P2 FA10 - Derive EOM for simple mechanical system

L San Andres (c) 2010

For the system shown in the figure, Z(t)= Zo cos( t) is a periodic displacement input (known). Perform the following tasks: a) Set X(t) as th e coordinate for the block motion; X=0 denotes the unstretched length of spring 1. Draw a free body diagram and derive the b) c) d)

+ C X + ( K + K ) X = K Z (Must demonstrate h ow you derive the equation) block equation of motion. You should get: M X 1 2 ( t) 2 (t )

For K 1=250 N/m, K 2=400 N/m, M =50 gram and C=0.75 N.s/m, find the system natural frequency (Hz) and damping ratio ( ). For Z o=15 mm, =20 Hz, determine the periodic forced response of the block X (t), i.e., calculate the amplitude of motion and the phase lag relative to Z(t). Please explain where operation is above or below or around the natural frequency; most importantly, is operation safe? BONUS: Determine the ex citation frequency range where the amplitude of motion |Xmax |=3 0 mm is not exceeded.

M := 0.05 kg N s C := 0.75 m

N K1 := 250 m N K2 := 400 m

X C M K2 Z(t)=Zo cos(t)

K1

(a.1) FBD and forces

X=0 means unstretched position of spring 1

For FBD diagram, assume a state of motion such that Z > X > 0
Dashpot force:

FD =C dX/dt M
Spring 1 force:

Spring 2 force:

FS2 =K2 (Z-X)

FS1 =K1 X

Normal from wall = weight

N=W

(1)

(a.2) derive EOM

From the FBD diagram, Newton's 2nd law states:

d2 dt

X = FS2 FS1 FD 2 N= W

(2)

d where F Damper = C x dt FS1 = K X FS2 = K2 ( Z X)

(3) is the dashpot force

wall reaction force

(4) Elastic forces from springs 1 and 2. Spring 2 drives motion of block.
where

Substitute Eqs. (3,4) into Eq. (2) to get

d M X + C X + Keq X = K2 Z ( t) 2 dt dt

d2

(5)

Z ( t) = Zo cos ( t)
and

Keq := K1 + K2

(b) Calculate natural frequency (fn) and viscous damping ratio (): Keq n := M
:= C
2 Keq M
.5

n = 114.018

sec

fn :=

n
2

fn = 18.146 Hz
(6)

.5

= 0.066 Zo := 0.015 m f := 20 Hz Z ( t) := Zo cos ( t) := f 2

(c) Calculate response for

The system response will be

Where Xop and are the amplitude and phase lag of the motion X(t)

X ( t) := Xop cos ( t + ) r := n r = 1.102

1

(7) operation just above natural frequency .5

Find frequency ratio

Using FRF formula:

Find Amplification factor

A ( r) :=

( r 2) 2 + ( 2 r) 2 1
Zo K2 Keq
2

A ( r) = 3.86

Find Amplitude at steady-state

note here that Fo=K 2 Zo

Xop := := atan

A ( r)

Find phase lag

Xop = 0.036 m
since r>1

Xop = 2.375 Zo

1 r

Let
Let

2 r

large amplitude ratio NOT very safe

= 2.548 rad
180

= 145.97 degrees

(d) Frequency range for max amplitude

from

K2 Xop := Zo A ( r) Keq = A ( r) =
2 2

Xmax := 30 mm Xmax Keq := Zo K2

2

then

( 1 r

2 2

= 3.25

+ ( 2 r)

expanding:

1 2 s + s
2
collecting like terms

) + (2 ) 2 s =1
2

where

s= r

s + 4 2 s + 1 = 0 a s + b s + c = 0 a :=
2 2

and identifying this polinomial with

solve for the two roots

b := 4 2 s1 = 0.694 s2 = 1.304

c := 1

s1 :=

.5 2 ( ) b b 4 a c

2 a

b + b 4 a c s2 := 2 a

) .5
0.5

hence, the two frequencies below and above which the amplitude ratio will be less than Xmax are:

f1 := s1 fn = 18.146 Hz

fn

f2 := s2

0.5

fn

f1 = 15.119 Hz

f2 = 20.724 Hz

For comparison purposes, lets plot the function

K2 Xop ( r_) := Zo Keq

1
2 2

( 1 r_
f2

+ ( 2 r_)

.5

f1
0.06

Xop

freq 0.04 f n
0.02 0

Xmax

10

15

20

25

30

35

40

freq

MEEN 363 - SP09 - Exam 2 - Problem 1 (Periodic Forced Response)

A machine weighing W=2200 lb is known to cause unacceptable vibrations in nearby equipment. The engineering modification consists of mounting the machine on four vibration isolators, as shown schematically in the figure. Upon installation, the static deflection of each isolator is 1 inch. Each isolator has a viscous damping coefficient equal to C=2.8 lb.s/in. There is a periodic load excitation acting on the machine, i.e. F=Fo cos(t), where F o=1000 lb and frequency f=(/2)=7 Hz, determine a) System natural frequency (fn [Hz]) and damping ratio (). [10] b) Amplitude [inch] and phase lag [degrees] of system motion Y(t) at the excitation frequency (f=7 Hz). [10] c) Sketch the response Y(t) versus time, label physical dimensions in graph. [5] d) The amplitude of motion if (for some unfortunate reason) a periodic force with a frequency equal to the system natural frequency, i.e. F=Fo sin(nt), is exerted on the machine. [5]

F(t)

EOM with Y origin from SEP:

M
Y(t) M

d Y + K Y + C Y = Fo cos ( t) dt dt2 M := 2200 lb in s g = 386.089 2 Ci := 2.8 lbf s in

d2

Isolation Mount

known parameters

(2 other on back) not shown

static deflection:

:= 1 in

Ki :=

M g
4

Ki = 550

lbf in

isolator stiffness

Each isolator supports 1/4 the machine weight. Total stiffness: Force magnitude:

K := 4 Ki Fo := 1000 lbf

C := 4 Ci rad := 7 2 s

K = 3.853 105

N m s m

frequency of forcing function: a) Natural frequency and damping ratio:

C = 1.961 103 N

K n := M

.5

:=

C
2 ( K M)
.5

rad n = 19.649 s

fn :=

n
2

fn = 3.127 Hz

static deflection:

Ys :=

Fo K

Ys = 0.455 in F ( t) := Fo cos ( t) Y ( t) = Yop cos ( t + ) Yop = Ys H ( r)

= 0.05
compare to response if M=C=0 (low frequency)

b) for harmonic excitation: The machine periodic response is

Fo YS ( t) := cos ( t) K

Find frequency ratio

r :=

n r = 2.238 Operation above natural

frequency

The amplification factor and phase angle are: 1

H ( r) :=

(1 r

2 2

) + ( 2 r)

.5

( r) := atan ( r)
180

2 r 2 1 r
degrees

since r >1

H ( r) = 0.249

( r) = 3.086 Yop := Ys H ( r)

= 176.804

Yop = 0.113 in

<< Ys

Ys = 0.455 in T :=
2

(c) graph of periodic response for two periods of motion

0.5 0.25

Y ( t) := Yop cos ( t + ( r) )

period

System response vs. time

T = 0.143 s
Note the response at t=0 shows a negative displacement. This is since the phase angle is close to -180 degrees. Broken-line shows curve of F(t)/K

Y(t) [in]

0 0.25 0.5

0.14

0.29

time (s) (d) for excitation at the natural frequency:

r := 1

= 4.544 in Yopn = 4.544 in Phase lag is -/2 rads (-90 degrees) 2 (e) Isolators are effective at frequency of operation (r>1.44), where Yop < Ys
150

Ys

H ( r) = 9.997 Yopn := Ys H ( r)

Amplitude (mm)

100

50

0.5

1.5

2.5

frequency ratio r=w/wn

MEEN 363 FALL 06 EXAM 2 PROBLEM 2

The figure shows a simple mechanical system (K-C-M) excited by a harmonic force F(t)=Fo sin(t), Fo=50 lbf. The frequency of the external force excitation ranges from 4 Hz to 20 Hz. Tests show the natural frequency is fn (10 Hz) and the viscous damping coefficient (C) equals 4 lbf.s/in. The mass of the system is M=120 lb. a) Determine the system steady-state amplitude of motion (in inch) and phase lag (in degrees) for excitation at 15 Hz. [10] b) Repeat (a) with stiffness equal to twice the original value. By how much the amplitude of motion is reduced or amplified? [6] c) Repeat (a) with a mass equal to twice the original value. By how much the amplitude of motion is reduced or amplified? [6] d) Based on the results (a-c) what system configuration will you select? Are there other options? Explain. [3]
DATA: F( t) = Fo sin( t) fn := 10 Hz sec in rad sec f := 15 Hz := f 2 Fo := 50 lbf W := M g Analyze excitation at frequency of operation between 4-20 Hz

F(t) M C K

KEY:

M := 120 lb C := 4 lbf n := fn 2

n = 62.832

First determine the ORIGINAL system stiffness (K) and damping ratio ( ): K := n M := C 2 M n = 0.102
2

K = 1.227 10

3 lbf

Let:

Korig := K M orig := M orig := for later use

in

Basic Knowledge as stated on Summary of formulas: The system periodic response is Y( t) = s H sin( t + ) where s = Fo K is the static displacement,

H is the amplification factor and the phase angle defined as a function of the frequency ratio (r= /n ): H( r , ) := 1
2 2 2 1 r + ( 2 r)

.5

( r , ) := atan

2 r 2 1 r

with

r=

(a) ORIGINAL system

r :=

f fn

r = 1.5

operation above natural frequency phase lag: ( r , )

since r>1 180 180 = 166.191 degrees

Yop_original :=

Fo K

H( r , )

amplitude:

Yop_original = 0.032 in

(b) If the stiffness is doubled:

NEW: n :=

K := 2 Korig C

K M

.5

Natural frequency increases and damping ratio decreases respect to original system rad sec = 0.072 n r = 1.061 >1

:=

2 M n

n = 88.858

r :=

Yop_b :=

Fo K

H( r , )

amplitude:

Yop_b = 0.103 in

phase lag: ( r , )

180

180 = 129.135 degrees since r ~ 1

Yop_b Yop_original

= 3.25 amplitude increases 3.25 times, because operation is too close to resonance. Furthermore, damping ratio is smaller than original.

(c) If the mass is doubled:

NEW: K n := M Fo K H( r , )
.5

K := Korig := C 2 M n

M := 2 M orig

Natural frequency decreases and damping ratio decreases respect to original system rad sec = 0.036 180 r := n r = 2.121 >1

n = 44.429

Yop_c :=

amplitude:

Yop_c = 0.012 in

phase lag: ( r , )

180 = 177.487 degrees nearly - 180 deg, since r >> 1

Yop_c Yop_original

= 0.367

Amplitude decreases to 0.37 times, because operation is well above natural frequency.

(d) Select case (c), i.e. double mass since it shows the lowest amplitude of response. The system effectively works as an isolator. Doubling the stiffness is certainly NOT a good choice since natural frequency is too close to excitation frequency.
Another option? Keep original system and DOUBLE damping? Not really, see below: orig = 0.102 f Yop := H , 2 orig Korig fn Fo Yop = 0.029 in compared to: Yop_original = 0.032 in fn = 10 Hz f = 15 Hz r= f fn = 1.5

very little reduction. Why? already operating above natural frequency. Damping does little to reduce vibration amplitude!

