1

“ANALYSIS OF TENSION
MEMBERS”
Module 3:
TENSILE STRENGTH OF
TENSION MEMBERS
I. LEARNING OBJECTIVES
At the end of the lesson, the student will be able to:
1. Compute the net and effective area of the structural shape used as tension
member.
2. Analyze and compute the tensile strength of a tension member based on the
controlling strength limit state in terms of
a. Yielding of he gross section
b. Fracture of the effective net area
3. Differentiate the ASD from LRFD in analysis of tension members.

II. OVERVIEW
A component of a structure resisting tension loads along its longitudinal axis is called a
tension member. Tension members are encountered in most steel structures. They
occur as principal structural members
• in bridge and roof trusses;
• in truss structures such as transmission towers; and
• in bracings systems in multistoried buildings.

They frequently appear as tie rods

• to stiffen a trussed floor system ,or
• to provide intermediate support for a wall girt system.

Tension members may consist of

• a round bar or flat bar;
• a single structural shape consist of angles, tees, channels, W shapes; or
• built-up from a number of structural shapes.

The cross section of some of the typical tension members are as shown in Figure 1.

round bar flat bar angle double angle

starred angle

Figure 1. Typical cross section

of tension members
Channel
built-up channel
 2

III. COURSE MATERIALS

1.0 NOMINAL STRENGTH
The strength of a tension member may be described in terms of the “limit states”
that govern. The controlling strength limit state for a tension member will be one of
the following:

a. yielding of the gross-section of the member away from the connection;

b. fracture of the effective net area (i.e. through the holes) at the connection; or
c. block shear fracture through the bolt holes at the connection.

gross-section
(a) (b) (c)

Figure 2. Strength of tension member. (a) yielding of the gross section. (b) failure
through the effective net area. (c) block shear

• When the limit state is general yielding of the gross section over the member
length without holes, the nominal strength 𝑇𝑛 may be expressed as

𝑇𝑛 = 𝐹𝑦 𝐴𝑔
Where: 𝐹𝑦 = yield stress
𝐴𝑔 = gross cross-sectional area

• When the limit state is a localized yielding resulting in a fracture though the
effective net area of a tension member having holes, such as for bolts, the
nominal strength 𝑇𝑛 may be expressed as

𝑇𝑛 = 𝐹𝑢 𝐴𝑒

where: 𝐹𝑢 = specified minimum tensile strength

𝐴𝑒 = 𝑈𝐴𝑛 = effective net area
𝐴𝑛 = net area
𝑈 = reduction coefficient (an efficiency factor)

Note: The nominal strength presented above is not applicable to threaded steel
rods or to members with pin holes (eyebars).

2.0 GROSS AREA and NET AREA

Whenever a tension member is to be fastened by means of bolts or rivets, holes
must be provided at the connection. As a result, the member cross sectional area at
the connection is reduced and the strength of the member may also be reduced
depending on the size and location of the holes.

• Gross area, 𝐴𝑔 , refers to the sum of the product of the thickness and the gross
width of a member, without holes or part removed, measured perpendicular to the
axis of the member.
𝑡

𝐴𝑔 = 𝑡 𝑥 𝑤
𝑤

gross area
 3

• Net area, 𝐴𝑛 ,refers to the reduced cross-section along the holes or the gross
cross-sectional area of a member minus the holes, notches and other indentions.
In considering the area of such items, it is important to realize that it is usually
necessary to subtract an area a little larger than the actual hole. This is the result
of punching, cutting or drilling holes.

𝐴𝑛 = 𝑤𝑛 𝑥 𝑡
𝑤 where: 𝑤𝑛 = 𝑤 − 𝑛(𝑑ℎ ) = net width
𝑤 = gross width
𝑑ℎ = hole diameter

net area

. Methods used to punch, cut or drill holes in a plate are:

a. It the plate thickness is less than the punch diameter, punch a standard hole.
• For rivet / bolt : total deduction is equal to the rivet/bolt diameter plus 1.6
mm ( 1/16”)
• For fasteners: total deduction is equal to the fastener diameter plus 3.2
mm (1/8 “)
b. If plate is much thicker than the bolt diameter,
• Holes may be subpunched – with diameter of 4.8 mm (3/16”) undersized –
and then reamed outto a full size after the pieces area are assembled.
• Drill holes to a diameter of the bolt or rivet plus 0.8 mm (1/32”).

