Sie sind auf Seite 1von 101

NEi Nastran and Femap

DYNAMICS ANALYSIS

BASIC DYNAMICS

REVIEW OF FUNDAMENTALS

BASIC DYNAMICS

TABLE OF CONTENTS

Page
SINGLE DOF SYSTEM

1-3

CREATING A SPRING ELEMENT

1-5

CREATING A MASS ELEMENT

1-8

SETTING UP A MODAL ANALYSIS

1-12

SPRING ELEMENT TYPES

1-18

UNDAMPED FREE VIBRATIONS SDOF SYSTEM

1-25

DAMPED FREE VIBRATION SDOF

1-30

DAMPING WITH FORCED VIBRATION

1-34

TEXT REFERENCES ON DYNAMIC ANALYSIS

1-44

BASIC DYNAMICS

SINGLE DOF SYSTEM

In this section we introduce the basics of Dynamic Analysis by considering a


Single Degree of Freedom (SDOF) problem
Initially a free vibration model is used to describe the natural frequency
Damping is then introduced and the concept of critical damping and the
undamped solution is shown
Finally a Forcing function is applied and the response of the SDOF is explored in
terms of time dependency and frequency dependency and compared to the
terms found in the equations of motion

BASIC DYNAMICS

SINGLE DOF SYSTEM


Creating a spring element : Step 1

BASIC DYNAMICS

SINGLE DOF SYSTEM


Creating a spring element : Step 2

BASIC DYNAMICS

SINGLE DOF SYSTEM


Creating a spring element : Step 3

BASIC DYNAMICS

SINGLE DOF SYSTEM


Creating a mass element : Step 1

BASIC DYNAMICS

SINGLE DOF SYSTEM


Creating a mass element : Step 5

BASIC DYNAMICS

SINGLE DOF SYSTEM


Creating a mass element : Step 5

BASIC DYNAMICS

10

SINGLE DOF SYSTEM


Creating constraints: Step 1

BASIC DYNAMICS

11

SINGLE DOF SYSTEM


Setting up Analysis: Step 1

BASIC DYNAMICS

12

SINGLE DOF SYSTEM


Setting up Analysis: Step 2

BASIC DYNAMICS

13

SINGLE DOF SYSTEM


Setting up Analysis: Step 3

BASIC DYNAMICS

14

SINGLE DOF SYSTEM


Reviewing Results: Step 1

k
= 100
m
k
= 10 rad/s
n =
m
f = n / 2 = 1.59 Hz

BASIC DYNAMICS

15

SINGLE DOF SYSTEM


Data defined in .NAS file: SOL, Case Control, Parameters and EIGRL

BASIC DYNAMICS

16

SINGLE DOF SYSTEM


Data defined in .NAS file: CELAS2, CONM2, Parameters and EIGRL

BASIC DYNAMICS

17

SINGLE DOF SYSTEM


Spring types

If Spring/Damper is used in Femap a ROD simulation of a spring results:

BASIC DYNAMICS

18

SINGLE DOF SYSTEM


Spring types

If Spring/Damper with BUSH formulation is used in Femap a CBUSH element results:

BASIC DYNAMICS

19

SINGLE DOF SYSTEM


Spring types
Set up in Femap:
CELAS2 DOF Spring
CROD
Spring/Damper
CBUSH
Spring/Damper
with BUSH
formulation

1 DOF spring stiffness with no Property in Femap


1 DOF element axial stiffness, links to PROD and MAT1 to define AE/L
6 DOF element linked to PBUSH

Manually input in Nastran:


CELAS1 1 DOF spring element with links to PELAS property definition
Format:

BASIC DYNAMICS

20

SINGLE DOF SYSTEM


Spring types
In general CBUSH is recommended in most cases. It will be used in more advanced
analysis later in the course.
The main advantages are:

All 6 DOF in a single element


Extendable to Frequency dependent and nonlinear
Arbitrary elastomeric center
No errors due to large rotation and other grounding effects
Special purpose 1D version for shock mounts

BASIC DYNAMICS

21

SINGLE DOF SYSTEM (Cont.)

SDOF System theory


p(t)

m = mass (inertia)

u(t)

b = damping (energy dissipation)


m

k = stiffness (restoring force)


p = applied force
k

u = displacement of mass

u& = velocity of mass


&u& = acceleration of mass

responses u& and u&& vary in time


load input p can vary in time
m, k and b are constant with time in linear analysis
BASIC DYNAMICS

22

SINGLE DOF SYSTEM (Cont.)

The equation of motion is:

mu&&(t ) + bu& (t ) + ku (t ) = p (t )

Inertia Force

Stiffness Force

Applied Force

Damping Force

BASIC DYNAMICS

23

SINGLE DOF SYSTEM (Cont.)

