IMPROVING THERMAL PERFORMANCE IN STRUCTURAL FENESTRATION PRODUCTS

ABSTRACT
The manufacturers of fenestration products (windows and doors) for use in a building facade
have many materials to choose from to produce their products. This paper discusses the use of
aluminum with a structural thermal barrier as the material of choice for fenestration worldwide.
It will offer fundamental facts about thermal barriers for commercial and residential applications.
The structural longevity and design flexibility of such products is discussed. The cost of
producing thermal barrier fenestration and processing techniques is presented.
Data will be provided which illustrates the ability to meet energy codes required for the
architectural building industry. Comparison studies are also given on various types of systems
that can be used for insulating aluminum windows and doors.

INTRODUCTION
Even today as we are securely into the twenty-first century the time proven attributes of
aluminum make it the unparalleled choice for use in the world’s structural windows. The
properties of aluminum allow for the ease of manufacturing to be combined with reliability of
structural performance. Aluminum windows have proven their worth in harshest of climates
from the frigid cold of the Siberian peninsula to the extreme heat of the many deserts
throughout the world to the some of the most stringent wind load requirements in the hurricane
zones of the tropical climates. While the initial energy requirement for producing aluminum
alloys is high, once the material is created the ability and number of times it can be recycled is
unlimited.

Thermal barrier history, types and properties

With all of its intrinsic accolades, aluminum, when used for fenestration has a shortcoming.
Thermal Conductivity is defined as “a measure of the rate at which heat flows through a
material…. an insulating material is a poor conductor of energy and thus has a low thermal
conductivity”. PVC, wood and aluminum make up 95% of the worlds windows, the thermal
conductivity of Vinyl/PVC is 0.17 W/m*K (1), that of a hardwood and/or maple is 0.16 W/m*K
(1) and the thermal conductivity of aluminum alloys is 160 W/m*K (1). With the impressive
insulating properties of these other materials and ever increasing energy requirements how can
aluminum compete and
maintain its popularity?
The answer comes from a Table 1: The thermal conductivity of materials used in
technology that has evolved commercial fenestration products, taken from NFRC
from its conception over forty 101.
years ago. The Alaska State
Courthouse in the United Aluminum Stainless Steel Polyamide Polyurethane
States was the first 160 W/m*K 14.3 W/m*K 0.30 W/m*K 0.12 W/m*K
documented time that a

Patrick Muessig, is global technical manager for Azon USA Inc., Kalamazoo, Michigan – USA

1
Thermal barrier history, types and properties - cont.
polymer thermal barrier was
inserted into an aluminum
window in a commercial Table 2: Structural properties of thermal barrier materials
application. Those crude
windows served as a Material Property Polyamide Polyurethane
forerunner for today’s thermal Shear 439 g/m^2 1171 g/m^2
barrier market. The engineers Impact Strength 1.0 ft/lbs/in 1.9ft/lbs/in
knew even at that day in age Flexural Strength 17,800 psi 19,000 psi
that with all of the benefits of Shore D Hardness 79/81 84/86
aluminum windows, energy Percent Elongation 8% 64%
efficiency was still the
missing ingredient. The
material used at this time was
a two part polyurethane system. The thermal conductivity of polyurethane is 0.12 W/m*K, that
makes it over a 1,000 times more insulating than aluminum (see Table 1). This adaptation to
the aluminum window now made it an energy efficient design; however, if aluminum windows
with a thermal barrier option were to become common, all the attributes that made aluminum
the material of choice for structural purposes had to be maintained.
It would have to have the tensile and flexural strength properties that would compliment that of
aluminum. It would require high shear strength properties so that a window could resist
deflection and support lateral loads. It would also require the ability to maintain its properties
when exposed to a wide range of temperature differentials. It would have to have a lifetime to
match that of the current aluminum fenestration products.
In the some forty years of thermal barrier design refinement two dominant systems have
prevailed. One is the polyurethane – pour and debridge system and the other is polyamide –
the strip system. Both have their places of dominance in the world. In North America
polyurethane is the material of choice and has over 90% of the thermal barrier market. In
Eastern Europe polyamide has become the choice and has close to the same market share.
Aside from the thermal conductivity other properties vary as well (Table 2).

