Clemson University TigerPrints All eses eses 5-2015 Investigation of Shear Capacity for Light-Frame Wood Walls Constructed with Insulated Oriented Strand Board Panels Ross Johnson Phillips Clemson University Follow this and additional works at: hps://tigerprints.clemson.edu/all_theses is esis is brought to you for free and open access by the eses at TigerPrints. It has been accepted for inclusion in All eses by an authorized administrator of TigerPrints. For more information, please contact [email protected]. Recommended Citation Phillips, Ross Johnson, "Investigation of Shear Capacity for Light-Frame Wood Walls Constructed with Insulated Oriented Strand Board Panels" (2015). All eses. 2506. hps://tigerprints.clemson.edu/all_theses/2506
121
Embed
Investigation of Shear Capacity for Light-Frame Wood Walls Constructed with Insulated Oriented
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Clemson UniversityTigerPrints
All Theses Theses
5-2015
Investigation of Shear Capacity for Light-FrameWood Walls Constructed with Insulated OrientedStrand Board PanelsRoss Johnson PhillipsClemson University
Follow this and additional works at: https://tigerprints.clemson.edu/all_theses
This Thesis is brought to you for free and open access by the Theses at TigerPrints. It has been accepted for inclusion in All Theses by an authorizedadministrator of TigerPrints. For more information, please contact [email protected].
Recommended CitationPhillips, Ross Johnson, "Investigation of Shear Capacity for Light-Frame Wood Walls Constructed with Insulated Oriented StrandBoard Panels" (2015). All Theses. 2506.https://tigerprints.clemson.edu/all_theses/2506
Modeling results for the sheathing with 1-in. insulation demonstrated a higher
capacity for shear walls constructed using screws as compared to those using nails for
35
both spacing schedules (Table 3.1). For all cases the drift limit controlled the design
value for this wall setup with ranges of 177-213 plf and 218-267 plf for the 4/12 and 3/6
spacing schedules, respectively.
Comparison of displacements at the maximum load and ultimate failure between
the two fastener schedules differed generally by less than 0.15 in. However, the
difference between displacements at maximum load and ultimate load for each fastener
showed a wider range depending on the stiffness and failure mode for the fastener
connection. For example, 8d and 0.148 smooth nails were characterized by withdrawal
failure in the connection tests and demonstrated much greater differences between
maximum displacement and yield displacement than all other fasteners.
The percent differences between 3/6 spacing and 4/12 spacing with respect to the
ultimate limit and drift limit were an average increase of 30 percent and 19 percent,
respectively.
36
Table 3.1. M-CASHEW2 results for 1 in. insulated shear walls.
Fastener Spacing
(in) Fastener Maximum Load (lbf)
Displacement at Maximum
Load (in.)
Displacement at Ultimate Failure
a (in.)
Ultimate Limit
b
(plf)
Load at 0.2 in. (lbf)
Drift Limit (plf)
4/12
0.131 smooth
nail 8160 5.35 7.40 313 1574 197
0.131 ring nail
6710 3.15 3.70 257 1413 177
0.148 smooth
nail 8000 3.80 5.30 307 1671 209
Stainless steel screw
9570 3.80 4.55 367 1697 212
Carbon screw
9100 3.35 4.00 349 1791 224
Prototype screw
10860 3.75 4.50 416 1706 213
3/6
0.131 smooth
nail 11010 5.40 7.30 422 1907 238
0.131 ring nail
9080 3.30 3.75 348 1746 218
0.148 smooth
nail 10810 3.90 5.40 414 2015 252
Stainless steel screw
12910 3.95 4.65 495 2038 255
Carbon screw
12340 3.45 4.10 473 2135 267
Prototype screw
14730 3.90 4.65 565 2039 255
a Ultimate failure of the wall was selected as 80% of the maximum load.
b Ultimate Limit was determined by dividing the maximum load by the wall length, multiplying by an
adjustment factor for specific gravity, and dividing by a safety factor of 3.
The 0.131 ring nail exhibited the lowest capacity of all fasteners given its small
head diameter, small shank diameter, and low initial stiffness. This fastener typically
failed in shear at the smaller section of the deformed shank (i.e. in-between the rings)
during the cyclic fastener connection tests. The carbon screw had a higher drift limit
which was associated with the higher initial stiffness identified in the fastener connection
37
tests, whereas the prototype screw demonstrated the highest ultimate limit which
appeared to be correlated with head diameter. Regression analysis of fastener properties
with the ultimate limit and drift limit more clearly illustrated these correlations (Figure
3.2).
Figure 3.2. Correlation of fastener head diameter with ultimate limit (a) and drift limit (b) and initial stiffness with ultimate limit (c) and drift limit (d) for 1 in. insulation shear walls. Note: Ultimate and drift limits represent 3/6 fastener spacing.
3.3.2 Results for panels with 2 in. insulation
Capacity for the large diameter fasteners used with the 2 in. insulation was
generally lower than that of the 1 in. insulation shear walls, as expected. The differences
between nails vs. screws were less dramatic in part because 3 nails and 1 screw tested
R² = 0.8477
0
100
200
300
400
500
600
0.000 0.200 0.400 0.600
Ult
imat
e L
imit
(p
lf)
Head Diameter (in) a)
R² = 0.2575
0
50
100
150
200
250
300
0.000 0.200 0.400 0.600D
rift
Lim
it (
plf
)
Head Diameter (in) b)
R² = 0.2532
0
100
200
300
400
500
600
0 2000 4000 6000
Ult
imat
e L
imit
(p
lf)
Initial Stiffness (lbf/in) c)
R² = 0.8793
0
50
100
150
200
250
300
0 2000 4000 6000
Dri
ft L
imit
(p
lf)
Initial Stiffness (lbf/in) d)
38
were designed specifically for this wall system. The shoulder nails were created to
increase the initial stiffness of the fastener connection by adding material below the nail
head making that portion of the shank (“the shoulder”) larger than the rest of the shank,
thus creating a tighter fit where the nail head contacted the sheathing. The newly
developed screw (referred to as New DSV screw) modified a previous screw by
increasing the head size and decreasing the length of screw thread so that it would not
extend into the framing-insulation slip plane (Figure 3.3). These changes yielded higher
drift limits as compared to ring- or smooth-shank nails or smaller head diameter fasteners
(0.162 nails and DSV screw).
