Measurement System

Behaviour
to Various Inputs

Each measurement system responds

differently to different types of inputs.
• General Model for Measurement Systems
– First-Order (Step, Sine Input etc)
– Second-Order (Step, Sine Input etc)
– Frequency Response
Overall Aim of Chapter:

Simulating measurement system behavior through mathematical modeling. Modeling

provides answers on how a measurement system’s design affects its response to an
input signal. Is the response sufficient, does it meet accuracy requirements (time
dynamic, frequency response and steady state,?
 Objectives
• Relate generalized measurement system models to dynamic
response
• Describe and analyze models of zero-, first-, and second-order
measurement systems and predict their general behavior
• Calculate static sensitivity, magnitude ratio, and phase shift for
a range of systems and input waveforms
• State the importance of phase linearity in reducing signal
distortion
• Analyze the response of a measurement system to a complex
input waveform
• Determine the response of coupled measurement systems
 Time-Domain Analysis of Measurement Systems
Static vs Dynamic
From Control Systems: Classification of Signals !!

y (t ) = yt (t ) + yss (t )

lim y (t ) = 0
t →
t Remains after transient
has died out
Can the measurement system respond fast enough to keep up with the input
signal?

Frequency-Domain Analysis of Measurement Systems

Vibration signals vary in amplitude and time and at one or more
frequencies. What is accuracy at an applied (measurement)
frequency?
 Dynamic Measurement
The ability of any measurement system to follow dynamic signals is a
characteristic of the design of the measuring system components.

Small Mass Reacts to

Motion, Places
Piezoelectrical crystal
into compression or
tension.
k=restoring force to
move accelerometer
back to equilibrium.
c=opposes any
displacement away
from equilibrium.
Model of Measurement system:

Modeling guides us in choosing specific instruments and measuring

methods by predicting system response to signals and in determining
the type, range, and specifics of calibration
Many measurement systems can be modelled by
zero-, first-, or second-order linear, ordinary
differential equations

Zero-Order System (sensor)

aoy=F(t) (zero order DE)
y(t) = K F(t) ie Y(s)/F(s) = K
K=1/ao is static sensitivity or steady
gain of system.
K=dy/dx during calibration @ value of x

Desirable response of a sensor:

• No delays
• Infinite bandwidth
• Only changes amplitude
• No energy storing elements
FIRST ORDER Measurement systems that contain
storage elements do not respond
SYSTEMS instantaneously to changes in
input, 1st order DE.
𝒂𝟏 𝒚ሶ + 𝒂𝒐 𝒚 = 𝑭(𝒕)
𝝉𝒚ሶ + 𝒚 = 𝑲 ∗ 𝑭 𝒕 = 𝑲𝑨𝑼(𝒕)
𝝉 = 𝒕𝒊𝒎𝒆 𝒄𝒐𝒏𝒔𝒕𝒂𝒏𝒕, K=Steady State
Gain or Static Sensitivity
Usually determined from the static
calibration curve. K= dy/dx

With initial value y(0) = yo.

 𝚽 Both are
B(𝝎) Represents amplitude of 𝜷 = 𝝎 frequency
steady state response and
𝚽(𝝎) Represents the phase shift dependent.
Magnitude
ration M(𝛚)

Dynamic Error
𝜹(𝛚)=M(𝛚)-1

𝑷𝒉𝒂𝒔𝒆 𝑺𝒉𝒊𝒇𝒕,
𝚽 𝝎 = − tan−𝟏 (𝝉𝝎)
 Summary: First-Order Sensors
• First-order sensors have one element that stores energy and
one that dissipates it
• Step response:
– y(t) = Ak(1-e-t/τ)
– A is the amplitude of the step
– k (=1/a0) is the static gain, which determines the static
response
– τ (=a1/a0) is the time constant, which determines the
dynamic response
• Ramp response
y(t) = Akt - Akτu(t) + Akτe-t/τ
• Frequency response
– Better described by the amplitude and phase shift plots
Source: http://research.cs.tamu.edu/prism/lectures/iss/iss_l2.pdf
 Second-Order Systems
Systems with inertia. Accelerometers and Pressure Transducers.
 Time Response of
The prototype Second-Order system
𝟏 𝟐𝜻
d2y dy 𝒚ሷ + 𝒚ሶ + 𝒚 = 𝑲𝑭(𝒕)
a2 2 + a1 + a0 y(t) = F(t) (𝝎𝒏 ) 𝟐 (𝝎𝒏 )
dt dt
d 2 y a1 dy a0 1
+ + y(t) = F(t)
dt 2
a2 dt a2 a2
d2y dy 1
where : 2 + 2zw n + w 2n y(t) = F(t)
dt dt a2