P2. Periodic forced response of a SDOF mechanical system. DESIGN COMPONENT

The signal lights for a rail may be modeled as a 176 lb mass mounted 3 m above the ground of an elastic post. The natural frequency of the system is measured to be 12.2 Hz. Wind buffet generates a horizontal harmonic F force at 12 Hz. The light filaments will break if their peak accelerations exceed 15g. Determine the maximum acceptable force amplitude |F| when the damping ratio =0.0 and 0.01.
Full grade requires you to explain the solution procedure with due attention to physical details
3m

The excitation force is periodic, say F(t)=Fo sin( t). then the system response will also be periodic, Y(t), with same frequency as excitation. Assuming steady state conditions:

STEADY RESPONSE of M-K-C system to PERIODIC Force with frequency

Case: periodic force of constant magnitude Define operating frequency ratio:

r=

F ( t) = Fo sin ( t) System periodic response:

where:

Y ( t) = s H ( r) sin ( t + )

(1)

s =

Fo Ke

H ( r) =

( 2 2 2 1 r ) + ( 2 r)
2

.5

tan ( ) =

2 r 1r
2

care with angle, range: 0 to -180deg From (1), the acceleration is

a ( t) = Y ( t) = A sin ( ( t + 180) A= Fo Ke
2 2 .5
or

the magnitude of acceleration is

( 1r

2 2

A=

+ ( 2 r)

Me

Fo

r
2 2

2 2 .5

(1 r

+ ( 2 r)

hence, define

Amax := 10 g

maximum allowed acceleration of filament

Me := 150 lb system mass fn := 12 HZ

natural frequency

HZ := 2

1 s

Let

f := 11.5 HZ excitation frequency due to wind buffets f close to natural frequency ro := ro = 0.958 fn

The maximum force allowed equals

2 ( 2 2 1 r ) + ( 2 r) Fmax( r , ) := Amax Me 2

.5

without any damping

Fmax ro , 0 = 133.27 lbf Note the importance of damping that leads to a

substantial increase in force allowed

with damping

:= 0.1

Fmax ro , = 340.231 lbf

For the force found the amplitude of acceleration is

Fmax ro , Fmax ro , 0

( (

) )

= 2.553

Fo := Fmax ro ,

A ( r , ) :=

Me

Fo

r
2 2

2 2 .5

(1 r

+ ( 2 r)

13 f fn Amplitude of acc/g

Amax g

1 fn = 75.398 s 1 2 =5

6.5

GRAPH NOT FOR EXAM

50

100 frequency (rad/s)

150

200

with damping without damping SInce ro~1, a simpler enginering Amax Me 2 = 300 lbf formula gives
which gives a very good estimation of the maximum wind force allowed

b) a system with damping =0.1 will produce a 255 % increase in allowable force Hence, the rail lightsystem will be more reliable, lasting longer. Fmax ro , = 2.553 Fmax ro , 0

( (

) )

c) Posts are usually hollow for the cables to be routed. These posts have layers of elastomeric material (~rubber-like) inside to increase their structural damping. Modern posts are wound up fro composites that integrate damping layers. Clearly, adding a "true" dashpot is not cost-effective

MEEN 363/617 Example Base motion - Frequeny response

The figure displays a schematic view of a read/write head in a hard disk. The support arm holding the R/W head is represented by structural stiffness Ks=105N/m, and damping coefficient Cs [Ns/m] to be determined. The R/W head mass equals 25 gram. Ka=106N/m represents the stiffness of the air film between the rotating disk and the R/W head. When the disk spins, micro-asperities in the disk induce a wavy-like motion represented as Xd=A cos(t), where A= 1 micrometer. The frequency (f=/2 ) of the excitation varies from 500 Hz to 4000 Hz, depending on the rotational speed of the disk and the radial position of the R/W head support structure. The origin of the coordinate systems, X=Xd=0, represents the static equilibrium position. a)
R/W head mechanism

Support arm

Cs
Read/write head
M

Ks

X
Hard disk

Ka

Xd =A cos(t)

Hard disk rotating drum

Draw a free body diagram, derive the equation of motion for the R/W head system, and calculate the system natural frequency (Hz) [5, 5, 5] b) State a formula for the motion X(t), i.e. amplitude and phase angle as function of frequency [5] c) When =n, find the magnitude of the damping coefficient Cs such that the amplitude of motion for the R/W hear does not exceed 2 micro-meter. [5] d) Sketch the amplitude and phase for the frequency response of the R/W head. Your graphs must show physical dimensions and explanatory sentences explaining relevant information related to the system motion [2 x 5]

DATA:

Ka 10

6 N

Ks

Ka 10

M 0.025 kg

: TBD

KEY:
(a) Draw FBD, derive EOM, find n

Assume X >Xd
FD= Cs dX/dt FS= Ks X - W

Define coordinate systems with origin at static equilibrium position. Upon assembly, the arm support spring holds the weight of the R/W head. Note that in actuality, the air film stiffness is zero when the disk does not spin. Apply Newtons law M d
2 2

X
Fa= Ka (X Xd)

dt

X = FD Fs Fa W

The Equation of motion for the system is

M Ke Ka Ks d

d X Ke X Cs X = Ka Xd = F( t) dt dt
2

where

for periodic motions

Xd = A cos t A 1 10
6

then F( t) = A Ka cos t

natural frequency

Ke n M

0.5

fn

n 2

fn 1.056 10 Hz

n 6.633 10

3 rad

(b) Establish system periodic response The system periodic response is: (from cheat sheet)
X( t) = H sin t = XA sin t where A Ka Ke XA = H( r)
7

is the "static" displacement

9.091 10

H is the amplification factor and is the phase angle defined as a function of the frequency ratio (r=/n):
H( r) = 1
2 2 2 1 r 2 r .5

tan =

2 r 1r
2

with

r=

(c) Calculate damping coefficient needed for amplitude of motion NOT to exceed 2 microns
XAmax 2 10 Define ratio
6

m XAmax 2.2 =
2

and from equation for amplification factor H: for operation at the natural frequency r=1 1 2