One of the shoulder nails was also heat treated to increase its bend yield strength
with the idea that its stiffness and shear capacity would increase as well. For this fastener
(0.148 hard shoulder nail), the heat treatment increased the ultimate load as well as the
drift load as compared to the untreated 0.148 shoulder nail.
As seen with the 1 in. insulation modeling, M-CSAHEW2 predicted a higher
ultimate limit than drift limit for all fasteners for both spacing schedules. The SDWS224
screw had the highest capacity of all fasteners with a maximum loading of 11,670 lbf
(and highest ultimate limit of 447 plf) as well as the highest drift limit of 192 plf (Table
3.2). The 0.148 shoulder nail had the lowest maximum load (5180 lbf) while the 0.162
ring nail showed the lowest drift limit of 137 plf.
39
Figure 3.3. Modifications made to the DSV screw to enhance its shear capacity when used with 2 in. insulation OSB.
With respect to wall displacement, the 0.162 smooth nail had the largest wall
displacement at maximum load and ultimate failure indicating that these walls were more
ductile than walls with other fasteners. On the contrary, the light-duty structural insulated
panel (SIPLD) screw had the least amount of displacement and the smallest displacement
difference between the maximum load and ultimate failure suggesting that these walls
would experience sudden, brittle failure.
Trends for 3/6 fastener capacities (Table 3.3) were similar to those observed in the
modeling using 4/12 spacing. The differences between the two spacing schedules were 30
percent for the ultimate limit and 21 percent for the drift limit.
40
Table 3.2. M-CASHEW2 results for 2 in. insulated shear walls 4/12 fastener spacing.
Fastener Maximum Load (lbf)
Displacement at Maximum
Load (in.)
Displacement at Ultimate Failure
a (in.)
Ultimate Limit
b
(plf)
Load at 0.2 in. (lbf)
Drift Limit (plf)
0.148 shoulder
5180 5.60 7.00 199 1319 165
0.148 hard shoulder
7360 4.95 7.10 282 1364 171
0.162 shoulder
6280 4.60 7.40 241 1453 182
0.162 smooth
7200 7.30 9.65 276 1430 179
0.162 ring 5740 5.00 6.80 220 1093 137
0.203 ring 8200 4.80 6.55 314 1406 176
New DSV screw
7040 4.70 5.90 270 1338 167
DSV screw 6270 4.80 6.50 240 1259 157
Stainless steel screw
8570 5.75 7.55 329 1459 182
SIPLD screw 8700 3.60 4.30 334 1480 185
SDWH194 screw
10470 5.15 6.15 401 1522 190
GRK-RSS screw
10900 5.20 6.30 418 1496 187
SDWS224 screw
11670 5.75 7.00 447 1533 192
a Ultimate failure of the wall was selected as 80% of the maximum load.
b Ultimate Limit was determined by dividing the maximum load by the wall length, multiplying by an
adjustment factor for specific gravity, and dividing by a safety factor of 3.
41
Table 3.3. M-CASHEW2 results for 2 in. insulated shear walls 3/6 fastener spacing.
Fastener Maximum Load (lbf)
Displacement at Maximum
Load (in.)
Displacement at Ultimate Failure
a (in.)
Ultimate Limit
b
(plf)
Load at 0.2 in. (lbf)
Drift Limit (plf)
0.148 shoulder
6980 5.65 6.85 268 1639 205
0.148 hard shoulder
9920 5.05 7.05 380 1692 212
0.162 shoulder
8470 4.65 7.45 325 1790 224
0.162 smooth
9710 7.40 9.65 372 1760 220
0.162 ring 7750 5.05 6.90 297 1393 174
0.203 ring 11050 4.90 6.75 424 1730 216
New DSV screw
9490 4.80 5.90 364 1671 209
DSV screw 8460 4.85 6.45 324 1580 198
Stainless steel screw
11560 5.80 7.55 443 1799 225
SIPLD screw 11880 4.35 4.45 455 1808 226
SDWH194 screw
14130 5.20 6.25 542 1854 232
GRK-RSS screw
14710 5.30 6.40 564 1823 228
SDWS224 screw
15750 5.85 7.10 604 1861 233
a Ultimate failure of the wall was selected as 80% of the maximum load.
b Ultimate Limit was determined by dividing the maximum load by the wall length (per unit area),
multiplying by an adjustment factor for specific gravity, and dividing by a safety factor of 3.
Regression analyses for the 2 in. fasteners showed positive associations of shank
diameter (R2=0.616) and head diameter (R
2=0.860) with ultimate limit – increasing shank
diameter or head diameter correlated with a higher ultimate limit (Figure 3.4). However,
fastener properties that one might assume to be correlated with drift limit (e.g. shank
diameter, initial stiffness, or bend yield strength) showed no distinct relationship with this
parameter.
42
Figure 3.4. Regression analyses for 2-in. fasteners with respect to ultimate limit and drift limit predicted by M-CASHEW2: a) and b) shank diameter; c) and d) head diameter; e) and f) bend yield strength; g) and h) initial stiffness.
