www.northwestlighting.com

Anatomy of a Mock-Up

Beginning of construction of the North Clackamas mock-up
(see larger photo)

Introduction:

A full scale mock-up of a prototypical classroom for the North Clackamas High School was constructed to evaluate the proposed architectural design and lighting strategies. Both daylight and electric lighting were evaluated, individually and as an integrated approach.

This project was conducted in the mock-up facility at the Lighting Design Lab, Seattle WA. Because of the classroom's complex ceiling and its interior and exterior lightshelves, this was a much more elaborate mock-up than standard for the Lab.

Thus the mock-up was accomplished as a collaboration of the Lighting De-sign Lab and Portland General Electric (PGE). The Lab provided its standard mock-up services for the project; Portland General Electric funded the incremental costs (labor and materials) of constructing the inclined ceiling plane and lightshelves.

This report documents the construction, evaluation and costs of the mock-up.

MOCK-UP DESCRIPTION:

The classroom mock-up was constructed in the east side of the Lab's mock-up facility with the following details:

Size: The classroom was modeled as 30'-5"deep (from window wall to backwall) by 26' wide (length of the window wall). Note that the depth was slightly less than the actual classroom; this was necessary so that the moveable ceiling would clear the HVAC ducts in the back of the mock-up room.

Ceiling: The design of the proposed classroom ceiling incorporates a suspended "cloud" portion that covers most of the room but is pulled back from the walls by 2 to 4 feet on all sides. This part of the ceiling is sloped and drops to a low point about 2/3 of the wayback from the window wall. (see drawing) This sloped portion of the ceiling was mocked-up with white "blackout" fabric stretched over a wood frame. The light reflectance of the blackout fabric was measured at 71% which was close to that measured for the proposed standard white ceiling tile (68%) and within the tolerance of the mock-up. This suspended "cloud" portion of the classroom ceiling was larger than our moveable ceiling. So a wood frame extension to our moveable ceilingwas constructed to represent this larger area.

Window wall: The south facing Lab windows were masked down to simulate the proposed window area for the classroom. To ensure appropriate daylight distribution from the window, we aligned the major horizontal mullion in the mock-up space with the planned mullion in the classroom. The existing mullion in the mock-up space is centered at 9'-7", and the planned mullion for the classroom is centered at 8'-3". So we increased the floor to ceiling height by 14" to accomodate this.

We kept the window openings as designed above and below this mullion. So in effect, the extra 14" occured between the floor and the lower window sill. All horizontal light level measurements were taken at 44" above floor height. (Note these measurements at 44" represented the light levels that would occur at 30" desk height in the room.) The ceiling was constructed so that it could be lowered to evaluate the spatial feeling of the room at the actual ceiling height. The electric lighting was suspended from the moveable ceiling so that it moved with the ceiling.

LDL installers mount perforated metal panels. Exterior lightshelves are visible behind them.
(see larger photo)

Lightshelves: The interior lightshelf was constructed of perforated metal (1/16" diameter holes on 1/8" staggered centers; 23% open area). The inside edge of the lightshelf was a wooden cove constructed per the architectural drawings to hold the electric lighting. The exterior lightshelf was constructed of foamcore. Foamcore has a relatively high reflectance and could over predict the interior light levels if the actual material is considerably less reflective. The exterior overhang above the clerestory window was not modeled.

Site: The mock-up window wall faces approximately due south. There is no control over exterior shading to the mockup room. Existing 2-3 story buildings across the street from the site partially block the lower portion of the sky dome. There is also a deciduous tree that shades the western half of the mock-up room. This tree was shading the west portion of the classroom as it was mocked-up. So we can expect measured daylight levels to be lower than for an unshaded window at the proposed school. Measurements under direct sun conditions were taken only in the east portion of the room during morning hours when most of the lightself received full sun.

Final appearance of the North Clackamas mock-up.
(see larger photo)

EVALUATION:

The mock-up was evaluated for five things: 1) spatial qualities, 2) daylight levels, 3) electric lighting design, 4) lighting quality, and 5) implications for lighting controls:

Spatial qualities: Both the architects and the school representatives were very pleased with the spatial qualities of the classroom. The sloped ceiling creates an interesting effect that reduces the scale of the room to "human height" in the center while leaving height at the perimeter for clerestory glazing. Note that with the current HVAC design this sloped ceiling no longer serves the function of enclosing HVAC ducts in the center of the room while allowing full ceiling height at the perimeter window area. So the ceiling shape must be evaluated for its spatial qualities since it no longer serves a clear daylighting function.

There was some discussion of the different heights of the lightshelves at the perimeter window and back wall. The back wall fixture is lower (7'-2") than the perimeter window lightshelf (8'-0"). There was some concern that the back wall shelf is too low for a school installation (kids will tend to hang on it!). However, it allowed more even light distribution from the cove mounted electric lighting. (See below.)

