Steel and Timber Structures

Chapter [Link] Members and Stability Problem.

If a 6 mm diameter steel rod 1 m long is placed in a resting machine subjected to a pull, as shown in a Figure below, it will be
found to carry a load of about 7 kn before failure occurs. If on the other hand this same rod had been subjected to compression,
then the maximum load, which would have been carried, would be about 0.035 kN, a very big difference.

Failure in the first test occurs by the fracture of the member; in the second it is due
to bending out of the line of action of the load, as indicated.

Since the load carrying capacity of a member in compression is very different from that of a similar member in tension, requires
special treatment. It is seen that failure takes place by bending. This can not occurs unless a moment acts on a member and this
moment results from a number of effects, which make an apparently axial load acts eccentrically. The causes are:
1. The fact that no member can be made perfectly straight.
2. Imperfection in manufacturing leaving some part of the member with slightly different mechanical properties from
the remainder.
3. Inability to ensure that the load actually acts along the centre of area of the cross-section.

3.1. Types of Equilibrium.

a) Stable: The body returns to its initial position after disturbing its condition of equilibrium.
b) Neutral: The body remains in the same apparent equilibrium in its new position.
c) Unstable: The body loss its initial condition of equilibrium.
Now lets apply a disturbing force F at mid-height acting as shown in the figure.

If the strut returns to its position prior to the application of F, then it is in stable
equilibrium.
If it remains in the deflected position, it is in neutral equilibrium.
If it continues to deflect, it is in unstable equilibrium and the strut loses its load carrying
capacity and fails.

We can see that for low value of P the equilibrium is stable, but that as P is increased a load value is obtained which causes the
strut to be in a state of neutral equilibrium. This load value is known as the critical or buckling load of a strut.

3.2. Critical load of a pin – ended strut.(Euler formula).

Lets consider the strut AB with length l as shown in the following figure.

The maximum deflection is a at mid span, and at distance x from the origin, the deflection is (a – y).
The differential equation of bending gives

11/10/2012 Page 11
 Steel and Timber Structures

The solution for this equation is: ; where A and B are constants of integration. To evaluate A
and B it is as follows:
1. When x = 0; y = 0

2. When x = 0; ; (angle of rotation).

, therefore ; is possible only if A = 0;

Finally the solution is:

Now, when x = ± l/2; y = a, and then , from which ,

Therefore , and ; now squaring , and finally we obtain the formula to calculate the critical

load, known as Euler Formula.

; where le = effective length.

Value for Ratio le / l for different end conditions.(Theoretically).

To write the Euler formula in terms of stress, divide the critical load over the area.

is the least radius of gyration.

. The relation is the Slenderness ratio. And

Limitation of the Euler Formula.

The formula show that depends only on the elastic modulus of the material and on the slenderness ratio, this value is true
only for a constant modulus of elasticity; i.e. within elastic limits of the steel.
The steel behaves elastic only up to Proportional Limit σ p. The Structural Stability Research Council (SSRC) of the USA accept
for σp = 0.5 σy, that is 0.5 the value for the yield limit to ensure perfectly elastic behaviour. Then for mild steel like A – 36, σy
= 24.82 kN/cm2 and E = 2 x 104kN/cm2:

. Therefore, for values of the slenderness ratio

less than 126, Euler’s formula is not valid, as shown in the following figure.

11/10/2012 Page 12
 Steel and Timber Structures

As we see, the Euler’s buckling load can only represents column behaviour at higher values of the slenderness ratio. For lower
values of the le / r, empirical formulae are used.
The SSRC of the USA recommended the following formula, which is accepted by the AISC (American Institute for Steel
Construction) code.

. In which , and for mild steel like A – 36 Cc = 126.

3.3. Design of Axially loaded Columns.

According with EBCS – 3. Design of Steel Structures, section 4.5.4,1; the compression resistance of cross section is as follows:
1. For member in axial compression, the design value of the compressive force Ncom,Sdat each cross-section shall satisfy:

.
Where Ncom,Rd is the design compression resistance of the cross-section, taken as the smaller of:

a) The design plastic resistance of the gross section, , (for classes 1 – 3 cross-sections)

b) The design local buckling resistance of the gross section, where Aeff is the effective

area of the cross section (for class 4 section).