R² = 0.7267
0
200
400
600
800
0.000 0.100 0.200 0.300
Ult
imat
e L
imit
(p
lf)
Shank Diameter (in) a)
R² = 0.3663
0
50
100
150
200
250
0.000 0.100 0.200 0.300
Dri
fit
Lim
it (
plf
)
Shank Diameter (in) b)
R² = 0.8404
0
100
200
300
400
500
600
0.000 0.200 0.400 0.600 0.800
Ult
imat
e L
imit
(p
lf)
Head Diameter (in) c)
R² = 0.397
0
50
100
150
200
250
0.000 0.200 0.400 0.600 0.800D
rift
Lim
it (
plf
)
Head Diameter (in) d)
R² = 0.2434
0
200
400
600
800
0 100 200 300
Ulit
imat
e L
imit
(p
lf)
Bend Yield (ksi) e)
R² = 0.0586
0
50
100
150
200
250
0 100 200 300
Dri
ft L
imit
(p
lf)
Bend Yield (ksi) f)
R² = 0.2421
0
200
400
600
800
0 1000 2000 3000 4000
Ult
imat
e L
imit
(p
lf)
Initial Stiffness (lbf/in) g)
R² = 0.5822
0
50
100
150
200
250
0 1000 2000 3000 4000
Dri
ft L
imit
(p
lf)
Initial Stiffness (lbf/in) h)
43
3.4 Discussion and Summary
Validation of M-CASHEW2 for shear wall modeling has been addressed in
previous research and showed good agreement between model results and actual shear
wall tests. A comprehensive analysis of M-CASHEW2 was performed by Shirazi (2012)
where he compared model results from four different shear wall experiments. In all cases,
M-CASHEW2 did well in predicting the general trend of the shear walls’ backbone
curves for monotonic tests, but the accuracy of absolute values for the model and tests
varied, which were attributed to different materials used (e.g. 0.131 smooth nails vs. 8d
box nails) or variability in wood properties
Other studies demonstrated differences between 6-10 percent for the displacement
at ultimate load while the ultimate load itself was 5-18 percent different using the
CASHEW model (Judd and Fonesca 2005) and indicated initial stiffness may be
overestimated while peak load and post-peak behavior displacement may be
underestimated (Li 2007). However, the overall reliability of the models for predicting
shear wall response under different shear wall test scenarios has been proven to be
relatively accurate.
The results presented here showed that shear walls constructed with nails had
lower capacities than those constructed with screws for both insulation thicknesses.
Decreasing the fastener spacing (and thus increasing the number of fasteners per wall)
yielded increases in shear capacities of approximately 20-30 percent. The overall capacity
may actually be overestimated where values of more than 8160 lbf and 5180 lbf were
predicted for the 1-in. and 2-in. walls, respectively. Compare these values to 3911 lbf
44
(Folz and Filiatrault 2001), 4586 lbf (Judd and Fonesca 2005) and 6901 lbf (Fonesca et
al. 2009) for conventional shear wall construction. It would be expected that including a
layer of insulation between the sheathing and framing members would decrease the shear
capacity of the wall; however, the different fastener spacing schedules and non-traditional
fasteners used could be contributing to the higher than expected shear wall strengths, as
was initially hypothesized. Performing actual shear wall tests with this wall system and
comparing those results with the predicted values demonstrated that the model used for
this research showed good agreement between the overall values and behavior of the wall
(see Chapter 4).
Comparing the amount of material for each fastener and the results from the
modeling scenarios can provide a better idea as to which fastener may be more efficient
in its performance. For example, in the 1-in. insulation tests the difference between the
amounts of steel used for stainless steel screw was 12 percent less than for the prototype
screw yet the difference in drift limit (controlling value) between these two fasteners was
negligible. Therefore, it could be argued that the stainless steel screw would be a better
choice. The same could be said for the carbon screw vs. the prototype screw where there
as a 5 percent difference in steel but the carbon screw achieved a higher drift limit than
the prototype screw (with less material). For the 2-in. insulation, the New DSV screw had
approximately 40 percent less material than the SDWS screw but only a difference of 11
percent in the drift limit. The SIPLD screw also contained less steel than the SDWS
screw (14 percent) and had only a 3 percent difference in drift limit values. However,
45
other factors must be considered (i.e. cost of each fastener, failure type of the fastener,
rate of installation, etc.) when deciding which fastener to use.
Identifying fastener properties that affect wall behavior would be important to
help better understand the dynamics of the structure and allow one to make certain
changes based on objectives of that particular situation. As discussed previously in
Chapter 2, the geometry of the fastener can be important for influencing the shear
capacity of walls. According to results presented, the shank diameter and head diameter
of the fastener appeared to be correlated with the ultimate limit for shear wall design and
initial stiffness of the connection had a positive correlation with the drift limit. Trends
identified for fastener properties and design values in single-fastener tests also need to be
verified for full-scale shear wall tests.
46
Chapter 4. Full-Scale Shear Wall Tests
4.1 Introduction
Full-scale shear wall tests were performed to evaluate the capacity of the 2-in.
insulated OSB panels and to compare the test results with predictions from M-
CASHEW2 models previously developed from small-scale fastener connection tests.
Preliminary tests were performed using 1-in. insulated OSB with 0.131 smooth nails to
compare with previous tests and ensure that the setup and results were comparable.
Assuming the model can accurately predict shear wall capacity and behavior under static
loads, different scenarios with a variety of fasteners could be used to help maximize the
performance of this wall system providing better performance of these walls under lateral
loading for actual structures. This would give design professionals another option for
increasing light-frame wood construction energy efficiency while not modifying the
existing construction techniques for this structure type.
A comprehensive evaluation of wood shear wall testing and modeling was
provided by van de Lindt (2004); however, new products, materials, and technologies
continue to be introduced in this field. Research on these materials must be performed to
better understand how these innovations affect the structures in which they are
incorporated. The results presented are intended to help expand the knowledge of how
continuous insulation can affect shear wall capacity when placed between the framing
and sheathing of light-frame wood shear walls.