Daylight levels: Daylight measurements were made under both overcast and clear sky conditions. The clear sky measurements were taken with both the full window open and just the clerestory portion open to differentiate the contribution of the clerestory glazing and evaluate the impact of closing blinds in the lower window. The following conditions prevailed during the measure-ments:

Clear sky (about 7000 footcandles total exterior horizontal) daylight values ranged from 190 footcandles near the window to 15 footcandles at the back of the room for the full window open.They reduced to about 30 footcandles near the window to about 9 foot-candles at the rear of the room when the bottom window was blocked off. Overcast sky (about 2600 footcandles total exterior horizontal) values ranged from about 114 footcandles near the window to about 5 foot-candles at the rear of the room.

The attached illustrations show the measured values in both footcandles and daylight factors for clear and overcast sky conditions with the bottom window open and covered. (Day-light factors are the interior light levels normalized by the horizontal exterior light level. Here they are reported as a percent of the exterior level.)

Note that daylight factors are surprisingly similar for the clear andovercast sky conditions. This, how-ever,may be somewhat fortuitous be-causethe tree which shades the westsection of glazing has a greater impacton the overcast data than on the directsun data (taken at morning hourswhen the entire lightshelf receives fullsun). Also note that covering the bot-tomwindow gives the most even illu-minationacross the space,dramatically reducing the levels nearthe window but causing less of a re-ductionat the rear of the space.

The suspended ceiling (cloud) stops short of the lightshelf and it appeared from viewing the mock-up that this compromises the function of the lightshelf and the indirect electric lighting mounted in it. Too much of the daylight is "thrown away" into the recess above the cloud. After viewing the mock-up we recommended extendingthe "cloud" over at least part of the interior lightshelf.

There are times when direct sun will penetrate through the clerestory glazing into the classroom. This was not observed in the mock-up because of the high summer sun angles at the time of viewing. A shading mask was done earlier (at the time of the scale model study) to show the possible times when direct sun can penetrate the clerestory glazing. Blackout shades proposed for the clerestory glazing can be used to eliminate this potential serious glare problem. Duringthe mock-up viewing, there was some discussion of whether these shades are necessary for AV purposes.The dual function of these shades for both AV and direct sun control should be evaluated in this decision.

The original classroom design shows computers lining the back wall of the room with the monitors facing the window wall. We brought a monitor into the mock-up to demonstratethe glare problems created with monitors in this orientation. We suggested creating small "alcoves" along the back wall allowing the computers to be lined up back-to-back with the monitor screens perpendicular to the window wall.

Barbara Erwine brought up the possibilityof reducing the visible transmission of the glazing below the lightshelf to minimize glare from the lower glazing.

Electric lighting: Two different electric lighting schemes were evaluated: 1) the pro-posed scheme with side lighting mounted in the lightshelves at the window wall and back wall and 2)pendants running parallel to the window wall and hung under the sloped ceiling about 1/3 and 2/3 of the way back from the window wall (per the architectural drawings). In addition, two different fixture types were evaluatedfor each of these schemes as noted below:

1) Lightshelf mounted scheme:

2) Pendants:

Electric lighting design: The group who viewed the mock-up quickly agreed that they preferred the indirect lighting scheme with luminaires mounted in the lightshelf better than the pendant schemes. It was less distracting in the space and gave less of a hot spot on the ceiling.However, the Eliptipar luminaires specified for the project did not perform well in the lightshelf cove. Their reflectors were too visible and caused a source of glare. They also did not illuminate the sloped ceiling as evenly as expected. The Lightolier Lyteflood39 luminaires could be adjusted to perform better for glare and even illumination but this was accomplished at the expense of efficiency (some of the light had to be spilled into the cove).They are also not designed to be hidden in a cove like this and were only mocked up to give an idea of how another type of fixture might solve the glare problem.

Clearly another alternative fixture must be identified and designed into the cove. This may involve a redesign of the lightshelf and cove to efficiently and evenly illuminate the sloped ceiling. For both cove mounted indirect luminaires there was an obvious luminance gradient across the ceilingplane. Luminance levels dropped significantly at the low part of the ceil-ing. As expected, the lower cove for the back wall lightshelf provided more even electric lighting illumination on the ceiling than the higher cove at the window wall.

Quick measurements showed that the electric lighting could provide about 30 footcandles of light fairly evenly across the space which the group agreed seemed sufficient for classroom work.

Lighting quality: Overall lighting quality was evaluated empirically in the space and documentedwith slides. All participants were pleased with the quality of the indirect use of daylight and electric light and how they worked together. The ceiling gradient was noted as a problem however.

