Skip to main content Skip to footer

Roof Expansion And Control Joints

July 3, 1997

Roof Expansion And
Control Joints

 

B> Iris Zieiontaj R Eng.
Introduction
Roof expansion and control joints frequently pose a set of difficult if not unusual problems to the roof designer
and installer. By their very nature, an expansion or control joint on a roof has to have the ability to meet and
accommodate movement (horizontal, vertical and in shear) while remaining waterproof for the life of the roof.
Traditionally, the approach has been to raise the expansion joint above the plane of the roof and then install flash¬
ing roof membrane around it, with the objective of moving the actual waterproofing of the joint above the water
line. This approach, while having some advantages, can also introduce problems to the roof s performance over its
service life.
This article will examine the various methods and approaches of designing and detailing expansion and control
joints in roof assemblies. Design parameters considered include the amount of movement required in a joint and
the detailing and installation procedures involved, while taking into consideration the impact on the performance
of the roof. Presented also will be leading edge technology in roof joints that use a flat type of expansion joint
which accommodates tri-dimensional movements of the roof and roof membrane while remaining waterproof. An
actual case study of an installation with a flat expansion joint will be examined. The advantages and disadvantages
of this type of expansion and control joint will be reviewed.
Building Expansion Joints
Buildings readily move by contraction and expansion
of components resulting from changes in temperature.
The degree of this movement is dependent upon the
type of material from which they are constructed. The
extent of movement is dependent upon building dimen¬
sions and the ambient temperature (variation) change.
In order for a building and its components to accom¬
modate these movements, the extent of this contraction/
expansion is usually limited to a maximum of 2″ (50
mm). Any movement that is greater than 2″ (50 mm)
within a building can result in damage to the building and
its external finishes (cracking, dislodgement of exterior
panels). In order to limit the expansion and contraction of
the building, the building’s designer (architect or structur¬
al engineer) incorporates a periodic and intentional break
in the building structure. These breaks or gaps are usually
referred to as expansion joints. An expansion joint is in
essence a joint that passes through the entire building’s
cross section as if slicing through it. The idea is to provide
a gap or space for the building to move. To further clarify
this terminology, when reference is made to an expansion
joint, this refers to the actual gap in the building struc¬
ture. This paper is concerned with waterproofing the
expansion joint gap.
The presence of an expansion joint poses some very
unique problems in the building envelope design as such
joints are not always straight and yet must keep out the
elements, air, water, moisture, dust, etc., while allowing
for the building’s movement. The primary concern is
water and moisture, as this causes the most damage to a
building’s interior and roof components. Water can enter
the building’s interior through the expansion joint at
either below or above grade. Many remedies to this prob¬
lem have been sought. Each method has had it drawbacks
or problems. A properly designed expansion joint ideally
should remain waterproof and weather tight.
Location of Expansion Joints
Expansion joints are usually found in the following
places: on a roof, within a building’s wall, above and
below grade. Reasons for their requirement at these loca¬
tions can include the following:
• Changes in deck direction.
• Changes in deck type, e.g., a concrete deck abutting
8 • Interface July 1997
a steel deck.
• Re-entrant building corners.
• Long buildings exceeding 200 feet (61 m) or more
in length.
• Additions built next to existing structures.
• Areas requiring isolation from vibration.
• Buildings that experience extreme thermal move¬
ment, i.e., freezer buildings.
Roof Control Joints
Expansion joints are considered to be the main type of
joint in a building and they are deliberately installed to
accommodate movement. Another type of joint that is
introduced to provide room for movement is a roof control
joint. These roof control joints can be thought of as
expansion joints within the roof assembly. Their function
is to provide room for the movement of the roof mem¬
brane material and relieve stress that may build up. This
is especially critical on very large roofs.
The reasoning behind this is that roof membranes are
materials that have the physical ability to contract and
expand just like any other material (a fundamental law of
natural physics). The degree of expansion and contraction
is measured through what is known as the coefficient of
thermal contraction or expansion. The coefficient of ther¬
mal contraction or expansion is described as unit change
in length per degree change in temperature. For example,
concrete has a coefficient of thermal expansion of 5.5 x
10-5 in/in°F (10 x 10-5 m/m°C). For the purposes of this
discussion, material contraction and expansion are syn¬
onymous as they describe movement due to temperature
change.