47
4.2 Methodology
Monotonic shear wall tests were performed on eight, 8 ft. x 8 ft. wood shear walls
sheathed with insulated OSB panels to identify the shear capacity of this wall system.
Four different fasteners were used for these tests with two replicates per fastener. The
amount of lateral load applied to the wall was recorded until the wall reached failure
(defined as 80 percent of the maximum load) and wall deflection was measured. Behavior
of the wall components (framing members, OSB sheathing, and fasteners) were
documented and compared to an existing shear wall model to examine how well the
model could predict the performance of this shear wall system.
4.2.1 Test Setup
The shear wall test frame was intended to be used for a variety of different wall
types and various shear wall tests and has the capacity to accommodate up to 50 kips of
applied lateral load. The configuration of the frame was such that a lateral force can be
applied to the wall through a “spreader bar” attached directly to an actuator by a 0.5-in.
thick base plate (Figure 4.1). The spreader bar was a 6 in. x 6 in. x 0.25 in. hollow steel
section that was kept in place by two brackets with rollers which ensured smooth,
straight-line horizontal movement. Two upright I-sections provided out-of-plane stability
and were used as reference points for measuring wall displacement. These uprights
served as the mounting locations for the brackets. The spreader bar was attached to the
wall specimen using 0.625-in. threaded rods through which the force was transferred
from the actuator to the wall.
48
Figure 4.1. Setup for shear wall test frame with wall specimen.
The bottom of the frame was an 18-in. I-section with 8-in. flanges turned on its
side forming a channel where the base of the wall was attached to the frame by a timber
spacer that was permanently affixed to the frame. Four footers (W6x25 sections) attached
to the frame bottom were anchored directly to the concrete slab floor.
Tie down rods using high-strength, 1-in. diameter steel coil rods, 11 ft. long were
attached directly to the flanges of the bottom I-section and passed through a roller sitting
on top of the spreader bar. This roller allowed the tie rods to remain vertical as the
spreader bar moved horizontally when a lateral force was applied to the wall. The
purpose of the tie rods was to serve as hold-downs as the wall was not anchored to the
bottom of the frame with hold down devices (ASTM 2013).
49
A single ended actuator with a capacity of 146 kips was used for this test setup.
The actuator had a stroke length of 42 in. and an internal force transducer with a capacity
of 110 kips. It was attached to the shear wall frame in a horizontal position using four, 1.5
in. bolts and suspended with two large chains from turnbuckles with jaw fittings on each
end. The actuator was controlled through a computer interface which controlled the rate
of load application and recorded time, displacement of the actuator hydraulic piston, and
the force applied.
4.2.2 Wall Construction
Shear wall construction followed ASTM E72 (ASTM 2013) specifications.
Framing members were kiln-dried 2 x 4 nominal Douglas fir lumber, structural grade,
Class C. Specific gravity was calculated by measuring the dimensions of each timber,
obtaining its weight, and determining its moisture content with a moisture meter. Only
those lumbers that had a specific gravity within ±0.3 of the accepted standard for Douglas
fir (0.5) were used (ASTM 2013). Moisture content of the lumber averaged 8.2 percent.
Each wall consisted of a single sill plate, double top plate, and studs spaced 24 in.
on-center. Additional studs were placed 0.75 in. from the end studs to which they were
“stitched” using six 16d common nails that were 3.5 in. long and spaced 6 in. apart at the
top, middle, and bottom of the stud. This configuration created a space-column providing
additional stiffness to the wall while minimizing the amount of material used. The studs
were connected to the sill plates using two 16d common nails. The double top plate was
connected using two 10d common nails spaced 4 in., 18 in., and 34 in. from each end of
the wall (Figure 4.2). Pilot holes were drilled in the two end studs for each nail to prevent
50
splitting in the wood given the large number of nails concentrated at these locations. All
nails for the framing members were installed using a pneumatic palm-nail tool.
For the 1-in. insulated OSB panels, a 3/6 fastener spacing was used for 0.131
smooth nails, 3.25 in. long. The nails were placed 0.375 in. from the panel edges and
offset 1 in. along the seam where the two panels abutted each other. This configuration
replicated tests performed previously to ensure that the test frame and wall setup used for
this project produced results comparable to previous tests. All panel-to-frame nails were
installed using a standard, pneumatic framing nail tool.
51
Figure 4.2. Shear wall configuration for full-scale shear wall tests. A slight modification to this setup was used as studs were spaced 24 in. on-center rather than 16 in. on-center. Modified from ASTM E72-13a (2013).
52
The 2-in. insulated panels where installed using the same technique as the 1-in.
panels except, the fastener spacing was 4/12 and pilot holes were drilled for each panel-
to-frame fastener to ensure the fastener would be oriented at the correct angle to penetrate
the framing member properly. Based on the modeling results and discussions with
experts familiar with the materials used for this wall system, it was decided to use the
0.148 shoulder nail, 0.148 hardened shoulder nail, New DSV screw, and the SIPLD
screw for full-scale shear wall testing with the 2-in. insulated panels. As mentioned
previously, the shoulder nails and New DSV screw were developed specifically for these
insulated OSB panels for shear wall construction. These fasteners were designed to
increase the initial stiffness of these walls and increase the maximum load the walls could
sustain. The SIPLD screw is used frequently for rigid foam insulation and is readily
available; therefore, it was included in the test series.
Nails were installed using the palm nail tool whereas screws were installed with a
standard electric drill. The heads of the fasteners were placed flush with the sheathing
face except for the SIPLD screw, which would cause excessive damage to the panel
around the fastener head if it had been driving flush with the sheathing face.
