School of Engineering

Solid Mechanics 3

Laboratory Experiment

Structural Buckling of Columns

Course Coordinator: Dr. Monir Takla

Tutor’s Name: Tamer Botros Ataalla

Student Name: Low Guo Seng

Student Number: S3523224

Date of Experiment: 1-MAY-2017

 SUMMARY

Buckling is a failure mode which is caused by excessive compressive stress

on to a structure. In this experiment, Euler buckling is conducted to show how the
effective length of a strut affects the critical buckling load of a column, and how
the boundary of the column affects the effective length and the buckling load. In
this experiment, all the columns used are assumed to be straight before
compression starts on the axial direction. The critical buckling load is then
measured. The critical buckling load represents the maximum load that can be
exerted onto the column before the column is not able to go back to its original
shape.

In this experiment, it is observed that as the column’s effective length

increases, the critical buckling load decreases, hence proving the Euler’s equation.
By comparing the experimental result with the theoretical result, the result follows
the same trend line but the values of the experimental result are slightly off. This
may be caused by errors that may have occurred throughout the experiment. The
experiment also show that the effect of boundary conditions of the columns to the
buckling load. This experiment proves very useful to understand the how if
buckling affected and the effects of it.

School of Engineering, RMIT University 2

1. Learning Objectives of the Experiment

This experiment offers an introduction to column buckling. A column (or a strut) is a

slender structural member, which is mainly loaded by axial compression. Failure
(lateral deformation) of thin-long columns under excessive compressive loads is called
bucking. The experiment aim is to visualize the concept of column buckling and
increase your knowledge in the topic.

Specific technical objectives are:

 Demonstration of Euler buckling

 Relationship between strut length and collapse load
 Relationship between various end-fixing conditions and collapse load
 Nature of deflection and deflected shapes with various end-fixing conditions
 Writing up a technical report
 Review of theoretical knowledge

2. Introduction to Experiment

In this experiment you are asked to load struts of various lengths to establish the
relationship between the strut length and the buckling load, then compare the results to
the simple Euler theory.

You are asked to examine the Euler Buckling equation and select an appropriate
parameter to establish a relationship between the buckling load and the length of the
strut. You should plot a graph to prove the relationship between the buckling load and
the reciprocal of the strut length. You should conclude whether the Euler formula
predicts the buckling load.

This experiment also examines the effect of changing boundary (end) conditions at the
ends of the column on the buckling load.

You should realise the relationship between the buckling load and the second moment
of area as well as the type of material. The buckling load is directly proportional to the
second moment of area and the Young's modulus.

The magnitude of buckling loads would suggest that the flat thin beams were selected
to give relatively small loads over a wide range of end conditions and strut lengths.

School of Engineering, RMIT University 3

3. Description of Equipment
The experiment hardware fits onto a multi-purpose “Structures Test Frame”. You
compress aluminium columns (struts) using a screw mechanism. The equipment uses
chucks to hold the struts and allows different end-fixing conditions. An integral load cell
connected to a Digital Force Display displays the load on the strut as it is compressed.
A magnetic deflection scale shows how much the strut buckles. You continue
compressing the strut until reaching the critical buckling load. You then repeat the
experiment using different strut lengths or fixing conditions, analysing your results. The
equipment includes strut storage space and five different sizes of aluminium strut.

Figure 1 The STR12 Hardware (Shown without the STR1 Frame)

School of Engineering, RMIT University 4

This experiment is based on the 'STR12 - Buckling of Struts' Structures Hardware,
shown in Figure 1. It consists of a back plate with a load cell at one end and a device to
load the struts at the top. There are five aluminium alloy struts included in a holder on
the back plate.

The general equations used in this experiment are:

More detailed analysis are available below in section 4

School of Engineering, RMIT University 5

4. Theoretical Background

School of Engineering, RMIT University 6

Finally,

School of Engineering, RMIT University 7

Reference:

Hibbeler, R.C., Mechanics of Materials, Sixth International Edition, Prentice Hall, 2005.

School of Engineering, RMIT University 8

5. Results and Calculation
Length (mm) Pinned/Pinned (N) Pinned/Fixed (N) Fixed/Fixed (N)
320 92 92
400 102 77
420 38 38
480 140 145
500 55 70
520 19 19
Table 1: Experiment result

If Pinned/Pinned, length is multiply by a factor of 1, for effective length.

