United StatesDepartment ofAgriculture

Forest Service

ForestProductsLaboratory

Research

Strength of BoltedTimber Connectionswith Steel Side

PaperFPL-RP-513 Members

T. L. Wilkinson

Abstract

This study investigated the properties of bolted tim-ber connections with steel side members. The intentof this study was to determine (1) how well the Euro-pean Yield Model (EYM) predicts experimental results(2) if there is a difference in results from tension- andcompression-loaded specimens, and (3) what the effectis of steel side member thickness. Three bolt diame-ters and four ratios of main member thickness to boltdiameter were investigated. Connections were loadedboth parallel and perpendicular to the gram. Resultswere compared to predicted results from the EYM. Thestudy revealed that the EYM predicts connection yieldload with acceptable accuracy.

Keywords: Bolts, wood connections, European YieldModel, yield load, dowel embedment strength

July 1992

Wilkinson, T. L. 1992. Strength of bolted timber con-nections with steel side members. Res. Pap. FPL-RP-513. Madison, WI: U.S. Department of Agriculture, ForestService, Forest Products Laboratory. 10 p.

A limited number of free copies of this publication areavailable to the public from the Forest Products Labo-ratory, One Gifford Pinchot Drive, Madison, WI 53705-2398. Laboratory publications are sent to more than 1,000libraries in the United States and elsewhere.

The Forest Products Laboratory is maintained in coopera-tion with the University of Wisconsin.

Strength of Bolted Timber ConnectionsWith Steel Side Members

T. L. Wilkinson, Research General EngineerForest Products Laboratory, Madison, Wisconsin

Introduction

Historically, design values for bolted timber connec-tions (NFPA 1986) have been based on research con-ducted by Trayer (1932). Trayer’s recommendationswere based on an empirical fit of experimental data andthus are limited to the connection geometries evaluated,although Trayer’s recommendations have been extrapo-lated to other geometries by rule of thumb type proce-dures. Trayer empirically related bolt bearing stress atproportional limit with the ratio of main member thick-ness to bolt diameter. He further related the bolt bear-ing stress with the small, clear compressive strengthof the wood. His sample sizes were four and fivespecimens.

Doyle and Scholten (1963) evaluated bolted connectionsin Douglas-fir with sample sizes of three specimens.Soltis and others (1986) evaluated bolted Douglas-firconnections with wood side members using sample sizesof 16 specimens. Both of these studies found that theprevious recommendations from Trayer were still valid.

In recent code developments, the European YieldModel (EYM) proposed by Johanson (1949) has beenadopted as the base for predicting the strength ofbolted connections. The EYM can predict the strengthfor various connection geometries and material com-binations for two- and three-member connections.Soltis and others (1986) found that the EYM explainsmuch of the observed behavior of bolted timber con-nections. McLain and Thangjitham (1983) also foundthat the EYM predicts bolted connection performance.These studies indicated that the EYM can be a usefulmodel to predict connection behavior for single-boltedconnections.

Most investigations of bolted timber connections haveused small sample sizes that do not allow the calcula-tion of reliable statistical properties of the distributionof connection properties. In the future, design loads forbolted connections may be derived using reliability-

based design procedures, which would require distribu-tion properties.

Most studies have reported only proportional limitloads for bolted timber connections. For future designprocedures, other connection properties may need to beknown. These properties include maximum load andconnection deformation information. Deformation datawould be useful in setting serviceability limits for con-nections in a reliability-based design procedure. De-formation data are also needed for predicting the loaddistribution among bolts in multiple bolt connections inprocedures such as that developed by Lantos (1969).

Objective

This paper presents data for single-bolted timber con-nections with steel side members. Input data for usewith the EYM are also presented and EYM-predictedloads are compared to experimental results. The objec-tives were to determine (1) how well the EYM predictsexperimental results, (2) if results from tension-loadedspecimens differ from results from compression-loadedspecimens, and (3) what the effect is of steel side mem-ber thickness.

