Content uploaded by Max Howard Sherman
Author content
All content in this area was uploaded by Max Howard Sherman on Aug 05, 2015
Content may be subject to copyright.
LBNL 43382
Evaluation of PEGIT Duct
Connection System
I.S. Walker, D.E. Brenner, M.H. Sherman and D.J.
Dickerhoff
Environmental Energy
Technologies Division
August 2003
This work was supported by the Assistant Secretary for Energy
Efficiency and Renewable Energy, Building Technologies, of the US
Department of Energy (DOE) under contract No. DE-AC03-76SF00098.
ERNEST ORLANDO LAWRENCE
BERKELEY NATIONAL LABORATORY
2
Disclaimer
This document was prepared as an account of work sponsored by the United
States Government. While this document is believed to contain correct
information, neither the United States Government nor any agency thereof, nor
The Regents of the University of California, nor any of their employees, makes
any warranty, express or implied, or assumes any legal responsibility for the
accuracy, completeness, or usefulness of any information, apparatus, product,
or process disclosed, or represents that its use would not infringe privately
owned rights. Reference herein to any specific commercial product, process, or
service by its trade name, trademark, manufacturer, or otherwise, does not
necessarily constitute or imply its endorsement, recommendation, or favoring
by the United States Government or any agency thereof, or The Regents of the
University of California. The views and opinions of authors expressed herein
do not necessarily state or reflect those of the United States Government or any
agency thereof, or The Regents of the University of California.
Ernest Orlando Lawrence Berkeley National Laboratory is an equal
opportunity employer.
3
Evaluation of PEGIT duct connection system
INTRODUCTION
Most air duct system components are assembled in the field and are mechanically
fastened by sheet metal screws (for sheet metal-to-sheet metal) or by drawbands (for flex
duct-to-sheet metal). Air sealing is separate from this mechanical fastening and is usually
achieved using tape or mastic products after mechanical fastening. Field experience has
shown that mechanical fastening rarely meets code or manufacturers requirements and
that sealing procedures are similarly inconsistent. To address these problems, Proctor
Engineering Group (PEG) is developing a system of joining ducts (called PEGIT) that
combines the mechanical fastening and sealing into a single self-contained procedure.
The PEGIT system uses a shaped flexible seal between specially designed sheet metal
duct fittings to both seal and fasten duct sections together. This study was undertaken to
assist PEG in some of the design aspects of this system and to test the performance of the
PEGIT system.
This study was carried out in three phases. The initial phase examined the performance
of a preliminary seal design for the PEGIT system. The second phase began testing a
second seal design. The third phase performed more detailed testing to optimize the
production tolerances of the sheet metal fittings to be used with the second seal design.
This report summarizes our findings from the first two phases and provides details about
the third phase of testing.
Phase 1.
A detailed report on Phase 1 can be found in Appendix 1. In Phase 1, the initial PEGIT
design was tested for leakage and evaluated for equivalence to mechanical code
requirements. The results of this phase showed that the PEGIT system had the ability to
produce duct fitting seals that were as good as a conventional taped connection (less than
0.5 cfm at 25Pa) and considerably better than an unsealed connection (1.7 cfm at 25 PA).
However, the testing revealed two problems that were both related to the inability of the
inner duct’s formed lip to maintain the position of the seal on the inner duct. The first
problem was that the seal was not securely held from moving longitudinally by the lip.
Combined with the friction between the rubber seal and duct surfaces, this led to the seal
being pulled out of the crimped lip in the round sheet metal as it was inserted in another
duct section. When the seal was incorrectly seated in this way, it resulted in additional
connection leakage. The second problem with the seal was that it was not held at the
correct angle (as shown in Figure 1). This angle is critical because it is a factor in
determining the sealing force and the forces required to assemble or dissemble the duct
sections.
4
Figure 1. Illustration of seal deformation showing how seal pulling out of fitting
changes the angle of the seal that leads to reduced pressure on the inside of the outer
fitting
Seal pulled out of
fitting
Seal correctly
seated in fitting
Seal angle
Seal angle
5
With regards to mechanical security of the connection, the test samples were adjustable.
A longitudinal section of the sheet metal was removed (about 0.25 inch (6 mm)) for
several inches at the end of the sample. The resulting gap allowed us to adjust the
circumference of the fitting using a screw adjuster (see Figure 2). This adjustability
meant that it was not possible to evaluate the security of a production fitting. However,
the basic concept of having a rubber seal snap into a groove in a sheet metal fitting
provides a good mechanical connection that is very difficult to pull apart. This means
that the design meets the intent of the Uniform Mechanical Code (UMC (1994)
1
), which
states that sheet metal duct connections must be secured using three sheet metal screws
“... or an equivalent fastening method.”. Given that the UMC does not give any
specifications for the strength of the duct connection, there is no reason to believe that the
PEGIT system should not be acceptable.