4.2.3. Testing Protocol
Each wall was attached to the test frame using four, 0.5-in. bolts spaced 24 in.
apart for the sill plate and three, 0.625-in. threaded rods spaced 32 in. apart for the top
plate. The top plate was connected to the spreader bar, through which the force from the
actuator was transferred to the wall. String potentiometers were attached to the top right
corner, the bottom right corner, and bottom left corner of the wall. A total of four string
53
potentiometers were used to measure drift (Δ1), sliding (Δ2), uplift (Δ3), and compression
(Δ4) of the wall allowing for accurate measurement of total wall deflection (Figure 4.3).
Tie rods were used as hold downs as per ASTM E72 (ASTM 2013) and tightened to the
standard’s specifications.
The loading protocol required a force to be applied at a uniform rate of 400
lb./min. until reaching 3 different target loadings: 790 lbf (Stage 1); 1570 lbf (Stage 2);
and 2360 lbf (Stage 3). After attaining each target loading, the force was removed from
the wall at the same rate as it was applied until no load remained acting on the wall. The
specimen was then allowed to “rest” for 5 minutes before the next stage began (ICC-ES
2013). At the completion of Stage 3, the wall specimen was loaded until failure, which
was defined as 80 percent post-peak of the maximum load. Data points were recorded
every 0.25 sec. for the actuator force, actuator displacement, and string potentiometers for
the entirety of the test.
Once failure of the wall was achieved, documentation of the wall’s condition was
recorded by noting the locations of high stress (locations of fastener failure), identifying
failure mode of each wall fastener and its location, and photographic evidence of unusual
wall behavior (e.g. splitting of sill plate or buckling of framing members).
Two walls were tested for each fastener type unless the variation between the
specimens was greater than 15 percent, at which point a third wall would be added to the
test matrix. Test results presented below showed none of the wall specimens exceeded
this limit; therefore, only two walls were required.
54
Comparisons between the test results and the models created using M-CASHEW2
were made to identify how well the model predicted each wall’s shear capacity.
Regression analysis was performed to examine relationships between fastener properties
and wall shear capacity and stiffness.
Figure 4.3. Shear wall test setup showing locations of string potentiometers (Δ1, Δ2, Δ3, and Δ4). Reproduced from ASTM E72-13a (2013).
55
4.3 Results
4.3.1 Shear Wall Testing
Preliminary tests to verify the equipment and test procedure were working
properly yielded results similar to those obtained in previous shear wall tests performed
by the Huber Engineered Woods using 1-in. insulated sheathing. Therefore, 2-in.
insulation shear wall tests were subsequently performed without making any changes to
the testing protocol discussed above.
The 0.148 shoulder nail demonstrated the lowest capacity attaining a maximum
load of 4520 lbf whereas the SIPLD had a significantly greater capacity of 7370 lbf
(Table 4.1). These values translated into ultimate limits of 173 plf and 223 plf,
respectively. These results were expected as the 0.148 shoulder nail had the lowest bend
yield strength, smallest head diameter, and smallest shank diameter. Increasing the bend
yield strength resulted in a slightly higher capacity (5811 lbf) as evidenced by comparing
the unhardened nail with the hardened nail. The final wall deflection at failure was
identical between the two nails (4.538 in.) but the drift limit differed, likely due the
stiffness of the fastener itself. For both nails, the ultimate limit was lower than the drift
limit and therefore would be considered the design value for these walls.
Comparison between the two screw fasteners showed the smaller diameter screw
(New DSV) had a lower capacity and greater drift than the SIPLD screw. Ultimate limits
for these screws were 255 plf for the New DSV screw and 283 plf for the SIPLD screw.
56
In contrast to the nail fasteners, the screws had lower drift limits than ultimate limits
giving them design values of 193 plf (New DSV screw) and 218 plf (SIPLD screw) based
on these test results. Shank diameter, head diameter, and bend yield contributed to the
differences in capacity. However, the bend yield strength probably had less influence on
the overall capacity as once that value exceeds 145,000 psi there is little difference in
shear wall performance (Anderson et al. 2007). But the bend yield strength did affect the
failure mode of the fastener and thus the overall failure of the wall.
Table 4.1. Full-scale shear wall test results for maximum load, deflection, ultimate limit and drift limit for each fastener. Average of two walls presented with standard deviation in parentheses.
Fastener Maximum Load (lbf)
Deflection at
Maximum Load (in.)
Ultimate Limit (plf)
a
Ultimate Limit COV
Drift Limit (plf)
b
Drift Limit COV
Design Value (plf)
0.148 shoulder
nail
4520 (303.337)
4.538 (0.256)
173 (11.6) 0.07 196 (23.2) 0.12 173
0.148 hard shoulder
nail
5811 (580.132)
4.538 (0.923)
223 (22.2) 0.10 230 (2.5) 0.01 223
New DSV screw
6662 (267.119)
5.926 (0.355)
255 (10.2) 0.04 193 (23.8) 0.12 193
SIPLD screw 7370
(397.894) 4.472
(0.388) 283 (15.2) 0.05 218 (28.7) 0.13 218
a – Ultimate limit was calculated by dividing the maximum load by the length of the wall, multiplying by a
specific gravity adjustment of 0.92 (for Douglas fir) and dividing by a safety factor of 3. b – Drift limit was determined as the load present when the wall reached 0.2 in. deflection.
As demonstrated in the fastener connection tests (Chapter 2), the failure
mechanisms for these fasteners were different. Table 4.2 summarizes the failure modes
observed for all fasteners for the full-scale shear wall tests. The nails experienced
withdrawal, pull-through, and edge tear-out failure but differed in the dominant failure
mode as a result of bend yield strength. The unhardened 0.148 shoulder nail
57
predominantly failed by pulling through the sheathing and experienced more shank
deformation than observed for the hardened 0.148 shoulder nail, which failed mostly by
withdrawal from the framing members (Figure 4.4a and b).