European Yield Model

The EYM provides an analysis method to predict thestrength of a two- or three-member dowel-type connec-tion. It was originally developed in Europe (Johanson1949) and is based on equilibrium equations resultingfrom the free body diagram of a bolt in a wood mem-ber. The method has been examined by McLain (1983)and Soltis (1986). Both found general agreement be-tween predicted and experimental data sets.

The EYM assumes that the capacity of a bolted con-nection is attained when either (1) the compressivestrength of the wood beneath the bolt is exceeded(Mode I or II yielding) or (2) one or more plastichinges develop in the bolt (Modes III or IV yielding).

These assumptions provide for several modes of yielding diameter. Steel side member thickness was 1/4 in.depending on connection member dimensions, member (6.4 mm), 3/8 in. (9.5 mm), and 1/2 in. (12.7 mm)strength, and bolt strength. For three-member connec- for the 1/2-in. (12.7-mm) bolts; 3/8 in. (9.5 mm) fortions, the yield load Zy is the smallest value given by the 3/4-in. (19.0-mm) bolts; and 1/2 in. (12.7 mm) forthe following equations. the 1-in. (25.4-mm) bolts.

Mode I,,, (1a) The wood main members were cut from Douglas-firglulam beams constructed of L2 or better grade lami-nating stock. Specimens with 1/2-in. (12.7-mm) boltshad two laminations, those with 3/4-in. (19.0-mm)bolts had three laminations, and those with 1-in. (25.4-mm) bolts had four laminations. All laminations were1-1/2 in. (38.1 mm) thick. Bolt holes 1/16 in.(1.6 mm) greater than the bolt diameter were drilledparallel to the glue lines. All wood members had amoisture content of approximately 12 percent.

Mode I, ( lb)

Mode III.

(1c)

Mode IV(1d)

where

Re is Fem /Fes,

Rt tm/ts ,

tm thickness of main member (in.),

t s thickness of side member (in.),

Fe m dowel bearing strength of main member(lb/in2),

Fes dowel bearing strength of side member (lb/in2),

Fy bending yield strength of bolt (lb/in2), and

D bolt diameter (in.).

(Conversion factors are 1 in. = 25.4 mm; 1 lb/in2 =6.89 kPa.)

M e t h o d s

Specimens

Three-member connections loaded both parallel andperpendicular to the grain were evaluated. Each con-nection consisted of a wood main member and steelside members fastened with a single bolt.

Three bolt diameters, 1/2 in. (12.7 mm), 3/4 in.(19.0 mm), and 1 in. (25.4 mm), were used. The steelbolts were SAE Grade 2. Bolts were long enough toprevent bearing on the threads. Nuts were finger tightat time of test.

Ratios of main member thickness to bolt diameter(tm /D) were 3.0, 5.125, 6.75, and 10.25 for each bolt

Twenty replications were created for each combinationof variables for a total of 240 parallel-to-grain loadedspecimens and 240 perpendicular-to-grain loadedspecimens.

The connection configurations were made to exceedNational Design Specification (NFPA 1986) require-ments for end and edge distance. The distance fromthe bolt to the end of the main member was 7.5 timesthe bolt diameter for parallel-to-grain loading. The dis-tance from the bolt to the loaded edge exceeded 4.25times the bolt diameter for perpendicular-to-grain load-ing. Edge distance equaled or exceeded 1.5 times thebolt diameter for both loadings. For perpendicular-to-grain loading, the distance between supports was 9 in.(230 mm) for specimens with 1/2-in. (12.7-mm) bolts,13.5 in. (340 mm) for specimens with 3/4-in. (19.0.mm) bolts, and 18 in. (460 mm) for specimens with1-in. (25.4-mm) bolts. These distances were 3 timesthe depth of the main member, as specified by ASTMD1761-88 (ASTM 1990).

In addition to the main series of teats, 20 parallel-to-grain loaded specimens were evaluated using compres-sion loading instead of tension loading. These speci-mens used 3/4-in. (19.0-mm) bolts and a tm/D of 6.75.Two additional groups of 20 parallel-to-grain loadedspecimens each were evaluated with 1/2-in. (12.7-mm) bolts and steel side member thicknesses of 3/8 in.(9.5 mm) and 1/2 in. (12.7 mm). For these two groupsthe tm /D was 5.125.

