Shear Strength and Stiffness of Diaphragms From Test Results
Shear Strength and Stiffness of Diaphragms From Test Results
Shear Strength and Stiffness of Diaphragms From Test Results
net/publication/271431778
CITATIONS READS
0 269
1 author:
Gary Anderson
South Dakota State University
53 PUBLICATIONS 281 CITATIONS
SEE PROFILE
All content following this page was uploaded by Gary Anderson on 05 August 2015.
Author(s)
Affiliation
Affiliation
Publication Information
The authors are solely responsible for the content of this technical presentation. The technical presentation does not necessarily
reflect the official position of the American Society of Agricultural Engineers (ASAE), and its printing and distribution does not
constitute an endorsement of views which may be expressed. Technical presentations are not subject to the formal peer review
process by ASAE editorial committees; therefore, they are not to be presented as refereed publications. Citation of this work should
state that it is from an ASAE meeting paper. EXAMPLE: Author's Last Name, Initials. 2005. Title of Presentation. ASAE Paper No.
05xxxx. St. Joseph, Mich.: ASAE. For information about securing permission to reprint or reproduce a technical presentation, please
contact ASAE at [email protected] or 269-429-0300 (2950 Niles Road, St. Joseph, MI 49085-9659 USA).
An ASAE Meeting Presentation
Abstract. Twenty nine light-gage metal diaphragms were constructed and tested to determine the
strength and stiffness of the diaphragms. The strength and stiffness determined by testing for 12
diaphragm assemblies was compared to the strength and stiffness predicted by the procedure
outlined in the Metal Construction Associations (MCA) diaphragm design primer. The MCA method
predicted seam/corner fastener strength within 6% of test values and predicted the correct failure.
The method did not accurately predict sheet buckling strength (within 65%) and over estimated
stiffness by a factor of 2.5. The test diaphragm knuckling strength was 7438lb. Seam strength
without stitch screws was determined as well as with one stitch screw over the purlin. The effect of
purlin spacing was also evaluated. The simple beam test may not be adequate for determining
diaphragm structural properties because of load transfer into the diaphragm without blocking.
Keywords. Light-gage metal diaphragm, strength, stiffness, diaphragm test, MCA predictive
equations
The authors are solely responsible for the content of this technical presentation. The technical presentation does not necessarily reflect the
official position of the American Society of Agricultural Engineers (ASAE), and its printing and distribution does not constitute an
endorsement of views which may be expressed. Technical presentations are not subject to the formal peer review process by ASAE
editorial committees; therefore, they are not to be presented as refereed publications. Citation of this work should state that it is from an
ASAE meeting paper. EXAMPLE: Author's Last Name, Initials. 2005. Title of Presentation. ASAE Paper No. 05xxxx. St. Joseph, Mich.:
ASAE. For information about securing permission to reprint or reproduce a technical presentation, please contact ASAE at [email protected] or
269-429-0300 (2950 Niles Road, St. Joseph, MI 49085-9659 USA).
Introduction
Lateral loads (wind and earthquake) applied to post-frame buildings are resisted by the post-
frame system (posts and truss) and by the light gage metal diaphragms that comprise the roof,
walls, and ceiling of the building, ANSI/ASAE EP484.2 (2003). Often the diaphragm system is
substantially stiffer than the post-frame system resulting in the majority of the lateral load being
carried by the diaphragm system. The diaphragm system can be designed to resist the applied
lateral loads if the strength and stiffness of the diaphragm subassemblies is known. One
subassembly of interest is light gage metal sheeting over 2x4 purlins that are attached to the
truss top chord dimension lumber. These subassemblies span from truss-to-truss and when
connected together appropriately they act as a ‘deep’ beam carrying lateral load applied to the
building to the end wall diaphragms which transfer the load to the foundation (post foundation).
The same type of diaphragm used to resist lateral loads in post-frame buildings may be used to
resist lateral load in buildings in which Structural Insulated Panels (SIPs) are used as walls with
trusses spanning from wall-to-wall. Steel sheathing is attached to purlins that span from truss-
to-truss and the purlins are attached to the top chord of the truss. The basic difference between
the two buildings is that in post-frame construction, the truss heels are attached to the posts
making the load path into the diaphragm through the post-frame system while the truss heel is
attached to the top plate of the SIP in the SIP construction making the load path through the SIP
to the diaphragm. The load path out of the diaphragm for both types of buildings is usually
through an end wall or shear wall.