For the screw fasteners, the New DSV screw showed the highest percentage of
edge tear-out among all fasteners and a predominant failure mode of pull-through.
Deformation of the screw primarily occurred as a single hinge located at the framing-
insulation interface (Figure 4.4c). In contrast, the SIPLD screw had the highest
percentage of non-failure but the screws that did fail usually did so in shear. This
behavior resulted in a sudden wall failure whereas the other fasteners demonstrated a
more gradual failure response.
Table 4.2 Failure mode of fasteners for full-scale shear wall tests.
Fastener Failure Mode (%)
None Withdrawal Pull-through Edge Tear-out Shear Total
Number
0.148 shoulder nail
10.1 15.2 58.9 2.8 0.0 158
0.148 hard shoulder nail
23.1 63.9 21.5 5.1 0.0 158
New DSV screw
9.5 0.0 69.9 19.9 0.0 158
SIPLD screw 66.1 0.0 0.0 2.2 31.6 158
58
Figure 4.4. Failure for each fastener from full-scale shear wall testing: a) 0.148 shoulder nail; b) 0.148 hardened shoulder nail; c) New DSV screw; and d) SIPLD screw.
Regression analysis of fastener properties with shear wall results for ultimate limit
and drift limit indicated positive relationships with the ultimate limit for initial stiffness
(R2 = 0.820) and fastener head diameter (R
2 = 0.685) and a positive relationship between
the drift limit and bend yield (R2 = 0.652) (Appendix C). While conclusions drawn from
these comparisons may be tenuous given the small sample size, the relationships seem
logical as it would be expected that increasing head diameter could increase shear wall
capacity by requiring greater force to pull a larger fastener head through the sheathing;
and higher bend yield strengths would be associated with higher loads necessary to reach
the 0.2-in. drift limit criterion.
The behavior of the sheathing panels as the walls were subjected to lateral loads
demonstrated rigid-body rotation where most of the stress was concentrated at the panel
59
corners as the panel rotated around its centroid. Failure first occurred at the bottom corner
closest to where the load was being applied to the wall. As this fastener failed, stresses
were redistributed to other fasteners. The seam between the two panels also experienced a
higher amount of stress compared to other portions of the wall.
For the SIPLD screw, the panel corners typically failed as the screw would tear
out of the sheathing; but the screws adjacent to the corner fasteners and the majority of
the screws along the middle seam all failed in shear. The remainder of the fasteners did
not fail or showed only slight yielding at the framing-insulation interface. For the other
fasteners, failure occurred around the perimeter of both sheathing panels with the
fasteners at the corners and along the top of the wall tending to showed tear-out failures.
Fastener pull-through more commonly occurred along the sides and bottom of the walls.
In the case of the 0.148 hardened nail the areas of highest stress were characterized by
withdrawal failure.
The rigid insulation offered little resistance to the lateral forces applied to the
wall. In all specimens, tears in the insulation were observed where the fasteners
experienced high stress. Compression of the insulation at locations of high fastener stress
was also observed resulting in a dimpled-looking appearance along the edge of the panel.
While the polyiso insulation has a compressive strength of 22 psi, it contributed little if
any shear resistance.
Regarding the behavior of the framing members during full-scale shear wall
testing, the majority of the deflection occurred along the top plate where the load was
being applied. Very little movement occurred in the sill plate but some uplift was
60
observed on the end stud closest to the load application. Separation of interior studs from
the sill plate and top plate were observed during most tests but little if any deformation of
the studs was noted. Splitting of the sill plate and/or middle stud (seam between the
sheathing panels) did not occur for any of the tests.
4.3.2. Comparison of Test Results with Modeling
Models created using M-CASHEW2 were compared to full-scale shear wall test
results to see how well the existing model predicted wall behavior. In all cases the
backbone from M-CASHEW2 had a similar shape as the test results; however, values at
0.2-in. wall deflection were under-predicted by the model while the maximum loads of
the model were higher than the shear wall tests results (Figure 4.5). For the 0.148
shoulder nail, the differences between the actual tests and the model drift limit and
maximum load were 14 percent; the hardened 0.148 shoulder nail had a difference of
approximately 16 percent for both drift limit and maximum load; the New DSV screw
differed by 11 percent and 6 percent for the drift limit and maximum load, respectively;
and the difference for the SIPLD screw was 16 percent for both values. In all cases the
model was within 20 percent of the actual test results.
The discrepancy in drift limit values between the model and test results was
probably a result of the variability in the initial stiffness for these connections. As seen in
the fastener connection tests (Chapter2), the variability in initial stiffness can be fairly
significant (Shirazi 2012). For this wall system the variability would probably be even
more significant given the compression of the insulation between the sheathing and
61
framing members. Depending on how much force was applied when connecting the
sheathing to the framing, the initial stiffness values could be affected.
Over prediction of the model most likely resulted from the manner in which the
fastener connection tests were performed. In the small-scale tests, the force was applied
strictly in the vertical direction; therefore no connection could fail with the fastener
tearing through the side of the sheathing since the end distance was considerably larger in
the connection tests (3 in.) versus the shear wall tests (0.375 in.). However in the full-
scale shear wall tests, the panels rotated and the trajectory of force could be in the
direction of the shorter distance to the edge of the panel resulting in tear-out failure,
which was documented to occur. Additionally, M-CASHEW2 does not take into account
tear-out as a failure mode; therefore, it would reasonable for the model to over-predict the
0.148 shoulder hard E72-Test1 0.148 shoulder hard E72-Test2
M-CASHEW2b)
Drift Limit (plf)
Maximum Load (plf)
63
Figure 4.5. Backbone curves for full-scale shear wall tests and M-CASHEW2 models: a) 0.148 shoulder nail; b) 0.148 hardened shoulder nail; c) New DSV screw; and d) SIPLD screw.
deformations to the shank of the fastener and increasing the head diameter increased the
connection shear capacity.