Figure 2. Screw adjustable PEGIT fitting
Phase 2.
In the second phase of testing, the PEGIT design had been changed based on the findings
in Phase 1. A new seal shape and a different profile for the sheet metal fittings was
developed that would keep the seal in place. An improved leakage test apparatus using
high precision orifices was used for the leakage testing. The forces required to assemble
and pull apart the fittings were estimated using simple procedures. The remainder of this
section summarizes the procedures and results of our Phase 2 leakage tests and our
assembly and disassembly force tests. Appendix 2 provides more details about the Phase
2 tests.
Phase 2 Leakage Testing
The Phase 2 leakage tests used the apparatus illustrated in Figure 3. The leakage is
determined by pressurizing the test sample over a range of pressure differences from
1
Uniform Mechanical Code. 1994. International Conference of Building Officials. Whittier, CA.
Paragraph 601.5.1
6
about 10 Pa to 50 Pa and measuring the airflow required to maintain these pressure
differences. For comparison purposes, a reference pressure difference of 25 Pa was
chosen; this is a common reference pressure used in duct leakage testing. Based on our
tests, the leakage flow at 25 Pa pressure difference is 0.26 cfm (0.12 L/s). This is very
close to the 0.3 cfm (0.14 L/s) at 25 Pa measured in Phase 1. This is a very low level of
leakage: it was about one half the leakage measured for a well-taped connection of a
similar configuration.
Figure 3. Phase 2 test apparatus for leakage testing.
Phase 2 Assembly and disassembly forces
We used two different methods to assess the force required to push the two parts of the
connection together.
Assembly Force, Method1: Placing weight on the sample
A simple method of loading the connection that allows large cumulative weights, while
allowing very small increments is to add water to a bucket placed on top of the sample
connection. This method also ensures that the round sheet metal section was loaded
evenly around the circumference. However, we found that even with a full bucket of
water the connection was not pushed together. The bucket and water exerted a force of
45 lbf (200 N).
Assembly Force, Method 2: Direct loading on a scale
In this second test, the sample was placed on a scale and was loaded by having a
technician push on the end of the sample (as shown in Figure 4). The assembly force was
145 lbf (645 N). Even at this high loading, it is necessary to wiggle the fitting from side
to side to help slide the outer duct section over the rubber ring seal.
Sample
Connection
Airflow
measurement
venturi
Airflow
from fan
7
Figure 4. Assembly force measurement using a scale
Disassembly Force
To pull the connection apart, the joined ducts were suspended from a scale to record the
applied force. A lever was used to apply the disassembly force by pulling down on the
other end of the connection. This allowed us to safely place large loads on the
connection. Figure 5 is a photograph of this experimental apparatus. The connection was
loaded with up to 110 lbf (490 N), but did not separate. We stopped testing at this load
because it is far in excess of any separation force we can reasonably imagine would be
applied in a field installation.
8
Figure 5. Disassembly testing using scale and lever
Phase 2 Summary
The results of the second phase of testing showed that the redesigned seals and sheet
metal components solved the problems found in Phase 1. The resulting air seal and
required disassembly forces were acceptable, but the assembly forces were too high. We
considered 20 lbf (90 N) to be a reasonable limit for assembly force based on the strength
and deflection of plenum and other duct assemblies made from duct board or sheet metal.
This led to the final phase of testing to investigate optimizing the sheet metal part
dimensions to balance leakage and assembly force requirements.
Phase 3
In Phase 3, an adjustable-diameter inner and sheet metal duct section was used to
systematically vary the gap between the inner and outer ducts. This adjustability allowed
us to optimize the assembly force and air tightness, and to investigate the effects of
production tolerances on system performance. To make the testing more repeatable and
controllable, one end of the outer duct section was welded onto a 9”9”0.188” (230
mm230 mm4.8 mm) steel plate with a 5” (127 mm) diameter hole aligned with the
inside diameter of this duct. The size of the plate coincides with the leakage test fixture
on LBNL’s duct component air leakage test apparatus and allows us to easily conduct
leakage tests on the assembled duct. The duct welded to the steel plate formed a fixed
Scale
Test sample
Lever arm
9
inner section. The adjustable outer duct section used metal adjustable screw clamps to
adjust the circumference (and thus the diameter) of the test section relative to this fixed
section. Figure 6 is a photograph showing the end of a sample connected to the square
steel plate.