Note: The effective hole diameter based on specification is 𝑑𝑒𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒 = 𝑑ℎ + 3

3.0 STAGGERED HOLES (Effect on the Net Area)\

Figure 3. Paths of failure on net section

Whenever there is more than one hole and the holes are not lined up transverse to
the loading direction (Figure 3 b), more than one potential failure line exists. The
controlling failure line is that which gives the largest stress on an effective
net area. In many cases, the critical failure path is also the path that has minimum
net area.
To determine the critical net area, V.H. Cochrane proposed an empirical formula
to account for the “length correction”.
The method is to take the gross width of the member, regardless of the line
along which failure might occur, subtract the diameter of the holes along
zigzag section being considered and add for each inclined line the quantity by
the given expression 𝑺𝟐 ⁄𝟒𝒈. This value is known as the Cochrane’s empirical
formula.

g (gage)

Figure 4

s (pitch)
 4

From Figure 4,
𝑆2
Cochrane’s Formula =
4𝑔

where: 𝑠 = longitudinal spacing (or pitch) of any two holes.

𝑔 = transverse spacing (or gage) of the same holes.

The path at which failure might occur is the one which has the smallest net width
for a particular joint. The smallest net width obtained is multiplied by the plate
thickness to give the minimum net area, 𝐴𝑛 .

𝐴𝑛 = 𝑤𝑛 𝑥 𝑡
𝑆2
where: 𝑤𝑛 = 𝑤 − 𝑛(𝑑ℎ ) + ∑ = smallest net width along failure line
4𝑔

Important: For the purpose of design calculation, the net area shall be taken not
greater than 85% of the gross area, thus

𝐴𝑛 ≤ 0.85 𝐴𝑔

This assumes that the connection has an efficiency of only 85%. This is applied to
riveted and bolted splice and gusset plate.

1) Net Sections for Angles with Staggered Holes

(a) The gross width of the angle is the sum of the width of the legs , 𝐿, less the
thickness 𝑡.

(a) (b) (c)

Figure 5 Net width of angle with staggered holes

𝑤𝑔𝑟𝑜𝑠𝑠 = 𝑤 = 𝐿1 + 𝐿2 − 𝑡

(b) When holes are staggered on two legs of an angle, the gage length 𝑔 for use in
the 𝑆 2 ⁄4𝑔 expression is obtained by using a length between the centers of the
holes measured along the centerline of the angle thickness, i.e., the distance A-B
in Figure 5 (a).

Thus, the gage distance 𝑔 is the sum of the gages from the back of angles less
the thickness
𝑡 𝑡
𝑔 = 𝑔𝑎 − + 𝑔𝑏 −
2 2

𝑔 = 𝑔𝑎 + 𝑔𝑏 − 𝑡
 5

(c) The critical width of the section with staggered holes is similar for that for plates.

𝑆2
𝑤𝑛 = 𝑤 − 𝑛(𝑑ℎ ) + ∑
4𝑔

2) Net Sections for Rolled Shapes

The 𝑆 2 ⁄4𝑔is merely an approximation or simplification of the complex stress

variation that occur in members with staggered arrangements of bolts/rivets. The net
area is computed by getting the net width first, applying𝑆 2 ⁄4𝑔 if the holes are
staggered and then multiplying with the thickness. This procedure is true when
computing the net width for plates and angles. However, for channels, W-sections,
S-sections and others the web and flange thickness are not the same. As a
result, it is necessary to work with the net areas, 𝐴𝑛 , rather than the net width, 𝑤𝑛 .
The net section is determined by calculating the gross area (𝑤𝑥𝑡), deducting the
sum of the areas across the holes (𝑛𝑑𝑛 𝑥𝑡)and adding the terms (𝑆 2 ⁄4𝑔 𝑥 𝑡) for each
change in direction of the path.