The strength and stiffness of light gage metal on dimension lumber diaphragm subassemblies
are generally determined by testing in accordance with ANSI/ASAE EP 558 (ASAE, 2004)
and/or ASTM E455-04 (ASTM, 2004). Both standards allow testing of cantilever or simple beam
diaphragm subassemblies to determine strength or stiffness. The cantilever test has been used
successfully to test post-frame diaphragm subassemblies. Post-frame construction lends itself
to cantilever testing. Lateral load is transferred to the post frame either directly via wall girts or
indirectly by side sway of the frame system. The shear load that is transferred to the diaphragm
subassembly between frames is basically established by the difference in lateral displacement
between the two post-frames. The post-frame diaphragm subassembly resembles a cantilever
test assembly with the load rafter being the truss top chord that displaces the most and the
support rafter is the truss top chord that displaces the least (Note that the truss top chord that is
the support rafter in this case is the load rafter in the next diaphragm subassembly.). Posts are
often spaced 8’-10’ apart leaving a reasonable test span which can accommodate at least two
seams with 36” cover panels.
In the SIP building, trusses sit on top of the SIP and are generally spaced closer together; 4’ o-
c. The close spacing of the trusses does not allow for two or more seams in a cantilever tests
assembly and the SIP walls do not provide a distinct frame system to divide the building into
units even though the SIP wall transfers load from the wall to the panel top plate which in turn
transfers the load to the diaphragm subassembly via the truss top chord. The shear load in the
diaphragm subassembly is a function of the difference between displacement of the two trusses
rather than the post frames.
Predicting diaphragm subassembly strength and stiffness would be beneficial to design
engineers. However, past attempts to predict strength and stiffness of light gage metal
2
diaphragms on dimension lumber frames have not been fruitful, Boone and Manbeck (1985)
unless extensive finite element models with thousands of degrees of freedom are used, Wright
(1992). Models to predict the strength and stiffness of light gage metal on steel frames has also
been pursued in the past with more success, Atrek and Nilson (1980), Davies and Bryan (1982),
Luttrell (1981), and Luttrell and Mattingly (2005). These models are generally applied to steel
framed structures. The method presented by Luttrell and Mattingly (2005) is stated to be for
light-gage metal sheeting on dimension frame diaphragms as well as light-gage aluminum
sheeting on steel or aluminum frames. A limited amount of test data is presented for verification
of the model relative to wood frames.
Objectives
The objective is to determine the strength and stiffness of the light gage metal sheeted
diaphragms on dimension lumber frames:
1. Standard fastening pattern with purlins 24” o-c, without blocking (shear connectors)
used by EPS
2. Standard fastening pattern with purlins 24”o-c, with blocking
3. Addition of a stitch screw at the purlin to #2
4. Addition of a stitch screw at the purlin and between purlins to #2
5. Standard fastening with blocking and purlins 30”o-c
6. Compare test results to the Luttrell and Mattingly (2005) predicted strength and
stiffness
The test apparatus was either the cantilever test or the simple beam test of ANSI/ASAE 558
(2004) and/or ASTM E455-04 (2004) except the loads were tensile loads rather than
compressive for the loaded rafters, see figures 1 and 2. Load was applied to the diaphragm in
increments of 200lb. A waiting period of 2 minutes was allowed to elapse before loading
continued to the next increment. Load increment and time were controlled by the load
apparatus. Deflection measurements were made with 4 LVDTs positioned against the rafters in
the direction of load, 2” throw ( 1B and 3B of figure 3), or against the purlin for measurements
perpendicular to the load, 1” throw (2A and 4A of figure 3). The net deflection (∆T) of the
diaphragm is:
∆T=∆1B-∆3B+(a/b)(∆2A+∆4A) 1
where,
∆1B-measured displacement of the load rafter, in
∆2A-measured displacement of the purlin at corner 2 (figure 3), in
∆3B-measured displacement of the rafter supported rafter, in
∆4A-measured displacement of the purlin at corner 4 (figure 3), in
3
a-width of the diaphragm, 104.5”
b-length of the diaphragm, 142.5”
The LVDTs for the simple beam test were placed with the 2” throw LVDTs against the load
rafters at the same rafter end as the load was applied. The 1” throw LVDTs were placed against
the supported rafters at the end opposite of the load. The net deflection (∆T) is:
∆T=1/2(∆2+∆3-∆1-∆4) 2
where,
∆1-measured displacement of support rafter, in
∆2-measured displacement of load rafter closest to ∆1, in
∆3-measured displacement of load rafter closest to ∆4, in
∆4-measured displacement of support rafter opposite of ∆1, in
The shear deflection (∆S) is found by subtracting the bending deflection (∆B) from net
displacement (∆T).