Thermal barrier comparisons

The processing of the two systems varies as much as
their material properties. The polyamide or strip system Figure 1: Aluminum polyamide –
comes to the manufacturer in set lengths and widths. It strip system thermal barrier profile
is a pre-formed material with a clear adhesive strip on
either side. The polyamide thermal barrier is used as a
connector; two separate pieces of aluminum are slid
onto two polyamide strips that sit on a top of one
another (see Figure 1). Once the two pieces are joined
by the polyamide a knurling wheel is run upon the top
of each strip and it forces the aluminum clamps into the
polyamide strips thus giving the product its shear
strength. After this process the product is heated to
melt the adhesive strips and prevent water infiltration
and help secure the polyamide strips.
The polyurethane or pour and debridge system actually
comes to the manufacturer in a liquid form. A mixing
machine with two tanks, one Iso and one Resin hold
the required materials separate until the time of
processing. A single aluminum extrusion passes through the mixing machine and liquid
polyurethane is poured in the desired cavity size. The material then hardens exothermically into
a solid structural material.

2
Thermal barrier comparisons -cont
The next step is using a mill to ‘debridge’ or remove the
bottom of the aluminum cavity thus creating a thermal Figure 2: Aluminum polyurethane-
pour and debridge thermal barrier
barrier (see Figure 2). The exact cost of the two profile.
systems varies. The polyamide material tends to be
more costly to produce. For either system the total
material cost added to a standard 1.2m x 1.2m
aluminum window is close to 12 Euro. Using two
computer programs one called Therm 5.0 and one
called RESFEN 3.1 – Residential Fenestration the
effects of a thermal barrier can be demonstrated in
window performance and how that translates into
annual energy savings.
Therm 5.0 is a state-of-the-art computer program
developed at Lawrence Berkeley National Laboratories
(LBNL) for use by building component manufacturers,
engineers, architects and others interested in heat
transfer. Using Therm one can model two-dimensional
heat transfer effects in building components, this
analysis allows the evaluation of a product’s energy
efficiency. The program uses a finite element analysis
method that examines every material within a fenestration product and produces a rating of
heat transfer (U-factor) and the solar heat gain effects (SHGC). It is through the use of this
software that the effects of the low thermal conductivity of the thermal barrier materials are
demonstrated.
By importing AutoCAD drawings and assigning material properties, Therm allows you to model
almost any frame profile. Once the material properties are assigned, an Insulating Glass Unit is
imported from another LBNL software program Window 5.0. Window allows the operator to
design an IG unit with any low-e coating, gas fill, or size desired. Next the proper boundary and
environmental conditions are assigned then the program assesses how much, how fast, and

Table 3: Requirements for Document ‘J’ in Scotland

Frame Type With gas or oil central With electric heating, solid fuel central
heating having an efficient heating, or gas or oil central heating
boiler having an inefficient boiler
Aluminum or steel 2.2 W/m*K 2.0 W/m*K
windows
PVC-U or timber 2.0 W/m*K 1.8 W/m*K
windows

where the heat would flow through the profile. The operator is then given a rating for U-frame
and U-edge, which are transferred back to Window 5.0, where the size and type of window are
entered. Once this is accomplished two very important results are derived, the total fenestration
product’s U-Factor and SHGC (Solar Heat Gain Coefficient). These two ratings dominate the
industry and when code officials decide what requirements need to be met these are the two
that are chosen.
Therm provides more insight than just a number rating. It allows the operator to view how the
heat flows through a profile using heat flux magnitude, infrared heat flow, flux vectors and
isotherms. This ability to simulate versus physical testing allows the designer to optimize the
fenestration product including placement, size and material required for thermal barriers before
a single unit is ever built.

3
Thermal barrier comparisons -cont
Figure 3: Current Energy Star Criteria
Since many fenestration manufacturers
now have global operations, it is
imperative that they be able to meet all of
the different and ever changing energy
codes for given regions. In Scotland they
are required to meet Document “J” (see
Table 3), in England it is Document “L”
and in fact its requirements are the first
column of Document “J” and in the United
States ‘Energy Star’ (see Figure 3) is
dominating. (Note the conversion is 1.0
US unit = 5.7778 W/m*K)
When these requirements first appeared
many aluminum manufacturers where
concerned about the ability of their
products to meet the required
performance levels. They had to compile
a myriad of information and
interchangeable options that would allow
them to comply and still be profitable.
During this phase the thermal barrier for
aluminum windows changed from an
option to a requirement.
The manufacturers found that a standard
aluminum window with a clear IG unit had
an average U-factor of around 3.5 W/m*K.
Adding a low-e hard coat to the IG unit
took them down to around 2.7 W/m*K, which was still not enough. Adding a structural thermal
barrier increased the window’s performance to below 2.2 W/m*K, thus meeting the Document
‘L’ requirements.