Figure 6. Five inch diameter test sample connected to square steel plate
Assembly/disassembly test procedure
This steel base plate was used to improve the assembly/disassembly force measurements
using a test stand fabricated from 1” square steel tube. Assembly force was measured by
assembling the joint on the stand and adding weights in increments of 2.5 lbf (11 N) until
the joint pressed together completely or 50 lbf (220 N) was reached. Figure 7 shows a
test sample on the test stand.
Disassembly force was measured by placing four sheet metal screws at 90 degree
increments through the exposed end of the inner duct, with the screw points protruding
through the inside diameter. A wooden disk attached to a 0.75 inch (19 mm) diameter
piece of PVC pipe with appropriate threaded fitting was screwed into the disk and
secured with a nut. A flange was added to the opposite end of the PVC pipe so that
weights could be placed high enough not to touch the test stand. The test section was
turned upside down (compared to assembly force testing) so that the inner duct was at the
bottom and the PVC test assembly was inserted from the top of test section until it was
stopped by the sheet metal screws. Disassembly force was measured by adding weights
in increments of 2.5 lbf (11 N) until the joint separated or 50 lbf (220 N) was reached.
Figure 8 illustrates the test apparatus used for disassembly testing.
Square
base
Test
sample
10
Figure 7. Apparatus used for assembly force testing
Figure 8. Apparatus used for disassembly force testing
Weights
Test stand
Square base
Test
sample
Fixed outer duct
section
Adjustable
inner duct
section
Gear clamp adjuster
PVC pipe
Weights
Flange
Fixed outer
duct section
Adjustable
inner duct
section
Force
11
Assembly/disassembly force test results
The first set of force tests were done with a fixed 5 inch (127 mm) I.D. outer duct welded
to the 9” flange and an adjustable inner duct. Proctor Engineering supplied the inner and
outer duct sections. Adjustability was achieved by cutting a longitudinal section out of
the inner duct and welding on two metal adjustable gear clamp sections (shown in Figure
7). Testing these adjustable duct sections adds uncertainty to the test results because the
testing showed that the clamps flexed under loading and the seal was not in contact with
the duct in the longitudinal gap. The flexing under loading could change the required
assembly and disassembly forces. The longitudinal gap made air leakage testing of the
assembly difficult because this gap had to be blocked using duct tape and therefore it
added a potential leakage site that was hard to control. On the other hand, if this gap was
well sealed, it eliminated a section of the seal area from being tested. For these reasons,
further tests used non-adjustable solid duct sections that were built to specific sizes.
Table 1 summarizes the test results with the adjustable inner ducts. As expected, as the
gap between the two parts gets smaller, the assembly and disassembly forces increase.
The two smallest outside circumferences of the inner duct gave reasonable assembly
forces. The largest two circumferences of the inner duct required unreasonably large
assembly forces. In all cases, the disassembly pulled the seal out of the groove in the
inner duct section. The variation shown in one case is because small side-to-side
wiggling motions can strongly affect the assembly forces for these tight fittings and the
large forces used in the test result in flexing of the adjustable clamp and resulting changes
in required force.
Table 1. Assembly and disassembly test results for adjustable Inner Duct into 16.47
inch Internal Circumference Fixed (IC) Outer Duct
Inner Duct
Outside
Circumference
(OC) (in.)
lbf to assemble
lbf to disassemble
Notes
15.50
5
25
Pulls seal out of groove on
disassembly.
15.69
5
32.5
Pulls seal out of groove on
disassembly.
15.88
7.5
37.5
Pulls seal out of groove on
disassembly.
16.06
Inconsistent: 10
to 50
45
Assembly force varies over
a wide range
16.25
50
>50
12
Because of the unreliability of the adjustable fitting, the next tests were performed using
a series of fixed size inner ducts. Five fixed diameter inner ducts were supplied and
tested with the same fixed outer duct. Table 2 shows similar results to the adjustable
inner duct tests in Table 1, except that the test at 15.90 OC that is close to the adjustable
test at 15.88 OC required greater forces. This is most likely because the fixed size
fittings are more rigid than the adjustable fittings.
Table 2. Assembly and disassembly test results for Fixed Inner Duct into 16.47 inch
Internal Circumference Fixed Outer Duct
Inner Duct
Outside
Circuference
(in.)
lbf to assemble
lbf to disassemble
Notes
15.31
2.5
15
15.51
2.5
15
15.70
5
45
15.90
>50
>50
16.10
Could not assemble
Since many new and replacement duct systems use flexible ducting, the force it took to
disassemble flexible ducting from a conventional inner duct and a PEGIT inner duct was
also tested. The flex duct core was pulled over the PEGIT sheet metal inner duct with the
rubber seal removed. A 0.3175 in. (8.1 mm) wide nylon tie-wrap was placed over the
flex duct core and tightened into the groove (that the seal normally sits in) in the PEGIT
sheet metal fitting. A conventional (non-PEGIT) duct fitting was also tested. The results
in Table 3 show that the PEGIT sheet metal fitting is as good as a conventional duct
fitting for attaching flex duct cores.
