|
|
|
CURTAIN WALL PERFORMANCES: LABORATORY
VS. REAL BUILDING |
PDF
version |
|
|
Written
By Raymond Ting
|
|
|
I.
INTRODUCTION
In designing a curtain wall system, in addition
to the appearance parameter, there are three major
performance parameters to be considered as listed
below:
1. Watertightness
2. Structural Safety
3. Energy Efficiency
In designing a new building, a building life of
at least 40 years is normally considered. Therefore,
it is apparent that the above three major performance
parameters must be of long-term in nature. The art
of devising a valid test method for evaluating a
long-term performance parameter is a very complicated
and challenging task, especially when one parameter
is affected by the other parameter. A few examples
are listed below:
1. Structural failure caused by rusting bolts and/or
connection clips due to water leakage problem.
2. Significant loss of energy efficiency due to
the wetting of insulation caused by water leakage
problem.
3. Loss of energy efficiency due to loosening of
air seal caused by structural movements.
4. Water leakage due to failure of watertightness
system caused by structural movements and thus leading
to the chain reactions of Item 1 and/or Item 2 above.
The main objective of conducting a mock-up test
is to use the test result for predicting the long-term
performance of the curtain wall on the real building.
Based on this objective, the purpose of this paper
is to discuss and to recommend modifications to
the commonly used test methods and procedures in
the industry.
II. WATERTIGHTNESS PERFORMANCE TEST
The commonly used watertightness test methods and
procedures in the industry are listed below:
1 Static Water Test (ASTM E331)
2. Dynamic Water Test (AAMA 501)
3. Structural Tests
a. ASTM E330, Negative & Positive
Wind Load Test (one cycle).
b. Vertical Interfloor Deflection
Tests (3 cycles).
c. Story Drift Tests (3 cycles).
4. Static Water Test (ASTM E331)
During each stage of the water test, if water leakage
occurs, it is permissible to repair and to retest
till it passes the test.
Attempting to use the short-term laboratory test
for the evaluation of the long- term performance
of the building would necessitate valid simulations
in the aspects of natural phenomena, field workmanship,
and exposure/aging effects. The laboratory results
would fail to reflect the real building performance
if either one of the following two conditions happens:
1. There was no water leakage in the test while
water leakage happened in the real building. This
has been the unsolved problem in the industry for
a long, long time. Apparently, the test methods
and procedures are inadequate to simulate the natural
phenomena, the field workmanship, and the exposure/aging
effects. Therefore, the purpose of the mock-up test
has been greatly reduced to serve for devising a
proper repair method.
2. There was water leakage in the test while no
water leakage happened in the real building. This
condition is most likely caused by wrong simulations
of the natural phenomena.
The logic for the three aspects of simulations
are discussed as follows:
A. Simulation of Natural Phenomena
Water leakage problem always happens on the windward
wall. Therefore, the wind and the rainwater behaviors
on the windward wall are logically explained below:
Amount of Water: Rainwater on the wall will accumulate
as it runs down along the wall surface, therefore,
the lower the level, the larger the amount of water
running on the wall.
Phenomena of Wind: When the air mass in the wind
impacts the wall, it must be rebound and diverted
around the wall. Due to this air mass rebounding
behavior, the majority of the raindrops in the air
mass would be carried away from the wall surface,
therefore, the stronger the wind, the lesser the
amount of water on the wall. This phenomenon can
be readily verified by the fact that the observed
amount of running water on the wall during a dynamic
water test is much lesser than that during a static
water test. It is also easy to verify that the observed
amount of running water on the wall of a high-rise
during a rainstorm is similar to that during a dynamic
water test.
Based on the above analyses of the natural phenomena,
it can be concluded that the Static Water Test (ASTM
E331) without the rebounding air mass behavior is
only suitable to be conducted at low differential
air pressure (low wind condition). This author recommends
using a differential air pressure of 1.6 psf (25-mph
wind) for the Static Water Test (ASTM E331). However,
because almost all water leakage problems happen
in strong wind condition, the Static Water Test
may be eliminated if simulating the natural phenomena
is the only purpose for conducting the test.
For the Dynamic Water Test (AAMA 501), AAMA recommends
to use 20% of the maximum positive design wind pressure
with a maximum of 12 psf as the test pressure. This
is reasonable since the stronger the wind, the lesser
the amount of water on the wall. However, the testing
duration of 15 minutes is questionable since the
duration of critical rain and wind combination during
a typical rainstorm could last much longer than
15 minutes. It is recommended to extend the test
duration to 60 minutes.
B. Simulation of Field Workmanship
It is well recognized that the workmanship on a
test mock-up is much better than that on the real
building. In most curtain wall systems, there are
some critical seal locations. If poor workmanship
is executed at one of the critical seal locations,
it would very likely result in water leakage during
the test. On the contrary, if poor workmanship is
executed at a non-critical seal location, it would
very likely pass the test. Therefore, simulating
poor workmanship by predetermined location is not
feasible. Therefore, most of the laboratory tests
conducted to date failed to simulate the field workmanship
and thus, supervising the field quality control
becomes the major function of a curtain wall consultant.