∆S=∆T-∆B 3
The bending deflection (∆B) for the cantilever and simple beam tests are given by equations 4
and 5 respectively.
∆B=Pa3/3EI 4
where,
P-load at which the bending deflection is being calculated, lb
a- width of diaphragm (see equation 1), in
E-modulus of elasticity of the edge purlins, 1.4x106psi
I-moment of inertia of diaphragm, in4
I=1/2(Ab2)
A-area of edge purlin, 5.25in2
b-length of diaphragm (see equation 1), in
∆B=23PL3/684EI 5
4
where,
P-applied load, lb
L-width of diaphragm (3a), in
a-center-to-center spacing of rafters, in
E-modulus of elasticity of edge purlins, 1.4x106psi
I-moment of inertia of diaphragm, in4
I=1/2(Ab2)
A and b defined under equation 5
Displacements were recorded after the 2 minute waiting period before load was moved to the
next increment.
The cantilever diaphragms were loaded to failure twice. The first failure was a bending moment
failure (usually a purlin-to-rafter connection failure) and the second was a shear failure. Steel
brackets were attached to the rafter and the purlin to transfer load out of the diaphragm by
passing the purlin-to-rafter connections. The simple beam test does not yield a bending failure
so they were tested once.
Luttrell and Mattingly (2005) presented a procedure for predicting the strength and stiffness of
light gage metal sheathed diaphragms. The strength of the diaphragm is governed by the field
fastening, the panel corner fastener, or the buckling strength of the sheet. The lowest capacity
determined is the strength. Luttrell and Mattingly’s (2005) strengths are equivalent to sheet-to-
sheet load transfer and sheet buckling, Anderson and Bundy (1992). Luttrell and Mattingly
(2005) strengths depend on the strengths of the structural fasteners attaching the sheet to the
purlin, and the seam fastener strength attaching two adjacent sheets together (stitch fasteners).
The starting point for the procedure is determining the strength of the fasteners. Often, Luttrell
and Mattingly (2005) give fastener equations by size. It should be noted that screw sizes may
very somewhat from manufacturer to manufacturer and standard to standard. The stitch
fastener strength is:
QS=115dt 6
where,
QS-in plane strength of stitch fastener, k
d-fastener diameter, in
t-base metal thickness of sheet, in
5
Equation 6 should be modified by (t/0.028)1/2 if the base metal thickness is less than 0.028”.
The structural fastener strength for sheathing attached to wood purlins may be governed by the
strength of the sheeting material or the strength of the fastener in wood. The sheet strength is:
where,
Qf-in plane shear strength of fasteners attaching sheet to
the purlin, k
Fu-ultimate tensile strength of sheet material, ksi
d-fastener diameter, in
t-sheet metal base thickness, in
Fy-yield strength of sheet base metal, ksi
If the fastener should fail in the wood, then the strength is:
The equations for Qfw are valid for fasteners that penetrate the purlin by 7d or more. The
minimum penetration of the fastener is 4d. Fasteners penetrating less than 4d have no strength
and those penetrating less than 7d but 4d or more shall have the strength adjusted linearly by
multiplying Qfw from equation 8 by the fastener penetration depth divided by 7d. Again when the
screws are #12 or #14, Qf will be adjusted by multiplying the result from equation by (t/0.028)1/2
when the sheet metal base thickness is less than 0.028”.
Once the fastener strengths have been established the diaphragm strength can be found.
Equation 9 gives diaphragm strength when sheet buckling controls.