Another need for roof control joints in a roof assembly
may also be as a result of the roof membrane being laid
down on a substrate that has a different coefficient of
expansion than that of the roof membrane itself, i.e. insu¬
lation boards. As the air temperature changes, these two
materials, which are in direct contact, undergo a dimen¬
sional change which is not uniform. The membrane may
dimensionally change more than the substrate to which it
is attached or adhered.
It should also be noted that a surface area of a roof will
change dimensionally in the two principal plane direc¬
tions. The roof surface movement can be established
through a thermal coefficient known as the coefficient of
thermal surface expansion. The coefficient is approxi¬
mately twice that of the coefficient of linear expansion.
To accommodate surface movements on large roofs it is
common to introduce roof control joints in the roof mem¬
brane, in both directions. The joints are usually perpen¬
dicular to each other.
Design Methodology of
Expansion and Control Joints
Expansion joints, as mentioned, by their own very
nature interrupt the building’s envelope and compromise
the waterproofing integrity of the roof. Frequently,
expansion joints are overlooked in the design process and
are considered to be an appendage listed in specifications
as “to suit site conditions” or at worst, an afterthought
during the installation of the roof. The fact is that expan¬
sion joints in roof assemblies are a leading cause of water
infiltration and leakage. Lack of proper consideration and
detail will result in compromising the roof s waterproofing
integrity, either immediately after installation or during
the service life of the roof.
Approaches to the design of expansion joints are many,
ranging from the crude to the sophisticated. The most
simple approach is to provide additional roof membrane
material at the expansion joint location specifically to
accommodate the movement. There are a number of “off
the shelf’ expansion joint solutions. These usually com¬
prise a variety of rigid plastic membrane sheets pre¬
formed into a bellows type shape (the bellows allowing
for material movement). The more sophisticated systems
in place are also bellows types manufactured to site spe¬
cific conditions to accommodate predetermined move¬
ments. The various solutions require that these expand¬
able materials be raised above the roof level on a curb.
Once the expansion joint is elevated above the roof level,
the roof membrane is then flashed around the base of the
curb and the waterproofing expansion joint material spans
the gap. The entire expansion joint detail is usually
capped with metal flashing as a protection from the ele¬
ments and mechanical damage.
The approach described above has been used in roof¬
ing for a long time; in fact, if one were to look back at a
building roofed in the 1940s, this was the “modus operandi.”
This approach, although elevating the expansion joint
assembly above the roof membrane, has a number of dis¬
advantages which can be detrimental to the roof mem¬
brane and the roof assembly. Outlined are some of the
factors that may be considered to have a negative impact
on the roof itself.
Primarily, the elevated curb obstructs water flow across
the roof, and, in fact, has a damming effect. Water cannot
flow to a drain, which may be located on the other side of
the expansion joint. The roof owner is then left to make
modifications to the expansion joint by installing scuppers
through the expansion joint to allow water flow.
Ponded water on a roof, as conventional wisdom dic¬
tates, is not a good thing and every effort has to be made
to remove standing water from the roof. In addition, a
raised expansion joint on a curb is exposed to roof traffic
and damage. This is an important consideration, especial¬
ly when there is no metal coping or cap on the expansion
joint curb. Another problem associated with expansion
July 1997 Interface • 9
Typical conditions encountered with expansion joints:
The above diagrams describe some very typical expansion joint conditions. Fig. A: Two expansion joints coming from two different
elevations joined together at a point. Fig. B: Three expansion joints, from three different elevations, coming together to a point. Fig.