Experimental Procedure

The test procedures for parallel-to-grain loading(Fig. 1) and perpendicular-to-grain loading (Fig. 2)generally followed those given in ASTM D1761-88(ASTM 1990). Tension loading was used with theparallel-to-grain loaded specimens except for one group,which used compression loading. The applied load was

2

Figure 2—Experimental arrangement for loadingbolted connections perpendicular to grain.

Figure 1—Experimental arrangement for loadingbolted connections parallel to grain.

deformation controlled at either 0.035 or 0.050 in/min(0.89 or 1.27 mm/min). The tests were terminatedwhen failure occurred.

Deformation was measured using linear variable dif-ferential transducers. Load-deformation values werecontinuously recorded until failure occurred. Defor-mations were measured to an accuracy of ±0.001 in.(0.025 mm).

Minor Tests

A wood property needed for input in the EYM is thedowel bearing strength. This property was determinedby uniformly loading a bolt in a half hole (Fig. 3).One specimen was obtained for each connection froma section of the glulam beam directly adjacent to themain member. Bolt holes were located in the samelaminations as the bolt hole in the connection. A load-deformation curve was obtained using the movement ofthe movable head of the testing machine as a measureof deformation.

Bolt bending yield strength was determined from bend-ing tests of bolts (Fig. 4). Bolts were loaded at thecenter of the span. Spans of 2-13/16 in. (71 mm),3-3/4 in. (95 mm), and 5-15/32 in. (139 mm) wereused with 1/2-in. (12.7-mm) bolts; 3-3/4 in. (95 mm),5-1/4 in. (133 mm), and 8-3/16 in. (208 mm) with3/4-in. (19.0-mm) bolts; and 7 in. (177 mm) and 11 in.(279 mm) with l-in. (25.4-mm) bolts. These werethe longest spans that could be used for the variouslength of bolts without bearing on the threads. A load-deflection curve was obtained using the movement ofthe movable head of the testing machine as a measureof deflection. Approximately 10 bolts of each diameterand length were tested.

Specific gravity and moisture content were determinedfor each bolted connection. Specimens were taken froma section of the glulam beam that was directly adjacentto each main member and that contained only thoselaminations in which the bolt holes were drilled.

Definition of Yield Load

The yield load as given by the EYM may be definedas any load on the load-deformation curve. One pro-posed definition of yield load is the maximum load.Another approach, originally suggested by Harding andFowkes (1984), defines the yield load as the intersec-tion of the load-deformation curve with a straight lineparallel to the initial portion of the load-deformationcurve and offset a distance of 5 percent of the fastenerdiameter from the origin of the load-deformation curve.This 5 percent offset approach has been selected as thedefinition of yield load for this study.

3

Figure 3—Experimental arrangement for deter-mining dowel embedment strength.

The 5 percent offset yield load has several advan-tages over using the maximum load. For connectionswith large tm /D ratios, the EYM equations predict aMode III. yielding. The typical load-deformation curve(Fig. 5) shows three stages of yielding. The third stageis caused by the restraint imposed by the bolt head andnut, thus producing a Mode IV yielding. By using the5 percent offset load, yield load is defined in the regionof Mode IIIs yielding.

For perpendicular-to-grain loading, there is often adrop in load before the maximum load is reached(Fig. 6). This is due to a fracture failure mode, whichis not handled by the EYM. The 5 percent offset ap-proach defines yield at a load before this initial drop inload occurs.

With yield load defined on a 5 percent offset basis, theother input properties in the EYM equations shouldalso be defined on the same basis. In this study, thedowel bearing strength of the connection members andthe bending yield strength of the bolts have been deter-mined on the basis of a 5 percent offset load.

Figure 4—Experimental arrangement for determining bolt bending yield strength. The spanwas the longest possible without bearing onthreads.

Results

Results of the connection teats for specimens loadedin tension are summarized in Table 1. Values of pro-portional limit load, yield load (based on 5 percentoffset), and maximum load are presented along withthe deformation at each load. The initial slope of theload-deformation curve is also presented. These resultsindicate that the yield load was less variable than theproportional limit load and generally less variable thanthe maximum load.