6
1/ 4
3250 I x3t 3 p
Su= 9
Lv wt + wb 2 ww
where,
Su-ultimate strength of diaphragm, k/ft
Lv-purlin spacing, ft
Ix-moment of inertia sheet, in4/ft
t-sheet base metal thickness, in
p-corrugation pitch, in
wt-width of top of corrugation, in
wb-width of bottom of corrugation, in
ww-length of corrugation web, in
The strength of the fasteners in the field attaching the sheets to the purlins to include stitch
fasteners transferring load sheet-to-sheet is:
(2(λ − 1) + B )Q f orQ fw
Su= 10
L
where,
Su-see equation 9
λ-corner fastener strength reduction fastener
d d Lv
λ=1-
240 t
dd-height of corrugation, in
Lv-purlin spacing, ft
t-sheet metal base thickness, in
L-length of diaphragm parallel to load, ft
Qf or Qfw-smaller value from equations 6, 7, and 8, k
B-fastener contribution term
B=ns(Qs/Qf or Qfw)+2npΣ(xp/w)2+4Σ(xe/w)2
ns-number of stitch screws between purlins
Qs-see equation 6
7
Qf or Qfw-see equations 7 and 8
np-number of intermediate purlins
xp-distance a sheet-to-intermediate purlin fastener is from
the sheet centerline, in
xe-distance a sheet-to-edge purlin is from the sheet
centerline, in
w-cover width of sheet, in
The sheet-to-purlin fasteners on the edge purlin are subjected to shear flow across the end of
the sheet as well as shear along the panel side/edge. The fastener on the panel corner will need
to resist the largest shear force along the panel edge because it is the farthest from the sheet
centerline thereby carrying the largest fastener load parallel to the corrugation and it will resist
the same shear force from shear along the end of the panel if it is assumed this shear
distributes equally among the sheet-to-edge purlin fasteners. Equation 11 gives the corner
fastener strength as:
N 2B2
Su=Qf or Qfw 11
L2 N 2 + B 2
where,
Su-see equation 7 and 8
N-number of sheet-to-edge purlin fasteners in sheet flat per foot of
sheet width
B-see equation 10
L-see equation 10
EtK
G’= 12
2.6( s / p ) + φDn + C
where,
8
E-modulus of elasticity of the steel sheet, 29500ksi
p-corrugation pitch, in
s-flat width to form one corrugation, in
s=p+2ww
ww-see equation 9
φ-modifier for number of spans in the length of a diaphragm
φ Number of spans
1.0 1
1.0 1
0.9 3
0.8 4
0.71 5
0.64 6
0.58 7 or more
Dn-panel warping coefficient
C-fastener flexibility coefficient
Equation 12 has three distinct terms that represent three different types of distortion that lead to
the overall diaphragm displacement. The first term (2.6(s/p)) is the displacement due to the
shear strain of the sheeting. The second is displacement due to the end warping of the sheet
due to shear flow across the end of the panels. This is the same shear flow that may lead to
corner fastener strength reductions, see equations 10 and 11. In effect, Dn is the average
amount of corrugation warp across the sheet end. Each corrugation may warp a different
amount depending on the rib geometry and sheet-to-edge purlin fastening. The individual
corrugation warping terms (Dni) are:
d d wt2 1
1.5
0.94 pV 2
Dni= 12
wt 25 L t
where,
V-number of corrugations between sheet-to-edge purlin fasteners
Rest of terms previously defined
1.5
d d wt2 1
Dni= 13
25 L t
9
Equation 13 should be used with caution. Post-frame buildings often use panels with profiles
that have narrow ribs with wide flats between them. Placing a sheet-to-edge purlin fastener on
one side of each rib in the flat will not contain warping on both ends of the diaphragm. Shear
flow pushing the rib towards the fastener will not warp the end of the panel nearly as much as
shear flow pushing the rib away from the fastener into the flat. The rib may become flat and the
flat portion of the panel will rise up in a parabolic fashion. The shear flow on the ends of the
diaphragm is in opposite direction so that one end will warp if the fastener pattern is consistent.
Also, it does not appear that the way ‘V’ is determined will account for this behavior. It is
recommended that equation 13 be used when there are sheet-to-edge purlin fasteners on each
side of the rib and they are next to the rib.
EtS f 24 L
C= 14
w 2∑ ( x e / w) + n p ∑ ( x p / w) + 2n s S f / S s
where,
Sf-flexibility of field fastener, in/k
Ss-flexibility of seam fasteners, in/j\k
Rest of terms previously defined
Ss=3/(1000t1/2) 15
Sf=1.5/(1000t1/2) 16
The fastening pattern used in this test program is shown in figure 4. A sheet-to-sheet-purlin
fastener is used. Therefore, B in equation 10 and C in equation 14 had to be modified as
follows.