C: An expansion joint going around a circular column. Fig. D: An expansion joint condition at the base of a parapet wall. Fig. E: A
four-way expansion joint junction. Fig. F: An expansion joint following the contours of an irregular exterior wall. Fig. G: An
expansion joint bordering with square columns. This is a fairly typical condition when an expansion joint abuts a roof curb or col¬
umn. Fig. H: Changes in elevation of a continuous expansion joint, a common condition found on multilevel roofs or in transition of
plane surfaces from the horizontal to the vertical
joints that is even more fundamental to their function is
that of expansion and contraction. The materials used to
waterproof expansion joints today are rigid or allow very
little material elongation. The movement of these materi¬
als is accomplished through the provision of extra material
(e.g. bellows, folds). This in itself has disadvantages. The
material does not always return to its original shape and
position, and the material can also suffer physical damage
during the movement. An even more critical aspect is the
material’s ability to be joined or seamed together. Even
straight expansion joint runs cannot be manufactured from
one piece of material; two or more pieces have to be
joined together to make up the length required. The two
most common methods used for joining material pieces are
adhesives or hot air welding/fusing. Though initially pro¬
viding a watertight connection, the material at the actual
joint does not always exhibit the same mechanical proper¬
ties (elongation, tensile) as the rest of the expansion joint
material. These issues can be of critical significance as to
the performance of an expansion joint, especially when a
complex shape or configuration is required, examples of
which are shown below.
A Different Approach
to Expansion Joints
As described above, traditional expansion joints rely on
the advantage of being raised above the roof level.
Though this is a logical precaution, there are other
approaches. In other countries, the philosophy is to pro¬
tect and make integral any roof protrusions or projections
on a roof by not exposing them to the elements or roof
traffic. This is to prevent unnecessary damage. This
thinking has found its way to roof expansion joint design,
too. The approach to designing such waterproofed expan¬
sion joints is to make them as level as possible with the
roof membranes. This has evolved a new concept of the
flat waterproofed expansion joint — an idea that has
found application in parking garages and blind-side water¬
proofing for the last three decades.
A Case Study
In September of 1996, a unique problem was encoun¬
tered on a reroofing project. The owner of an eight floor
building decided to reroof. The existing roof was an
inverted BUR Protected Membrane Roof Assembly. On
the roof were eight skylights which allowed light into the
building’s central atrium. An expansion joint ran across
the building adjacent to the skylights (see Figure 1). The
expansion joint had been repaired at least three times and
was a source of continual water leakage and aggravation to
the building’s occupants, (see Figure 2).
Particular areas of failure were at the corners of the
10 • Interface July 1997
Fig.: 1 View of roof with skylights and expansion joint.
skylight curb where the expansion joint butted the sky¬
light curb (see Figure 3). In addition, the roof was drained
at the perimeter, thus water flow was being obstructed by
the raised existing expansion joint detail. The roof con¬
sultant and roofing contractor were faced with a signifi¬
cant dilemma—how to provide an expansion joint thatdid
Fig.: 3 Close up of existing expansion
joint at skylight curb corner.
two things: 1) reli¬
ably waterproof the
corner of the sky¬
light curb; and, 2)
allow the flow of
water across the roof.
This all had to be
accomplished in an
economical and cost
effective manner.
The determina¬
tion was made to try
a flat “zero profile”
expansion joint
waterproof material.
Although this may
be thought of as an
unorthodox approach
to the problem, it
seemed to fit the
bill. The total length
of the expansion
joint was 210 feet (61 m), terminating at a vertical expan¬
sion joint down the building’s side elevation.
Product Description
The system is comprised of 10″ (250 mm ) or 13″ (325
mm) wide material strips which are compounded out of a
specially formulated EPDM elastomer with polyester
fleece embedded in flaps on either side. The expansion
Fig.: 2 Close up of existing expansion joint.
joint material is manufactured by an extrusion process
from an EPDM elastomer. During the manufacturing
process, the polyester fleece is embedded into the gelling
EPDM matrix on both sides, top and bottom, of the
material. The intent is for the fleece to engage the plies
of the bitumen roof membrane. The joint material has no
fleece material attached over the actual expanding joint
part. The EPDM elastomer used in the formulation is
very unique, as it has a number of specific chemical prop¬
erties suited to this application. The EPDM elastomer
used is an Ethylene-Propylene-Diene-Methylene with a
saturated polymethylene chain. The saturated polymeth¬
ylene chain in the elastomer renders it inert by not allow¬
ing for chemical reactions to occur. The material is resis¬
tant to the effects of UV, ozone, high temperatures,
chemicals such as alkalis, acids, saline solutions, alcohol
and ketones. The high quality and purity of the EPDM
elastomer also make vulcanization of expansion joint
pieces possible. This makes the construction of details
around unique shapes possible and watertight without rhe
use of glues or tapes.