Table 3. Force required to remove flex duct from inner sheet metal duct
Duct Type.
lbf to Disassemble
Notes
Conventional Inner Duct
25
PEGIT
25
No seal. Tie-wrap aligned and
tightened in groove
Air leakage test procedure
We used our duct component air leakage test apparatus to determine the air leakage of the
PEGIT fittings. It is similar to the system used in Phase 2, and uses the same high
precision venturis. However, the apparatus used in Phase 3 differed by having a plenum
to which test samples were connected. The plenum has a single square (9”9” (230
mm230 mm)) opening to which the square based samples were connected. This
opening has a closed cell foam air seal and the box has a built in-pressure port. A
baseline test was performed with the test plenum opening blocked to determine the
background leakage of the system. The background leakage was subsequently subtracted
from the total leakage measurements to obtain the leakage of the sample. This apparatus
13
and the associated calibration procedures made the leakage testing more precise and
repeatable and also made comparisons to other duct fittings easier to carry out.
The flow venturis were calibrated using a high precision mass flow controller. A set of
flow venturis were used to cover a range of flow rates from about 0.05 cfm (0.02 L/s) to
about 200 cfm (95 L/s). The test procedure uses a fan to pressurize the sample, with the
flow meter located between the fan and the test sample. The leakage is measured at a
pressure difference of 25 Pa. Because we could not always exactly match 25 Pa for each
test, the airflow results from each test are corrected to the airflow at 25 Pa using Equation
1. We used a pressure exponent of 0.6 based on previous experiments (see Appendix 2)
where similar samples were tested over a wide range of pressures and flow to determine
their leakage characteristics. This pressure matching correction and the background
leakage subtraction gave very small corrections, typically 0.1 cfm or less.
6.0
25
)25(
measured
measured
P
QPaQ
(1)
The leakage testing was conducted right after the joint was assembled using the assembly
force testing procedure and before the disassembly testing. This method streamlined the
testing process and made sure that the assembly that was leak tested was the same as the
one force tested.
Air Leakage Test Results
Table 4 summarizes the air leakage test results for the fixed diameter samples. As
expected, the tighter the fit, the less the air leakage. For comparison, the previous tests
conducted for this study showed that a conventional duct fitting carefully sealed with tape
has leakage of about 0.5 cfm (0.22 L/s); this should be the maximum target leakage for
the PEGIT system. The results show that this target is only met by one combination: the
tightest one that could be assembled. However, this combination required greater then 50
lbf (220 N) of assembly force. All the combinations that could be assembled with 50 lbf
or less force had too much leakage to be acceptable.
Table 4. Fixed Inner Duct into 16.47 inch Inner Circumference Fixed Outer Duct
Inner Duct
Outer
Circumference,
in.
P Leak
(Pa)
Venturi flow,
cfm (L/s) at P
Leak
Adjusted flow cfm (L/s)
at 25Pa
15.31
24.3
7.4 (3.5)
7.5 (3.6)
15.51
25.4
3.2(1.5)
3.2 (1.5)
15.70
24.8
1.4 (0.7)
1.3 (0.6)
15.90
29.1
0.1 (0.1)
0.0 (0.0)
16.10
Could not
assemble
14
Discussion and Conclusions
During this study, the PEGIT system has been evaluated for air leakage and ease of
assembly and disassembly. Ideally, the system would have as little leakage as possible,
be easy to assemble and difficult to take apart. Significant changes were made by Proctor
Engineering to the design of the PEGIT system to improve the air seal and ensure that it
stayed in the sheet metal fitting. However, the final results indicate that achieving air
leakage equal to or less than a conventional duct fitting sealed with tape requires such a
tight fit between the inner and outer duct sections that the resulting assembly forces are
greater than 50 lbf. This assembly force seems too large if systems are to be field
assembled. The test results are also an indication of acceptable production tolerances.
Changes in circumference of 0.2 in. (5 mm) gave significantly different air leakage and
assembly force requirements. This implies that production tolerances for circumference
need to be better than 0.1 in. (2.5 mm) for consistent performance. This corresponds to
diameter tolerances of 0.03 in. (0.8 mm).