Figure 4: RESFEN calculations for annual heating and cooling costs of

a dwelling in Chicago, Illinois USA

4
Thermal barrier comparisons -cont
But why is saving energy through improved fenestration products important and just how much
energy is saved going from 3.5 U-factor to 2.2? To answer the first part, today it is widely
accepted that 1/3 of all energy is consumed by commercial buildings of that 38% of that is
attributed to the heating and cooling. To answer the second part we will utilize the previous
mentioned LBNL software program RESFEN.
RESFEN or residential fenestration is a software program that allows the operator to calculate
the heating and cooling costs of a residential dwelling in North America. By entering the
location, square footage of the house, the square footage of the fenestration products, their
orientation and the heating and cooling costs for the region RESFEN gives a simple dollar
amount for the dwelling (see Figure4). In this example a U-factor of 0.61 was used, the
equivalent of 3.50 W/m*K, representing a standard aluminum frame with a clear IG unit. The
total annual heating and cooling costs for this home is $812.97 USD. If all of the windows are
replaced with ones that meet the Document ‘L’ requirements of 2.2 W/m*K or .38 in IP units,
the annual heating and cooling costs are now $716.44 USD. That represents a saving of
$96.53 USD, for a single story home, when this scenario is applied to some of the taller
commercial buildings having 80-100 stories the savings can reach into the millions.
Unfortunately, at present time the software is limited to larger North American cities and only
residential applications. There have been discussions to make a much more comprehensive
commercial version ‘COMFEN’ for global use but the development period is forecasted to be at
least 3-5 years.
We saw the savings of $96.53 when going from windows with a U-factor 3.5 W/m*K to a 2.2
W/m*K, but how much of that can be attributed to the use of a thermal barrier system?
Previously we mentioned that with a low-e coating the windows performance was around 2.7
W/m*K, so actually the thermal barrier improvement was taking the window from 2.7W/m*K to
below a 2.2 W/m*K. If the same RESFEN simulation was performed with windows whose U-
factor was 2.7W/m*K the total annual heating and cooling costs were $755.21 USD. From
before, we know that windows with a U-factor 2.2 W/m*K cost $716.44 USD. Therefore the
annual heating and cooling savings attributed to the thermal barrier are $38.77 USD.
If you are following the calculations you probably have noticed that the addition of a low-e
coating taking the window from 3.5 to 2.7 saved $57.76 USD which was more than the thermal
barrier. While it is understood that both are required to meet energy standards, the true savings
of each of the thermal options must be noted. In speaking with three aluminum pour and
debridge thermal barrier window manufacturers in the United States, the average material cost
per window to add a low-e coating was $55 USD. To add a thermal barrier to a window was
$11 USD. Since our test home in Chicago, IL had 20 windows the total additional cost to add
low-e to the windows was $1,100 USD and to add a thermal barrier was $220. When these
numbers are divided by their respective annual energy savings the payback period becomes 19
years for the low-e coating and 5.7 years for the thermal barrier. As shown the thermal barrier
becomes the most energy efficient material for the return on investment.

Improving insulated glass with warm-edge spacers

Once a structural aluminum window incorporates a thermal barrier, a low-e coating and gas fill
now the spacer for the insulated glass (IG) unit thus becomes the weakest link. Warm-edge
technology started making inroads in the mid-1990s. It was promoted as a way to reduce
condensation and improve the comfort level associated with fenestration products used in the
home.
Today there are many residential warm-edge spacers on the market from pre-desiccated foam,
to butyl with corrugated steel to many PVC designs. They add not only the benefits from
reduced condensation but also they improve the overall energy efficiency of the unit, while
improving acoustical properties and ease of manufacturing. Due to these improvements it is
estimated that the warm-edge market now represents 75% of all spacer material used in
residential applications.