2
𝐴𝑛 = (𝑤 𝑥 𝑡) − (𝑛𝑑ℎ 𝑥 𝑡) + (∑ 𝑆 ⁄4𝑔) 𝑡

3) Effective Net Area

For a flat plate loaded in axial tension until failure occurs across its net area, the
actual tensile failure stress is equal to the tensile strength of the steel. However, for
the common structural shapes (W, S, C, L and M), the actual tensile failure stress is
probably less than the tensile strength of the steel. This reduced strength of the
member is caused by the concentration of shear stress called shear lag, in the
vicinity of the connection. Shear lag is a nonuniformity of stresses in rolled sections
when tensile load is applied nonuniformly. As a result, the flow of the tensile stress
between the original member cross-section and the net section is not entirely 100%.

To take into account, the AISC specifies that the net area,𝐴𝑛 ,should be multiplied by
a reduction coefficient 𝑈, to obtain the effective net area, 𝐴𝑒 , that is

𝐴𝑒 = 𝐴𝑛 (𝑈)
where:
𝑈 = reduction coefficient
𝐴𝑛 = net area

The value of the reduction coefficient is affected by the cross-section of the member
and by the length of its connection. It is expressed empirically as

𝑥
𝑈 =1− ≤ 0.90
𝐿
where:
𝑥 = distance from centroid of the member being connected eccentrically to
plane of load transfer.
𝐿 = length of the connection in the direction of the loading. It is equal to the
distance between the first and the last bolt in line.

4) Value of 𝒙 in the Eccentricity in Joint

a. Bolted Members
• For wide flange section, W-shape, connected by its flange only, assume that
the section is split into two structural tees, The value of 𝑥 will be the distance
from the outside edge of the flange to the c.g. of the structural tee. (see
Figure 6)
 6

(a) (b)

Figure 6

• For the channel (see Figure 7) connected through it web, 𝑥 is the distance
from the out outside face of the web to the channel centroid.

Figure 7

• For angle bar (see Figure 8) connected through one of its leg, 𝑥 is the
distance from the out outside face of the leg to the angle centroid.(Note: the
tensile strength of unequal angle is appreciably increased if the longer leg is
connected, since the smaller the value of 𝑥 , the larger is the effective area
and thus the larger is the design strength).

Figure 8

(a) (b)

b. Welded Members

AISC rules for 𝐴 and 𝑈 when tension loads are transferred by welds;

• Load is transferred only by longitudinal welds to other than a plate member,

or by longitudinal welds in combination with transverse welds, 𝐴 is equal to
the gross area of the member 𝐴𝑔 .

T w

𝐴 = 𝐴𝑔

𝐴𝑒 = 𝑈𝐴𝑛 = 𝑈𝐴𝑔

• Tension load be transferred only by transverse welds, 𝐴is to equal the area of
the directly connected element and 𝑈 = 1.0.

T
Transverse weld
 7

𝐴𝑒 = 𝑈𝐴𝑛 = 𝑈𝐴𝑐𝑜𝑛𝑛𝑒𝑐𝑡𝑒𝑑
• Should load be transmitted to a plate by longitudinal welds along both sides of
a plate spaced apart such that 𝑙 ≥ 𝑤

T
𝑤

𝐴𝑒 = 𝑈𝐴𝑔

where: 𝑙 = length of weld along one side of the plate

𝑤 = distance between longitudinal welds (i.e. plate weld)
𝑈 = 1.00 For 𝑙 ≥ 𝑤
𝑈 = 0.87 For 2𝑤 > 𝑙 ≥ 1.5𝑤
𝑈 = 0.75 For 1.5𝑤 > 𝑙 ≥ 𝑤

4.0 PHILOSOHIES OF DESIGN AND ANALYSIS

1) Allowable Stress Design (ASD) Method

Stress
FU

stress-strain region for effective net area

FY

stress-strain region for gross member area

The stress-strain diagram shows that increasing the tension will result in a linear
increase in the strain until the yield stress 𝐹𝑦 is attained. Thereafter, inelastic strain
will occur and continue in the ultimate strain region.