2. Buckling Resistance of Axially Loaded Compression Members (Nb,Rd).

; Where

For the constant axial compression in members of constant cross-sections, the value of χ for the appropriate non -
dimensional slenderness may be determined from:

Where:

11/10/2012 Page 13
 Steel and Timber Structures

Notes:
- The imperfection factor α corresponding to the appropriate buckling curve shall be obtained from Table 4.8,
page 21 of EBCS-3.
- The selections for a buckling curve for a cross-section shall be obtained from Table 4.11, page 24 of EBCS-3.
- Values for the reduction factor χ for the appropriate non-dimensional slenderness may be obtained from Table 4.9,
page 21 of EBCS-3.

3.4. Buckling length of compression members.

For the basis about buckling length read [Link], EBCS-3. When the column belong to a building frame, the procedure is as
follow.

The frames are divided into 2 types, as shown in the figure above. The coefficient for buckling length ratio (k) depends of the
type of frames; as shown, if sway is not allowed, k < 1, other case if sway occur then k > 1.
According to Appendix A of EBCS-3,
1 The buckling length l of a column in non-sway mode may be obtained from Fig. A.2.1.
2 The buckling length l of a column in a sway mode may be obtained from Fig. A.2.2.

The distribution factors at the ends of the member η 1 and η2 are obtained from:

11/10/2012 Page 14
 Steel and Timber Structures

The symbol Σ includes only those members rigidly connected to the joint. For example:

The distribution factors are:

is the column stiffness coefficient

is the effective beam stiffness coefficient
Finally, the slenderness ratio shall be taken as follows:

Where r is the radius of gyration about relevant axis, determined using the properties of the gross

cross-section.
The values of the slenderness ratio λ shall not exceed the following:
1 For members resisting loads other than wind loads 180
2 For members resisting self weight and wind loads only 250
3 For any member normally acting as a tie but subject to
reversal of stress resulting from the action of wind 350

3.5. Design step for loading compression members:

1. Determine the axial load, Nsd.
2. Determine the buckling length, l, which is a function of the column length, L, and the statical system of the column.
3. Select a trial section (take into consideration economy, i.e. least weight per unit length).
4. Determine the Class of the section according to Section 4.3.2 and Table 4.1. If the cross-section is classified as Class
4, determine Aeff according to Section 4.3.4 and Table 4.4.
5. Determine the non-dimensional slenderness ratio (Section [Link]).
6. Using Table 4.11 determine the appropriate buckling curve.
7. Using Table 4.9 find the value of χ. Interpolation must be used to determine more exact values.
8. Calculate the design buckling resistance Nb,Rd of the member. Buckling about both principal axes must be checked.
9. Check the computed buckling resistance against the applied load. If the calculate value is inadequate or is too high,
select another section and go back to Step 4.

11/10/2012 Page 15
 Steel and Timber Structures

Example 1.
The column B – E on the Figure shown below is under the action of NSd = 2800 kN. Both sides are pinned. Check the resistance
of the column. Steel grade Fe 430 is used.

Solution:
Step 1: Axial load NSd = 2800 kN.
Step 2: Buckling length L = 4000 mm (pinned end both sides. Frame non-sway mode).
Step 3: The section is given.
Step 4: Determine the class of the cross-section and check for a local buckling. The section is subjected to uniform compression.
For the section to be classified as at least class 3, in order to avoid any modification to the full cross sectional area due to local
buckling, the limiting width to thickness ratio for class 3 section are (See Table 4.1 EBCS-3).
Outstand element of compression flange: c / tf ≤ 15 ε.
Web subject to compression only: d / tw ≤ 39 ε.
For Fe 430 steel grade fy = 275 N / mm2. Thus
This gives the following limiting values:
Outstand element of compression flange: c / tf = (254/2) / 16.3 = 7.78 < 14 x 0.92 = 12.88 OK.
Web subject to compression only: d / tw = (310-2 (33)) / 9.1 = 26.8 < 42 x 0.92 = 38.64 OK.
Therefore, the section belongs to at least Class 3. Thus, β A = 1.0

Step 5: Determine the non-dimensional slenderness ratio.