2 2 1 r 2 r

Hence

1 2

0.227 Cs 2 Ke M
0.5

Cs 75.378 N

s m

Value of damping coefficient

Functions for graphs

(d) FRF graphs (amplitude and phase) of R/W head versus frequency range of interest

3 10 Amplitude of response (m)

FRF
180 Phase of response (degrees)

FRF

120

1.5 10

60

1000

2000 frequency (Hz)

3000

4000

1000

2000 frequency (Hz)

3000

4000

Note that at low frequencies, amplitude equals

9.091 10

and phase lag is 0 degrees

at high frequencies, amplitude approaches null values and phase lag approaches 180 degrees at the natural frequency, amplitude of motion is NOT for exam: Relative motion is of importance, i.e. Y=X-Xd Y X A XAmax 2 10
6

YA Y FRF

180

arg Y

3 10 Amplitude of response (m)

180 Phase of response (degrees)

FRF

120

1.5 10

60

1000

2000 frequency (Hz)

3000

4000

1000

2000 frequency (Hz)

3000

4000

Y n 0 9.091 10

A 9.091 10

Phase angle cannot be positive

MEEN 363 - QUIZ 3

Names:

The rotor of an electric generator weights 750 lb and is attached to a platform weighing 7750 lb. The motor has an imbalance eccentricity (a) of 2.5 mils. The platform can be modeled as shown in the figure. The equivalent stiffness (K) of the platform is 1 million-lb/in, and the equivalent damping is C=100 lb-s/in. The operating speed of the generator is 1800 rpm. (a) determine the response of the platform (amplitude and phase) at the operating speed. (b) determine the response of the platform (amplitude and phase) if the rotor spins with a speed coinciding with the system natural frequency. (c) if the platform stiffness is increased by 25%, determine the allowable amount of imbalance (a) that will give the same amplitude of motion as determined in (a). Assume that the mass of the platform and the damping do not change appreciably by performing the stiffening.
a rotor platform k c

KEY: System excitation due to rotating imbalance Given K := 10

6 lbf

in

M rotor := 750 lb

M platform := 7750 lb sec in M = 8.5 10 lb

3

M := M rotor + M platform

C := 100 lbf

calculate the system natural frequency and damping ratio: n :=

M
rad s

.5

:=

C 2 M n

n = 213.125

= 0.011

little damping

and operating frequency ratio (r) for rotor speed: RPM := 1800 := RPM 2 rad 60 s = 188.496 rad s r := n r = 0.884
3

The system response (amplitude and phase) for imbalance excitation a := 2.5 10 are: 2 M rotor r 2 r Yop( r) := a ( r) := atan M .5 2 1r 2 2 2 1 r + ( 2 r)

in

Thus, at

r = 0.884 <1 Yop( r) = 7.894 10 Let: Yoper := Yop( r)

4

in

( r)

180

= 4.947

[degrees]

(b) If the rotor should spin with a speed coinciding with the system natural frequency, r := 1 := n RPM := 60 2 rad s RPM = 2.035 10
3

the system response is Yop( 1 ) = 0.01 in also determined from: a M rotor M 1 2 = 0.01 in Koriginal := K to maintain Yoper = 7.894 10 in
4

:= 90

degrees

Yop( 1 ) Yoper

= 13.111

(c) If the platform increases K by 25%, K := 1.25 K

calculate the NEW system natural frequency and damping ratio: n := K

.5

M
rad s

:=

C 2 M n = 9.531 10
3

n = 238.281

small change in damping ratio

and operating frequency ratio (r) for rotor speed: RPM := 1800 := RPM 2 rad 60 s = 188.496 rad s r := M rotor M n r = 0.791 r
2 2 .5

from relationship: Yop( r) := a determine the (new) allowable imbalance: a := Yoper

2 Mrotor r .5 M 2 2 2 1 r + ( 2 r)

2 1r

+ ( 2 r)

a = 5.354 10

in

i.e. ~ twice as original imbalance (eccentricity) displacement,

Example: Base motion excitation

An instrumentation sensor with mass 0.5 lb is attached to the casing of a steam turbine running at 3600 rpm. The casing motion has an amplitude of 0.005 in. Test show that the mass supported by the sensor attached has a natural frequency of 70 Hz. How much damping (lb.s/in) is needed to keep the sensor steady state amplitude below 0.015 in? Calculate the sensor stiffness (lb/in). KEY: This is a typical problem in which the source of excitation for the sensing element is base motion. The sensing element is a simple spring-mass-damper system. a) calculate frequency ratio (r) giving us information on the regime of operation (below, around or above the natural frequency) excitation frequency: RPM := 3600 := RPM natural frequency: 2 rad 60 s rad s n = 439.82 W := 0.5 lbf = 376.99 rad s rad s M := 0.5 lb

n := 70 2 n

frequency ratio:

r :=

r = 0.86 operation near to resonance:

b) The amplitude of motion (A) for the casing and desired (Yop) for the sensing element at frequency ratio (r) are: A := 0.005 in Yop Yop := 0.015 in
2 1 + ( 2 r) = A 2 2 2 1 r + ( 2 r) .5

and related by:

[1]

Let's define G as the ratio

G :=

Yop A Yop A G=3 [1]

From the expression for amplitude ratio:

2

G 1 r

+ ( 2 r)

2 = 1 + ( 2 r)

we can determine the damping ratio () from: G 1 r

2

2 2 ) 1 = ( 2 r) ( 1 G ) 2

or:

2 2 2 G 1 r 1 := 2 2 ( 2 r) 1 G

.5

Thus, at

r = 0.86

= 0.12

is the damping ratio needed to produce the desired amplitude ratio. G = 3 C := 2 W g n s m

c) The desired damping coefficient is equal to:

C = 0.14 lbf

sec in

C = 24.91 N

The stiffness of the sensor element is:

K := n M K = 4.39 10
4N

K = 250.52

lbf in

MEEN 363 - Torsional vibrations

A device to mix painting is composed of the paddles and hub connected through a stepped steel shaft to an electric motor. The mass moment of inertia (I) of the hub and blades is 2 kg.cm2, and the painting introduces a viscous damping (D) equivalent to 3 N.cm.s/rad. The stiffnesses of Stepped steel the connecting shafts equal 10 and 30 Nm/rad, respectively. shaft: lengths a) Derive the equation of motion for the paddles angular displacement (l1,l2) and (t) as a function of the motor displacement e(t)=m sin(t). diameters a) Calculate the system natural frequency, critical damping and (d1,d2 ) damping ratio b) Calculate & graph the FRF (amplitude and phase) of the paint mixer c) At what motor speed the amplitude of response is the largest? d) The engine operates at a frequency of 25 Hz with amplitude m = 22 degrees. Determine the twist angle and moment on the drive shaft. Are the calculated values reasonable (acceptable)?

ORIGIN := 1

Schematic view of paint mixer

e(t)
Hub & paddles

(t)

a) Derive equation of motion: Sum of moments

I d 2 dt2 = Torquedrive Torque drag

d Torque drag = D dt Torque drive = K e m rad

[1]

I := 2 kg cm

mass moment of inertia s rad

drag torque from viscous fluid (paint)

D := 3 N cm

drive torque from transmission shaft

where the stiffness of the stepped shaft equals (springs in series) K1 := 10 N K2 := 30 N m rad K :=

1 + 1 K 1 K2

K = 7.5 N

m rad

Thus, the equation of motion is: I d + D + K = K e = K M sin( t) dt dt d

2 2

[2]

the natural frequency and damping ratio equal:

K n := I
:= D 2 I n

.5

n = 193.649

rad s fn :=

n 2

fn = 30.82 Hz = 0.387

based on the periodic force response of a second order system, the impeller dynamic response is given by: ( t) = M H( r) sin( t ) [3]

where H and are the amplification ratio and phase lag, defined as: H( r) := 1.

2 2 1 r + ( 2 r)
2

.5

( r) :=

atan

2 r 2 1 r
180 in degrees

[4]

+ if r > 1 return

with

r=

as the frequency ratio.

Graphs of the amplitude ratio and phase angle follow:

2 Amplitude ratio H Phase lag (degrees) 0 1 2 3 frequency ratio 4

200

100

2 frequency ratio

since there is a fair amount of damping, the peak amplitude of motion does NOT happen at r=1, i.e. when the engine frequency coincides with the natural frequency. Reviewing our knowledge, recall that the maximum amplitude occurs are (See Handout USES of FRF) rpeak := 1 2 H( rpeak) = 1.4

.5

rpeak = 0.837 Hpeak := 1 2 1

( 1 2)

.5

Hpeak = 1.4 for engine operation at f := 25Hz := f 2 H( r) = 1.398 r := f fn r = 0.811

( rpeak) = 65.16 degrees M := 22 degrees notice that this operating frequency is very close to the one giving the peak motion. r = 30.753 radians (phase angle)

( r) = 61.438 degrees let: r := M H( r) r := ( r) 180 degrees

Let:

period of motion:

T :=

1 f T = 0.04 s

thus, the drive forcing function and mixer responses are: e( t) := M sin( t) ( t) := r sin t r

Graph both angular responses vs. time (2 periods of motion)

50 Angle (degrees) e( t) ( t) 0

T = 0.04 s

50

0.02

0.04 t time (s)

0.06

engine mixer the twist angle is defined as twist( t) := e( t) ( t)

50 Angle (degrees)

twist( t)

50

0.02

0.04 t time (s)

0.06

i.e. approximatelty equal to.

TWIST := 27

180

The ampltude of the drive moment transmitted through the stepped shaft is just Torque drive := K TWIST Torque drive = 3.534 N m

Example: Base motion excitation

L San Andres (2007)

The figure shows a vehicle moving with speed V along a wavy road. Prior tests show that the vehicle weighing 2000 lbf has a natural frequency of 3 Hz. Neglect the influence of the tire's bouncing mode and determine: a) the car speed (V) in mph (miles/hour) that will cause the highest amplitude of motion for the vehicle? Explain your answer b) Find the damping ratio () and damping coefficient (C lin lbf.s/in) for the car speed in (a) such that the vehicle's absolute amplitude of motion is less than 5 ft. c) Using C found in (b), calculate the vehicle steady amplitude of motion (ft) at a cruising speed V of 70 mph?
V (velocity)

Y suspension system Yb

Data:

b=2 ft

M 2000 lb fn 3 Hz n fn 2 n 18.85 rad sec

L=50 ft

KEY:

L 50 ft path wavelength b 2 ft amplitude of wave

(a) the car speed must excite the car-suspension system natural frequency. The relationship between the natural period of vibration (Tn) and the time it takes the car to travel a full wave 5280 ft length (L) is mph 3600 sec 2 1 L = Tn = = fn Vn L fn V n Vn 150 ft Vn 102.27 mph sec

(b) The car motion amplitude (Yop) as a function of the amplitude of road wave amplitude (b), frequency ratio (r) and damping ratio is (from Cheat Sheet): r= n Yop = b G ( r) cos t Yop
2 1 2 r = G ( r) = 2 b 2 2 r 1 r 2 [1] .5
desired