B=4Σ(xe2/w2)+2npΣ(xp2/w2)+ns(Qs/Qf)+4(Qfseam/Qf)Σ(xeseam2/w2)
+2np(Qfseam/Qf)Σ(xpseam2/w2) 17
where,
Qfseam-strength of the sheet-to-sheet-purlin fastener, k
10
xeseam-distance sheet-to-sheet-to-edge purlin fastener is from
sheet centerline, in
epseam-distance sheet-to-sheet-intermediate purlin fastener
is from centerline of sheet, in
EtS f
C= X
w
24 L
2∑ xe / w) + n p ∑ ( x p / w) + 2n s ( S f / S s ) + 2( S fseam / S f )∑ xeseam / w) +n p ( S fseam / S f )∑ ( x pseam / w)
18
where,
Sfseam-flexibility of the sheet-to-sheet-to-purlin fastener, in/k
Diaphragm Construction
Five different cantilever and five different simple beam test construction were constructed.
Three replication of each construction were built except for the fifth construction of the simple
beam test, two replications were made for the fifth replication.
Cantilever Tests: The cantilever tests are labeled with a construction number of 1 through 5
followed by a ‘C’ indicating cantilever, then a the number 1-3 indicating the replication and then
the last number being a 1 (bending failure) or 2 shear failure. The diaphragms were nominally 9’
by 14’ with purlins 2’ o-c except for construction 5 which had purlins 30”o-c and was nominally 9’
by 12-5/8’. The purlins were all #2SPF 2x4s placed on edge over the rafters. The rafters were
MSR1650f 1.5E SPF 2x6s 16’ long. The purlins were attached to the rafters with a 60d threaded
hardened steel nails. Blocking (shear connectors) was used with all constructions except 1.
Blocking consisted of 2x4s nominally 22.5” long of the same material as the purlins and
attached to the rafters with the same 60d nails as used to attach the purlins. Construction 2 had
blocking between ever other purlin while constructions 3-5 had blocking between all purlins.
Figure 4 shows the sheet fastening used for all diaphragms. Three 1-1/2” long #10 screws were
used to attach the sheeting to the shear connectors. Construction 3 had a ¾” long #12 stitch
screw placed over the purlin through the ribs of the over lapping sheet. Construction 4 had ¾”
long #12 stitch screws placed the same as construction 3 with an additional stitch screw
centered between the purlins making the fasteners nominally 12”o-c.
Simple Beam Tests: The framing for the simple beam tests is shown figure 5. The rafters and
purlins are described in the previous section. The constructions were all nominally 12’ by 14’
with the rafter’s 4’o-c. The purlins were flat over the rafters and attached with two 0.131”
diameter nails that were 3-1/4” long. At purlin joints four nails were used, two in the end of each
11
purlin. The labeling was the same as for the cantilever test except an ‘S’ was used to denote
simple beam test and the fourth number is left out because the diaphragms only experienced a
shear failure. Blocking was used on the perimeter (support) rafters in all cases attached with
four nails like the ones used to attach the purlins to the rafters. The blocking was attached to the
sheet with three 1-1/2” long #12 screws same as the cantilever tests. Constructions 1-2 also
had blocking between the purlins on the load rafters. The blocking was not attached to the
sheeting though. The sheet fastening for 1S and 3S was the same as 1C, and 2S and 4S
fastening was the same as 3C. The sheet fastening for 5S is as shown in figure 4 with the
addition a ¾” long #12 stitch screw over each purlin and two additional stitch screws between
the purlins leaving the stitch screws 8”o-c.
Experimental Results: Table 1 gives the failure loads for all test run. The cantilever bending
moment failures were all purlin-to-rafter connections failures generally by the 2x6 rafter splitting
along the 60d threaded hardened steel nail. Table 2 gives the stiffness for the diaphragms and
the load that the stiffness was calculated at. Stiffness is defined as the load divided the
deflection it caused. The bending moment numbers would be useful if the purlins were not made
continuous over the rafters (truss top chords) and it was anticipated that the end wall/shear wall
would behave in a rigid manner making the diaphragm supports more fixed than pinned like a
simple beam. The load at which the stiffness was determined incorporates a safety factor of 2.5.
Test 1C second loading resulted in a combination failure of slip between the sheets at the
seams and transferring load into or out of the diaphragm (first line of fasteners along the load or
support rafter). The purlins also showed significant twist. The series average was 4283lb with a
COV of 15.9%. The stiffness average 7999lb/in with a COV of 14.3%.
