Wood Structural Panel to Cold-Formed Steel Shear

Connections with Pneumatically Driven Knurled Steel Pins

Reynaud Serrette, M.ASCE1; and David Nolan2
Downloaded from [Link] by THE UNIVERSITY OF NEWCASTLE on 02/26/17. Copyright ASCE. For personal use only; all rights reserved.

Abstract: In cold-formed light-frame steel construction, wood structural panels can be connected to framing members with adhesives,
staples, pins, and screws. Although screws are the most commonly specified fastener, steel pins have been used in construction for
many years. Unlike that of screws, the design of connections with pins is still largely based on manufacturer-provided data. This paper
examines the behavior of shear connections in which wood structural panels are attached to cold-formed steel framing using knurled
steel pins. The results of a series of tests are presented for a range of thicknesses of cold-formed steel and wood structural panels. On the
basis of observed behavior, expressions are proposed to compute connection peak-shear strength. The expressions are given in a form
that can be easily adapted for other pin-connection configurations. DOI: 10.1061/(ASCE)SC.1943-5576.0000321. © 2017 American
Society of Civil Engineers.
Author keywords: Cold-formed steel; Shear; Connection; Pins; Knurl; Wood structural panel.

Introduction documents is based on the proprietary nature of these fasteners.

This argument, although valid, can also be applied to screws;
In light-frame construction, the most common method for connect- yet, design provisions for screws are based on testing results for
ing structural panels to cold-formed steel (CFS) framing requires a variety of screws (Pekoz 1990). Nonetheless, when pins are
the use of screws (Ide 2009). As an alternative to screws, pneumati- used in design, reference to national evaluation reports (e.g., the
cally driven, hardened steel pins (with a ballistic style point) have International Association of Plumbing and Mechanical Officials
been used in light-frame construction for more than 30 years (Nolan Uniform Evaluation Services or the International Code Council
2009). Pins offer both a structurally sound and economically effi- Evaluation Services) is necessary. Unfortunate is that many of
cient connection, and they provide the same levels of performance these published evaluation reports do not include basic connec-
(including ductility) and reliability as screws. One illustration of tion design values, because their focus is typically on end use
performance similarity is presented in Fig. 1, which shows a com- (e.g., shear walls, diaphragms, roof uplift loads). When pro-
parison of normalized, reversed, cyclic response curves of CFS vided, connection values typically are given in terms of allow-
shear walls (1:1 aspect ratio) constructed with screws and steel pins; able values and are incomplete or missing information about
both walls were tested using the same protocol. As Fig. 1 shows, the safety factors. JAACO (2015), for example, provided allowable
normalized envelope curve for the wall with pins is similar to that short-term (wind and seismic design) connection shear strength
of the wall with screws. for smooth-shank 3.68-mm-diameter steel pins used for attach-
Steel pins are manufactured from hardened steel wire, and they ing 11.9-mm oriented strand board (OSB)–rated sheathing to 68-
are configured with a ballistic point and flat head. The ballistic point mil (1.73-mm) CFS framing, but the safety factor associated
is necessary for the penetration of steel materials, and the flat head with the recommended value was not provided. HILTI (2014)
provides bearing resistance. Manufacturers typically knurl the sur- produces a knurled 2.59-mm-diameter pin and recommends
face of the pin shank to allow for more efficient installation and allowable shear strengths for framing thicknesses from 27 to
pullout friction or holding strength. Four typical knurling configura- 68 mil (0.67 to 1.73 mm). HILTI has noted that the safety factor
tions/patterns are shown in Fig. 2. For highly specialized pins, knurl associated with the recommended allowable values were based
patterns can be combined. The most common knurl pattern cur- on calculations “in accordance with Section F1.1 of AISI 100,”
rently used for pins in CFS construction is the left- or right-hand but the safety factor was not provided. These are but two
knurl. examples.
Unlike screws, steel pins for attaching wood structural panels Baur and Suaris (2003) studied the tensile and bearing capacities
(WSPs) to CFS are not explicitly recognized in American Iron of pins in CFS construction. They utilized a cementitious sheathing
and Steel Institute (AISI) design specifications or standards. A material in their tests. Using the flexural tensile strength of the
main argument for the exclusion of pins from national design sheathing, Baur and Suaris (2003) concluded that AISI’s equations
for screws provided good (within 10%) predictions for shear
1
strength. Davis (2012) presented a series of calculations for com-
Professor, Dept. of Civil Engineering, Santa Clara Univ., 500 El puting the shear and tensile strengths of CFS-to-CFS connections
Camino Real, Santa Clara, CA 95120 (corresponding author). E-mail: using pins. The Davis calculations were based on equations similar
rserrette@[Link]
2 to those for screw connections and data provided by a
General Manager, ET&F Fastening Systems, Inc., Solon, OH 44139.
Note. This manuscript was submitted on August 5, 2016; approved on manufacturer.
December 16, 2016; published online on February 22, 2017. Discussion The following sections of this paper describe testing con-
period open until July 22, 2017; separate discussions must be submitted ducted to explore behavior and a method developed to estimate
for individual papers. This paper is part of the Practice Periodical on the connection strength of WSPs fastened to CFS shear connec-
Structural Design and Construction, © ASCE, ISSN 1084-0680. tions using steel pins. On the basis of the observed behavior,

