Enginyeria en Tecnologies Aeroespacials

Escola Superior d’Enginyeries Industrial, Aeroespacial i Audiovisual de Terrassa

Introduction to ANSA - Beam natural modes

Aerospace Vehicles
Assignment D - Report

Jordi Sancho Ticó

Aniol Picola Baré

Date . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13th October 2021

Instructor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oriol Casamor Martinell
Aerospace Vehicles

Contents
1 Aim and Scope 3

2 Description of the Structure 3

3 Methodology 4

4 Model description 4

5 Results 5
5.1 FE Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5.2 Analytical calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
5.3 Comparison with experimental data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2
Aerospace Vehicles

1 Aim and Scope

The aim of this deliverable is to understand the basics of a commercial finite element pre-processor software.
Concretely, this deliverable involves ANSA (used to generate the mesh and other structure characteristics), as
well as MSC NASTRAN and META, to solve and visualize results, respectively.
With reference to this, this report presents briefly the followed methodology, the model description and the dy-
namic analysis. The latter is carried out through a FE analysis, analytical calculations and plotted experimental
data.

2 Description of the Structure

The analysed beam is depicted in the following figure:

Figure 1: Set up of the beam experiment with accelerometer (red)

It is basically a beam with an L-shape section. The accelerometer to obtain experimental data is placed at the
tip of the beam and its length is 25 mm. On the other hand, the beam is 1 m long and its section is 24x24x1.5
mm. It is made out of steel and the whole structure (including the 30-gram accelerometer) weights 540g.

3
Aerospace Vehicles

3 Methodology
As it has been previously mentioned, the pre-process software used is ANSA, especifically ANSA Pre Processor
21.0.2 version. The software solver used to obtain the results is NASTRAN, concretely Nastran 2019 version.
On the other hand, the post-processing software used to visualise and analyse the results is META, concretely
META Post Processor 21.0.2 version.

Two different dynamic analysis have been made: FE analysis and analytical calculations. Regarding the FE
analysis, the first 4 eigenmodes have been computed. The beam has been firstly treated in ANSA defining
the FE model. Then, the deformed structure coloured with displacements magnitudes has been obtained using
META. Concerning analytical calculations, the development for a free-free bounday conditions configuration
beam has been followed, computing the centroid, neutral axis of the section and finally the eigenfrequencies.

4 Model description
It has been associated to the beam a mesh with squared elements sized 4 mm and the accelerometer, which has
been simplified using RBE3 type rigid elements. The image below shows these aspects.

Figure 2: Close-up vision of the FE model of the beam

The beam has a double free-end configuration that will be considered to develop the analytical section.

4
Aerospace Vehicles

5 Results
The aim of this section is to present the results obtained through both methods and compare them with the
experimental data obtained in class with the real structure and accelerometer.

5.1 FE Analysis

In this section are presented the 4 first eigenmodes of the structure coloured with the displacements magnitude
in [mm]. The associated frequencies of the corresponding eigenvalues are 78.3 Hz, 122.8 Hz, 138.1 Hz and 218.2
Hz, respectively. Note that the coloured magnitudes represent displacements in mm.

Figure 3: Representation of the 1st eigenmode

5
Aerospace Vehicles

Figure 4: Representation of the 2nd eigenmode

Figure 5: Representation of the 3rd eigenmode

6
Aerospace Vehicles

Figure 6: Representation of the 4th eigenmode

5.2 Analytical calculations

The equation used to calculate by analytical methods the four first natural frequencies is the following:

  12
λi E·I
fi = · (1)
2πL2 m

Where λ is an experimental value that depends on the boundary conditions and the vibration mode, L refers to
the length of the structure, E means the modulus of elasticity, I refers to the area moment of inertia of beam
about neutral axis and m is the total mass of the structure. In this case, the statement of the problem takes
E = 210 GP a and m = 540 g (mass of the structure + mass of the accelerometer). Furthermore, it is known
that the structure has an L-shape section with both high and width equal to 24 mm and thickness equal to 1.5
mm.

Making the hypothesis that in the experimental case the structure was only supported by an ideally slim thread,
it has been assumed that the boundary conditions are free-free at both sides when doing the analytical calcu-
lations. λ values corresponding to this case are the following:

λ1 4.73004074
λ2 7.85320462
λ3 10.9956078
λ4 14.1371655

Table 1: Values for λ corresponding to free-free boundary conditions

7
Aerospace Vehicles

Regarding the value of I, it is necessary to make the analytical calculation of the neutral axis in order to obtain
the required value.