Test 2C second loading failure was slip between the sheets with some waviness developing in
the flats of the sheets. The construction average was 5101lb with a COV of 3.9%.The seam
strength without stitch screws and with blocking controlling purlin twist and transferring load
into/out of the diaphragm is then 5101lb. The stiffness for the series was 20007lb/in and a COV
of 43.3%. Purlin twist and purlin-to rafter connection slip greatly impact stiffness as can be seen
by comparing the stiffness of 1C to 2C. The large COV was not expected and may be attributed
trying to determine small differences between relatively large numbers.
Test 3C failed again by seam slip with buckles developing along the ribs in the flats. The
average failure load was 6884lb (COV of 2.8%) showing that the stitch screws increased the
seam strength 35%. It is not known if the buckles influenced failure since ribs did not bend
upward and load did not drop off. The stiffness for the construction was 28996lb/in, but data was
not collected for the range tested except for one replication. Test 3C3-2 the tension purlin failed
in tension, but it was deemed that failure was imminent.
Test 4C failed by complete sheet buckling. At least five ribs across the diaphragm at a 30° from
the long side of the diaphragm bent upward with a loud metallic ping and the load dropped to
the point where the load apparatus shut down to prevent damage due to rapid acceleration of
12
the load apparatus. The average load was 7483lb with a COV of 0.1%. For the given sheet-to-
purlin fastening, this is the maximum capacity of the diaphragm. Additional stitch screws would
not be expected to add strength. The stiffness was 21738lb/in with a COV of 42.9%. The
stiffness is less than expected when compared to 2C, this may be due to the load at which the
stiffness is found being 40% of the average shear strength is well into the range of the load
deflection curve where deflection is increasing with little increase in load.
Test 5C was 12.625’ long with purlins 30”o-c. Failure was seam slip with some waviness of the
sheets in the flats between the ribs. This failure is very similar to that of 2C. The average failure
load was 4255lb with a COV of 2.0%. The average failure load is about 83% of 2C on an
adjusted length basis and is the same as 1C. The increased buckling length did not reduce the
buckling strength relative to the loss of seam strength due to loss of sheet fasteners to the point
where the diaphragm buckled. The stiffness was 18854lb/in with a COV of 42.9%. Again,
stiffness is affected more by purlin slip/twist than other diaphragm slip components, see 1C and
2C.
Test1S and 2S did not perform as expected. Constructions 1S1, 1S2, 1S3, 2S1, and 2S2 had
the rafters fail. 1S1 support rafter failed in tension and light gage metal plates were added for
the next replication. The plated increased the capacity as can be seen from table 1. Metal plates
were placed on each side of the rafter for construction 2 but 2S1 failed when the bolts pulled
through the rafter at the load device. Only 2S3 failed by sheet buckling at a load of 9806lb which
is about 2/3 of the expected load as indicated by 4C. The stiffness of 1S was 22262lb/in (COV of 70.7%)
and 26120lb/in (COV of 19.3%) for 2S. The stiffness values are comparable to 3C and 4C. It
should be pointed out that the tension edge purlin was discontinuous. A gap opened at the
discontinuity but the purlin did not pull apart. Load may have transferred to the continuous
intermediate purlin next to the edge purlin thereby preventing failure.
Tests 3S, 4S, and 5S all failed when the load rafter(s) pulled away from the purlins. Without the
blocking on the load rafters, the nailed purlin-to-rafter connections were insufficient to fail the
diaphragm in shear. This connection had a an average strength of 7398lb for 7 connection
across all eight tests. Since the purlin-to-rafter connection failed, no shear strength data was
obtained and no comparisons can be made relative to sheet fastening as desired. The simple
beam test may not lend itself well to determining diaphragm shear strength unless purlin-to-
rafter connections can be reinforced. The stiffness was higher for these constructions even
though purlin-to-load rafter slip was noticeably large. The stiffness was 27563lb/in (COV of
33.4%), 37218lb/in (COV of 8.7%), and 32444lb/in (COV of 31.4%) for 3S, 4S, and 5S
respectively. The COVs are in the anticipated range leading one to conclude that the loads were
still in the more linear portion of the load deflection curve. 3S3 and 4S2 had the discontinuity in
the tension edge purlin. Again, the joint opened but did not fail at loads of 7336lb and 7200lb
respectively.