If Pinned/Fixed, length is multiply by a factor of 0.7, for effective length.
If Fixed/Fixed, length is multiply by a factor of 0.5, for effective length.

Sample calculation for effective length:-

For 320mm, since it is Pinned/Pinned, factor is 1
Since 𝐿𝑒𝑛𝑔𝑡ℎ𝑒𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒 = 𝑙𝑒𝑛𝑔𝑡ℎ ∗ 𝑓𝑎𝑐𝑡𝑜𝑟
∴ 𝐿𝑒𝑛𝑔𝑡ℎ𝑒𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒 = 320𝑚𝑚 ∗ 1 = 320𝑚𝑚

For 400mm, since it is Pinned/Fixed, factor is 0.7

Since 𝐿𝑒𝑛𝑔𝑡ℎ𝑒𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒 = 𝑙𝑒𝑛𝑔𝑡ℎ ∗ 𝑓𝑎𝑐𝑡𝑜𝑟
∴ 𝐿𝑒𝑛𝑔𝑡ℎ𝑒𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒 = 400𝑚𝑚 ∗ 0.7 = 280𝑚𝑚

For 480mm, since it is Fixed/Fixed, factor is 0.5

Since 𝐿𝑒𝑛𝑔𝑡ℎ𝑒𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒 = 𝑙𝑒𝑛𝑔𝑡ℎ ∗ 𝑓𝑎𝑐𝑡𝑜𝑟
∴ 𝐿𝑒𝑛𝑔𝑡ℎ𝑒𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒 = 480𝑚𝑚 ∗ 0.5 = 240𝑚𝑚

Sample calculation for load:-

For buckling load, the load is averaged.
At 480mm, the load is 140N and 145N.
𝐿𝑜𝑎𝑑1 +𝐿𝑜𝑎𝑑2
Since 𝐿𝑜𝑎𝑑𝑎𝑣𝑔 = ,
2
140𝑁 + 145𝑁
∴ 𝐿𝑜𝑎𝑑𝑎𝑣𝑔 = = 142.5𝑁
2

By rearranging effective length in order,

Effective Length (mm) Average Buckling Load for Experimental (N)

240 142.5
280 89.5
320 92
350 62.5
420 38
520 19
Table 2: Average buckling load for experiment

School of Engineering, RMIT University 9

Theoretical result:-

Sample calculation for effective length of 240mm:-

Π2 𝐸𝐼
As for theoretical result, since theoretical critical buckling load, 𝑃𝑐𝑟 = 𝐿2

Where the properties of the struts are:-

Modulus of Elasticity, E = 69GPa
Width, b = 0.02m
Thickness, h = 0.00195m
1
And since 𝐼 = 12 𝑏 ∗ ℎ3 ,
1
∴𝐼= ∗ (0.02𝑚) ∗ (0.00195𝑚)3 = 1.2358125 ∗ 10−11 𝑚4
12

Since L = effective length, L = 240mm

Π2 ∗ (69 ∗ 109 𝑃𝑎) ∗ (1.2358125 ∗ 10−11 𝑚4 )
∴ 𝑃𝑐𝑟 =
(0.24)2
∴ 𝑃𝑐𝑟 = 146.1097𝑁

Effective Length (mm) Buckling Load for Theoretical (N)

240 146.1096621
280 107.3458742
320 82.18668494
350 68.70135949
420 47.70927742
520 31.12395169
Table 3: Theoretical result

Effective Length Average Buckling Load Buckling Load for Theoretical (N)
Length (mm) (mm) for Experimental (N)
240 480 142.5 146.1096621
280 400 89.5 107.3458742
320 320 92 82.18668494
350 500 62.5 68.70135949
420 420 38 47.70927742
520 520 19 31.12395169
Table 4: Theoretical result vs experimental result

School of Engineering, RMIT University 10

 Buckling Load against Effective length
160

140

120
Buckling Load (N)

100

80
Experimental Result
60 Theoretical Result

40

20

0
200 250 300 350 400 450 500 550
Effective Length (m)

Graph 1: Buckling Load against Effective length

School of Engineering, RMIT University 11

6. Discussion
Based on the experimental and theoretical results, it shows that as the effective
length increases, the buckling load decreases. This is also shown in the Euler’s formula,
1
where the bucking load is inversely proportional to effective length squared, 𝑃𝑐𝑟 ∝ 𝐿2 .
Since the experimental result’s graph’s trend follows the theoretical result’s graph’s
trend, hence it shows that the Euler’s formula is fairly accurate.