© ASCE 04017002-1 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., -1--1

Downloaded from [Link] by THE UNIVERSITY OF NEWCASTLE on 02/26/17. Copyright ASCE. For personal use only; all rights reserved.

Fig. 1. Reversed cyclic performance of shear walls (1:1 aspect ratio) constructed with pins and screws

The measured thicknesses and material properties of the steel

used in the connection tests are summarized in Table 1, and Table 2
lists the ratios of measured framing thickness (tm), measured yield
strength (Fy_m), and measured tensile strength (Fu_m) to their re-
spective minimum specified values.

Test Setup

A symmetric lap-shear test setup and loading scheme, as shown

in Fig. 4, was used in this test program. Strips of WSP (long
side parallel to the longitudinal axis of the full sheet and the
applied load) were connected to the flanges of back-to-back
CFS C-sections. Four pins connected the WSP strips to the
upper C-sections, and eight screws connected the opposite end
of each strip to the lower C-sections. Displacement transducers
were used to measure the average connection deformation at
Fig. 2. Steel pin and basic knurl patterns the pins.
Connection of the C-sections to the loading system was achieved
using a bolt and aluminum clevis. This connection permitted free
simple structural models are proposed and equations are devel-
rotation about an axis perpendicular to the web of the C-sections
oped to predict strength. Although the proposed equations are
under load, which was applied to each specimen at a fixed rate of
specific to the pins tested (helical knurled steel pins), the majority
2.54 mm/min until a decrease in load was observed or the connec-
of available pins have a similar configuration. The expressions
tion deformation became excessive.
also provide a framework for investigating pins with alternative
knurl patterns.
Test Results
Test Program Fig. 5 shows theoretically anticipated failure modes/limit states
for the shear connections tested. For the ranges of steel and WSP
The test program presented in this paper used 2.54-mm-diameter thicknesses tested with the knurled-pin connections, there was
helical knurled steel pins with several combinations of OSB-rated no evidence of fastener shear or fastener bending. In most cases,
sheathing and CFS-framing thicknesses. Fig. 3 provides a summary fastener pullthrough or pullout with tilting and bearing was the
of the scope of the test program. All fasteners were sourced from governing failure mode. These interaction modes are shown in
ET&F Fastening Systems, Inc. (Solo, Ohio), and sheathing was Fig. 6 and illustrated in Fig. 7. In some cases, the interaction/
sourced from local lumber yards. For each connection configuration combined failure modes shown in Figs. 6 and 7 were precipitated
(pin–sheathing–framing) identified in Fig. 3, three identical tests by an opening of the connection as pin tilt became more
were conducted (total of 36 tests). pronounced.

© ASCE 04017002-2 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., -1--1

Downloaded from [Link] by THE UNIVERSITY OF NEWCASTLE on 02/26/17. Copyright ASCE. For personal use only; all rights reserved.