Predicted Strength and Stiffness: The strength and stiffness of constructions 2C, 3C, 4C, and
5C were predicted using equations 6-18. Construction 1C strength and stiffness was not
predicted because the failure mode was transferring load into/out of the diaphragm which is not
considered in equations 6-18. Constructions 1S, 2S, 3S, 4S, and 5S were not predicted since
failures included rafter failure and purlin-to-rafter connection failure which are again not
addressed by the equations presented by Luttrell and Mattingly (2005). As stated, the
13
diaphragms tested had sheet-to-sheet-purlin fasteners which do not behave like sheet-to-purlin
or sheet-to-sheet fasteners, see figure 4. The strength of these fasteners was assumed to be
like elevated side lap fasteners. The strength (Qf in kips) is given by Luttrell and Mattingly (2005)
as:
d=0.145” Qf=17.3t
d=0.177” Qf=-31.8t
d=0.187 Qf=33.5t
Luttrell and Mattingly (2005) require a minimum penetration depth of 4d with linear interpolation
up to 7d when the full values given above are valid. This requirement appears to be
unnecessary. If the fastener is prevented from tilting, both sheets will bear on the fastener
causing a moment that tries to tilt the fastener of Qf times the coated sheet thickness (tc). The
force at the surface of the purlin required to resist this tilting moment is the tilting moment
divided by the rib height (Qf tc/dd). The rib height is usually a magnitude of order or greater than
the coated sheet thickness making the force at the purlin needed to resist fastener tilting small.
Construction 2C had a field fastener strength of 5317lb (Eq 10), a corner fastener strength of
5387lb (Eq 11), and a buckling strength of 11386lb (Eq 9). The test failure was seam failure at
5101lb. The model predicted the correct failure and the ratio of test to predicted strength was
0.9594. Construction 3C had a predicted field fastener strength of 7277lb, a corner fastener
strength of 7304lb while the buckling strength remained the same. Again, the seam failure was
predicted correctly and the strength ratio was 0.9460. Construction 4C had stitch screws 12”o-c.
The predicted field strength was 8992lb and the predicted corner fastener strength was 8958lb
very similar to the field fastener strength like 2C and 3C. The buckling strength is the same. The
test failure was sheet buckling at 7438lb which yield a strength ratio of 0.8303. However,
buckling was not the predicted failure mode, corner fastener failure was predicted. The last
construction 5C with purlins 30”o-c had a field fastener strength of 4203lb, a corner fastener
strength of 4276lb, and a buckling strength of 7289lb. The test failure was a seam failure at
4255lb giving a strength ratio of 1.0124. The method described by Luttrell and Mattingly (2005)
predicts strength within 5% of test and predicts the correct failure except for sheet buckling. This
difference may be accounted for by the load duration factor for the tests being approximately 1.5
not 1.6 which would be used for short term tests (10-15 minutes). Sheet buckling is the upper
limit of diaphragm strength for a given sheet profile and sheet-to-purlin fastening. Additional
seam fastening once sheet buckling is reached will yield no increase in diaphragm strength. The
buckles generally form by aligning in the flats along the major ribs. The buckling ribs will form at
some angle across the diaphragm that is affected by the sheet-to-purlin fastening. The buckle
does not go through sheet-to-purlin fasteners making it likely that more sheet-to purlin fasteners
will confine the buckle more increasing the diaphragm buckling strength. If a sheet were glued
to the purlins, then the buckles would be confined between purlins and would not form a pattern
between sheet-to-purlin fasteners across the diaphragm. This would yield the highest
diaphragm buckling strength. Equation 9 does not account for this affect.
Equation 12 was used to predict the stiffness of the four diaphragm constructions. The predicted
stiffness values were 49.0k/in, 60.1k/in, 66.7k/in, and 45.1k/in for 2C, 3C, 4C, and 5C
respectively. The test stiffness values in construction order were 20.0k/in, 29.0k/in, 21.7k/in, and
14
18.9k/in. The test stiffness values are 40.8%, 48.4%, 32.6%, and 41.8% of the stiffness values
predicted. The predicted stiffness values are running approximately 2.5 times the test values.