5
Improving insulated glass with warm-edge spacers -cont
As is often the case in the fenestration industry, products are introduced for residential use and
once they have proven their efficacy, commercial applicators start asking for the same
technology. There were few solutions for commercial use. Residential warm-edge spacers
weren’t designed to handle the deflection and windloads associated with a fifty-story building or
a 15-foot span on a curtainwall. The window and glazing producers had limited options to meet
the demands of commercial use.
Conveniently a warm-edge spacer is defined as anything less conductive than the standard
aluminum box spacer. So a natural choice was stainless steel. While stainless steel had all of
the intrinsic structural properties required it was still a highly conductive material. The thermal
conductivity of stainless steel is 14.3 W/m*K, when this material replaces a standard aluminum
spacer some benefits are noted, the sightline temperature increases 1-2C and the
condensation resistance is increased 2-3 points. It also decreases the units overall U-factor
around 0.02-0.05. While all of these are undoubtedly improvements, lost in the mix was an
improvement in acoustical properties and many companies within the industry could not justify
the increase in cost and the slower rates of production.
The slower rates of production stemmed heavily from two facts: The abrasive wheels required
to cut the stainless steel spacer had to be added to the production line and were slower and
more expensive than those used for aluminum. Additionally, many manufacturers who bend
their aluminum spacers require a tight .093” radius; they could not meet this requirement using
stainless steel. This
added to the
Figure 5: How an aluminum thermal barrier warm-edge
manufacturer’s frustration
spacer works
as they already viewed
the bending process a
bottleneck in assembly
line production.
The other option involved
taking an aluminum
spacer and incorporating
a polyurethane thermal
barrier in the middle of
the profile. This
technology ▬an
adaptation taken from
thermal barrier aluminum
frames (see Figure 2)
▬h ad already proven
itself in architectural
windows in commercial
markets. Through testing
and thermal simulation, aluminum thermal barrier polyurethane warm-edge spacer proved it
abilities. It demonstrated all of the attributes in commercial applications that have made warm-
edge technology so attractive for residential applications.

Table 4: Property comparison of aluminum, stainless steel and an aluminum polyurethane

warm-edge spacer in the same aluminum thermal barrier, Low-e, dual IG unit

Spacer type Aluminum Stainless Steel Aluminum

Polyurethane
Sight Line Temp (C) 6.2 C 6.7C 8.7C
Sight Line CRF 31.0 33.5 43.5
Total U-Factor 2.34 2.31 2.18

6
Improving insulated glass with warm-edge spacers -cont
The sight-line temperature increased 2-5C, the Condensation Resistance increased 8-15
points, the overall U-factor was improved by 0.12-0.22 W/m*K and the additional benefits of
acoustical improvements and overall comfort returned without the production difficulties
associated with stainless steel. (see Table 4 for comparison)
The improvements are such that when required to produce a more efficient product
manufacturers now have an option. Filling an IG unit with a low conductive gas such as Argon
or Krypton was once a viable way to meet this request. However as this practice increased
many manufacturers were displeased with added step of production or the cautions required for
having such material on a production line. Additional questions surfaced about the percent of
low conductance gas that was
actually in the IG unit initially and
what percent remained during the Figure 6: A window incorporating a thermal
lifecycle of these units. Many barrier frame and a structural thermal barrier
manufacturers once faced with warm-edge spacer
adding a low conductive gas fill to
the IG units have turned to warm-
edge technology as a practical
replacement. The use of a structural
thermal barrier warm-edge spacer
does not require an additional
production step and restrictive
material handling codes no longer
apply. The warm-edge spacer is
now accepted as a part of an
energy efficient commercial
building.
Both the thermal barrier technology
and the introduction of warm-edge
spacers (see Figure 6) have
allowed commercial window
manufacturers to produce an
energy efficient product that has
never before been matched without
sacrificing structural properties. It
was estimated that in the later part
of the twentieth century there were
enough aluminum thermal barrier
extrusions produced to encircle the
globe nine times. Aluminum thermal barrier and the structural warm-edge technology are now
combined in a fenestration product created for the commercial market that will allow the ever
increasing demands of high performance energy and possibly more important those of the end
consumer.
While the concept remains the same, many new developments in the chemical formulization of
thermal barriers and the design of structural warm-edge spacers continue to evolve and as this
happens these two products will remain a staple in the industry for all of the important reasons
this paper has discussed.

7
References

NFRC – (National Fenestration Rating Council), “NFRC 101: Procedure for Determining
Thermo-Physical properties of Materials for Use in NFRC-Approved Software Programs” (2002)
Therm 5.0, Windows and Daylighting Group, Energy and Environment Division, Lawrence
Berkley National Laboratory,-Berkley, California, USA
Window 5.0, Windows and Daylighting Group, Energy and Environment Division, Lawrence
Berkley National Laboratory,-Berkley, California, USA
RESFEN 3.1, Windows and Daylighting Group, Energy and Environment Division, Lawrence
Berkley National Laboratory,-Berkley, California, USA
DOE – (Department of Energy), Energy Star Requirements www.energystar.gov/default.shtml
Office of the Deputy Prime Minister, Building Regulations, “Approved Document L –
Conservation of fuel and power” www.safety.odpm.gov.uk/bregs/brad12.htm
COMFEN – Windows and Daylighting Group, Energy and Environment Division, Lawrence
Berkley National Laboratory, -Berkley California, USA (Under Research)