Technical Data as Manufactured
Appearance:
Color: Orange red with white fleece on the side flaps
Physical Properties:
Bonding Medium: 1.20 oz/ft2 (400 g/m2)
polyester fleece
Elongation: > 500 % at -40°F (-40°C)
Tensile Strength: 725 psi (5 N/mm2)
Hardness: 45 + 5 Shore A
Long term Operating
Temperature: -40°F (-40°C) to +194°F (+90°C)
Geometry
Width Thickness A B C
13 1/4″ (340 mm) 87 mils (2.2 mm) ±1 3/4″ (+40 mm) + 13/16″ (±20 mm) + 1 3/16″ (±30 mm)
July 1997 Interface • 11
Installation
Material for the entire 210-feet (61 m) expansion joint
was installed in two hours. The installation procedure is
described in brief as follows. The expansion joint mater¬
ial is shipped to the job site in a roll. Installation takes
place as roofing progresses (see Figure 5). The material
Fig. 5: The building expansion joint being
prepared for roof expansion joint installa¬
tion.
is laid down in
a flood coat of
asphalt/bitumen
(see Fig¬
ure 6). The
asphalt/bitumen
should be
at its EVT
when being
applied to the
fleece. Both
the fleece and
the roof ply
surface must
be covered in
asphalt/bitumen.
The
asphalt/bitumen must totally encapsulate and impreg¬
nate the fleece with a bleed out visible along the ply’s
edge. The next step is to strip in the top ply of the roof
membrane over the expansion joint material (see Fig-ure
7).
The top ply should be installed in a flood coat of
asphalt/bitumen as well. It is important that the
asphalt/bitumen be at the correct temperature at the
point of application. The asphalt/bitumen flood coat
must impregnate the entire polyester fleece material;
only the top fleece surface needs to be flood coated. It
does not matter if the entire joint is covered in
asphalt/bitumen. Particular attention is paid to the sky¬
light corner curb
detail (see Figure 8).
It should be noted
that the asphalt/bitu¬
men will not stick to
the material that has
no fleece. The instal¬
lation is finally com¬
plete once the para¬
pet flashings are
installed (see Fig-ure
9). The Protected
Roof Membrane is
then given a glaze
coat of asphalt and
insulation, slip sheet
and ballast are
installed and placed
on the roof (see Fig¬
ure 10).
Fig. 4: An illustration showing the
three principal movement directions
with a flat expansion joint.
Direction d is horizontal, Direction
B is transverse (shear) and Direc¬
tion C is vertical.
Advantages and Disadvantages of
Flat Expansion Joint Systems
One of the key advantages of the flat waterproof
expansion joint systems is that it can be tailored to spe¬
cific site conditions. This makes the waterproofing of
difficult joint details easy, (i.e., curved joints, inside/outside
corners, changes in joint elevations and planes.)
The case study project described above showed that the
installation of the flat expansion joint is simple and can
provide a solution to a specific situation.
The flat expansion joint system has a number of
advantages. These can be grouped broadly into two cate¬
gories, technical and economic advantages. These are as
follows:
12 • Interface
(Left): Fig 6: The roof expansion joint being installed in a flood coat of
asphalt.
(Above): Fig. 1: The roof expansion joint being stripped in with flashing
plies.
July 1997
Fig. 8: The roof expansion joint detail at the skylight curb corner.
Fig. 9: The completed installation of the flat roof expansion
joint. Note the verticalflashings.
Fig. 10: The completed roof with roof expansion joint installed.
Technical
1. Provides a flat, “zero profile” waterproofed joint with¬
out obstructing the flow of water.
2. The system is tailored to specific site conditions.
This includes such difficult roof areas to detail and
waterproof as inside and outside corners, “T” joints,
joint intersections, curved joints, multilevel joints,
etc.
3. The flat expansion joint system has the capacity to
move in all three planes simultaneously, i.e. horizon¬
tally (side ways), vertically (up and down) and trans¬
versely (shear), while remaining flat.
4. There are no complicated calculations or “guessti¬
mates” to be done, since the joint is flat and in the
plane of the expansion joint. The only measurements
required are clear distances between fixed points.