Table 2 presents the results of the dowel bearingstrength tests along with the average moisture con-tent and specific gravity of the bearing specimensand bolted connection specimens. The dowel bearingstrength was based on a 5 percent offset load. Appar-ently specimen thickness had no effect on the bearingstrength. The bearing strength perpendicular to thegrain decreased as the diameter increased. This sameeffect was noted by Trayer (1932). There was no diam-eter effect for parallel-to-grain loading.

4

Figure 5—Typical load-deformation curve for abolted connection loaded parallel to grain withtm /D ratio of 10.25. D = 1/2 in. (12.7 mm)(1 in. = 25.4 mm; 1 lb = 4.45 N).

Table 3 gives the bending yield strength of the bolts.The load for calculating the yield strength was obtainedby offsetting the initial slope of the load-deflectioncurve by 5 percent of the bolt diameter and taking theload where the offset slope intersected the curve. If theoffset did not intersect the curve, the maximum loadwas used. The yield strength was calculated by

where

Bending yield strength (2)

P is 5 percent offset load (lb),

L span (in.), and

D bolt diameter (in.).

This equation uses the plastic section modulus (D3/6).The results in Table 3 indicate that the bending yieldstrength decreased as bolt diameter increased.

Tables 4 and 5 compare the specimens loaded in com-pression with those loaded in tension. Apparently thetwo types of loading do not significantly differ exceptthat more deformation at maximum load was obtainedwith compression loading.

Table 6 shows the effect of steel side member thickness.For the 1/2-in. side members, 80 percent of one sidemember was bearing on bolt threads. This would ac-count for the slightly lower proportional limit and yieldloads for this plate thickness.

Figure 6—Typical load-deformation curve fora bolted connection loaded perpendicular tograin with tm /D ratio of 5.125. D = 3/4 in.(19.0 mm) (1 in. = 25.4 mm: 1 lb = 4.45 N).

Analysis

One purpose of the study was to see how well theEYM predicted the experimental results. For inputin the EYM equations, the experimental dowel bear-ing strength was averaged over all tm/D ratios for eachbolt diameter. The dowel bearing strength of the steelside members was not measured. The side memberswere low carbon steel with an assumed yield stress of36,000 lb/in2 (248.2 MPa). The yield stress was in-creased by 35 percent to account for stress concen-trations, as given by American Institute of Steel Con-struction, Inc. (AISC) (1970), to arrive at an allowablebearing stress of 48,600 lb/in2 (335.1 MPa). This al-lowable bearing stress was used as the dowel bearingstrength of the steel side members. The AISC presentlyrecommends an allowable bearing stress of 69,600lb/in2 (479.9 MPa), which would increase the predictedyield load by 5 percent for Mode III, yielding while notchanging the predicted yield load for Mode I, yielding.Other input variables were as measured.

Figures 7, 8, and 9 compare the EYM-predicted yieldloads with experimental yield loads for parallel-to-grainloading. Figures 10, 11, and 12 give comparisons forperpendicular-to-grain loading. In practically all casesthere appears to be general agreement between EYM-predicted loads and experimental results. Notice thatthe EYM predicts only Mode I, and Mode IIIs yield-ing. This was one of the reasons for selecting the 5 per-cent offset load as the definition of yield load becauseit eliminates the apparent Mode IV yielding caused byend restraint on the bolt.

5

Tab

le 1

—Su

mm

ary

of b

olte

d co

nn

ecti

on p

rope

rtie

s fo

r te

nsi

on-l

oade

d co

nn

ecti

ons

(tm

/D =

0.5

)a

Loa

d (l

b)D

efor

mat

ion

(i

n.)

Init

ial

Bol

tP

ropo

rtio

nal

lim

itY

ield

Max

imu

mP

ropo

rtio

nal

lim

itY

ield

Max

imu

msl

ope

(lb/

in)

diam

eter

(in

.)t m

/DA

vera

geb

Ave

rage

Ave

rage

Ave

rage

Ave

rage

Ave

rage

Ave

rage

1/2

3/4 1 1/2

3/4 1

3.00

05.