In undamped, free vibration analysis, the SDOF equation


of motion reduces to:

mu&&(t ) + ku (t ) = 0
Assume a solution of the form:

u (t ) = A sin nt + B cos nt
This form defines the response as being HARMONIC,
combinations of sine and cosine shape responses with a
resonant frequency of:

BASIC DYNAMICS

24

UNDAMPED FREE VIBRATION SDOF SYSTEM

For a SDOF system the resonant, or natural frequency, is given


by:

k
n =
m

We can solve for the constants A and B:

When t = 0, sin( n t ) = 0 thus B = u(t = 0)


Differentiating solution :
u& (t ) = An cos n t Bn sin nt
When t = 0, B n sin( n t ) = 0 thus
u& (t = 0)
u& (0)
A=
u (t ) =
sin nt + u (0) cos nt

BASIC DYNAMICS

25

UNDAMPED FREE VIBRATION SDOF SYSTEM


(Cont.)

The response of the Spring will be harmonic, but the actual form of the
response through time will be affected by the initial conditions:
If u ( 0) = 0 and u& ( 0)
or velocity
If

=0

u (0) = 0 and u& (0) 0

there is no response; no initial disturbance

response is a sine function magnitude

u&0
Ifu ( 0)
If

0 and u& (0) = 0

u (0) 0 and u& (0) 0

Depends on input velocity

response is a cosine function (180


phase change), magnitude u0
response is phase and magnitude
dependent on the initial values
BASIC DYNAMICS

26

SINGLE DOF SYSTEM UNDAMPED FREE VIBRATIONS

The graph is from a transient analysis of a spring mass system with Initial
velocity conditions only

k = 100
m=1

T
Amp

u&0 = 1

k
= 10 rad/s
m
f = n / 2 = 1.59 Hz
n

T = 1/f = 0.63 secs


Disp.

Amp = u&0 / n = 0.1

Time (Seconds)
BASIC DYNAMICS

27

SINGLE DOF SYSTEM UNDAMPED FREE VIBRATIONS

The results we saw from NEi Nastran had the correct frequency, but how do we
get the displacement value?
u&0

u0

BASIC DYNAMICS

28

SINGLE DOF SYSTEM UNDAMPED FREE VIBRATIONS

The displacement is indeterminate in an Eigenvalue Analysis

No loading or initial conditions are applied

The mathematical method balances Inertia energy and Stiffness


energy for a particular unique (or eigen) value solution

It is finding a characteristic of the system, not a response

The displacement is an arbitrary scaling, the NEi Nastran solution it is


(1.0)

We will see later that this is an eigenvector

BASIC DYNAMICS

29

DAMPED FREE VIBRATION SDOF

If viscous damping is assumed, the equation of motion becomes:

mu&&(t ) + bu& (t ) + ku (t ) = 0
There are 3 types of solution to this, defined as:
Critically Damped
Overdamped
Underdamped

A swing door with a dashpot closing mechanism is a good analogy


If the door oscillates through the closed position it is underdamped
If it creeps slowly to the closed position it is overdamped.
If it closes in the minimum possible time, with no overswing, it is
critically damped.

BASIC DYNAMICS

30

DAMPED FREE VIBRATION SDOF (Cont.)

For the critically damped case, there is no oscillation, just a


decay from the initial conditions:

u (t ) = ( A + Bt )e bt / 2 m
The damping in this case is defined as:

b = bcr = 2 km = 2m n
A system is overdamped when b > bcr
We are generally only interested in the final case - underdamped

BASIC DYNAMICS

31

DAMPED FREE VIBRATION SDOF (Cont.)

For the underdamped case b < bcr and the solution is the form:

u (t ) = e bt / 2 m ( A sin d t + B cos d t )
d

represents the Damped natural frequency of the system

d = n 1 2

is called the Critical damping ratio and is defined by:

b
=
bcr
In most analyses

is less than .1 (10%) so


BASIC DYNAMICS

d n
32

DAMPED FREE VIBRATION SDOF (Cont.)

The graph is from a transient analysis of the previous spring mass system with
damping applied

Frequency and
period as before
Amplitude is a
function of damping
Disp.
5% Damping
2%
Damping

Time
BASIC DYNAMICS

33

DAMPING WITH FORCED VIBRATION

We now apply a harmonic forcing function:


note that

is the DRIVING or INPUT frequency

p sin t

The equation of motion becomes

mu&&(t ) + bu& (t ) + ku (t ) = p sin t


The solution consists of two terms:
The initial response, due to initial conditions which decays rapidly in the presence of
damping
The steady-state response as shown:

u (t ) = p / k

sin(t + )

(1

) 2 + (2 / n ) 2

This equation is described on the next page


BASIC DYNAMICS

34

DAMPING WITH FORCED VIBRATION (Cont.)