All participants felt that with an indirect scheme like this, they could be comfortable with lower footcandle levels in the space. They felt that a level of 30 footcandles was comfort-able. 30 footcandles is within the range of the IES recommendations for a classroom. The lighting designer will check to see if there are any OregonState education requirements for footcandle levels that supersede the IES recommendations. There was also an acknowledgment that during the daytime, daylight would increase lightlevels.

We noted the need for blackout shades on the clerestory windows for AV purposes.

Implications for Lighting Controls: There was an animated and informa-tive discussion of potential lighting controls. Some of the options discussedwere:·Daylight dimming at the perimeter. Daylight switching at the perimeter and daylight dimming at the back wall. (Barbara Erwine questioned the cost effectiveness of dimming the back wall luminaires). One photocell in each classroom vs. two roof mounted photocells - one controlling all south facing classrooms and one controlling all north facing classrooms. Occupancy coupled schemes where the back wall lights come on automatically when someone enters the space but the window wall lights have to be turned on manually from a separate switch near the window. (These window wall lights could also be dimmed by a photocell.) All lights would turn off automatically when the space is vacated.

Daylight levels measured in thespace support pursuing daylight controls at least for the window wall luminaires.We discussed both dimming and switching options as noted above. The lighting designers will cost out some of these options and present them to the architects and owners for a decision.

Comparison of Mock-Up to Other Modeling Techniques: The intention of this mock-up was to get additional information about the space and what lighting quality and quantity to expect. It was not a research project designed to compare with other modeling techniques used.Several deviations from the actual design have already been noted above. However, it is compelling to try to relate the information from this mock-up to the LightScape and daylight model data previously generated. So this section will look at the mock-up in relation to the other approaches.

Qualitative Evaluation: Interior classroom images are available for the LightScape simulation, the daylighting model and the mock-up. Enclosed in this report are a set of slides documenting the mock-up and comparison images for the LightScape simulation and the mock-up. The school administration staff and architectural team who attended the mock-up were very enthusiastic about the ability to be "in" the space to evaluate it. Further interrogation of them would provide the best measure of the qualitative benefits of the different techniques.

Quantitative Evaluation: The following table shows daylightfactor measurements for the daylighting scale model and the mock-up for overcast sky conditions. A blank column has been left to include the estimates from the LightScape simulation. The daylight model values are the values measured in the scale model's south facing classroom, first floor with the corridor daylighting apertures covered. These values are shown in the attached drawing provided by the architect andwere corrected for a glazing transmissionof 68% (the value measured for the mock-up). The mock-up values were interpolated for the locations of measurements in the scale model.

Above (from left): Adam Griffen, LDL Stage Technician; Connie Reynolds,Interface Engineering Lighting Designer; Melanie Taylor, LDL CommercialLighting Specialist; Nacho Bravo, LDL Electrical Installer; (headless) Barbara Erwine, LDL Daylighting Specialist

Below: Final appearance of the North Clackamas mock-up, showing daylighting contribution only

 

 

COMPARISON OF MEASUREMENTS

(DAYLIGHT FACTORS, percent of exterior)

Ft. from window

MOCK-UP
Daylight Factor*

MODEL
Daylight Factor*

3

4.20

9.3

9

1.04

3.7

15

0.42

1.6

21

0.28

.9

27

0.20

.5

*interpolated for locations in model

*corrected for 68% glazing transmission

The scale model values are higher than the corresponding measurements in the mock-up by a factor of 2.2 to 3.8. Measurements in the center of the room show the greatest deviation. A more thorough analysis would be required to explain the change across the room; however, the following factors can explain the overall trend of the scale model values being higher than the mock-up measurements:

The tree which shaded the mock-up window and blocked the view of the sky could account for a factor of 2 difference between the readings, since it blocked the view of half of the sky dome. The lower window was redesigned in the time between the model study and the mock-up construction. The scale model had more lower glazing than the mock-up. Because of the presence of the tree in the west half of the window, the mock-up measurements were not taken in the center of the space but at 7'-4" from the east wall to get the best view out an unobstructed window. The model measurements on the other hand were taken down the centerline of the room which gives them a better view of the window and thus a higher potential reading.

MOCK-UP COST:

The mock-up was a collaboration of the Lighting Design Lab and PGE EarthSmart Program with staff support from BOORA Architects and Interface Engineering. Using the existing facilities at the Lighting Design Lab, the mock-up was accomplished at a minimal cost with high value for the expenses. Individual contributions are noted below:

Lighting Design Lab Contribution:

PGE Contribution:

 

 Download this report as a PDF file.

[Home] [Comments]

Northwest Lighting ©1999 Lighting Design Lab
Hosted by inter.Light