125

6.75

010

.250

3.00

05.

125

6.75

010

.250

3.00

05.

125

6.75

010

.250

3.00

05.

125

6.75

010

.250

3.00

05.

125

6.75

010

.250

3.00

05.

125

6.75

010

.250

3,06

0(8

.7)

4,07

0(1

3.8)

3,32

0(1

8.2)

3,53

0(1

0.0)

4,23

0(1

8.0)

6,63

0(1

1.8)

6,88

0(1

2.1)

6,76

0(1

3.9)

9,17

3(1

2.7)

11,2

90(1

5.9)

10,2

80(7

.0)

13,8

50(1

2.5)

4,56

0 (9

.3)

6,45

0 (1

2.7)

5,46

0 (1

0.9)

5,70

0 (1

1.9)

8,02

0 (1

1.8)

12,5

40 (

9.9)

13,0

40 (

10.9

)13

,030

(11

.6)

13,6

30 (

9.5)

24,5

50 (

11.9

)22

,350

(11

.3)

26,3

20 (

8.4)

4,85

0(8

.7)

0.02

2(1

9.1)

7,94

0(1

8.9)

0.04

5(3

9.6)

11,9

70(1

3.3)

0.02

6(2

3.6)

14,8

00(1

2.6)

0.03

9(2

5.1)

8,83

0(1

3.4)

0.02

2(2

3.9)

17,9

60(1

3.8)

0.03

6(2

4.2)

24,5

80(1

1.7)

0.04

2(1

8.9)

28,2

60(9

.2)

0.03

9(1

6.2)

13,7

10(9

.1)

0.05

4(2

2.2)

31,5

60(1

8.6)

0.04

8(2

5.1)

38,1

90(1

5.7)

0.05

6(1

6.3)

39,9

50(1

8.6)

0.06

2(1

8.0)

Per

pen

dicu

lar

to g

rain

0.06

0(1

1.7)

0.13

0(4

1.5)

0.09

4(1

5.0)

0.15

5(3

1.5)

0.06

8(9

.7)

0.37

6(3

2.5)

0.08

7(2

0.0)

0.62

8(2

1.6)

0.07

9(1

3.4)

0.20

9(5

0.0)

0.10

2(1

4.0)

0.27

0(3

8.2)

0.11

3(9

.5)

0.45

5(3

8.9)

0.10

7(9

.4)

0.65

6(2

0.7)

0.12

4(1

1.6)

0.12

4(2

6.5)

0.13

8(1

3.1)

0.23

7(3

9.8)

0.15

6(1

1.1)

0.61

2(3

5.3)

0.15

9(1

0.7)

0.47

7(4

2.5)

1,55

0(2

4.1)

2,83

0(2

1.3)

3,59

0(2

2.7)

0.03

4(3

3.0)

0.08

1(1

4.4)

0.18

6(3

9.4)

1,81

0(1

6.1)

3,13

0(1

5.7)

5,38

0(1

5.9)

0.04

1(2

9.4)

0.09

1(1

5.5)

0.30

5(3

5.3)

2,22

0(1

1.9)

3,58

0(1

3.8)

7,08

0(1

0.3)

0.05

5(2

4.2)

0.11

1(1

5.1)

0.37

2(2

8.2)

3,72

0(4

2.7)

5,79

0(3

3.6)

12,2

30(1

2.6)

0.11

4(4

0.3)

0.20

0(3

2.9)

0.61

0(2

5.2)

2,02

0(2

5.5)

3,82

0(1

9.5)

5,27

0(1

4.1)

0.03

6(3

8.0)

0.10

0(1

6.9)

0.27

1(3

4.1)

3,22

0(2

0.1)

5,81

0(1

6.4)

8,59

0(1

8.2)

0.04

7(2

7.0)

0.11

5(1

2.4)

0.32

6(3

2.1)

4,26

0(1

2.4)

7,48

0(1

0.6)

12,2

00(1

3.2)

0.06

4(2

2.6)