Fig. 3. Test program scope

Table 1. CFS Framing Properties

Designated Minimum Design

framing Minimum specified Minimum specified framing framing Measured Measured tensile Measured
thickness yield strength tensile strength thickness thickness yield strength strength thickness Elongation
(mil)a [Fy_min (MPa)] [Fu_min (MPa)] [tmin (mm)] [tdes (mm)]b [Fy_m (MPa)] [Fu_m (MPa)] [tm (mm)] (%)
33 228 310 0.84 0.88 311 374 0.88 34.10
54 345 448 1.37 1.44 400 455 1.41 30.62
97 276 345 2.45 2.58 284 353 2.55 38.93
a
1 mil = 0.0254 mm.
b
tdes = tmin/0.95
1.05tmin.

Table 2. Ratio of Measured to Minimum Steel Specifications mm) framing, the higher deformation COV values seem to be asso-
Designated ciated with a pullthrough limit state. The deformation values at 50%
framing peak strength showed a larger-than-expected deformation in the
thickness (mil)a Ry = Fy_m/Fy_min Rt = Fu_m/Fu_min tm/tmin tm/tdes 11.9-mm-OSB 54-mil (1.37-mm) connections than all the other
connections; for the three tests, the 50% peak-strength average de-
33 1.37 1.20 1.058 1.005
formation was 2.64 mm, and the individual deformation values
54 1.16 1.02 1.035 0.984
were between 2.51 and 2.74 mm, which are relatively consistent
97 1.03 1.02 1.038 0.986
measures.
a
1 mil = 0.0254 mm. For a given WSP sheathing grade, Table 3 shows that connection
shear strength increases with increases in framing thickness. This
increase seems to be associated with a change in failure mode from
Average peak strength, and deformation values at 33, 50, and pullout, with tilting and bearing in the CFS framing, to pullthrough
100% of peak strength, are summarized in Table 3. Also listed in or a combined pullthrough with bearing in the structural panel.
Table 3 are measures of variability in the test data, that is, standard Fig. 8 shows the relationship between the measured shear
deviations and coefficients of variation (COVs) (standard deviation strength and connection deformation for the three framing thick-
divided by average characteristic value). The COVs for the meas- nesses tested. As is evident in Fig. 8, the lower (10–15%) meas-
ured peak strength values range between 1.1 and 12.9%, and the ured peak strengths of the 11.9-mm-OSB connections, compared
COVs for the corresponding deformation range from 2.6 to 29.0%. to those of the 11.1-mm-OSB connections, were found in 54-mil
The greater variability in deformation is characteristic of failure (1.37-mm) and 97-mil (2.45-mm) framing. This result is some-
modes that involve bearing in WSP (APA–The Engineered Wood what counterintuitive, but the COVs of peak strength for these
Association 2013). In addition, except for the connections with connections are quite low, which suggests that the measured
11.1- and 11.9-mm OSB-rated sheathing attached to 33-mil (0.84- peak strength values are consistent. This observation requires

© ASCE 04017002-3 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., -1--1

Downloaded from [Link] by THE UNIVERSITY OF NEWCASTLE on 02/26/17. Copyright ASCE. For personal use only; all rights reserved.

Fig. 4. Test setup: (a) illustration; (b) actual

further study. Fig. 8 also shows that the deformation capability complex limit states observed in tests. At a basic level, three limit
(a pseudomeasure of ductility) of the connections beyond the states seem relevant, failure caused by (1) the pin, (2) pin bearing in
33 and 50% peak-strength values is quite good considering the the WSP, and (3) bearing in the CFS framing. However, as the test
characteristic variability in deformation measures; the lowest results reveal, failure typically involved an interaction between
values for DVpeak =D33%Vpeak and DVpeak =D50%Vpeak are 9.1 and bearing in the WSP and CFS when the fastener tilts or rotates. In
3.5, respectively. addition, tilting introduces a potential for either pin pullout (from
the framing) or pin pullthrough (the sheathing). From these obser-
Proposed Design Equations vations, it is recommended that the nominal shear strength of pin
connections with WSPs be computed as the lesser of the strengths
The development of equations for estimating the strength of pin derived from (1) bearing in the CFS framing and (2) combined tilt-
connections with WSPs attached to CFS framing should reflect the ing, bearing, and pullout or pullthrough. The shear strength of the

© ASCE 04017002-4 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., -1--1

Downloaded from [Link] by THE UNIVERSITY OF NEWCASTLE on 02/26/17. Copyright ASCE. For personal use only; all rights reserved.