C. Simulation of Exposure/Aging Effects
The factors causing the exposure/aging effects are
stated as follows:
1. Sealant material degradation due to UV-light
exposure.
2. Dislocation of sealant line or sealant adhesive
failure due to stress fatigue caused by repeated
cycles of structural movements. Structural movements
include thermal movements, vertical inter-floor
deflections, inter-floor story drifts, and wind
load deflection cycles.
The commonly used structural tests on the mock-up
are normally utilized as the simulation of the aging
effects. However, the UV-light effect is not included
and the number of structural load cycles is inadequate
according to this writer's opinion. Another factor
as explained below is not considered. Most curtain
wall systems use curable caulking such as silicone
to form the critical seals. The curable caulking
is quick to form a skin while the body of the sealant
normally requires a long time to solidify completely.
The tests are normally conducted with the seals
in the partially cured stage, which is very flexible
and can tolerate high degree of displacement or
distortion without breakage. Therefore, it becomes
invalid to use the structural tests for the aging
simulation. It is believed that story drift is more
critical for sealant distortion - thus, it is more
effective for aging simulation.
Summarizing from the above discussions, it is recommended
to use the following procedures for conducting the
water infiltration test.
1. Dynamic Water Test (AAMA 501): Use an equivalent
differential air pressure of 20% of the maximum
positive wind load for structural design with a
maximum of 12 psf. Use a test duration of 60 minutes.
2. Structural Tests
a. Positive and Negative Wind
Loads (one cycle).
b. Vertical Inter-floor Deflection
(3 cycles).
c. Story Drift at L/100 (50
cycles).
3. Repeat Dynamic Water Test as stated in Item 1.
The relationship between the system design principle
and the proper test method is further discussed
below:
1. System with Perfect Seal Principle (known as
Wet Seal System): In this type of system, any imperfect
seal location would become the water leakage source;
therefore, the nature of the test method, static
or dynamic, would have little or no effect on the
test result.
2. Pressure Equalized System with Delayed Drainage:
In this type of system, the intersections of the
horizontal and vertical members are normally the
critical seal locations. As long as the critical
seals are perfect, the nature of the test method,
static or dynamic, has little effect on the test
result except the following conditions: Depending
on the rain screen and the drainage design, the
dynamic test may be more critical in causing the
accumulation of water in the system. Regardless
of the test method, both the test pressure and the
test duration have significant effect on the outcome
of the test due to the limitation of gutter capacity
3. Pressure Equalized System with Instantaneous
Drainage: This type of system generally eliminates
the need of critical seal and the rebounding air
mass in the dynamic test is utilized to accomplish
the pressure equalization. Due to the instantaneous
drainage design, it is much less sensitive to the
test pressure or the test duration. This system
is most likely to be successful in the dynamic test.
However, in the static test without the effect of
the rebounding air mass, if the nozzle is located
in such a way that the water is constantly spraying
at the corner seam of the intersection point of
the horizontal and the vertical members, with the
steady suction in the static test, small amount
of water could infiltrate due to the capillary action
along the corner seam. If the system passed the
dynamic test and failed the static test, it is very
likely to create the strange situation of leakage
in the test and no leakage in the real building.
Due to this reason, it is recommended to use the
dynamic water test only unless the static water
test is conducted at a low differential air pressure.
III. STRUCTURAL TESTS
The design theory for the structural performance
include the determination of forces and the limitations
of material responses to the forces. Normally the
determination of forces is dictated by the governing
Building Code and the limitations of material responses
to the forces are dictated by the industrial associations
such as AISC and AAMA. These two items are separately
discussed below: 1. The Determination of Forces:
The forces considered in a curtain wall design include
thermal load, wind load, and seismic load. The thermal
load effects can normally be minimized or eliminated
by design. Therefore, the consideration of thermal
load in curtain wall design is normally not specified
in the Building Code. The determinations of wind
and seismic forces are clearly defined in the Building
Codes. However, for major high-rise projects, the
wind tunnel test is commonly used for more reliable
wind load determination.
2. The Limitations of Material Responses to the
Forces: There are three reasons for setting the
limitations as discussed below:
a. Structural Safety: Several examples are stated
as follows. Allowable Stresses or Safety Factors
for structural components; Gap Tolerance for vertical
inter-floor deflection; Strain Tolerance for inter-floor
story drift. The above items can be verified by
structural analyses and/or various structural test
methods. However, the effects of component corrosion
due to water leakage cannot be simulated. Lacking
the permanent solution to the water leakage problem,
the industry relies more on the corrosion resistant
material such as aluminum or stainless steel.