0.13

6(1

1.3)

0.40

0(2

1.8)

5,10

0(2

0.3)

9,44

0(1

4.7)

22,0

60(7

.9)

0.06

4(2

8.2)

0.15

3(1

5.4)

0.60

6(1

4.9)

3,58

0(1

7.4)

7,12

0(1

1.5)

10,5

20(1

1.8)

0.05

4(3

0.2)

0.14

5(1

2.7)

0.42

7(3

5.5)

5,43

0(2

0.8)

11,4

50(1

5.3)

16,2

80(1

0.4)

0.05

2(3

2.2)

0.15

1(1

4.5)

0.39

2(3

7.7)

7,23

0(2

0.3)

12,7

70(1

7.6)

20,9

70(1

0.0)

0.08

5(2

4.1)

0.18

6(1

3.1)

0.49

9(2

2.7)

7,75

0(2

1.4)

11,8

60(2

2.3)

29,7

80(1

1.6)

0.08

8(2

2.3)

0.19

2(2

4.6)

0.75

7(1

4.5)

Par

alle

l to

gra

in

129,

150

(22.

1)91

,810

(13

.1)

128,

810

(14.

0)94

,700

(16

.7)

214,

680

(27.

4)21

0,38

0 (2

0.0)

185,

940

(13.

4)20

4,58

0 (1

6.2)

237,

100

(21.

7)33

3,72

0 (7

.3)

237,

790

(8.9

)26

7,06

0 (9

.3)

56,7

60 (

23.2

)50

,460

(20

.9)

43,0

80 (

12.5

)35

,340

(30

.9)

69,4

10 (

26.3

)82

,910

(20

.7)

90,4

40 (

11.1

)81

,660

(13

.7)

84,4

60 (

15.8

)12

2,17

0 (2

2.5)

103,

980

(17.

1)85

,310

(25

.0)

a1

in.

= 2

5.4

mm

; 1

lb =

4.4

5 N

; t m

/D

is

rati

o of

mai

n m

embe

r th

ickn

ess

to b

olt

diam

eter

.bA

vera

ges

base

d on

sam

ple

size

of

20 s

peci

men

s; c

oeff

icie

nt

of v

aria

tion

in

par

enth

eses

, ex

pres

sed

as p

erce

nt.

Table 2—Summary of dowel bearing strength,moisture content, and specific gravitya

Dowelbearing

Average AverageBolt

strengthmoisture specific

diameter contentb(lb/in2)

(in.)gravity

tm /D (percent) (percent) Averagec

1/2

3/4

1

1/2

3/4

1

Parallel to grain

3.000 12.1 0.4715.125 10.3 0.4656.750 11.4 0.451

10.250 11.9 0.456

3.000 11.6 0.4565.125 11.5 0.4306.750 10.9 0.435

10.250 13.3 0.432

3.000 9.6 0.4455.125 10.5 0.4596.750 10.4 0.448

10.250 10.4 0.472

Perpendicular to grain

3.000 11.6 0.5025.125 10.9 0.4746.750 11.9 0.451

10.250 12.0 0.453

3.000 10.9 0.4585.125 11.0 0.4426.750 10.4 0.444

10.250 12.8 0.459

3.000 10.1 0.4415.125 11.1 0.4526.750 10.9 0.434

10.250 10.9 0.479

5,870 (11.7)6,130 (12.1)5,950 (9.4)5,750 (6.5)

5,340 (9.1)5,230 (8.1)5,490 (10.3)5,370 (12.9)

6,140 (11.1)6,310 (7.8)6,140 (6.8)6,160 (7.5)

3,240 (22.2)2,430 (14.2)2,420 (18.2)2,380 (14.5)

2,020 (17.7)2,000 (16.1)2,590 (18.0)2,730 (14.3)

2,000 (14.2)1,850 (10.8)1,790 (18.8)1,910 (13.0)

a 1 in. = 25.4 mm; 1 lb/in2 = 6.89 kPa; tm /Dis ratio of main member thickness to bolt diameter.

b Averages baaed on sample sizes of 20 specimens.cCoefficient of variation in parentheses, expressedas percent.