Data obtained from the fastener connection tests were then input in to a shear wall
model (M-CASHEW2) to predict full-size (8 ft. x 8 ft.) wall shear capacity and behavior.
Two different fastener spacing schedules were modeled (4-in. edge spacing with 12-in.
field spacing vs. 3-in. edge spacing with 6-in. field spacing). Ultimate limit and drift limit
were calculated for each spacing/fastener configuration for comparison. In all cases, the
drift limit was less than the ultimate limit, thus controlling the design value. The 4/12
spacing yielded lower results with drift limit values all less than 200 plf, whereas the 3/6
spacing had approximately 18 percent greater drift limits. The same trends were observed
for both insulation thicknesses although the capacity for the 2-in. insulation walls were
lower than the 1-in. walls. From these tests it was shown that screws had higher
capacities than nails and that certain properties of the fastener were related to increased
ultimate limit (head and shank diameters) and drift limit (initial stiffness). Results from
the modeling guided the selection of which fasteners to test for full-scale shear walls.
The full-scale shear wall tests examined the capacity of two nails and two screws
for the 2-in. insulated OSB. Three of these fasteners (0.148 shoulder nail, 0.148 hardened
shoulder nail, and the New DSV screw) were developed specifically for application using
this wall system. The modifications to the fasteners were intended to increase the
maximum load and/or the initial stiffness of these walls. The fourth fastener tested is
currently used for rigid insulation installation and is readily available for use.
71
The shear wall tests showed that the walls constructed with nails were governed
by drift limit with values of 173 plf and 223 plf for the 0.148 unhardened and hardened
shoulder nails, respectively. For the screws, the ultimate limit controlled with the New
DSV screw obtaining a design value of 193 plf and the SIPLD screw a design value of
218 plf. Results from these tests indicated the hardened 0.148 shoulder nail had the
highest design value.
Failure of each fastener type for the shear wall tests was also recorded. Different
dominant failure modes were identified for each fastener with the 0.148 unhardened
shoulder nail failing primarily by head pull through; the hardened 0.148 shoulder nail was
characterized by withdrawal failure; the New DSV screw failed primarily by pull
through, but also demonstrated the greatest number of edge tear failures; and the SIPLD
screw primarily failed by shear.
Results from these tests confirmed the validity of the gap mode equations for
predicting design values for light-frame wood construction connections when insulation
is placed between the exterior sheathing and framing members. The correct yield mode
was predicted for each of the fasteners using these equations.
This research showed that insulated OSB panels can be used for wood frame
shear walls following traditional construction techniques, but the capacity of walls built
using this system is lower than those without rigid insulation between the framing
members and sheathing. Using fasteners with different geometric properties helped to
recover some of the lost shear capacity, but there was still a reduction in capacity of
approximately fifty percent using the above specified fasteners at a 4/12 spacing.
72
Further research (e.g. cyclic shear wall tests) needs to be performed to better
predict loading conditions on these types of walls during wind or seismic events.
Additional modeling can be conducted to help optimize shear wall design by changing
fastener spacing schedules or using a combination of different fasteners in strategic
locations to resist areas of high shear stress. Slight modifications to the insulated panels
can also be made to further increase the shear capacity of this wall system without
significantly affecting the conventional construction techniques.
73
APPENDICES
74
Appendix A – Reference Displacement Values for Fastener Connection Tests
The method for obtaining the reference displacement values for the fastener
connection tests (Table A.1) followed Krawinkler et al. (2000). These values were then
input into a graphic user interface (GUI) in MATLAB (Figure A.1) that controlled the
displacement rate and magnitude for the UTM during the reverse cyclic fastener
connection tests. The sampling rate was set to 100 data points per second and the cross-
head stopping time was set to 0.5 seconds. This factor was used to coordinate the cross-
head motion with the change in displacement direction. Preliminary tests indicated that a
value of 0.5 seconds provided a close synchronization between the signal sent from the
data collection program and the UTM.
75
Table A.1. Reference displacement (Δr) for each fastener type used for connection tests. (Reference displacement was computed using the displacement associated with 80 percent of the post peak load multiplied by 0.60.)
Group #1: 1-in. insulation
Fastener Name 80% Post Peak
Displacement – Δm (in)
Reference Displacement - Δr
(in)
0.131 smooth nail 1.616 0.970
0.131 ring nail 1.007 0.604
0.148 smooth nail 0.895 0.537
Stainless steel screw
1.092 0.655
Carbon screw 0.973 0.584
Prototype carbon screw
1.062 0.637
Group #2: 2-in. insulation
0.148 hard shoulder nail
1.103 0.662
0.162 shoulder nail
1.173 0.704
0.162 head shoulder nail
1.135 0.681
0.162 smooth nail 1.277 0.766
0.162 ring nail 1.253 0.752
0.203 ring nail 1.232 0.739
New DSV screw 1.178 0.707
DSV screw 1.148 0.689
Stainless steel screw
1.103 0.662
SIPLD screw 1.155 0.693
SDWH194 screw 1.233 0.740
GRK-RSS screw 1.317 0.790
SDWS224 screw 1.477 0.886
76
Figure A.1. Graphic user interface for UTM displacement control during the reverse cyclic single fastener connection tests.
77
Appendix B – Backbone Curves for Monotonic and Cyclic Connection Tests
Figure B.1a. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 1-in. insulation for 0.131 smooth nail.