Step Re sponse if F(s) = AU(t)........K = 1 / a2 And, ωd

e-zwnt known as
y(t) = KA - KA sin(w n 1- z 2 t + cos-1 z )
1- z 2 ringing
note : frequency
0 < z < 1 and w d = w n 1- z 2 ........Damped (ringing)Frequency
 Time Response of
The prototype Second-Order Measurement System
when Sine Wave Input is Applied

d2y dy d 2 y a1 dy a0 1
a2 2 + a1 + a0 y(t) = F(t)....... 2 + + y(t) = F(t)
dt dt dt a2 dt a2 a2
d2y dy 1 Magnitude Ratio M(𝛚)
where : 2 + 2zw n + w n y(t) = F(t)
2

dt dt a2
Re sponse if F(t) = Asin w t......K = 1 / a2
KAsin[w t + F(w )]
y(t) = yh +
{[1- (w / w n ) ] +[(2zw ) / w n ]
2 2
}
2 1/2

(2zw ) / w n 𝑩(𝝎)
note : F(w ) = tan (-
-1
) 𝑴 𝝎 =
1- (w / w n ) 𝑲𝑨
1
Bandwith : M (w ) = ...and
{[1- (w / w n ) ] +[(2zw ) / w n ]
2 2
}
2 1/2

w R = w n 1- 2z 2 ........resonance frequency M (w) > 1.0

ωd = ringing frequency also damped natural frequency

𝟏
𝝉 = 𝒕𝒊𝒎𝒆 𝒄𝒐𝒏𝒔𝒕𝒂𝒏𝒕 =
𝜻𝝎𝒏
 Next- Second order frequency response
to input F(t)= A sin(𝛚t)

Source:
http://research.cs.tamu.edu/prism/le
ctures/iss/iss_l2.pdf
 The Measurement System and Frequency Domain
Concepts
Input Output

r (t ) = R sin 0t y (t ) = Y sin (0t +  )

R( s) Transfer Function Y ( s) = M ( s) R( s)
R( j ) , R ( j ) M ( s ) Y ( j ) = M ( j ) R ( j )
M ( j )
M ( j ) , M ( j )

For Sinusoidal Steady-state Y ( j ) = M ( j ) R ( j )

analysis, replace s-jw:
Y ( j  ) = M ( j  ) + R ( j  )
Magnitude, Phase Relation
 System Characteristics
For a system of zero
damping,ζ = 0, M(ω) will
approach infinity and
Φ(ω) jumps to −π in the
vicinity of ω = ωn. This
behavioris characteristic
of system resonance

Peak resonance in
underdamped systems occurs
at the resonance frequency:
 Measurement System Behavior
Example
• Determine the frequency response of a
pressure transducer that has a damping ratio of
0.5 and a ringing frequency (found by a step
test) of 1200 Hz.
• An accelerometer is to be selected to measure
a time-dependent motion. In particular, input
signal frequencies below 100 Hz (628 rads/s)
are of prime interest. Select a set of acceptable
parameter specifications for the instrument
assuming a dynamic error of +-5%.
Magnitude Ratio, Let 𝛇=0.7

𝒉𝒆𝒏𝒄𝒆, 𝝎𝒏 >1047 rads/sec

 Multiple-Function Inputs

The principle of superposition states that a

linear combination of input signals applied to a
linear measurement system produces an output
signal that is simply the linear addition of the
separate output signals that would result if each
input term had been applied separately.

and
 Example
Predict the steady output signal from a second-
order instrument having K = 1 unit/unit, ζ = 2,
and ωn = 628 rad/s, which is used to measure the
input signal F(t) = 5 + 10 sin 25t + 20 sin 400t

y(t) = 5 + 9.9 sin(25t − 9.1∘) + 7.8 sin(400t − 77∘)

 Coupled Systems
As instruments in each stage of a measurement system are connected
(transducer, signal conditioner, output device, etc.), the output from one
stage becomes the input to the next stage to which it is connected, and
so forth. The overall measurement system will have a coupled output
response to the original input signal that is a combination of each
individual response to the input.