Table 3—Bolt bending yield strengtha

Bolt diameter Bending yield strengthb

(in.) (1b/in2)

1/2 85,1253/4 79,260

1 47,000

a1 in. = 25.4 mm; 1 lb/in2 = 6.89 kPa.b Based on 5 percent offset load.

Table 4—Load comparison of tension-loaded andcompression-loaded bolted connections parallelto grain (D = 3/4 in.,tm /D = 6.75)a

Load (lb)

Tension Compression

Type of load Averageb Average

Proportional limit 6,880 (12.1) 6,850 (11.9)YieldMaximum

13,040 (10.9) 12,940 (15.8)24,580 (11.7) 26,070 (12.7)

a 1 lb = 4.45 N; tm /D is ratio of main memberthickness to bolt diameter.

b Averages based on sample sizes of 20 specimens;coefficient of variation in parentheses, expressedas percent.

Table 5—Deformation comparison of tension-loadedand compression-loaded bolted connections parallelto grain (D = 3/4 in., tm /D = 6.75)a

Deformation (in.)

Tension Compression

Type of load Average b Average

Proportional limit 0.042 (18.9) 0.045 (21.2)Yield 0.113 (9.5) 0.114 (14.0)Maximum 0.455 (38.9) 0.584 (24)

a 1 in. = 25.4 mm; tm /D is ratio of main memberthickness to bolt diameter.

b Averages based on sample sizes of 20 specimens;coefficient of variation in parentheses, expressedas percent.

Table 6—Effect of steel side member thicknesson bolted connection load when loaded parallel tograin (D = 1/2 in., tm /D = 5.125)a

Average load (lb)Steel

thickness Sample Proportional(in.) sizeb limit Yield Maximum

1/4 20 4,070 6,450 7,9403/8 19 3,980 6,780 8,3401/2 18 3,430 6,180 8,000

a 1 in. = 25.4 mm; 1 lb = 4.45 N; tm /D is ratioof main member thickness to bolt diameter.

b Number of specimens.

7

Figure 7—Experimental and EYM-predicted yield Figure 9—Experimental and EYM-pedicted yield

loads for bolted connections with 1/2-in. (12.7- loads for bolted connections with 1-in. (25.4-

mm) bolts loaded parallel to grain. tm /D is mm) bolts loaded parallel to grain. tm /D is

ratio of main member thickness to bolt diam- ratio of main member thickness to bolt diam-

eter. Fem = 5,920 lb/in2 (40.8 MPa); Fes = eter. Fem = 6.190 lb/in2 (42.7 MPa); Fes =

48,600 lb/in2 (335.1 MPa); Fy = 85,120 lb/in2 48,600 lb/in2 (335.1 MPa): Fy = 47,000 lb/in2

(586.9 MPa); and ts = 1/4 in. (6.4 mm). (For (324.1 MPa); and ts = 1/2 in. (12.7 mm). (For

yield load conversion, 1 lb = 4.45 N.) yield load conversion, 1 lb = 4.45 N.)

Figure 8—Experimental and EYM-predicted yieldloads for bolted connections with 3/4-in. (19.0-mm) bolts loaded parallel to grain. tm /D isratio of main member thickness to bolt diam-eter. Fem = 5,360 lb/in2 (37.0 MPa): Fes =48,600 lb/in2 (335.1 MPa); Fy = 79,260 lb/in2

(546.5 MPa); and ts = 3/8 in. (9.5 mm). (Foryield load conversion, 1 Ib = 4.45 N.)

Figure 10—Experimental and EYM-predicted yield loads forbolted connections with 1/2-in. (12.7-mm) bolts loaded per-

pendicular to grain. tm /D is ratio of main member thickness

to bolt diameter. Fem = 2,620 lb/in2 (18.1 MPa); Fes =

48,600 lb/in2 (335.1 MPa); Fy = 85,120 lb/in2 (586.9 MPa);and ts = 1/4 in. (6.4 mm). (For yield load conversion, 1 lb =

4.45 N.)

8

Figure 11—Experimental and EYM-predictedyield loads for bolted connections with 3/4-in.(19.0-mm) bolts loaded perpendicular to grain.tm /D is ratio of main member thickness to boltdiameter. Fem = 2,330 lb/in2 (16.1 MPa); Fe s

= 48,600 lb/in2 (335.1 MPa); Fy = 79.260

lb/in2 (546.5 MPa); and ts = 3/8 in. (9.5 mm)(For yield load conversion. 1 lb = 4.45 N.)