For Fe 430 steel grade, λ1 = 93.9 ε = 93.9 x 0.92 = 86.39
Slenderness ratio about y-axis: λy = L / iy = 4000 / 135 = 29.63
Slenderness ratio about z-axis: λz = L / iz = 4000/63.6 = 62.89
Hence, the non-dimensional slenderness ratio is determined as:

Step 6: Determine the appropriate column curves (Table 4.11 EBCS-3).

Use curve a for buckling about y-axis and curve b for buckling about z-axis.
Step 7: Determine value of χ. Using Table 4.9 and interpolating:
For y-axis: curve a for

For z-axis: curve b for

Therefore, buckling about the z-axis becomes critical.

Step 8: Calculate the design buckling resistance.

Step 9: Because 2800 kN> 2117.5 kN, the column do not resist.

11/10/2012 Page 16
 Steel and Timber Structures

Solution 1. Add an additional hinged support at mid-height to increase the resistance about the minor axis.

Go to Step 5.
Slenderness ratio about z-axis = 29.63 (don’t varies)
Slenderness ratio about z-axis = 2000 / 63.6 = 31.45
Non dimensional slenderness ratio don’t varies

Values of χ:
y-axis: χy = 0.97 don’t varies
z-axis: Curve b for

Hence buckling about the z-axis becomes critical

Solution 2: Add 2 plates 200 x 10 mm to reinforce the weak axis.

Now:

; then

For column curve b; χz = 0.77

And

11/10/2012 Page 17
 Steel and Timber Structures

Example 2
Determine the design buckling resistance of a 457 x 152 x 52 UB used as a pin-ended column. The column is 3.00 m long and
its steel grade is Fe 360.

Step 2: Buckling length = 3000 mm.

Step 3: The section is given.
Step 4: Determine the class of the cross-section and check for local buckling.

For Fe steel grade fy = 235 N / mm2. Thus,

These limiting values are:

Outstand element of compression flange: c / tf ≤ 15 ε = 15
Web subject to compression only: d / tw ≤ 39 ε = 39

For the 457 x 152 x 52 UB profile, the actual values are:

Outstand element of compression flange: c / tf = (152.4 / 2) / 10.9 = 7 < 14 OK.
Web subject to compression only: d / tw = (449.8 – 2 x 10.9 – 2 x 10.2) / 7.6 = 53.60 > 42
Therefore, the flange satisfies the Class 3 requirement, but the web is Class 4 section. Consequently, there must be a reduction
in the strength of the section to allow for the load buckling which will take place in the web. Therefore, the effective area, Aeff
must be determined for the web.
Explanation for the effect.

The effective width is beff = reduction factor x b = ρ x b.

The method to calculate the effective area (Aeff) is explained in section 4.3.4 of EBCS-3.
To calculate the reduction factor ρ is as follow.
a). ρ = 1; if

b). = p - 0.22 if > 0.673

2 p
p

Where

In which : t is the relevant thickness.

kσ is the buckling factor corresponding to the stress ratio ψ from Table 4.3 or 4.4 as appropriate.
for webs.
In our example, since the column is axially loaded the stress distribution is uniform, i.e. σ 1 = σ2. Table 4.3 is used to calculate
the effective width.
Thus, σ1/ σ2 = 1, and kσ = 4.0 (see lower part of table 4.3)

11/10/2012 Page 18
 Steel and Timber Structures

Therefore the area that should be ignored at the center of the web is:

And then

Step 5: Determine the non-dimensional slenderness ratio (axis-z govern).

Hence the non dimensional slenderness ratio

Step 6: Appropriate column curve.

For h / b = 449.8 / 152.4 = 2.95 > 1.2; and t f = 10.9 < 40 mm; use curve b for buckling about z-axis.
Step 7: Determine the value of χ.
Using Table 4.9 and interpolating, z-axis: curve b for
Step 8: Calculate the design buckling resistance.

Answer: The design buckling resistance

11/10/2012 Page 19