Overall system static sensitivity described by: K = K1K2…KH

The overall system magnitude ratio is the product: M(ω) = M1(ω)M2(ω)…MH(ω)
and the overall system phase shift is the sum: Φ(ω) = Φ1(ω) + Φ2(ω) + …+ ΦH(ω)
Question 3.4
 Question 3.4
A Measurement system is modeled as:
0.5 dy/dt + y = F(t)
With a steady state of 75 V a step input of 100 volt is
applied. dy
General Equation : a1 + a0 y = F (t )
dt
a1 dy 1
+ y = F (t ).....
a0 dt a0
Step Input Re sponse : F (t ) = AU (t )..hence
dy A 1
 + y = U (t )...... = time cons tan t...K =
dt a0 a0
A A
y (t ) = + ( yo − )e −t / ........ Amplitude = Gain = KA
a0 a0
 Solution
System Model: y(t) = y + (y(0) - y )e-t/;
y(0) = 75 units ,
y = 100 units

and  = 0.5 s .

then,
-t/0.5
y(t) = 100 + (75 - 100) e units

Or Use:

y(t) = 100 - 25 e-t/0.5

 Question 3.16
(examples on p.89 see all three options)
A measurement system has a natural frequency
of 0.5 rad/sec, a damping ration of 0.5 and a static
sensitivity of 0.5 m/V. Estimate its 90% rise time
settling time if F(t) = 2 U(t).
d2y dy d 2 y a1 dy a0 1
a2 2 + a1 + a0 y (t ) = F (t )....... 2 + + y (t ) = F (t )
dt dt dt a2 dt a2 a2
d2y dy 1
where : 2 + 2 n +  2 n y (t ) = F (t )
dt dt a2
Re sponse if F (t ) = AU (t )......K = 1 / a2

Solution for 0 =   1

− n t
  
y (t ) = KA − KAe  sin(n 1 −  t ) + cos(n 1 −  t )
2 2

 1 −  2

 Solution..see equation 3.15 p.89
−0.5(0.5)t 0.5
y(t) = 0.5(2) − 0.5(2)e [ sin0.43t + cos0.43t]
1− 0.5 2

= 1− e − t/4 [0.58sin0. 43t + cos0.43t]

where d = 0.43 rad/s. ; also known as ringing
Freq (also damped)

….from experimental inspection …

90% rise time = 2.74 s and

90% settling time = 9.5 s
 3.26
A Strain-gauge measurement system mounted
on airplane wing. The measurement system
has dynamic parameters.

Estimate dynamic error when operating at a

specific frequency (in flight) in wind gusts at
1Hz.

The characteristics is given as: Ringing

Frequency of 1200Hz, 90% rises time of
100msec, damping ratio of 0.8. Discuss??
fd = fn 1 − 2 f n = 1200 / 1 − 0.82 = 2000 Hz

The dynamic error is given by: (f ) = M(f) -1

1
M ( f = 1) = 1/ 2
  2
2

 ( n)  ( n) 
2
1 − f f + 2 f f
   
1
1/ 2 = 1.0 , so the dynamic error
 2
2 2
 1 − (1 2000 )  + ( 2(0.8)1/ 2000 )  at (1 Hz) = 0.0.
 

 
2  f f
 ( ) ( f )  − tan −1 n 
/ 2 f = 7.3 ms
1 = = =  
( )
2
 2 f  1 − f f n 
 3.32
An accelerometer to measure relative
displacement on a beam.