The test data is generally scattered which may account for some discrepancy between the test
and predicted values. However, it is not known at what load Luttrell and Mattingly (2005)
equations are predicting stiffness. The test stiffness is found at a load of 40% of the average
ultimate load which may not be consistent with Luttrell and Mattingly (2005). Lumber framed
diaphragms often exhibit nonlinear load-deflection curves. The blocking may not have
successfully eliminated all purlin-to-rafter connection slip and purlin twist which would also lead
to discrepancies in stiffness values.
Summary
Based on the test results presented in this paper and the modeling presented the following
summary of results can be made.
1. The design load that can be transferred into and out of the diaphragm without shear
connectors is 122plf for the standard fastening pattern and purlins 2’o-c.
2. The design seam strength of the standard fastening pattern is 146plf with purlins 2’o-
c and 122plf with purlins 30”o-c.
3. The design buckling strength of the sheet with purlins 2’o-c is 213plf with the
standard sheet-to-purlin fastening pattern.
4. An additional ¾” long #12 stitch screw at the purlin increases the design seam
strength to 197plf.
5. The strength equations of Luttrell and Mattingly predicted seam/corner fastener
strength within 5% of test values. The failure mode was also correctly predicted. The
buckling strength of the sheet for the standard fastening pattern is not correctly
predicted and the test buckling strength is 65% of the predicted.
6. The stiffness predicted by Luttrell and Mattingly (2005) is 2.5 times that determined
from testing.
15
References
_____.2004. Standard test method for static load testing of framed floor or roof diaphragm
constructions for buildings. ASTM E455-04. ASTM, West Conshohocken, PA.
______. 2004. Load tests for metal-clad, wood-frame diaphragms. ANSI/ASAE EP558. ASAE,
St. Joseph, MI.
Anderson, G.A. and D.S. Bundy. 1992. Estimating the stiffness and strength of roof panels from
test panels. Post Frame Building Design, ed Walker, J.N and F.E Woeste, ASAE Monograph
No. 11, ASAE, St. Joseph, MI.
Anderson, G.A. and D.S. Bundy. 1989. Characterizing diaphragm shear stiffness for diaphragm-
frame interaction analysis. Transaction of the ASAE. ASAE, St. Joseph, MI.
Atrek, E. and A.H. Nilson.1980. Nonlinear analysis of cold-formed steel shear diaphragms.
ASCE Journal of the structural Division 106(ST3):693-710.
Boone, G.R. and H.B. Manbeck. 1985. Using existing models to predict diaphragm behavior.
ASAE Paper No. 85-4501. ASAE, St. Joseph, MI..
Davies, J.M. and E.R, Bryan. 1982. Manual of stressed skin diaphragm design. John Wiley &
Sons, Inc., New York.
Luttrell, L.D. 1981. Steel Deck Institute diaphragm design manual. Steel Deck Institute, Canton,
OH.
Luttrell, L.D. and J.A. Mattingly. 2005. A Primer on Diaphragm Design. 1st ed. Metal
Construction Association, Glenview, IL.
16
Table 1. Diaphragm failure loads and moments.
17
5C1 1256 792.1 4351
5C2 1532 966.1 4205
5C3 1777 1121 4208
Mean 1522 959.8 4255
COV 17.1% 17.1% 2.0%
1S1 6001
1S2 8401
1S3 7449
Mean 7450
COV 15.3%
2S1 9001
2S2 7601
2S3 9806
Mean 8803
COV 11.3%
3S1 7569
3S2 6997
3S3 7336
Mean 7309
COV 3.7%
4S1 7570
4S2 7200
4S3 8112
Mean 7627
COV 5.4%
5S1 7403
5S2 6971
Mean 7187
COV 3.5%
18
Table 2. Design diaphragm shear capacity, moment capacity, and stiffness design
capacities from testing.
Stiffness Stiffness
Design at Design Design at Design
Moment Moment Shear Shear
Capacity Capacity Capacity Capacity
Test ft-lb/ft lb/in lb/ft lb/in
19
5C2 “ 18959 “ 15379
5C3 “ 14011 “ 28090
Mean 16356 18854
COV 15.2% 42.9%
20
Figure 1. Cantilever diaphragm test.
21
Figure 2. Simple beam diaphragm test setup.
22
Figure 3. Location of displacement measuring devices for testing.
23
Figure 5. Framing for the simple beam tests.
24
Figure 4. Standard sheet-to-purlin and sheet-to-sheet-to-purlin fastening used in test program.
25