapes, you should be able to retract these because they pose some difficulty in modeling.

If you have questions regarding the suitability of a candidate space just ask me.

Visit the space and document its dimensions and surface properties.  Determine the reflectance of all interior surfaces and the transmittance of any glazing.  You can determine reflectance by using a photometer (as outlined in our handouts) or by comparing the real space surfaces directly to candidate materials for the model. 

Build a model of the space that is large enough to use with a meter and a camera but small enough to carry to the site; a good scale is usually between 1/2" = 1'-0" and 1" = 1'-0".  The model should be simple, not painstakingly detailed.  It needs to be accurate only in terms of room proportions, opening sizes, and reflectance and transmission values (+ or - 10%).  Beware that some model materials such as Fomecor will transmit light and must be covered with chipboard, aluminum foil, or paint to become completely opaque.  It is really best to avoid Fomecor altogether.  If immediately adjacent surfaces such as roofs and pavement reflect enough light into the space to affect your measurements, they should be included in your model and have an accurate reflectance.  Note the plan orientation of the space in relation to the north.  Lastly, plan ahead for hatches in the floor and/or walls for taking measurements and photographs.

Note:  If you decide not to glaze the openings you will need to determine a factor to apply to your measurements.

Draw a plan of the space at 1/4" = 1'-0".  If the space has vertical planes that are visually important (for instance, blackboards or gallery hanging space) draw elevations of these at the same scale.  These drawings should be reproducible so that you can make “blank” copies on which to plot your various lighting measurements.

 1.  Stand outdoors and measure the illuminance in a horizontal plane below the overcast sky.  Be sure that you do not stand over or next to the sensor during this reading as you will block its view of the sky (it has a 180º fisheye field of view).  Record your reading and note the time of day.

2.   Remaining outdoors, orient the model the same direction as the real space and take measurements that will enable you to plot an isolux drawing of incident footcandles in plan (horizontal footcandles).  An isolux drawing is a contour map of the lighting level in a given plane.  Choose the plane you feel is most important for the activity in the space, such as the level of a desk top.  Take readings in a grid spacing that will allow you to plot contours in 10 fc increments.  During the course of your interior measurements you should take periodic measurements of exterior illuminance in a horizontal plane.  If these measurements remain relatively stable (equal to your original exterior measurement) then all is well.  If they vary widely then the sky is unstable and you should record these variations in exterior illuminance for use in later calculations. Under unstable skies, each interior measurement would ideally be matched to a corresponding exterior reading but this is often not practical.  How often you take the exterior readings is a judgment call based on sky stability.  If you have chosen a space that has an important vertical surface, take measurements for this surface also. 

3.  Photograph the model, taking at least two photos from opposite corners of the space.

4.   Repeat the measuring and photographing in the real space; be sure to turn off the electrical lighting before proceeding.  Photographs should be taken from the same viewpoints as in the model to permit direct comparison.  I tend to take quite a few photographs for several reasons.  First, it is usually difficult or impossible to go back later and repeat the process (you will have changed the model or it will have returned to the earth or whatever).  Second, there are several people in your group and many of you will desire a copy.  Finally, I often "bracket" my exposures by exposing a frame at 1 stop above and 1 stop below the indicated exposure.  This isn't absolutely necessary and depends on the trust you place in your exposure meter.  I think of it a welcome insurance particularly when using a camera with which I am not familiar. These strategies are particularly important with film cameras. Digital cameras allow you access exposures 'on the fly'.

Please take specific pains to establish close correspondence between the camera positions in the model and in the real space.  Knowing that I cannot trust my memory on this matter, I routinely make a sketch of the camera’s viewfinder noting what appears at the center and each corner of the frame.

5.  Compare the measurements obtained in the model with those from the space itself.  If there is a significant difference between them you will have to find the cause of the discrepancy or create a factor to account for the difference.  For instance, one obvious factor would be an adjustment for glazing transmittance when the real space has glazing and the model does not.  In such a case you reduce all readings in the model by the average transmission of the glazing (approximately 75% for double clear glass).  A similar adjustment might be required if the windows are soiled.  Please explain any changes or factoring that you do.  For this exercise achieving an average deviation of model values from the real values of 20% is sufficient.

1.  Slide or digital photographs of the model and the actual space from common vantage points under both cloudy and clear sky conditions. These paired photographs should be directly comparable in framing and exposure.

2.  Plan (and section) drawings showing isolux contours of footcandles under cloudy sky conditions.

3.  Plan (and section) drawings showing isolux contours of the model's daylight factors for a cloudy day.

4.  A brief written description of your findings for this phase of the project.