The measurement system has dynamic

parameters:
Damping ratio of 0.4,
Natural frequency of 18000Hz

Estimate dynamic error and phase shift when

receiving an impact frequency of 4500 Hz
 2nd order system: (f ) = M(f ) − 1 ; where
1
M( f ) = 1/ 2
  2
2

 1 − ( f f n )  + ( 2 f f n ) 
2

   
Hence; M(4500 Hz) = 1.04 > 1

So output amplitude B is 4% higher than input

amplitude A. This means the dynamic error
is (4500Hz) = + 0.04
−1 2 f f n −1 2(0.4)(0.25)
For this system, (4500) = − tan = − tan
1 − ( f fn ) 1 − ( 0.25 )
2 2

= - 0.21 rad
That is about a 12o
lag between input and output f R = 18000 1 − 2(0.4) 2
 Exam Questions
1.1 A force measurement system has a resonance
frequency of 82.5 rad/sec, a damping ratio of 0.4
and a static sensitivity (gain) of 2 V/N. If an input
signal of the form:

F (t ) = 3 + sin(8t ) + sin (165t ) Newton

is to be applied, write the expected form of the

output signal in volts. Show all calculations. (7)
For this system the natural frequency is:
n = R /(1 - 22 )1/2 = 82.5 r/s /(1 - 2(0.4)2)1/2 =100rad/s

With 1 /n = 0.08 and 2 /n = 1.65 and  = 0.4, use

equations M(ω) and Φ(ω) to calculate:

( ) / 7997
 ( ) = tan (−
−1
)
1 − ( / 7997 ) 2

1
M ( ) =

[1 − ( / 7997 ) 2 ]2 + [( ) / 7997 ]2 
1/ 2
 M(8 rad/s) = 1.004
M(165 rad/s) = 0.46
(8 rad/s) = -3.7o
(165 rad/s) = -142.5o

The transient response yh is given by equation:

y(t) = 6 + 2 sin[8t - 3.7o] + 0.92 sin[165t - 142.5o]
1.2 The moveable core of a linear variable
displacement transducer (LVDT) is connected
to a micrometer for calibration purposes. The
LVDT senses the micrometer displacement
and indicates a voltage output linear to the
input displacement.

• What is the independent variable?

• What is the controlled variable?
• What is the dependent variable?
• Suggest any extraneous variables?
(5)
Independent variables: micrometer setting.
Controlled variable : power supply input
Dependent variable: output voltage measured
Extraneous variables: operator set-up, zeroing
of system, and reading of micrometer ability to
set control variables
1.3 For the dynamic calibration of the LVDT
described in Question 1.2, a known
displacement is applied through a calibration
rig known as a shaker. The displacement is
setup to vary between 10 and 15 mm at a
frequency of 50 Hz. Express the input signal
as a one-term Fourier series. (2)

Ao=Average Value = (15+10)/2 = 12.5

C1= (15-10)/2 = 2.5
F1 = 50Hz
The displacement may be expressed
Y(t)=12.5 + 2.5 sin (100πt)……mm
1.4 Estimate any dynamic error that could result
from measuring the 50 Hz periodic waveform,
described in Question 1.3, using a first order
system having a time constant of 0.028
seconds. Comment on the resultant output
signal. (3)
By direct substitution into formula for 1st order
system Bandwidth,
M(f) = 1/[1 + (100*0.028)2]½ = 0.113
So that the dynamic error,  () = M() - 1
(50 Hz) = (100 rad/s) = - 0.887

Analysis
This result means that the output amplitude of the
50 Hz signal will be 88.7% smaller than the sensed
input amplitude. That is, the value will be
attenuated by a factor of 0.88.7
1.5 Determine the frequency response M(ω) and
Φ (ω) of a pressure transducer that has a
damping ratio of 0.5 and a resonance
frequency of 900Hz. Select four appropriate
values for ω. (6)
 R = 2 *  * 900 = 5655 rads / sec
R
n = = 7997 rads / sec
1 − 2 2

( ) / 7997
 ( ) = tan −1 ( − )
1 − ( / 7997 ) 2
1
M ( ) =
[1 − ( / 7997) 
] + [( ) / 7997 ]2
2 2 1/ 2

 ( rads / sec) M ( )  ( )
500 0.9979 − 3.14 o
2500
8000 1. 0 − 90 o
80000 0.01 − 175 o
 R = 2 *  * 900 = 5655 rads / sec
R
n = = 7997 rads / sec
1 − 2 2

( ) / 7997
 ( ) = tan ( −
−1
)
1 − ( / 7997 ) 2

1
M ( ) =
[1 − ( / 7997) ] + [( ) / 7997 ]
2 2

2 1/ 2

 ( rads / sec) M ( )  ( )
500 0.9979 − 3.14 o
2500
8000 1. 0 − 90 o
80000 0.01 − 175 o