This equation deserves inspection as it shows several important


dynamic characteristics:
Phase lead of the response relative to the input
(see next page)

u (t ) = p / k

This is the static loading


and dominates as tends
to 0.0

sin(t + )

(1

) 2 + (2 / n ) 2

At >> n both terms drive the


response to 0.0

At = n this term = (2 )^2 and controls


the scaling of the response
At = n this term = 0.0
With no damping present this
results in an infinite response
BASIC DYNAMICS

From this is derived the Dynamic


Magnification Factor 1/2
35

DAMPING WITH FORCED VIBRATION (Cont.)

is defined as a phase lead in Nastran :

= tan

2 / n

BASIC DYNAMICS

36

DAMPING WITH FORCED VIBRATION (Cont.)


Summary:
For

<< 1
n
Magnification factor
Phase angle

For

>> 1
n
Magnification factor
Phase angle

For

1 (static solution)
360 (response is in phase with the force)

0 (no response)
180 (response has opposite sign of force)

1
n
Magnification factor
Phase angle

1/2
270
BASIC DYNAMICS

37

DAMPING WITH FORCED VIBRATION (Cont.)

To explore the response of our spring mass system to the


forcing function we will use a Frequency Response
Analysis
This method allows us to compare the response of the
spring with the input force applied to the spring over a wide
range of input frequencies
It is more convenient in this case than running multiple
Transient Analyses, each with different input frequencies
We will apply the input load as 1 unit of force over a
frequency range from .1 Hz to 5 Hz
Damping is 1% of Critical
BASIC DYNAMICS

38

DAMPING WITH FORCED VIBRATION (Cont.)

Magnification Factor = 1/2 = 1/G = 50


Static Response = p/k = .01
Peak Response = .5 at 1.59 Hz
Note:
Use of a Log scale helps identify low
order response
Displacement

Frequency
BASIC DYNAMICS(Hz)

39

DAMPING WITH FORCED VIBRATION (Cont.)

There are many important factors in setting up a Frequency Response Analysis


and we will cover these in a later section
For now, note the response is as predicted by the equation of motion
At 0 Hz result is p/k
At 1.59 Hz result is p/k factored by Dynamic Magnification
At 5 Hz result is low and becoming insignificant

The Phase change is shown here:


In phase up to 1.59 Hz
Out of phase180 Degrees after 1 .59 Hz

BASIC DYNAMICS

40

DAMPING WITH FORCED VIBRATION (Cont.)

We now try a Transient analysis with a unit force applied to the spring at 1.59
Hz
Again damping of 1% Critical is applied
The result is shown on the next page:
The response takes around 32 seconds to reach a steady-state solution
After this time the displacement response magnitude stays constant at .45
units
The theoretical value of .5 is not reached due to numerical inaccuracy (see
later) and the difficulty of hitting the sharp peak

BASIC DYNAMICS

41

DAMPING WITH FORCED VIBRATION (Cont.)

Transient analysis with a unit force applied to the spring at 1.59 Hz

Displacement

Time

BASIC DYNAMICS

42

DAMPING WITH FORCED VIBRATION (Cont.)

If we plot input and output at the steady-state period, we can see that the input signal is
not very accurate hence our problem finding the exact magnification factor
We can also see the phasing between input and output is around 90 degrees as expected
at resonance

T input = 1/1.59 = 0.629 sec

output
input
Lead = 0.18 sec (approx)
=103 degrees (approx 90)
BASIC DYNAMICS

43

TEXT REFERENCES ON DYNAMIC ANALYSIS

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.

W. C. Hurty and M. F. Rubinstein, Dynamics of Structures, Prentice-Hall, 1964.


R. W. Clough and J. Penzien, Dynamics of Structures, McGraw-Hill, 1975.
S. Timoshenko, D. H. Young, and W. Weaver, Jr., Vibration Problems in Engineering,
4th Ed., John Wiley & Sons, 1974.
K. J. Bathe and E. L. Wilson, Numerical Methods in Finite Element Analysis, PrenticeHall, 1976.
J. S. Przemieniecki, Theory of Matrix Structural Analysis, McGraw-Hill, 1968.
C. M. Harris and C. E. Crede, Shock and Vibration Handbook, 2nd Ed., McGraw-Hill,
1976.
L. Meirovitch, Analytical Methods in Vibrations, MacMillan, 1967.
L. Meirovitch, Elements of Vibration Analysis, McGraw-Hill, 1975.
M. Paz, Structural Dynamics Theory and Computation, Prentice-Hall, 1981.
W. T. Thomson, Theory of Vibrations with Applications, Prentice-Hall, 1981.
R. R. Craig, Structural Dynamics: An Introduction to Computer Methods, John Wiley
& Sons, 1981.
S. H. Crandall and W. D. Mark, Random Vibration in Mechanical Systems, Academic
Press, 1963.
J. S. Bendat and A. G. Piersel, Random Data Analysis and Measurement Techniques,
2nd Ed., John Wiley & Sons, 1986.
BASIC DYNAMICS

44

NORMAL MODES ANALYSIS

BASIC DYNAMICS

45

TABLE OF CONTENTS

Page
2 DOF EQUATIONS OF MOTION

1-47

MULTI DEGREES OF FREEDOM

1-73

LANCZOS METHOD

1-83

MASS REPRESENTATION

1-84

RIGID BODY MOTION

1-86

WHY CALCULATE NORMAL MODES?