First, the centroid of the L-section has to be calculated:

24−1.5


24 · 1.5 · 0.75 + (24 − 1.5) · 1.5 · 1.5 +
P
Ai · x i 2
x̄ = P = = 6.56 mm
Ai 24 · 1.5 + (24 − 1.5) · 1.5

P
A ·y 24 · 1.5 · 12 + (24 − 1.5) · 1.5 · 0.75
ȳ = Pi i = = 6.56 mm
Ai 24 · 1.5 + (24 − 1.5) · 1.5

Secondly, the value of inertia moments with respect the principal axes of inertia are obtained:

1 1
Ixx = · 1.5 · 243 + 24 · 1.5 · (12 − 6.56)2 + · (24 − 1.5) · 1.53 + (24 − 1.5) · 1.5 · (6.56 − 0.75)2 = 3938.9661 mm4
12 12

 2
1 1 24 − 1.5
Iyy = ·24·1.53 +24·1.5·(6.56−0.75)2 + ·(1.5)·(24−1.5)3 +(24−1.5)·1.5· 1.5 + − 6.56 = 3938.9661 mm4
12 12 2

0
 
24 − 1.5
Ix0
Ixy =  *

y 0 +24·1.5·(12−6.56)·(0.75−6.56)+(24−1.5)·1.5· 1.5 + − 6.56 ·(0.75−6.56) = −2351.6120 mm4
2

Note that Ix0 y0 is equal to zero since rectangular sections are symmetric.

Once the previous results are known, the neutral axis is obtained as follows:

Iyy · y − Ixy · x
σz = 2
· Mx −→ σz = 0 (according the definition: no extension or compression occurs)
Ixx · Iyy − Ixy

Thus:

Ixy 2351.6120
0 = Iyy · yN A − Ixy · xN A −→ yN A = · xN A = − · xN A = −0.597 · xN A
Iyy 3938.9661

It can be seen the obtained neutral axis is inclined an angle θ = −30.84o with respect to the X-axis of inertia.
Then, to calculate the are moment of inertia of beam about neutral axis, the following equations which allow
to calculate the moment of inertia for an area about inclined axes are used:

Ix0 x0 = Ixx · cos2 θ + Iyy · sin2 θ − Ixy · sin2θ

8
Aerospace Vehicles

Iy0 y0 = Ixx · sin2 θ + Iyy · cos2 θ + Ixy · sin2θ

 
Ixx − Iyy
Ix0 y0 = · sinθ + Ixy · cos2θ
2

So that, the area moment of inertia of beam about neutral axis (θ = −30.84o ) is:

 
3938.9661 − 3938.9661
Ix0 y0 = · sin(2 · (−30.84o )) − 2351.6120 · cos(2 · (−30.84o )) = −1115.5923 mm4
2

It has to be mentioned that this value is replaced by its module since a negative frequency cares of physical
meaning.

Finally, replacing the corresponding values at equation 1, the following frequencies are obtained:

f1 [Hz] 74.1678
f2 [Hz] 204.4464
f3 [Hz] 400.7967
f4 [Hz] 662.5373

Table 2: Obtained results of analytical four first natural frequencies

5.3 Comparison with experimental data

The configuration of the in-class experiment was set to convert the vibration of the beam to data stored in a
MatLab file. Afterwards, those results were plotted in order to see the experimental frequencies of the eigenmodes
of the beam. The mentioned plot is presented on figure 7.

9
Aerospace Vehicles

10-3
4.5

4
X 210
Y 0.00401601
3.5
X 140
Y 0.00292288
3
X 75
Y 0.00247663
2.5

X 106
2 Y 0.00180705

1.5

0.5

0
0 50 100 150 200 250

Figure 7: Experimental data results

There are peaks of amplitude at some specific frequencies that correspond to the first 4 eigenfrequencies. These
4 peaks have been labeled in order to identify more easily the frequency value.

Finally, the following table contains the different values obtained for each vibration mode classified by the
followed methodology (experimental, FEA, analytic).

Experimental FEA Analytic

f1 [Hz] 75 78.3 74.1678
f2 [Hz] 106 122.8 204.4464
f3 [Hz] 140 138.1 400.7967
f4 [Hz] 210 218.2 662.5373

Table 3: Obtained natural frequencies classified by methodology

10
Aerospace Vehicles

On table 4 the errors of the FEA and analytical methods respect the experimental results are displayed.

Error %
FEA Analytic
4,4 1,1
15,9 92,9
1,4 186,3
3,9 215,5

Table 4: Table of errors in %

Regarding the FEA, percentage errors are relatively low, meaning the finite element method is a quite accurate
solution of the system. These errors may decrease if a refinement of the mesh of the model is carried out.

On the other hand, errors on the analytical case are excessively big, a fact that lead to think the analytical
method is not valid to solve the vibration modes of the element of study. The huge difference obtained for the
eigenfrequencies may be due to the fact that the accelerometer is not considered in the computations, neither
its position or its mass. The mass of the accelerometer is a 6% of the whole structure mass i what’s more, it is
located on the tip of the beam, so it have a great effect on the obtained eigefrequencies. However, it can be seen
that the higher the eigenmode, the greater the error, so this may be a case study to solve and to understand
the reason of this behavior.

11