Fig. 5. Basic/theoretical failure modes

Fig. 6. Observed failure modes: (a) pullout as well as tilt and bearing in CFS; (b) pullthrough; (c) pullthrough with bearing in the WSPs and through-
thickness cracking

Fig. 7. Combined failure modes: (a) combined pullout as well as tilt and bearing in the WSP; (b) combined pullthrough as well as tilt and bearing in
the WSP; (c) combined pullthrough and through-thickness cracking

pin must be evaluated also; this information is typically provided by

the manufacturer. Pnsb ¼ 2:7dtFu (1)

For the limit state of bearing in the CFS framing, Eq. (1), AISI screw- Pns1 ¼ 1:3dtFu (2)
bearing strength [AISI S100-12, Section E4.3.1 (AISI 2012)] was
adopted. Eq. (1) applies to screws regardless of thread depth and spacing. where d = nominal diameter of pin shank (in millimeters); Fu =
Pending further research, a conservative 50% reduction in the nominal tensile strength of framing (in megapascals); Pnsb = nominal bear-
bearing capacity of pins, as depicted in Eq. (2), is recommended. ing strength per AISI; Pns1 = recommended nominal bearing

© ASCE 04017002-5 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., -1--1

 strength for pins; and t = design thickness of framing (in

DPpeak (%)
COV
millimeters).

4.6
3.7
4.5

2.6

6.4
8.0
22.1
15.4
11.0

15.5
29.0

13.3
As noted previously, pin tilt/rotation results in bearing in both
the WSP and CFS framing but no bending in the pin. Tilting was
SD DPpeak accompanied also by either the pin pulling out from the CFS fram-
(mm) ing or pulling through the WSP. For bearing in the WSP with fas-
0.25
0.29
0.30
1.34
1.42
0.80
0.14
1.29
1.92
0.37
0.82
1.30
tener tilt/rotation without bending of the fastener, the National
Design Specification (NDS) for Wood Construction (AWC 2014)
state that connection shear strength is a function of fastener diame-
COV Ppeak

ter, the bearing length in each connected member, the bearing

3.3
6.4
9.1
8.5
3.4
3.3
1.1

2.0
2.8
12.9
10.3

10.2
(%)

strength of each connected member, and the grain direction in the

Downloaded from [Link] by THE UNIVERSITY OF NEWCASTLE on 02/26/17. Copyright ASCE. For personal use only; all rights reserved.

wood member. The NDS also gives the bearing strength of dowel-
type fasteners in WSP sheathing as a function of the specific gravity
SD Ppeak

of the sheathing and the fastener diameter. When these NDS models
0.029
0.114
0.177
0.075
0.053
0.058
0.011
0.215
0.208
0.020
0.056
0.242
(kN)

are adopted, the shear strength of the pin connection, when failure
results from tilting with pin pullthrough in the WSP, is taken to be a
function of pin diameter, thickness of the WSP, specific gravity of
D33%Ppeak D50%Ppeak
(mm)

the WSP, and tensile strength of the CFS framing. Using regression
0.88
1.90
0.66
1.26
2.63
0.58
0.76
1.71
0.62
0.99
1.96
0.75
analysis, the expression given by Eq. (3) was developed to estimate
the nominal shear strength Pns2 of connection failure caused by tilt-
ing and bearing with pin pullthrough.
(mm)
0.46
0.66
0.27
0.56
1.02
0.33
0.38
0.55
0.26
0.49
0.90
0.23

Pns2 ¼ 1:528d1:185 tWSP

0:255 1:84
G Fu (3)

where G = specific gravity of WSP; and tWSP = nominal thickness

DPpeak

5.43
7.97
6.72
6.04
9.25
7.33
5.27
8.32
6.65
5.76

9.78
10.24
(mm)

of WSP (in millimeters).