78
Figure B.1b. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 1-in. insulation for 0.131 ring nail.
79
Figure B.1c. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 1-in. insulation for 0.148 smooth nail.
80
Figure B.1d. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 1-in. insulation for stainless steel screw.
81
Figure B.1e Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 1-in. insulation for carbon screw.
82
Figure B.1e. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 1-in. insulation for prototype screw.
83
Figure B.2a. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for 0.148 shoulder nail.
84
Figure B.2b. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for 0.148 hardened shoulder nail.
85
Figure B.2c. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for 0.162 shoulder nail.
86
Figure B.2d. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for 0.162 smooth nail.
87
Figure B.2e. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for 0.162 ring nail.
88
Figure B.2f. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for 0.203 ring nail.
89
Figure B.2g. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for new DSV screw.
90
Figure B.2h. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for DSV screw.
91
Figure B.2i. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for stainless steel screw.
92
Figure B.2j. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for SIPLD screw.
93
Figure B.2k. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for SDWH194 screw.
94
Figure B.2l. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for GRK-RSS screw.
95
Figure B.2m. Comparison of backbone curves for monotonic test and cyclic model (averaged for all tests performed – 3 monotonic, 10 cyclic) of fasteners with 2-in. insulation for SDWS224 screw.
96
Table B.1. Hysteretic parameters from 1-in. and 2-in. fastener connection cyclic tests.
Appendix C – Regression Analyses: Fastener Properties/Shear Wall Tests
R² = 0.3801
0
50
100
150
200
250
300
0.000 0.050 0.100 0.150 0.200
Ult
imat
e L
imit
(p
lf)
Shank Diameter (in) a)
R² = 0.1492
190
200
210
220
230
240
0.000 0.050 0.100 0.150 0.200
Dri
fit
Lim
it (
plf
)
Shank Diameter (in) b)
R² = 0.6851
0
100
200
300
400
500
600
0.000 0.200 0.400 0.600 0.800
Ult
imat
e L
imit
(p
lf)
Head Diameter (in) c)
R² = 0.0267
190
200
210
220
230
240
0.000 0.200 0.400 0.600 0.800
Dri
ft L
imit
(p
lf)
Head Diameter (in) d)
R² = 0.3342
0
50
100
150
200
250
300
0 100 200 300
Ulit
imat
e L
imit
(p
lf)
Bend Yield (ksi) e)
R² = 0.6516
0
50
100
150
200
250
0 100 200 300
Dri
ft L
imit
(p
lf)
Bend Yield (ksi) f)
98
Figure C.1. Regression analyses for the 2-in. fasteners with respect to ultimate limit and drift limit results from full-scale shear wall tests: a) and b) shank diameter; c) and d) head diameter; e) and f) bend yield strength; g) and h) initial stiffness.
R² = 0.82
0
50
100
150
200
250
300
0 1000 2000 3000 4000
Ult
imat
e L
imit
(p
lf)
Initial Stiffness (lbf/in) g)
R² = 0.0354
190
200
210
220
230
240
2800 2900 3000 3100 3200 3300
Dri
ft L
imit
(p
lf)
Initial Stiffness (lbf/in) h)
99
Appendix D – Sample Calculations for Fastener Yield Modes
European Yield Mode Equations – Sample Calculation
100
101
Gap Yield Mode Equations – Sample Calculation
102
103
104
References
Anderson, EN; Leichti, RL; Sutt Jr., EG; Rosowsky, DV (2007). Sheathing nail bending-
yield stress: effect on cyclic performance of wood shear walls. Wood and Fiber
Science, 394: 536-547.
APA (2007). Diaphragms and shear walls: design and construction guide. Tacoma, WA.
APA – The Engineered Wood Association, 32 p.
ASTM (2013). Standard Test Method for Determining Bending Yield Moment of Nails
(ASTM F 1575 – 03). West Conshohocken, PA. ASTM International, 4 p.
ASTM (2013). Standard Test Methods of Conducting Strength Tests of Panels for
Building Construction (ASTM E72-13a). West Conshohocken, PA. ASTM
International, 12 p.
ASTM (2014). Standard Test Methods for Mechanical Fasteners in Wood (ASTM
D1761-12). West Conshohocken, PA. ASTM International, 10 p.
Aune, P., Patton-Mallory, M. (1986a). Lateral load-bearing capacity of nailed joints
based on the yield theory: theoretical development. Research Paper FPL 469.
Madison, WI. U.S. Department of Agriculture, Forest Service, Forest Products
Laboratory, 20 p.
Aune, P., Patton-Mallory, M. (1986b). Lateral load-bearing capacity of nailed joints
based on the yield theory: experimental verification. Research Paper FPL 470.
Madison, WI. U.S. Department of Agriculture, Forest Service, Forest Products
Laboratory, 32 p.
AWC (2008). Special design provisions for wind and seismic. Leesburg, VA. American
Wood Council, 80 p.
AWC (2012). National Design Specification (NDS) for Wood Construction. Leesburg,
VA. American Wood Council, 282 p.
AWC (2014). General dowel equations for calculating lateral connection values.
Technical Report 12. Leesburg, VA. American Wood Council, 24 p.
Dinehart, DW; Shenton III, HW (1998). Comparison of static and dynamic response of
timber shear walls. Journal of Structural Engineering, 124: 686-695.
Dolan, JD; Gutshall, ST, McLain, TE (1995). Monotonic and cyclic tests to determine
short-term load duration performance of nail and bolt connections volume 1:
105
summary report. Virginia Polytechnic Institute and State University Timber
Engineering Report No. TE-1994-00. Blacksburg, VA, Virginia Polytechnic
Institute and State University, 33 p.
Ekiert, C; Hong, J (2006). Framing-to-sheathing connection tests in support of