Yop 5 ft

and working with eqn [1]: G 1 r

2

Define G as

Yop b

G 2.5

2 2

2 r = 1 2 r

determine the damping ratio () from:

at

G 1 r

2 2

1 = 2 r 2 1 G2
.5

r 1

2 2 2 G 1 r 1 or 2 2 ( 2 r) 1 G

1 2 4 1 G

.5
[2]

0.22

is the damping ratio needed to produce the desired amplitude ratio,G 2.5

The needed damping coefficient equals: C 2 M n C 42.62 lbf (c) At a cruising speed of V 70 mph f V L f r fn V 102.67 f 2.05 Hz r 0.68 ft sec
excitation frequency

sec in

fn 3 Hz

frequency ratio

2 1 2 r Yop ( r) b 2 2 2 1 r 2 r

.5

Yop ( r) 3.42 ft (d) For completeness, graph the amplitude of response for multiple car speeds
6 amplitude of response (ft)

Vn 150
4

ft sec

50

100 150 200 car speed (ft/sec)

250

300

204 mph

EXAMPLE - EXAM 2 TYPE: Dynamic measurements were conducted on a mechanical system to determine its FRF (frequency response function). Forcing functions with multiple frequencies were exerted on the system and a digital signal analyzer (FFT) recorded the magnitude of the ACCELERATION/FORCE ([m/s2]/N) Frequency Response Function, as shown below. From the recorded data determine the system parameters, i.e. natural frequency (wn:rad/s) and damping ratio (z), and system stiffness (K:N/m), mass (M:kg), and viscous damping coefficient (C:N.s/m). Explain procedure of ANALYSIS/INTERPRETATION of test data for full credit.
1 Acceleration/Force (m/s^2/N) 1 0.9 0.8 0.7 0.6 A( x) 0.5 0.4 0.3 0.2 0.1 0 0 0 0 50 x frequency (Hz) 100 150 150

Test data showing amplitude of (acceleration/force)

Magnitude of FRF for mechanical system

Solution: Recall that for an imposed external force of periodic form: the system response Y(t) is given by: where the amplitude of motion (Yop) and phase angle (Y) are defined as:
Fo

F( t) = Fo sin( t)

[1] [2]

Y( t) = Yop sin( t + )

Yop( r) =

K 2 2 2 1 r + ( 2 r)

[3a]
.5

= atan

2 r 2 1 r

[3b] n [5] [4]

with
2

r=

from [2], we find that the acceleration is given by: aY( t) = Yop sin( t + ) = aop sin( t + 180 ) where: aop( r) = Fo 2 r M
2 2 2 1 r + ( 2 r)

[6]
.5

since:

Fo K

Fo 2 Fo 2 = r M 2 K n

thus, the magnitude of amplitude of acceleration over force amplitude follows as:
2

aop( r) Fo

2 2 1 r + ( 2 r)
2

.5 M

1 [7]

The units of this expression are 1/kg = m

s
2

For excitation at very high frequencies, r>>1.0 From the graph (test data): 1 M = 0.1 m 2 s N

1 M Thus

aop( r) Fo M := 10 kg

The system appears to have little damping, i.e. amplitude of FRF around a frequency of 50 Hz is rather large and varying rapidly over a narrow frequency range. Thus, take the natural frequency as expressed in rad/s as: fn := 50 Hz n := fn 2 n = 314.159 rad s

We can estimate the stiffness (K) from the fundamental relationship: K := n M

2

K = 9.87 10

5N

for excitation at the natural frequency (r=1), the ratio of amplitude of acceleration to force reduces to aop( 1 ) fo 2 M from the graph (test data), the ratio is approximately equal to one (1/kg). Thus. the damping ratio is determined as 1 := 1.0 2 M = 0.05 kg That is, the system has a damping ratio equal to 5%. This result could have also been easily obtained by studying the ratio of (amplitude at the natural frequency divided by the amplitude at very high frequency, i.e.) 1 1 = = 10 0.1 2 Once the damping ratio is obtained, the damping coefficient can be easily determined from the formula: C := 2 M n C = 314.159 N s m = 1

The number of calculations is minimal. One needs to interpret correctly the test data results, however.

Example: system response due to multiple frequency inputs

Consider a 2nd order system described by the following EOM
L San Andres (c) 2008

d Y + C Y + K Y = K z ( t) dt dt2

d2

where

z ( t) := a1 cos 1 t + a2 sin 2 t + a3 cos 3 t

is an external excitation displacement function Find the forced response of the system, i.e, find Y(t)

Given the system parameters

calculate natural frequency and physical damping

M := 100 kg
0.5

K := 106 n
2

N m

:= 0.10

K n := M

fn :=

fn = 15.915 Hz C = 2 103 s N m

C := 2 M n
2 .5 d := n ( 1 )

fd :=

d
2

Td :=

damped natural period

fd

Td = 0.063 s

Set frequencies and amplitudes of the excitation z(t) are: a1 := 0.05 m 1 :=

assemble:

a2 := 0.05 m 2 := n 10
9

a3 := 0.05 m 3 := 2.2 n

n
2

z ( t) := a1 cos 1 t + a2 sin 2 t + a3 cos 3 t excitation displacement

0.2

z(t) [m]

0.2

0.051

0.1

0.15

0.2

0.25
4 periods of damped natural motion

time (s)

SYSTEM RESPONSE is:

Y ( t) := ai H cos i t + i

)
2 r 2 1 r

The system frequency response function: amplitude and phase angle are H ( r) :=
1

( 1 r2) 2 + ( 2 r) 2

.5

( r) :=

atan

if r > 1 return graphs of frequency response function n

Amplitude and phase lag as a function of

Amplitude of FRF
6

r=

frequency ratio

Amplitude H(r)