1-87

ACCURACY IN NORMAL MODES

1-89

NORMAL MODES CHECK LIST

1-101

BASIC DYNAMICS

46

Normal Modes Analysis

Define the following problem in Femap and solve in NEi Nastran


K

DOF: 1

2M
3

K
4

K = 1000 lbf/in
M = 20 lb

BASIC DYNAMICS

47

OVERVIEW

In the previous section we looked at a SDOF problem


of a spring mass system.
In this section we will look at Normal Modes analysis
of Multi Degree of Freedom problems
The steps we will follow are:
Building a 2 DOF equation of motion using engineering
approach.
Summarizing some important ideas about Normal Modes
that emerge.
Setting the same problem using a Matrix approach.
BASIC DYNAMICS

48

2 DOF EQUATION OF MOTION USING AN


ENGINEERING APPROACH

Consider the system with 2 masses and 3 spring stiffnesses as shown.


x1
k

x2
k

2M

Use an engineering approach to solve the equations of motion.


First, set up free body diagrams for the masses.
kx1

k(x2-x1)

k(x2-x1)

2M

m&x&1

kx2

m&x&2

Equating the Inertia and Elastic terms we have,


For 1st mass:

M&x&1 = kx1 + k ( x2 x1 )

2 M&x&2 = k ( x1 x2 ) kx2

For 2nd mass:


BASIC DYNAMICS

49

2 DOF EQUATION OF MOTION

We assume the motion of x1 and x2 is harmonic so


x1 = A1 sin t
This means they vibrate at the same frequency but have
different
amplitudes A.
x2 = A2 sin t
We want to find what the frequency is, and the amplitudes.
Now

&x&1 = 2 A1 sin t

&x&2 = 2 A2 sin t

Then putting the harmonic terms into the free body equations.
For the 1st mass:
So

M 2 A1 sin t = kA1 sin t + k ( A2 sin t A1 sin t )


(2k M 2 )A1 kA2 = 0

For the 2nd mass: 2 M A2 sin t = k ( A1 sin t A2 sin t ) + kA2 sin t


2

so

kA1 + 2k 2 M 2 A2 = 0

BASIC DYNAMICS

50

2 DOF EQUATION OF MOTION

If we assemble these two equations in matrix form,


we have:
(2k 2 M )

A1 0
=
(2k 2 M ) A2 0

So we have 3 unknowns; 2 and the pair of


amplitudes
A
1

A2

BASIC DYNAMICS

51

2 DOF EQUATION OF MOTION

We can solve this by using the determinant of the


above equation, letting 2=.
Two roots of the equation are found as 1 and 2.
These roots are called Eigenvalues.

k
1 = 1 = 0.634
m
2

2 = 2

k
= 2.366
m

So the two frequencies where the inertia and elastic terms


balance are 1 and 2.

BASIC DYNAMICS

52

2 DOF EQUATION OF MOTION

The amplitudes are investigated by substituting back


into the equations of motion.
In turns out we can only solve for the ratio of the
amplitudes.
1

A1
= 0.731
A2

A1
= 2.73
A2

This is an important physical point in the analysis of


normal modes. We do not know the absolute
amplitudes, only relative amplitudes.
BASIC DYNAMICS

53

2 DOF EQUATION OF MOTION

If we arbitrarily call A2 = 1.00, then we can express


the relative amplitudes as Mode Shapes or
Eigenvectors.
2

A1 2.73
=

A
1
.
00

A1 0.731
=

A
1
.
000

2
0.731

1.000

-2.731

Mode 1

1.000

Mode 2

BASIC DYNAMICS

54

2 DOF EQUATION OF MOTION

The motion of all displacements is assumed


harmonic.
Resonance is found at a set of Natural Frequencies
where the Inertia terms balance the Elastic terms.
The Natural Frequencies are calculated by an
Eigenvalue Method
The relative amplitude, or mode shape, is found for
each natural frequency.

BASIC DYNAMICS

55

2 DOF EQUATION OF MOTION

Now Consider the system in terms of a Matrix


Solution.
k

DOF: 1

2M

The individual element stiffness matrices [K1],[K2],


and [K3] are:
1 1

[K ] = [K ] = [K ] = k
1

BASIC DYNAMICS

56

2 DOF EQUATION OF MOTION

To derive the model stiffness matrix [K], we assemble


the individual element stiffness matrices [K1],[K2],
and [K3]:
1
2
3
4
DOF:
1 1

1 1 + 1 1

[K ] = k
1 1 + 1 1

1
1

We constrain out DOFs 1 and 4 as they are set to


0.0
2 1
[K ] = k

1
2

and

1 0
[M ] = m
0 2

BASIC DYNAMICS

We lump masses at DOF

57

2 DOF EQUATION OF MOTION

The equation of motion in matrix form is:

[M ]{&x&} + [K ]{x} = 0
If we substitute in

{x} = { }e
{&x&} = { }e
[M ]{ } + [K ]{ } = 0
i t

And
Then
So

it

This means we have a mode shape,


{} , which varies sinusoidally with a
frequency .