Although the experimental result’s graph follows the theoretical result’s graph,
but it generally lies below the theoretical result except when effective length is 320mm.
effective length 320mm does not lie on the trend line of the experimental graph. The
difference where effective length 320mm does not lay on the trend line and
experimental results generally lies below theoretical results may have been caused by
errors. There are two types of possible errors that may occur in the experiment,
systematic and random error. One of the systematic errors that may have occurred is
instrumental error, where since the STR12 hardware is old and have been used many
times, it may not be able to accurately show the buckling load, hence making the result
to be below the graph. Another possible systematic error is materials property. The
material of the column used is aluminium with the given modulus of elasticity of 69GPa,
but since the theoretical graph is above the experimental graph, and buckling load is
directly proportional to modulus of elasticity, hence there is a possibility that the given
modulus of elasticity was too high which increases the theoretical values.

As for random error, since the experiment was done by inexperience students,
the column may not have been inserted properly sometimes, hence making the result to
be out of place, example, when effective length is 320mm. This error can be avoided by
asking multiple people to inspect before doing the experiment. Another random error
that may occur is that the columns are imperfect; this means that the column may have
been pre-buckled, since the column had been buckled many times. To avoid this error,
a new column could be used. Besides, another possible random error could occur when
zeroing the instrument. Since the instrument is zeroed every time it is testing a new test
or when doing a second try, and it may not be zeroed properly, hence making an error
in the result. To reduce this error, the experiment could be done multiple times and
averaging the result for higher accuracy of result.

As for when changing boundary conditions, since changing boundary conditions

would affect the buckling load. In table 4, it shows that length of 480mm column, not the
shortest column, may have been used but the effect length of the column is only
240mm, the shortest effective length. This is due to both ends of the column is fixed,
hence decreasing the effective length of the column by a factor of 0.5. And since
bucking load is inversely proportional to effective length squared, hence the smaller the
effective length, the larger the buckling load, therefore if both ends of the column is
fixed, it would increase the bucking load by a factor of 4. Since during the experiment,
pin joints were not actually used, but the column was just fit into the slit of the
instrument, hence this may have caused inaccuracy to the results obtain. To prevent
this error, actual pin joints should be used. Another possible error that may cause the
result to be inaccurate is when trying to fix the ends; the column is not properly
tightened, hence the column’s fixed ends are not properly fixed.

School of Engineering, RMIT University 12

 During experiment, when buckling was in progress, the column was observed to
be always buckling toward the wider side, and since it is the side with the smaller
second moment of area, hence the column buckles towards the side with the smallest
second moment of area. From the Euler formula, since buckling load is directly
proportional to second moment of area, hence the smaller the second moment of area,
the smaller the buckling load, and since the wider side has smaller moment of area, the
column starts to buckle at that side before being able to buckle towards the narrower
side. In this experiment, flat thin beams were used as it helps understand better how
second moment of area will affect the direction of buckling. If the column were to
change to a cylinder, is may not buckle as such as there is no side with smaller second
moment of area unless due to imperfections, hence there is not telling how the cylinder
might buckle.

The material used in the experiment is aluminium with a modulus of elasticity of

69GPa, since buckling load is directly proportional to modulus of elasticity, hence if the
material of column were to change to material with a lower modulus of elasticity, the
buckling load will decreases, or a material with higher modulus of elasticity, the buckling
load will increase accordingly.

7. Conclusions
In conclusion, the objectives of the experiment were achieved, showing the
relationship between critical buckling load and the effective length of struts. The
experiment shows that the longer the effective length of struts, the smaller the load for
critical buckling. And it also show that two fixed end has the highest critical buckling
load, one fixed end has the second highest and having both end pinned has the lowest
critical buckling load.

8. Student Comments
This experiment was conducted in a simple and interesting manner. From this
experiment, it shows that buckling theory applies in the real world. To improve the
experiment, one of the suggestions would be to allow the comparison of buckling load
with the same length but with different boundary conditions. Another suggestion would
be to use new column for better accuracy in the result.

School of Engineering, RMIT University 13