5 4 3 2 1 0 0 0.5 1 1.5 2 2.5 3

Q factor
1 2

=5

Frequency ratio

Phase angle of FRF

0

Phase angle (degree)

45 90 135 180

0.5

1.5

2.5

Frequency ratio

The response of the system is given by the superposition of individual responses, i.e

Y ( t) := Y1 cos 1 t + 1 + Y2 sin 2 t + 2 + Y3 cos 3 t + 3

where for first excitation:

)
r1 = 0.5

r1 :=

1 n H r1 = 1.322

( )

1 := r1 1

( )
180

Y1 :=

K a H r1 K 1 2 n

( )

Y1 = 0.066 m

= 7.595

degrees

for second excitation:

r2 :=

H r2 = 3.821

( )

2 := r2 2

( )
180

r2 = 0.9 = 43.452
degrees

Y2 := a2 H r2
for third excitation:

( )

Y2 = 0.191 m H r3 = 0.259

r3 :=

3 n

( )

3 := r3 3
180

( )

r3 = 2.2

Y3 := a3 H r3
Assemble physical response:

( )

Y3 = 0.013 m

= 173.463 degrees

Y ( t) := Y1 cos 1 t + 1 + Y2 sin 2 t + 2 + Y3 cos 3 t + 3 Now graph the response Y(t) and the excitation z(t): excitation & response
0.4

0.2 0 0.2 0.4

Note: The response Y shows little motion at the highest excitation frequency (3). There is an obvious amplification of motion with second frequency (2 ~ n).
0.2 0.25

z( & Y(t) [m]

0.051

0.1

0.15

time (s)

To understand better, let's plot the actual FRF:

Z Y

Td = 0.063 s

Amplitude of response
0.3 0.25

a1 = 0.05 m r1 = 0.5 r2 = 0.9 r3 = 2.2 Y3 = 0.013 m

Note how response amplitude for largest frequency is largely attenuated

Y1 = 0.066 m Y2 = 0.191 m

Amplitude [m]

0.2 0.15 0.1 0.05 0

0.5

1.5

2.5

Frequency ratio

Y 3 < a3

all r1 r2 r3 Phase angle of response

0

while amplitudes for first two frequencies are amplified, in particular for 2 which is close to the natural frequency

Y2 a2 Y1

= 3.821

Phase angle (degree)

45

a1

= 1.322

90

135

180

0.5

1.5

2.5

Frequency ratio

all r1 r2 r3

Example: system response due to periodic function

Consider a 2nd order system described by the following EOM L San Andres (c) 2008

d M Y + C Y + K Y = K z ( t) dt dt2 Given the system parameters

calculate natural frequency and physical damping

d2

ORIGIN := 1
where z (t) is an external periodic excitation function

M := 100 kg
0.5

K := 106 n
2

N m

:= 0.10

K n := M

fn :=

fn = 15.915 Hz C = 2 103 s N m

C := 2 M n
2 .5 d := n ( 1 )

fd :=

d
2

Td :=

damped natural period

fd

Td = 0.063 s

Define periodic excitation function: z ( t) := amp zo if t < T

2

Example - square wave T := Td

.33333

zo := 0.1 m

amp zo if t > amp :=

2

T
2
0.2

T d

fundamental frequency

amplitude (m)

z ( t)

= 0.333

0.2

0.063

0.13

0.19

NF := 7

number of Fourier coefficients

t time(s)

Find Fourier Series coefficients for excitation z(t)

mean value

a0 :=

z ( t) dt T 0

a0 = 0 m

j := 1 .. NF
coefs of cos & sin

a j :=

z ( t) cos ( j t) dt T 0
T

b j :=

z ( t) sin ( j t) dt T 0

a = ( 0 0 0 0 0 0 0)m Build z(t) as a Fourier series ZF ( t) := a0 +

b = ( 0.127 0 0.042 0 0.025 0 0.018 ) m

j=1

(ajcos( j t) + bjsin( j t))

NF = 7

NF

2 2 zaF := a j) + ( b j) ( j

.5

Amplitude

0.2

amplitude (m)

z ( t) ZF ( t)
0 0.2

Fourier coefficients
amplitude

0.2

zaF j

0.063

0.13

0.19

0.1

t time(s)

actual Fourier Series

0 1 2 3 4 5 6 7 8 9 10

Find the forced response of the system, i.e, find Y(t) ( Ycmcos ( m t ) + Ysmsin ( m t)
NF

SYSTEM RESPONSE is:

Y ( t) := Yo + K Yo := a0 K

m=1

m := 1 .. NF
(a) set frequency ratio

f m :=

m n

(b) build denominator

2 denm := 1 ( f m) + 2 f m 2

)2

(c)

build coefficient of cos()

2 am 1 ( f m) K 2 f mbm Ycm := K denm

(d) build coefficient of sin()

2 bm 1 ( f m) + 2 f m am K Ysm := K denm
(e) for graph of components

2 2 YF := Ycm) + ( Ysm) ( m

.5

Y ( t) := Yo +

m=1

(Ycmcos(m t) + Ysmsin(m t))

NF

Now graph the response Y(t) and the excitation (Fourier) z(t): excitation & response
0.4

T =3 Td
0.2

= 0.332

0.2

Fourier coefficients

z( & Y(t) [m]

0.2

amplitude
0 0.11 0.23 0.34 0.45 0.57

0.13

0.067

0.4

time (s)

0 1 2 3 4 5 6 7 8 9 10

Z Y
3 periods of fundamental excitation motion

input Z output Y

Note: obtain response for inputs with increasing frequencies (periods decrease)
=================================================

excitation & response

0.4

T = 5 Td
0.2

0.2

= 0.199 n Fourier coefficients

z( & Y(t) [m]

0.2

amplitude
0 0.24 0.47 0.71 0.95

0.13

0.067

0.4

time (s)