For the connection to flex duct cores, conventional tie-wraps were used and the groove in
the inner PEGIT duct fitting was used as the mechanical stop for the drawband, in place
of the bead on conventional duct fittings. The tests showed that the PEGIT duct fitting
gave as strong a mechanical connection as a conventional fitting.
The following are three suggestions to improve the PEGIT fittings in the future:
Use a softer formulation for the seal.
Change the seal profile to have a thin tip profile that will maintain an air seal
with looser fit between the inner and outer duct sections.
Use a lubricant on the seal (and possibly on the outer duct fitting) to make
assembly easier.
15
APPENDIX 1. Report on Initial PEGIT Design
Report for Proctor Engineering Group:
Evaluation of PEGIT Duct Connection System
June 1998
Iain S. Walker
Energy Performance of Buildings Group
Lawrence Berkeley National Laboratory
Berkeley, CA 94720
16
Introduction
This report outlines the laboratory measurements performed by LBNL on sample PEGIT
duct fittings. In addition to discussing the measurement procedure and the test results,
some comments on construction of the duct fittings will also be given.
The duct fittings were designed by Proctor Engineering Group (PEG) as a method of
connecting sections of forced air duct systems. The purpose of the fitting is to combine
the mechanical connection and air seal. The mechanical connection is provided by the
friction and interlocking of the rubber flange around the duct. The seal is provided by the
pressing of the rubber flange against the inside of the mating duct surface. The tests in
this report concentrate on the air sealing.
The connections designed by PEG are for round metal duct connections. For connecting
to flex duct, the round metal connection incorporates a collar that holds the inner liner of
the flex duct against the rubber flange on the male part of the connection. After the collar
is attached to the flex duct, it may be connected to other sheet metal components:
plenums, register boots, other flex duct collars or sheet metal duct.
Sample Construction
The test prototypes were prepared in LBNL sheet metal shops to PEG specifications.
The major differences between these prototypes and those that would be used in mass
production for commercial purposes are:
The prototypes had a different flange connection. Because of the difficulty in
forming the stainless steel to the complex shapes required of the fittings, the crimped
lip that retains the rubber flange was of a different shape.
The prototypes were fabricated from stainless steel. Production ducts and fittings
would of galvanized steel.
The prototypes were designed to be removable. To accomplish this, the clamping
mechanism used a hose clamp style of fastener that could be undone and reused.
Production fittings would more likely have overcenter/one time clamps for the collars
and no clamp for the duct to duct connection.
Although these are major construction differences, the results of these measurements can
still be used to test the concept of the PEGIT system. In addition, the prototype duct
sections had a single longitudinal seam that is common in sheet metal duct systems. This
was chosen for convenience because it was easier to manufacture than a spiral duct for
these prototypes that were made from scratch (flat sheet metal stock).
17
Leakage Test Outline
The sample connections were tested for leaks by pressurizing the duct section using a fan.
The sample duct sections were sealed at one end and an orifice flow meter was placed
between the fan and the test section at the other end. The orifice flow meter was
specially designed for low flows and had a small (0.25 inch) diameter orifice and was
calibrated using a mass flow controller and has an estimated uncertainty of 5% of flow
reading. The samples were then pressurized over a range of pressures typical of
residential systems, and the leakage flows were measured. The test apparatus is
illustrated in Figures 1 through 4.
In addition to the PEGIT connections, standard duct connections were also tested. This
standard connection was a round to round galvanized sheet metal duct connection with an
over center clamp. This connection was tested with just the clamp and then with the
connection taped.
Figure 1. Fan and flowmeter test apparatus
18
Figure 2. Whole test section
Figure 3. Duct to duct connection for PEGIT prototype
Figure 4. Duct to collar PEGIT prototype (for flex or plenum connection)
19
Measured Test Results
The measured leakage data shown in Figures 5 through 8 have been least squares fitted to
a power law relationship of the form:
n
PCQ
where Q is the flow rate (cfm), C is the flow coefficient (cfm/Pa
n
),
P is the applied
pressure difference (Pa) and “n” is the pressure exponent. The lines in the figures
correspond to the results of the least squares fit. In addition, the leakage has been
calculated in terms of flow at 25Pa (cfm25). The following table summarizes the results
of this analysis.
Connection type
Flow Coefficient
(cfm/Pa
n
)
Pressure Exponent
cfm25
PEGIT Duct-duct
0.0139
0.953
0.3
PEGIT Duct-flex
0.0646
0.843
1.0
Untaped Standard Duct-Duct
0.0942
0.894
1.7
Taped Standard Duct-Duct
0.0367
0.885
0.6
These results show that the PEGIT duct fittings have little leakage at the typical pressures
seen in residential duct systems. The PEGIT Duct-Duct connection is at least as good as
a taped standard connection, and considerably better than an untaped connection.