([K ] [M ]){ } = 0
2

2 1
0
2 1
k

1 2 m 0 2 { } = 0


BASIC DYNAMICS

This means we can find a mode


shape, {}, and frequency where
the inertia terms and elastic terms
balance

The Eigenvalue problem

58

2 DOF EQUATION OF MOTION

If we have a set of n physical degrees of freedom (2


in our case)
Then we have n sets of unique Eigenvalues i2 and
eigenvectors {i}
where i = 1 to n

For each of these sets, the inertia terms balance the elastic
terms and this is the definition of resonance.

BASIC DYNAMICS

59

2 DOF EQUATION OF MOTION

So at

k
m

1 = 0.634 , the motion is defined by:

0.731
{1} =

1
.
000

is in balance at this first resonant or natural frequency.

And at = 2.366 k , the motion is defined by: { } = 2.731


1.000
2

is also in balance at this second resonant or natural


frequency

BASIC DYNAMICS

60

2 DOF EQUATION OF MOTION

Let us add some values in and check out the numbers:


Let k = 1000 units of force / length
Let m = 20 units of mass

Then

k
1 = 0.634 = 5.629 rads s = 0.896 Hz
m
k
2 = 2.366 = 10.875 rads s = 1.731Hz
m

Notice the conversion of Frequency from Radians/s to Cycles/s


(Hertz)

f
2

BASIC DYNAMICS

61

2 DOF EQUATION OF MOTION

If we load this model with a time dependant set of


forces at DOF 2 and 3 we will get a displacement
response which is a combination of the two mode
shapes we calculated.
n
So {x(t )} = {i }i(t)
i =1

in our case

{x (t )} = 11 (t ) + 2 2 (t )

The scaling factors i for each mode shape i are


called the Modal Displacements. We will return to
this when we apply loading in later sections.
0.731

{x(t )} =

1.000

1(t)
BASIC DYNAMICS

-2.731

1.000

2(t)
62

2 DOF EQUATION OF MOTION

Now let us model the system using NEi Nastran

DOF: 1

2M
3

K
4

K = 1000 lbf/in
M = 20 lb

BASIC DYNAMICS

63

2 DOF EQUATION OF MOTION

The NEi Nastran .nas file will contain the connectivity


definition of the CELAS2 elements

BASIC DYNAMICS

64

2 DOF EQUATION OF MOTION

The NEi Nastran nas file will contain the CONM2


elements, there is one for each mass point.

BASIC DYNAMICS

65

2 DOF EQUATION OF MOTION


The NEi Nastran Solution , Case Control,
Parameter and EIGRL definition:

BASIC DYNAMICS

66

2 DOF EQUATION OF MOTION


The NEi Nastran results showing first and second
natural frequencies

BASIC DYNAMICS

67

2 DOF EQUATION OF MOTION

The values agree with those we calculated earlier.


The meaning of the Generalized Mass and Stiffness
will be discussed in the next few pages.
1 = 0.796

k
= 5.629rad / s = 0.896 Hz
m

2 = 1.538

k
= 10.875rad / s = 1.731Hz
m

BASIC DYNAMICS

68

2 DOF EQUATION OF MOTION


The NEi Nastran mode shape results:

BASIC DYNAMICS

69

2 DOF EQUATION OF MOTION

The difference in the normalization methods between the hand


calculation and Nastran is interesting:
In the hand calc we normalized displacement at Grid 3 to 1.000
quite arbitrarily

2.731
{2 } =

1
.
000

The Nastran method we selected in this run was Mass. This


means the maximum value in the e-vector list is set to modal mass
1.000 for each e-vector, so
.19858
{2 } =

.07268
This emphasizes that all we know about the e-vectors in a Normal
Modes analysis is their relative values we know the shape, but
not the amplitude.
The commonly used normalization method called Mass
normalization is discussed over page.
BASIC DYNAMICS

70

2 DOF EQUATION OF MOTION

Mass normalization is a useful way of normalizing an


Eigenvector because it can be thought of as a
universal standard.
We scale {} so that for each mode:
{ }T [M ]{ } = [I ]

{i }T [M ]{i } = 1.0

If we wish to compare modes between different


analyses, or even to test data, then it becomes
meaningful to compare the mass normalized
eigenvectors as:

{MODEL1}T [M ]{MODEL 2 }
{MODEL }T [M ]{TEST }

BASIC DYNAMICS

71

2 DOF EQUATION OF MOTION


The term {}T [M]{} is called the Generalized Mass.
It is clearly orthogonal if we can equate it to an Identity
matrix in Mass Normalisation.
In general, orthogonality is defined as:

{i }T [M ]{i } = mi
{i }T [M ]{ j }= 0.0

Generalized Stiffness, {}T [K]{} can be defined in a


similar way:

{i }T [K ]{i } = ki
{i }T [K ]{ j }= 0.0
BASIC DYNAMICS

72

EXTENDING TO MULTI DOF PROBLEMS


Normal Modes of Multi Degree of Freedom (MDOF)

Set up the example in Femap

Use mesh size of 1 inch

BASIC DYNAMICS

73

EXTENDING TO MULTI DOF PROBLEMS


Beam Section properties:

BASIC DYNAMICS

74

EXTENDING TO MULTI DOF PROBLEMS (Cont.)


Analysis Results:
First 10 Natural Frequencies range from 166 Hz to 2764 Hz
Generalized Mass is 1.0 in all modes due to Mass Normalization

BASIC DYNAMICS

75

EXTENDING TO MULTI DOF PROBLEMS (Cont.)


Analysis Results:
It is very important to identify each mode by shape as well as by frequency. Only by
doing both will an unambiguous definition of the response be made.
For example the frequencies could easily shift to switch Mode 9 and 10. Only by
description or plot can we confirm which is which
Mode

Frequency

Description

166.75

xz plane bend 1

409.17

xy plane bend 1

454.60

xz plane bend 2

878.67

xz plane bend 3

1033.28

xy plane bend 2

1427.49

xz plane bend 4

1842.85

xy plane bend 3

2089.58

xz plane bend 5

2761.51

axial

10

2764.00

xy plane bend 4

BASIC DYNAMICS

76

XY plane Modes

BASIC DYNAMICS

77

XZ Plane Modes

BASIC DYNAMICS

78

Axial Mode

BASIC DYNAMICS

79

EXTENDING TO MULTI DOF PROBLEMS (Cont.)


Remember, we defined the contribution of each mode as the modal
displacement:
So

{x(t )} = { }
i =1 to n

In the 2 DOF case

{x(t )} = 11 + 2 2

For the beam example, it may be possible to represent the response to


loading in the XZ plane using the first three modes. The assumption is that
the higher modes do not contribute significantly to the solution.
This is a significant advantage of modal methods, the response of the
beam, {x(t)}, has 222 physical degrees of freedom.
But we can represent the response by 3 Modal DOF, i
EXTREME CARE must be taken when assuming which modes contribute and
we will discuss this more in later sections.
As a taster consider the following over page

BASIC DYNAMICS

80

EXTENDING TO MULTI DOF PROBLEMS (Cont.)


Possible Problems with 3 modes only:
If we have loading in the XY plane, the response will be represented by first
order bending and will therefore be limited in accuracy
If the XZ loading excites at multiple inputs as shown, then the first two
modes may not represent the response, and the third order mode may be
needed

BASIC DYNAMICS

81

EXTENDING TO MULTI DOF PROBLEMS (Cont.)


Repeat the example with different element sizes:

mesh size = 0.5 inch


Frequency Description

Mode

mesh size = 4 inch


Frequency Description

mesh size = 12 inch


Frequency Description

1
2
3
4
5
6
7
8
9
10

BASIC DYNAMICS

82

LANCZOS METHOD

Real Eigenvalue Extraction Data, Lanczos Method


Defines data needed to perform real eigenvalue (vibration or buckling)
analysis with the Lanczos method.
Many advanced settings, basics covered here

BASIC DYNAMICS

83

MASS REPRESENTATION

A model uses a value of 0.00259 as the weight mass conversion


parameter - what does this mean?
Nastran requires consistent units.
Some systems of Units (including the US system) define density as
being a Weight Per Unit Volume (eg. lbs/in3).
This is not a consistent unit if used with loads of lbf and dimensions of
inches.

The weight mass conversion parameter converts weight mass


units to mass units by scaling by the appropriate units of
acceleration due to gravity.

So for our model defined in lbs and inches, g = 386.4 in/s2


PARAM,WTMASS,0.00259 converts the mass of the structure to
the correct units of (lbf/in/s^2).
Some industries also mix SI units for convenience, so density may
be given in N/m3 instead of the correct term Kg/m3
In this case PARAM,WTMASS,0.102 will scale by g = 9.81m/s^2

BASIC DYNAMICS

84

MASS REPRESENTATION

Another common method used in industry is to have length


measured in mm for convenience, but still want to apply forces
in N.
When converting a non-standard system of units the golden rule
is to apply Newtons Law of Motion and then dimensional
equivalence
Force = Mass * Acceleration
N = (mass units) * mm/s2
Kg *m/s2 = (mass units) *mm/s2
Kg *m/s2 = (Kg*103) * mm/s2

So mass is in Kg*103 or Metric Tonnes and density is then


Tonnes/mm3 and these units should be used in the model
In this case if units of Kg mass and Kg/mm3 density are used in the
model PARAM,WTMASS,0.001 will scale the mass and density
units correctly.