When tilting triggers pullout from the CFS framing, the thick-
ness of the framing becomes another limiting variable in strength
Ppeak (kN)
Measured

prediction. The following semiempirical equation, Eq. (4), was

0.878
1.787
1.934
0.885
1.550
1.742
0.966
1.673
2.015
1.014
2.036
2.384

developed to address this limit state:

Pns3 ¼ 2:098t1:05 d0:675 tWSP
0:275 1:84
G Fu (4)
Span rating in US Customary Units; (x/y) indicates support framing spacing on the roof (x) and floor (y).
CFS tensile strength

Eqs. (3) and (4) can be written in general form as

b
[Fu (MPa)]

Pns ¼ r t b 1 d b 2 tWSP
3
G1:84 Fu (5)
374
455
353
374
455
353
374
455
353
374
455
353

where the parameters r , b 1, b 2, and b 3 can be determined from

test data and/or numerical analysis. Therefore, Eq. (5) provides a
basic framework for exploring the effects of alternative connection
configurations.
thickness [tm (mm)]

Table 4 compares the measured connection peak strength with

Measured CFS

corresponding predicted strengths when G is taken to be 0.56. With

0.88
1.41
2.55
0.88
1.41
2.55
0.88
1.41
2.55
0.88
1.41
2.55

an average measured-to-predicted strength ratio of 1.05 and a 7.2%

COV, the proposed equations seem to provide reasonably good esti-
mates of connection shear strength. In practical terms, when the
minimum specified strength of CFS is used, the predicted values in
Table 3. Connection Strength and Deformation Measures

Table 4 will be relatively low.

Designated framing

The scope of the reported test program meets the criteria to per-
thickness (mil)b

mit an evaluation of data in accordance with AISI S100-12, Chapter

F (“Tests for Special Cases”) (AISI 2012). Thus, the accuracy of the
33
54
97
33
54
97
33
54
97
33
54
97

nominal shear strength of the connection predicted by the proposed

design equations can be evaluated in terms of the required resistance
factor according to the provisions of AISI S100-12, Chapter F
(AISI 2012). Using the factors and coefficients listed in Table 5,
the resistance factor ( f ) associated with the reported tests was
Note: SD = standard deviation.
11.1-mm OSB-rated sheathing

11.9-mm OSB-rated sheathing

15.1-mm OSB-rated sheathing

18.3-mm OSB-rated sheathing

0.62. For screw connections, AISI S100-12 (AISI 2012) requires

f to be 0.50, but it notes that higher values can be used if deter-
mined from tests.
1 mil = 0.0254 mm.

The 0.15 COV for the fabrication factor (VF) listed in Table 5
Sheathing gradea

might be unrealistically high for the pin connection in shear; that is,
a value of 0.10 is used for screws and power-actuated fasteners, and
the quality control concern for installation of pins compared to that
(24/16)

(32/16)

(40/20)

(48/24)

of screws is not inherently greater than it is for power-actuated fas-

teners. If VF is 0.10, then f = 0.67. To achieve the same resistance
b
a

© ASCE 04017002-6 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., -1--1

Downloaded from [Link] by THE UNIVERSITY OF NEWCASTLE on 02/26/17. Copyright ASCE. For personal use only; all rights reserved.

Fig. 8. Connection shear strength versus connection deformation: (a) 33-mil (0.84-mm) framing; (b) 54-mil (1.37-mm) framing; (c) 97-mil
(2.45-mm) framing

factor specified for screws in AISI S100-12, Chapter E (AISI 2012), strength of helical knurl steel pins for attaching WSPs to CFS
the target reliability index in Table 5 must be increased from 3.50 to framing. One pin diameter, three different thicknesses of CFS
4.22 (with VF = 0.15). framing, and four different thicknesses of OSB were included in
the study. Failure modes observed in the tests included pin pullout
and a combination of tilting, bearing, and pin pullout or
Conclusion pullthrough.
Based on the observed failure modes, analytical expressions
Data from a series of shear connection tests were presented to were developed to estimate the peak-shear strength of tested
illustrate behavior and develop expressions for estimating the connections. The analytical expressions fit the test data quite