Duct to Duct PEGIT connection
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 10 20 30 40 50 60 70 80
Pressure Difference, Pa
Leakage Flow, cfm
Figure 5. Measured pressurization test results for the duct-duct connection.
20
Duct to Flex PEGIT connection
0
0.2
0.4
0.6
0.8
1
1.2
0 5 10 15 20 25 30
Pressure Difference, Pa
Leakage Flow, cfm
Figure 6. Pressurization test results for the duct-flex collar connection.
Standard Connection
Taped
0
0.2
0.4
0.6
0.8
1
1.2
0 10 20 30 40 50
Pressure, Pa
Flow, cfm
21
Figure 7. Pressurization test results for a standard sheet metal duct connection (taped).
Figure 8. Pressurization test results for a standard sheet metal duct connection (untaped)
Standard Connection
Untaped
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 5 10 15 20
Pressure, Pa
Flow, cfm
22
An additional comparison can be made to the data given in the ASHRAE Fundamentals
Handbook (ASHRAE 1997, p. 32.16-32.17). The handbook gives duct leakage values
for longitudinal seams and per unit surface area and NO information regarding
connections (where most of the leaks actually are). For the purposes of this comparison
we will assume that the leakage for a 5m (16ft) duct run of 15 cm (8 inch) diameter round
sheet metal duct is all at the connection. The leakage rate is given as 0.15 l/sm
2
at 250
Pa for sealed and 1.5 l/sm
2
for unsealed. The total surface area is 2.4 m
2
, so the
corresponding leakage rates are 0.36 l/s and 3.6 l/s for sealed and unsealed cases.
Converting to 25 Pa from 250 Pa (and assuming a pressure exponent of 0.7) gives 0.07 l/s
(0.15 cfm25) and 0.7 l/s (1.5 cfm25). Thus the tested PEGIT connection is much better
than the ASHRAE requirement for unsealed ducts and almost as good as the “sealed
case”, particularly given that the ASHRAE data is not for connections per se.
Comments
The biggest problem with the prototypes tested here was the connection of the rubber
flange to the round duct. The lip to hold the flange was not well formed and resulted in a
couple of problems. The first problem was that the flange was not securely held in the
lip. Combined with the grippiness of the rubber, this led to the flange being pulled out of
the crimped lip in the round sheet metal as it was inserted in another duct section or the
clamping mechanism. If the flange is incorrectly seated in this way, it can result in
additional connection leakage. A possible solution to this problem is to have a more
positive lock to keep the rubber flange in the crimped socket by pinching (using a punch
mechanism) the sheet metal together at locations around the crimp. A stronger version of
this may have a hole punched all the way through. This would add another step to the
manufacturing process, however, it is necessary because the integrity of both the seal and
the mechanical connection depend on the flange being properly connected. Alternatively,
a lubricant could be applied (at the factory, not by the installer) to the flange so that the
connection slides together with less friction. The problems to be overcome with this
method would be finding a lubricant that does not dry out in storage and does not interact
with the rubber flange.
The second problem with the flange was that it was not held at the correct angle. This
angle can be critical for achieving a proper seal because it is a factor in determining the
outward pressure that rubber flange against the metal surface. In addition, this angle is
critical in determining the forces required to assemble or dissemble the duct sections.
With regards to mechanical security of the connection, the test samples have adjustable
clamps so the it was not possible to evaluate the security of a production fitting, however,
the basic concept provides a good mechanical connection that is very difficult to pull
apart (probably not possible with flex duct without destroying the flex duct itself). This
means that it meets the intent of the Uniform Mechanical Code (UMC (1994)) which
states that sheet metal duct connections must be secured using three sheet metal screws
23
“... or an equivalent fastening method.”. Given that the UMC does not give any
specifications for the strength of the duct connection there is no reason to believe that the
PEGIT system should not be acceptable. It should be noted that the UMC requirement is
rarely met in residential duct installations. In addition, the PEGIT system is simpler to
install than attaching three sheet metal screws, particularly in the limited access spaces
that HVAC systems are commonly installed in for residences.
References
ASHRAE. 1997. 1997 ASHRAE Handbook of Fundamentals. ASHRAE. Atlanta. GA.
p. 32.16-32.17
Uniform Mechanical Code. 1994. International Conference of Building Officials.
Whittier, CA. Paragraph 601.5.1.
24
Summary of Fire testing for duct seals.