a. Tensile Stresses - 𝑭𝒕

To prevent yield failure, the ASD allowable tension stresses of members (except
pin-connected members) are limited to:

a) Yielding of the gross area without rupture. 𝐹𝑡 = 0.60𝐹𝑦 (gross area)

b) Fracture on the effective net area. 𝐹𝑡 = 0.50𝐹𝑢 (net area)

where:
𝐹𝑦 = specified minimum yield strength of steel
𝐹𝑢 = specified ultimate stress of steel

b. Tensile Strength : 𝑷𝒕

a) Yielding of the gross area without rupture. 𝑃𝑡 = 𝐴𝑔 𝐹𝑡 = 0.60𝐹𝑦 𝐴𝑔

b) Fracture on the effective net area : 𝑃𝑡 = 𝐴𝑒 𝐹𝑡 = 0.50𝐹𝑢 (𝐴𝑛 𝑈)
 8

2) Load and Resistance Factor Design (LRFD)

a. General Philosophy

For tension members, the design strength must be equal or greater than factored
load

∅𝑡 𝑇𝑛 ≥ 𝑇𝑢
where:
∅𝑡 = resistance factor relating to tensile strength
𝑇𝑛 = nominal strength of a tension member
∅𝑡 𝑇𝑛 = design strength of a tension member
𝑇𝑢 = factored load on a tension member

b. Tensile Strength

The design strength ∅𝑡 𝑇𝑛 according to LRFD is the smaller of that based on

a. yielding in the gross section:

∅𝑡 𝑇𝑛 = ∅𝑡 𝐹𝑦 𝐴𝑔 = 0.90𝐹𝑦 𝐴𝑔

b. or, fracture in the net section:

∅𝑡 𝑇𝑛 = ∅𝑡 𝐹𝑢 𝐴𝑒 = 0.75𝐹𝑢 𝐴𝑒
.
Where ; ∅𝑡 = 0.90 = the resistance factor for the yielding limit state
∅𝑡 = 0.75 = the resistance factor for the fracture limit state

3) Reduction Factor , ∅𝒕 , and Safety Factor, Ω

The margin established between resistance and load in real cases is intended to
reduce the probability of failure or unserviceability to an acceptably small value. The
LRFD and ASD approach achieves the goal when there usually is insufficient
information for a complete mathematical description of either load and resistance.
a. A safety factor Ω𝑡 , is a number usually greater than 1.0 used in ASD method.
The nominal strength for a given limit state is divided by Ω𝑡 and the result
compared with the applicable service load condition.
b. A resistance factor ∅𝑡 , is a number usually less than 1.0 used in the LRFD
method. The nominal strength for a given limit state is multiplied by ∅𝑡 and the
result is compared with the applicable factored load condition.

The relationship between the factor safety Ω𝑡 (ASD) and the resistance factor ∅𝑡
(LRFD) in general is

1.50
Ω𝑡 =
∅𝑡

For the design strength

a. Gross section yielding

𝑇𝑛 = 𝐹𝑦 𝐴𝑔

LRFD : ∅𝑡 = 0.90 : ∅𝑡 𝑇𝑛 = ∅𝑡 𝐹𝑦 𝐴𝑔
 9

𝑇𝑛 𝐹𝑦 𝐴𝑔
ASD : Ωt = 1.67 =
Ωt Ωt

b. Tensile rupture strength

𝑇𝑛 = 𝐹𝑢 𝐴𝑒 = 𝑈𝐹𝑢 𝐴𝑛

LRFD : ∅𝑡 = 0.75 : ∅𝑡 𝑇𝑛 = ∅𝑡 𝑈𝐹𝑢 𝐴𝑛

𝑇𝑛 𝐹𝑢 𝐴𝑒
ASD : Ωt = 2.00 =
Ωt Ωt

IV. ACTIVITIES / ASSESSMENT