© ASCE 04017002-7 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., -1--1

 Table 4. Comparison of Measured and Predicted Connection Peak-Shear procedures recommended by AISI for special case testing, a re-
Strength sistance factor of 0.63 was determined at a 3.50 target reliability
Measured Predicted Measured/predicted index.
Sheathing gradea Ppeak (kN) Ppeak (kN) Ppeak The derived analytical expressions were presented in a
form that easily facilitates an evaluation of other steel pin-
11.1-mm OSB rated 0.878 0.862 1.02 connection configurations (particularly pin diameter, knurl
(24/16) 1.787 1.622 1.10 pattern, CFS framing thickness, and sheathing thickness and
1.934 1.746 1.11 grade).
11.9-mm OSB rated 0.885 0.879 1.01
(32/16) 1.550 1.650 0.94
1.742 1.777 0.98 References
Downloaded from [Link] by THE UNIVERSITY OF NEWCASTLE on 02/26/17. Copyright ASCE. For personal use only; all rights reserved.

15.1-mm OSB rated 0.966 0.938 1.03

(40/20) 1.673 1.753 0.95 AISI (American Iron and Steel Institute). (2012). “North American specifi-
2.015 1.888 1.07 cation for the design of cold-formed steel structural members.” AISI
18.3-mm OSB rated 1.014 0.988 1.03 S100-12, Washington, DC.
(48/24) 2.036 1.840 1.11 APA–The Engineered Wood Association. (2013). “Allowable bear-
2.384 1.982 1.20 ing stress for APA wood structural panels.” TT-001B, Tacoma,
Average 1.05 WA.
SD (Sample) 0.076 AWC (American Wood Council). (2014). “National design specification
COV (%) 7.2 for wood construction.” NDS-2015, Leesburg, VA.
Baur, S., and Suaris, W. (2003). “Evaluation of tensile and bearing capaci-
Note: SD = standard deviation. ties of cold-formed steel connections attached with pneumatically
a
Span rating in US Customary Units; (x/y) indicates support framing spac- driven pins.” Proc., 2003 Architectural Engineering Conf., ASCE,
ing on the roof (x) and floor (y). Reston, VA, 1–8.
Davis, K. (2012). “Pin connections for cold-formed steel with spreadsheet.”
PDHonline Course S241. PDHonline, Fairfax, VA. h[Link]
Table 5. Statistical Parameters .com/courses/s241/s241_new.htmi (Jan. 9, 2017).
HILTI. (2014). “HILTI X-GPN 37 MX power-driven fasteners used to
Factor attach wood structural panels to cold-formed steel framing.” Rep. No.
Parameter Material Fabricationa Professional ESR 3059, ICC Evaluation Service, Brea, CA.
Ide, B. (2009). “Connections with cold-formed steel framing.” Civil Struct.
Mean value 1.08 1.00 1.05 Eng. h[Link]
COV 0.10 0.15 0.07 -steel-framingi (Feb. 19, 2014).
Calibration coefficient 1.52 — — JAACO. (2015). “Power-driven pins for shear wall and diaphragm assem-
Target reliability index 3.50 — — blies with steel framing and wood structural panels.” Rep. No. ESR
Correlation coefficient 0.97 — — 2961, ICC Evaluation Service, Brea, CA.
Load effect factor 0.21 — — Nolan, D. (2009). “Pneumatically driven pins for wood based panel
attachments.” CFSEI Technical Note F300-09, Cold-formed Steel
a
Recommended for connections not listed in AISI S100-12, Chapter F Engineers Institute, Washington, DC. h[Link]
(AISI 2012). .com/pdf/[Link] (Jan. 9, 2017).
Pekoz, T. (1990). “Design of cold-formed steel screw connections.”
well with an average of the measured-to-predicted peak strength Proc., 10th Int. Specialty Conf. on Cold-Formed Steel Structures,
ratio of 1.06 and an associated 6.7% COV. Using statistical Univ. of Missouri-Rolla, Rolla, MO, 575–587.

© ASCE 04017002-8 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., -1--1