The following UL standards apply:
UL 181 “Standard for Factory Made Air Ducts and Connectors”, (Underwriters
Laboratories, Inc. Northbrook, Illinois, USA, 1994).
UL 181A “Standard for Closure Systems for use With Rigid Air Ducts and Air
Connectors”, (Underwriters Laboratories, Inc. Northbrook, Illinois, USA., 1993).
UL 181B “Standard for Closure Systems for use with Flexible Air Ducts and Air
Connectors”, 1
st
Edn (Underwriters Laboratories, Inc. Northbrook, Illinois, USA, 1995).
Includes ratings for 181B-FX Flexible Tape and 181B-M Mastic.
UL 214 “Standard for tests for Flame Propogation of Fabrics and Films”
UL 723 “ Standard for test for Surface Burning Characteristics of Building Materials”
UL 181
This standard is for factory assembled duct systems and has more fire resistance testing
than 181A or 181B.
There are four tests in UL 181:
1. Surface Burning Characteristic. The surface burning test is performed according to
another UL standard: UL 723 “ Standard for test for Surface Burning Characteristics
of Building Materials”. This test does not look like it applies to the PEGIT
connection because the test uses a sealant on a piece of rigid cement board.
2. Flame Resistance. The flame resistance test is another UL test UL 214 “Standard for
tests for Flame Propagation of Fabrics and Films”. This is unlikely to apply to the
PEGIT connection because there is no “fabric or film”.
3. Flame Penetration. A section of duct wall material is used to form one wall of a
furnace.
4. Burning. Samples are exposed to a naked bunsen burner flame. The samples are held
at an angle such that any melted seal will run out of the sample and be caught on a
cotton cloth. The cotton cloth must not ignite when the seal falls onto it. The bunsen
flame is half yellow and so it is not the hottest it could be (although the actual
temperatures used are not controlled!).
The attached pages from the UL standard show how these tests are performed. If the
PEGIT system does not include the air ducts and is just the connectors, then, as table 4.1
in the attached material shows, the burning and surface burning test apply to air
connectors and the flame resistance test applies to joining materials. The flame
penetration test is for flat sections of duct wall and will not apply to the PEGIT system.
25
UL 181A
UL 181A only has the Burning test from UL181. However, the test is adapted to be
specific for three types of sealants. Section 11 describes the test for metal foil tapes on
cement board, Section 22 is for metal foil tapes on Duct Board and Section 34 applies to
mastic. In each case the sealant is directly exposed to the flame and is not applied to a
duct seam, it is just applied to a flat surface. For the PEGIT system a section of duct
containing a joint would have to be exposed to the flame because the seal itself is
between sheet metal surfaces and not directly exposed to the flame. Or it could be argued
that this test is irrelevant BECAUSE none of the seal is directly exposed to flame.
UL 181B
UL181B has the UL723 surface burning characteristic test. AS with the burning test the
selant is simply applied to flat piece of cement board. This is also a situation that is not
really applicable to the PEGIT system.
Summary
The only test with any relevance for the PEGIT system is the burning test. We can try to
evaluate the materials used in the PEGIT system by doing this test in a fume hood where
we expose a PEGIT seal to a naked bunsen flame and see what happens. This has been
done previously for the aerosol sealant.
26
Appendix 2.
Status report on Evaluation of PEGIT Connection
System
July 2002
The sample PEGIT connection used for these tests is a 5-inch round-to-round sheet metal
connection. The connection has two parts: a smaller diameter male section including a
rubber sealing/locking ring that slides into a second female section of larger diameter.
The results for the test sample indicate that large assembly forces are required (greater
than 600 N). For ease of field assembly, these forces will need to be reduced. This can
be achieved by making the rubber seal smaller or more flexible (particularly as the force
required to pull the connection apart is also very high). If the gap between the male and
female sheet metal sections were made larger, this would also allow for lower connection
forces.
Assembly Force
We used two different methods to assess the force required to push the two parts of the
connection together.
1. Placing weight on the sample
A simple method of loading the connection that allows large cumulative weights, while
allowing very small increments is to add water to a bucket placed on top the sample
connection. This method also ensures that the round sheet metal section was loaded
evenly around the circumference. This even loading allows the testing to be both safe
and stable. However, we found that even with a full bucket of water the connection was
not pushed together. The bucket and water had a combined weight of 20.35 kg or 45.0lb
(determined using a digital scale) representing an assembly force of 200 N.
2. Direct loading on a scale
In this second test, the sample was placed on a scale and was loaded by having a
technician push on the end of the sample (see Figure 1). It was found that a load of (65.6
kg) 145 lb was required representing an assembly force of 644 N. Even at this high
loading it is necessary to wiggle the fitting from side to side to ease the sliding of the
female section over the rubber ring seal.