Figure 12—Experimental and EYM-predictedyield loads for bolted connections with 1-in.(25.4-mm) bolts loaded perpendicular to grain.tm /D is ratio of main member thickness to boltdiameter. Fem = 1,890 lb/in2 (13.0 MPa);Fes = 48,600 lb/in2 (335.1 MPa); Fy =47.000 lb/in2 (324.1 MPa); and ts = 1/2 in.(12.7 mm). (For yield load conversion, 1 lb =4.45 N.)

Figure 13 compares the experimental effect of steelside member thickness with the effect predicted bythe EYM. For this ratio of tm /D, the EYM predictsa Mode III. yielding and an increase in yield load with

Figure 13—Experimental average yield loadand range compared to EYM-predicted ef-fects of steel side member thickness. Fem =6,290 lb/in2 (43.4 MPa); Fes = 48,600 lb/in2

(335.1 MPa); Fy = 85,120 lb/in2 (586.9 MPa);D = 1/2 in. (12.7 mm); and tm /D = 5.125(1 in. = 25.4 mm; 1 lb = 4.45 N).

increased side member thickness. The results in Fig-ure 13 seem to be in general agreement except for the1/2-in. (12.7-mm) thickness. One of the side membersin these connections was bearing primarily on the boltthreads.

Conclusions

This study investigated the properties of bolted con-nections with steel side members. Three bolt diame-ters and four tm /D ratios were investigated. Connec-tions were loaded both parallel and perpendicular tothe grain. Some findings of this study are as follows:

1. A yield load based on a 5 percent offset load is gen-erally less variable than either proportional limit ormaximum load.

2. The European Yield Model (EYM) predicts boltedconnection yield load with acceptable accuracy. Inputproperties in the EYM equations should also be basedon a 5 percent offset basis.

3. Yield load slightly increases with increased steelmember thickness, except when the increased thicknesscauses the side member to bear on the bolt threads.

4. Connections loaded parallel to the grain by tensionor compression loading give the same results.

9

References

AISC. 1970. Manual of steel construction. 7thed. American Institute of Steel Construction, Inc.,101 Park Ave., New York, NY.

ASTM. 1990. Standard test methods for mechanicalfasteners in wood. ASTM D1761-88. American Societyfor Testing and Materials, Philadelphia, PA.

Doyle, D.V.; Scholten, J.A. 1963. Performance ofbolted joints in Douglas-fir. Res. Pap. FPL 2. Madi-son, WI: U.S. Department of Agriculture, Forest Ser-vice, Forest Products Laboratory.

Harding, N.; Fowkes, A.H.R. 1984. Bolted timberjoints. In: Proceedings, Pacific Timber EngineeringConference; 1984 May; Auckland, New Zealand. 3:872-883.

Johanson, K.W. 1949. Theory of timber connections.International Association for Bridge and StructuralEngineering. 9: 249-262.

Lantos, G. 1969. Load distribution in a row of fastenerssubjected to lateral load. Wood Science. 1(3): 129-136.

McLain, T.E.; Thangjitham, S. 1983. Bolted wood-jointyield model’. Journal of Structural Division, ASCE.109(8): 1820-1835.

NFPA. 1986. National Design Specification. NationalForest Products Association, Washington, DC.

Soltis, L.A.; Hubbard, F.K.; Wilkinson, T.L. 1986.Bearing strength of bolted timber joints. Journal ofStructural Engineering, ASCE. 112(9): 2141-2154.

Trayer, G.W. 1932. The bearing strength of wood un-der bolts. Tech. Bull. No. 332. U.S. Department ofAgriculture, Washington, DC.

10