ME 244L Department of Mechanical Engineering

Dynamic Systems and Controls Lab The University of Texas at Austin

Prof. R.G. Longoria Spring 2000
Note on Accelerometers

1 Introduction
Sensors that detect acceleration are called accelerometers. These devices take advantage of
instrumented spring-mass seismic structures that can be understood using Newtons law. Ac-
clerometers detect the motion of an object or point in inertial space. Note that this is different
from a displacement-type sensor that senses the relative motion between two points. Interestingly,
a displacement-type sensing element is required inside most accelerometers to provide a measure
of displacement or force, which can then be used to infer acceleration.
The basic element of any accelerometer is a seismic mass, m, which is sprung within a
casing that is attached to a test object. The elastic structure (e.g., spring) that supports the mass
and a displacement-sensing element are used to infer the force required to give the seismic mass
the same motion as the test object.

x = relative position
m
x
b vo
k
u(t)

Figure 1: Schematic diagram of a typical seismic device.

By sensing force or displacement (and thus force through use of elastic element) the sensing
arrangement gives an indication of the acceleration of the object to which the accelerometer casing
is attached (through F = ma). Measured accelerometer signals can be integrated to provide velocity
or displacement information as well.

2 Modeling
The basic structure and operation of a translational accelerometer may be modeled as a spring-
mass-damper seismic structure as shown in Figure 2. The displacement of the seismic mass

1
ME 244L: Dynamic Systems and Controls Lab, Spring 2000 2

in response to an input case acceleration is modeled as a second-order system. If we assume

that the input motion is harmonic (e.g., a sinusoid), X W XR V LQ  W , then the model of the
accelerometer can be found either from a direct application of Newtons law or from the bond
graph shown. Note that the variable of interest here, [, is the relative displacment. The base
acceleration is the input in the equation,
[   tQ [B  Q [ X  W 
S
position, X W , is the sinusoidal base motion, t
where [ is the relativeT E  NP is the damping
coefficient, and Q N P is the natural frequency.

x I: m C: 1/k
m .
V x
k b F . 0 1 R: b
u(t) u(t)

x = relative position

Figure 2: Accelerometer model as a base-excited mass-spring-damper system.

A transfer function is found by letting V replace the derivative operators, and then,
;
8
t V u t u
Q   t VQ  
If we assume that a sensing element can ideally convert the relative motion to voltage, or
c
YR 6[ [, then the magnitude ratio between the output voltage signal and acceleration is,
9R
8
t V u t6[ V u
Q   t Q  
where 6[ is a sensitivity constant in volts/(units displacement).
The magnitude ratio can be found from the transfer function by substituting the frequency
operator M for V. From a plot of the frequency response function (shown below) it is clear that an
accelerometer accurately responds to all frequency components up to approximately    Q .
Beyond this point, there can be significant difference between the input acceleration and the
output signal. The phase also increases. Note that the sensing element has not been included in
this analysis.
ME 244L: Dynamic Systems and Controls Lab, Spring 2000 3
Bode Diagrams

From: U(1)
20

10

Phase (deg); Magnitude (dB) 10

20

30

40

50
To: Y(1)

100

150

200
1 0 1
10 10 10

Frequency (rad/sec)

Figure 3: Frequency response plots for accelerometer model.

3 Sensing Elements for Accelerometers

The elastic sensing element, which functions to transfer deflection into an electrical signal, can
be made of a variety of materials exhibiting resistive, capacitive, inductive, piezoelectric, piezore-
sistive, or even optical effects. In piezoelectric accelerometers, a crystal sensing element is used
and this material can perform a dual function, acting as a precision spring as well as a sensing
element.
Servo accelerometers use an internal, closed-loop feedback control systems that behaves like
a mechanical spring. Measurement of acceleration occurs by virtue of a signal that generates a
restoring force required to keep the seismic mass at its original position.

4 Applications
Absolute acceleration measurements play a vital role in many feedback control systems and pro-
cesses. In autonavigators (inertial guidance systems) accelerometers sense the general motion of
spacecraft, airplanes, and missiles. In active dampers, they supply stabilizing feedback signals.
In environmental and simulation tests, they help control the level and spectral content of motion.
In testing and perfecting the structural behavior and performance of products, accelerometers
measure the shock, vibratory, and general motion experienced. Accelerometers also fumish valu-
able infor- mation for monitoring the health, testing the behavior, and checking the quality of
ME 244L: Dynamic Systems and Controls Lab, Spring 2000 4

machines and structures.

5 Rotational Accelerometers
Rotary accelerometers for measuring the rotational components of motion operate on similar
principles. They employ either an instrumented seismic mass having a moment of inertia and a
rotational degree of freedom or use two opposite polarity linear accelerometers spaced apart on
a diameter to measure the tangential components of acceleration. However, these are difficult to
actually build because the tranlational components can often make the rotational components
difficult t o m e a s u r e . T h i s e f f e c t i s r e f e r r e d t o a s c r o s s - t a l k .

6 Further Reading
Figliola and Beasley, Ch. 12, p. 545 (Section 12.5).