Force required to pull the connection apart
To pull the connection apart, it was suspended from a scale to record the applied force. A
lever was used to apply a load by pulling down in on the other end of the connection.
This allowed us to safely place large loads the connection. When the connection
separates, it will do so rapidly, and this system does not place the technician directly in
line with the connection. Figure 2 is a photograph of this experimental apparatus. The
27
connection was loaded with up to 110 lb (49.8 kg) or 488 N of separation force, but did
not separate. We stopped testing at this load for safety reasons and because this is far in
excess of any separation force we can reasonably imagine would be applied in a field
installation.
Figure 1. Pressing together the adjustable PEGIT duct fitting on a scale
Figure 2. Separation load testing of PEGIT duct fitting
Load
Scale
Sample
Connection
Lever to apply
load
28
Leakage Test
The leakage of the connection was measured using LBNL’s standard test apparatus for
measuring low leakage flow of duct connections. It consists of a variable speed fan, a set
of accurate orifices, with different sized orifices used for different flow rates. The
connection is capped at one end and the other end is connected to the test apparatus – as
shown in Figure 3. The sample was tested over a range of pressures, with the leakage
flow recorded at each pressure. The results of these tests are shown in Figure 4. The
leakage flow at 25 Pa pressure difference is 0.26 cfm. This is very close to the 0.3 cfm at
25 Pa measured for a previous iteration of the test connection about three years ago. This
is a very low level of leakage and is about one half the leakage measured for a well taped
connection of a similar configuration.
Figure 3. Test apparatus for leakage testing.
Fitted data:
Flow (cfm) = 0.0347Pressure(pa)
0.6254
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0 10 20 30 40 50 60
Imposed Pressure Difference, Pa
Leakage Flow, cfm
Figure 4. Results of leakage testing
Sample
Connection
Air flow
measurement
orifice
Air flow
from fan
29
Part II: Testing adjustable size ducts
The air leakage and assembly force depend on the dimensions of the male and female
components. The above test results showed that too much assembly force was required
by the initial rigid prototype. An adjustable female collar was used to investigate if this
force could be reduced without reducing the effectiveness of the air seal. In addition, in
production there will be variance in the size of fittings and we need to estimate the effect
of production tolerance on air sealing and assembly requirements.
The adjustable collar was tested in four configurations, each one progressively larger in
perimeter and diameter, starting with a base case that was the same as the rigid prototype
discussed above. The base case perimeter is 16 6/16 in (416 mm) (measured on the
outside of the adjustable section at the location of the screw adjuster – the nominal I.D.
for these samples was 5 inches). This was increased by 3/16 in (4.8 mm) for each test.
The perimeter was adjusted using a hose clamp screw (see Figure 5) to make for easy and
repeatable adjustment in small increments.
Figure 5. Screw adjustable PEGIT fitting
The assembly forces for the two smallest settings were greater than 550N (125 lb.) (the
test limit we imposed) and therefore we were not successful in assembling the
connection. The second largest setting had an assembly force of 290 N (65 lb.), and this
was reduced to 45 N (10 lb.) at the largest (loosest) setting. In all cases more than 100 N
(~20 lb.) was required to take the sections apart. In each case, therefore, the sections
were separated by completely loosening the adjustable screw.
30
different duct sizes
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0 10 20 30 40 50 60 70
Imposed Pressure Difference, Pa
Leakage flow, cfm
perimeter4 : 16 15/16 in
perimeter3 : 16 13/16 in
perimeter 2 : 16 9/16 in
perimeter 1 : 16 6/16 in
Figure 6. Dependence of air leakage on fitting tightness
The air tightness of the adjustable fitting was determined using the same method and
apparatus as for the rigid prototype discussed above. Figure 6 shows the expected result
that the larger perimeter results in greater air leakage, and the 25 Pa air leakage flows are
summarized in Table 1. The two larger perimeters that had more reasonable assembly
forces both have more than 1 cfm of leakage at 25 Pa – which is probably unacceptably
large. The smallest perimeter tested had about the same leakage at 25 Pa as the rigid
prototype.
Table 1. Air Leakage Flow Results For Adjustable Collar
Perimeter
(inches)
16 6/16
16 9/16
16 13/16
16 15/16
Air Leakage
Flow at 25 Pa
0.36
0.75
1.22
2.70
It appears, that with this particular seal shape that we cannot get a reasonable
compromise between assembly force and sealing. After inspecting the seal there is a
possibility that changing the seal shape to make the material thinner and/or more flexible
may improve this situation.