BASIC DYNAMICS

85

RIGID BODY MODES

For every DOF in which a structure is not totally constrained, it


allows a Rigid Body Mode (stress-free mode) or a mechanism.
There should be six Rigid Body Modes.

The natural frequency of each Rigid Body Mode should be zero.


A later section is devoted to discussion of Rigid Body Modes
BASIC DYNAMICS

86

WHY CALCULATE NORMAL MODES

We now consider some reasons to compute natural frequencies


and normal modes of structures
To assess the dynamic characteristics of a structure.
For example, if a structure is going to be subject to rotational or cyclic
loading input, to avoid excessive vibrations, it might be necessary to
see if the frequency of the input is close to one of the natural
frequencies of the structure.

Passing blade frequencies of a helicopter


Rotational speed of an automobile wheel
Rotational speed of a lathe
Vortex shedding or flutter of bridge and deck structures
Assess the possible dynamic amplification of the loads.
If a structure is loaded near a natural frequency with an input that
matches that frequency then the dynamic amplification can be
significant for a lightly damped structure, perhaps being an order of
magnitude higher than an equivalent static loading

Dynamic response of aircraft structure due to landing loads


can exceed static loading
Dynamic response of Tacoma Narrows bridge, runaway
loading
BASIC DYNAMICS

87

WHY CALCULATE NORMAL MODES (Cont.)

Use the Modal Data (natural frequencies and mode shapes) in a


subsequent dynamic analysis
We will see later that we have a class of transient and frequency
response analysis methods that use modal techniques, using
Modal data.

Assess requirements of subsequent dynamic analysis


For Transient response, calculate time steps based on the highest
frequency of interest
For Frequency response, calculate the range of frequencies of
interest

Guide the experimental analysis of structures


Identify optimum location of accelerometers, etc.
Avoid overstressing of components

Evaluate the effect of design changes


A normal modes analysis will give a clear indication of of frequency
shifts, changes in mode shapes to allow an early judgment on
effect of design changes to be made

BASIC DYNAMICS

88

HOW ACCURATE IS THE NORMAL MODES ANALYSIS?

The following section explores the key factors:


mesh density
element type
mass distribution
detail of constraints
detail of joints

BASIC DYNAMICS

89

MESH DENSITY
Mesh Density
The mesh must be fine enough to permit a representation of the of the
highest mode considered
In the case of the beam, we assumed the second order mode was
sufficient. The mesh is adequate for this.
However, if the higher order mode shown is required, then the mesh is
inadequate.

BASIC DYNAMICS

90

NORMAL MODES ANALYSIS WITH VARIOUS MESH


SIZE
Rectangular plate.
The coarse mesh (2 x 1)is unable to represent any higher order bending or torsional
modes, so it drops these and finds higher frequency in plane extensional and
shearing modes
Mode

133.1

Mesh 1a (10 x 4)
Description

Mode

1st order Bending

120.1

348.7

1st Order Torsion

821.4

2nd Order Bending

2043

Mesh 1b (2 x 1)
Description

Mode

1st Order Bending

133.6

1st order Bending

395.7

1st Order Torsion

689.6

1st Order Torsion

624.5

2nd Order Bending

832.8

2nd Order Bending

2nd Order Torsion

1003.

2nd Order Torsion

2133.

2nd Order Torsion

2278.

3rd Order Bending

2144.

1st Order Shear

2332.

3rd Order Bending

2358.

1st Order Shear

8722.

2nd Order Shear

2358.

1st Order Shear

3705.

3rd Order Torsion

9988.

1st Order Extension

4051.

3rd Order Torsion

4344.

4th Order Bending

16667

2nd Order Shear II

4552.

4th Order Bending

4763.

1st Order Axial


Bending

20793

2nd Order Extension

5633.

1st Order Axial


Bending

10

5569.

2nd Order Axial


Bending

10

22799

2nd Order Shear III

10

6433.

4th Order Torsion

BASIC DYNAMICS

Mesh 1c (50 x 20)


Description

91

ELEMENT TYPE
Element Type
The type of element chosen is very important in dynamic
analysis, in that it can control the stiffness representation
and to a lesser extent the mass distribution of the structure.
Examples of poor choices are:
Using TET4 elements to model solid structures. If they are used
in relatively thin regions that have plate or shell the results can
be very poor. TET10 or preferably HEXA are a better choice.
If RBE2 is used instead of an RBE3 on a flexible structure such
as a satellite platform then it may over stiffen the structure and
influence the frequencies badly.

In the next workshop a classic structure is analyzed using


different elements.