5. The flat expansion joint material is installed between
the roofing plies of BUR and Modified Bitumen in a
flood coat of asphalt. The result is that the water¬
proofed joint is an extension of the actual roof mem¬
brane, with no flashing plies or special detailing
required.
6. The flat expansion joint system forms a continuous,
uninterrupted waterproofing solution from the start of
the expansion joint to its termination, which can be in
a different part of a building — for example, under¬
ground. The “zero profile” of the tape allows it to tra¬
verse any building surface, from inside to outside,
above or below ground.
7. The expansion joint can be made continuous with the
same expansion capability along its entire length. Any
seams that are in the tape are as flexible and have the
same elongation as the expansion joint material itself.
8. The seams found in the expansion joint system are all
vulcanized under factory-controlled conditions
(although site seaming is possible through the use of
a special portable vulcanizing press system). Since
there is no site seaming, this prevents any weak
seams that can result from being done in adverse
weather or poor site conditions.
9. Being flat, the expansion joint end termination can
accommodate movement in all three directions.
There is no warping or wrinkling at the joint termina¬
tion.
10. The flat expansion joint is laid down at the roof
membrane level over insulation, leaving no void
space where condensation can form.
11. There are no metal fasteners used in the installation
of the flat expansion joint; thus, there is no possibility
of thermal bridging or additional thermal heat loss.
12. The expansion joint’s “zero profile” forms no trip
hazard on the roof, nor is it affected by normal roof
traffic.
Economic
1. Eliminates the need for expensive wood curbs and
roof membrane and metal flashings.
2. The expansion joint system can be laid down and
installed in minutes, with minimum preparation time
and skill required, resulting in significant labor sav¬
ings to the contractor.
3. The expansion joint is supplied in one piece for the
July 1997 Interface • 13
entire project; it is rolled and ready to be installed
immediately.
Conclusion
The case study project offered an insight into a differ¬
ent way of tackling an old problem. At last report, the flat
expansion joint is performing well. The roof now has
proper drainage and ponding water has been eliminated.
The difficult skylight corners which seemed to defy flash¬
ing and waterproofing have remained watertight. An
added benefit is that the expansion joint is hidden from
roof traffic and does not require any maintenance. The
opinion of the author is that, though every application has
its own solution, it would appear that flat expansion joints
demonstrate a place in roofing and provide a fresh and
innovative approach to an old problem.
Acknowledgments
I would like to acknowledge the contribution to this
paper of the following individuals and companies:
• Marc Allaire, Industrial Roof Consultants Group
Inc., Mississauga, Ontario
• Doug Fishburn, Fishburn Building Sciences Group,
Hornby, Ontario
• Chris Love, Halsall Associates, Burlington, Ontario
• Roof Consultants Institute 12th International
Convention and Trade Show, March 22/27, 1997
About The Author
Kris Zielonka is a professional
engineer who has been involved in
roofing and building since 1986.
He is a specialist in waterproof¬
ing of above- and below-grade structures. Zielonka is a
director of CRCA and an active member of its National
Technical Committee, as well as a member of both the
Canadian and American Association(s) of Civil
Engineers. This paper was presented at RCI’s 12th
Annual Convention and Trade Show in Anaheim.
Infrared Roof Moisture Surveys
1. Scan entire roof surface using high
resolution infrared camera.
2. Thermograms with matching photographs
of all problem areas.
3. Electronic moisture probes to confirm
accuracy of scan.
4. Core samples to determine make up of roof
system of each roof area.
5. Scaled roof drawing showing all problem
area dimensions and locations.
We do not sell any roofing supplies or repair services
Service Nationwide
Since 1985
Infrared Inspections, Inc.
1-800-543-2279
LABORATORY TESTING
Built Up Roofing
Insulation
Single Ply
Shingles
PUF
Jim D. Koontz & Associates, Inc.
1800 645 7663
Publish A
Business Card Ad
In Interface For Only $150.
Phone RCI Headquarters.
CALL FOR PAPERS! All RCI members and nonmembers are encouraged to submit technical papers and articles.
Interface will also consider certain reprints from other publications and proceedings and welcomes all suggestions. Please
contact Lyle Flogan (910) 768-7185 or call Kris Ammerman at RCI Headquarters.
14 • Interface July 1997