b. Preventing Adverse Effects On Other Performance
Parameter: It is well know that cycles of sealant
stress and strain due to structural movements could
cause sealant failure or sealant disposition. Therefore,
certain limitations on deflection or displacement
must be established to prevent sealant failure or
sealant disposition. If the system design relies
on the stiffness of the structure to minimize the
sealant line strain (Case 1 Design), then, it is
desirable to limit the deflection or displacement
as much as economically feasible. If the system
design utilizes space or releases stiffness or sealant
adhesion to tolerate the component relative movements
or to minimize sealant strain (Case 2 Design), then,
apparently the deflection or displacement limitations
can be greatly relaxed. Applying the limitations
for Case 1 Design on Case 2 Design would adversely
affect the performance of Case 2 Design and vice
versa. For this reason, the deflection or displacement
limitations should not be applied uniformly to all
systems. Other type of consideration is to prevent
deflection or displacement induced interference
with other interior components such as ceiling and
partition wall. However, this type of interference
can be easily prevented by a slip joint design and
no specific deflection limitation on the curtain
wall is required.
c. Preventing Uneasy Feeling of Occupants: The widely
used deflection limitations in the industry seem
to be the product of the mixed concern of Items
b and c. For example, the allowable wind load deflection
for glass or aluminum plate is L/60. The allowable
wind load deflection for the curtain wall frame
is L/175 and sometimes with an upper limit of 3/4".
The upper limit of 3/4" seems to be adopted from
the floor design. It is recommended to eliminate
this upper limit due to the reasons explained as
follows. People walking on the floor can directly
feel the floor deflection, therefore, placing an
upper limit on the floor deflection is reasonable.
The wind induced deflections on the curtain wall
are seldom noticeable by human eyes due to the rare
occurrence of extreme wind and short duration, therefore,
no upper limit is necessary and in fact, the limitation
of L/175 can be further relaxed unless the factor
in Item b is an overriding concern.
IV. THERMAL PERFORMANCE TEST
The energy efficiency of a curtain wall system is
mainly affected by the following four factors:
1. The U-Value of the vision glass.
2. The U-Value of the spandrel panel with the backside
insulation.
3. The U-Value of the curtain wall supporting members.
4. The air leakage rate of the windward wall.
The above Items 1 and 2 constitute the major exposed
surface area of the curtain wall, therefore, they
should be considered as the dominating factor in
the energy efficiency design of the curtain wall
system. Normally, the exposed surface area of the
curtain wall supporting members is a small percentage
of the curtain wall surface, therefore, its effect
on the energy should be relatively small. For Item
4 above, an allowable air leakage rate of 0.06 cfm
per square foot of wall area is commonly specified
in the industry. Unless massive air seal failures
occur on the real building, the air leakage should
have minor impact on the energy efficiency.
The thermal performance and energy efficiency can
be calculated and/or tested. However, if wetting
of the insulation due to water leakage occurs, the
energy efficiency of the curtain wall will be significantly
reduced. Since both the extent of water leakage
problem and its impact on energy cannot be predicted,
the long-term energy efficiency must be relied on
the elimination of water leakage.
V. CONCLUSIONS
1. If the objective of a water test is to evaluate
the long-term watertightness performance of the
real building, then, the water test must be able
to simulate the natural rainstorm phenomena, the
field workmanship, and the exposure/aging effects.
Based on the test methods commonly used in the industry,
it is recommended to exclusively use the dynamic
water test (AAMA 501) for simulating the natural
phenomena of rainstorms. To simulate the effects
of field workmanship and the exposure/aging factor,
this writer recommends to use repeated cycles of
story drift tests in addition to the normal cycles
of wind load and inter-floor deflection tests. The
field water test procedures (AAMA 501) should be
modified in accordance to the system design principle.
For a wet seal system, no modification on the stated
procedures is needed. For a pressure equalized open
joint system, purposely shooting a great amount
of water into the joint gap is not proper since
not only it can not represent the effects of natural
rainstorms but also it destroys the design function
of the rain screen for repelling the majority of
the water from entering the joint. When the field
water test is conducted in the above improper manner,
water infiltration could happen due to localized
overflow of water inside the joint. This would produce
the strange situation of leakage in field test and
no leakage in the real building. If this field test
is to be applied to the open joint system, then,
the nozzle should be constantly moving along the
joint without complete stop at any location and
without intentionally seeking for the gap.
2. In respect to the structural performance, the
method of analyses and the safety factors or the
stress limitations commonly used in the industry
are adequate. However, to assure long-term structural
performance, the stress fatigue problem and the
structural component corrosion due to water leakage
must be minimized. The deflection or displacement
limitations for the purpose of minimizing the sealant
stress or strain should be evaluated in accordance
with the system design principle.
3. In respect to the energy efficiency, the U-Values
of the vision panel and the spandrel panel should
be the primary consideration. However, the impact
of water leakage could be so significant to render
the theoretical energy calculation meaningless.
To encourage energy savings, giving tax incentives
to buildings designed with adequate calculated energy
efficiency with the assumption of no water leakage
impact is not productive. It is recommended that
tax incentives should be given with the combination
of calculated energy efficiency and the proof of
sustained long-term performance. In summary, long-term
building energy performance must be relied on the
elimination of water leakage problem.
*This article was published by Glass
Magazine in October 2002 |
|
|
|
|
|
|
|
|
|
|
|
|
|