BASIC DYNAMICS

92

Bracket Example
Bath Tub Fitting ( Tension to Shear Load transfer )
The geometry is found in the Femap examples folder.
Constrain as shown
Apply Material as shown
Mesh with TET10s; fine, coarse TET4s; fine , coarse

BASIC DYNAMICS

93

Bracket Example
Bath Tub Fitting ( Tension to Shear Load transfer )
Fine mesh TET10 model is assumed as baseline.
A very coarse TET10 mesh has maximum errors of 15.8%
A very coarse TET4 mesh has maximum errors of 135%
A fine mesh TET4 has maximum errors of 7.7%
TET10
fine
Mode

coarse

TET4

v coarse

fine

coarse

v coarse

691101 DOF 18945 DOF difference 2187 DOF


difference 105259 DOF difference
3291 DOF difference 1216 DOF
difference
1
793
808
1.98
918
15.80
854
7.69
1274
60.77
1867
135.53
2
873
885
1.46
1007
15.42
935
7.15
1349
54.60
1917
119.70
3
2628
2653
0.95
2805
6.73
2770
5.40
3493
32.93
4760
81.13
4
3949
4065
2.92
4503
14.03
4209
6.57
5783
46.45
8115
105.49
5
5696
5613
-1.47
5772
1.32
5884
3.30
7171
25.88
10708
87.98
6
5860
6078
3.71
6770
15.53
6277
7.12
8065
37.63
10981
87.38
7
6910
6960
0.72
7094
2.66
7063
2.22
9302
34.61
12932
87.16
8
6995
7207
3.03
7819
11.78
7324
4.71
9535
36.31
14794
111.49
9
9491
9740
2.63
10562
11.29
9925
4.58
13279
39.92
18702
97.06
10

10046

10250

2.03

11294

12.43

10679

BASIC DYNAMICS

6.30

14778

47.10

20044

99.53

94

Tuning Fork Example


We are interested in the first elastic mode of the tuning fork. The theoretical value is
440 Hz, and is known as the A above middle C musical note.
As an aside, it is interesting to note that the vertical translation of the stem
is what excites an instrument or another object that the tuning fork is placed
against.
Import the Femap geometry
Test the first natural frequency
Adjust the length of the tuning fork to get a better match

BASIC DYNAMICS

95

MASS DISTRIBUTION
Mass Distribution
A poor stiffness representation can influence a structure
badly and a poor mass representation can also have the
same effect.
The mass values may be wrong due to user error. The
values can be checked in the Nastran output file
There are two forms of mass representation in Nastran
lumped and coupled. Differences may occur in the analysis
depending on which is selected.

BASIC DYNAMICS

96

DETAIL OF JOINTS
Detail of Joints:
Is the joint flexibility correct?
For example a corner of a formed sheet structure will have an
internal radius which increases its torsional stiffness. It may be
important in this case to include the torsional stiffness via ROD
element.

If bolts are used to connect components together then the


bolt stiffness may play an important role in dynamic analysis.

CQUAD4

CROD

CQUAD4

BASIC DYNAMICS

97

DETAIL OF CONSTRAINTS
Detail of Constraints
When we idealize a structure we always make assumptions about
the connection to an adjacent structure or to ground.
Hence if a panel is surrounded on all sides by reinforcing structure, do
we represent that as fully built in, simply supported, or model an
equivalent edge stiffness using CELAS or CBUSH elements?

A particularly difficult case is where the connectivity is ill-defined,


such as the push-fit and snap connectors of a typical car dashboard
assembly.
Remember there is no such thing in nature as an infinitely stiff
connection or structure.
The mode shapes of the tower shown on the next slide are
significantly changed by the fact that the connection to ground is
not rigid. Errors will occur if it is assumed fully fixed.

BASIC DYNAMICS

98

ANALYSIS OF A TOWER WITH SOFT GROUND


CONNECTION
Objectives:
Account for the soil-base interaction using CBUSH
elements.

Soil Stiffness modeled with CBUSH


Elements

Tower Leg

RBE2

BASIC DYNAMICS

99

HAND CALCULATIONS

Hand Calculations
Manual checking of the frequencies in an analysis to make
sure answers are in the right ballpark can involve:
Using simple analogies of the structure to match standard
solutions in Roark or Blevins
Applying a 1g load in relevant directions and using the resultant
displacement at the cg. to calculate an equivalent SDOF
frequency.
Using idealization techniques to create simple FE models to
verify important modes of a complex model.

Remember the frequency is proportional to (k/m)1/2, therefore


consider whether stiffness or mass dominates errors and
that frequency can be relatively insensitive to errors in both.

BASIC DYNAMICS

100

CHECK LIST FOR NORMAL MODES PRIOR TO


DOING FURTHER ANALYSIS

RBMs - are they as expected


Is the frequency range adequate (we will discuss this more in
the section on modal effective mass)
Are the modes clearly identified
Is Mesh Density adequate
Is the Element Type appropriate
Is the Mass distribution correct
Is coupled vs. lumped mass important
Are the internal joints modeled correctly
Are the constraints modeled correctly
Do the results compare with hand calcs, previous experience or
test
BASIC DYNAMICS

101