0 1 2 3 4 5 6 7 8 9 10

Z Y

input Z output Y

excitation & response

0.4

T = 2 Td
0.2

0.2

n Fourier coefficients

= 0.497

z( & Y(t) [m]

0.2

amplitude
0 0.13 0.25 0.38

0.13

0.067

0.4

time (s)

0 1 2 3 4 5 6 7 8 9 10

Z Y

input Z output Y

excitation & response

0.4

T T d
0.2

= 1

0.2

n Fourier coefficients

= 0.995

z( & Y(t) [m]

0.2

amplitude
0 0.063 0.13 0.19

0.13

0.067

0.4

time (s)

9 10

Z Y

input Z output Y

excitation & response

0.4

T = 0.5 Td
0.2

= 1.99

0.2

Fourier coefficients

z( & Y(t) [m]

0.2

amplitude
0 0.024 0.047 0.071 0.095

0.13

0.067

0.4

time (s)

0 1 2 3 4 5 6 7 8 9 10

Z Y

input Z output Y

================================================= fastest Z (smallest period)

excitation & response

0.4

T = 0.2 Td
0.2

= 4.975

0.2

Fourier coefficients

z( & Y(t) [m]

0.2

amplitude
0 0.013 0.025 0.038

0.13

0.067

0.4

time (s)

0 1 2 3 4 5 6 7 8 9 10

Z Y

input Z output Y

EXAMPLE TORSIONAL VIBRATIONS:

2

Z d m e X
0

A cantilevered steel pole supports a small wind turbine. The pole torsional stiffness is K (N.m/rad) with a rotational damping coefficient C (N.m.s/rad).
Y d
k c

Me
0

c os

m Z X

The four-blade turbine rotating assembly has mass mo, and its center of gravity is displaced distance e [m] from the axis of rotation of the assembly. Iz (kg.m2) is the mass moment of inertia about the z axis of the complete turbine, including rotor assembly, housing pod, and contents.

k= T orsional T iffness e= T orional Damping c oefficient

The total mass of the system is m (kg). The plane in which the blades rotate is located a distance d (m) from the z axis as shown. For a complete analysis of the vibration characteristics of the turbine system, determine: a) Equation of motion of the torsional vibration system about the z axis. b) The steady-state torsional response (t) (after all transients die out). c) For system parameter values of k=98,670 N.m/rad, Iz=25 kg.m2, C= 157 N.m.s/rad, and mo=8 kg, e=1cm, d=30 cm, present graphs showing the response amplitude (rad) and phase angle as the turbine speed (due to wind power variations) changes from 100 rpm to 1,200 rpm. d) From (c), at what turbine speed should the largest vibration occur and what is its magnitude? e) Provide a design recommendation or change so as to reduce this maximum vibration amplitude value to half the original value. Neglect any effect of the mass and bending of the pole on the torsional response, as well as any gyroscopic effects.

MEEN 363 Example: Torsional vibrations Wind Turbine

40

Note: the torque or moment induced by the mass imbalance is T( ) = d Fu = d mo e 2 cos(t ) , i.e., a function of frequency (the rotational speed of the

turbine)

The equation describing torsional motions of the turbine-pole system is:

+ C + K = m e d 2 cos(t ) = T Iz o ( ) cos( t )
Note that all terms in the EOM represent moments or torques.

(1)

(b) After all transients die out, the periodic forced response of the system is

(t ) =

ss

(1 f

2 2

) + (2 f )

1/2

cos (t )

(2)

where

mo e d 2 mo e d 2 = = f ss = K K Iz T( )

(3)

1 2 f K C tan = 1 f 2 ; = ; = , and with f = n n Iz 2 K Iz

(4)

Eq. (3) in (2) leads to

(t )

2 m ed f = o cos(t - ) 1/2 2 2 Iz 2 (1 f ) + ( 2 f )
(6) &

(5)

m ed Let = o Iz

B=

f2 (1 f
2 2

) + ( 2 f )

1/ 2

(7)

and rewrite Eq. (5) as:

(t ) = ( B ) cos (t )

(8)
41

MEEN 363 Example: Torsional vibrations Wind Turbine

(c) for the given physical values of the system parameters: K = 98,670 N-m/rad Iz = 25 kg m2 C = 157 N-m/(rad/s)

K rad = 62.82 sec Iz C = = 0.05 2 K Iz

n =

mo e d 8 kg 0.01m 0.3 m 0.024 = = = 96 10-5rad 2 25 kg m 25 Iz

And the turbine speed varies from 100 rpm to 1,200 rpm, i.e., 10.47 rad/s to 125.66 rad/s, i.e.

= rpm /30 =

f =

= 0.167 to 2.00, n

thus indicating the system will operate through resonance. Hence, the torsion or twist angle (system response) is

(t ) = ( 96.4 105 rad ) B cos (t )

(d) Maximum amplitude of response Since the damping ratio is small, << 1, the maximum amplitude of motion will occur when the turbine speed coincides with the natural frequency of the torsional system, i.e., at f = /n = 1, thus B 1

2 1 (t ) = cos t 2 2

; and

the magnitude is max =

= 0.964 x 10-2 rad , i.e. 10 times larger than . 2

The figures below depict the amplitude

( B ) (degrees) and phase angle

(degrees) of the pole twist as a function of the turbine rotational speed (RPM)

MEEN 363 Example: Torsional vibrations Wind Turbine

42

amplitude of torsion (rad)

0.01

0.005

200

400 600 800 rotor speed (RPM)

1000

1200

180 Phase angle (degrees)

120

60

200

400

600 800 rotor speed (RPM)

1000

1200

(e) Design change: DOUBLE DAMPING but first BALANCE ROTOR!

MEEN 363 Example: Torsional vibrations Wind Turbine

43