Vibration Basics - Designing

Designing an elastic suspension.

In order to design a mounting system, it is essential to know precisely the main characteristics of the equipment to be suspended as well as the environmental conditions and excitation to which it will be subjected.

The following questionnaire sets out and identifies the parameters that should be known if possible before starting to design the mounting system.

The other paragraphs illustrate how to design a mounting system for sinusoidal and/or random excitation or shock.

Specification for anti-vibration / shock mounting system


Performance required

- Natural frequency or attenuation required
- Maximum permitted acceleration
- Maximum displacement

Physical characteristics

- Mass (kg or lb)
- Dimensions (m or in)
- Position of center of gravity (if possible provide drawing or sketch)
- Orientation of axes
- Moments of inertia about each of the three axes ( kg.m2 )
- Connections to the equipment that might change the mechanical response of the mounting system.
- Space available for the mountings.
- Maximum sway space around equipment
- Preferred mounting points.

Vibration and shock specification

Type of ambient vibration :
- Sinusoidal vibration
     Frequency (Hz)
     Amplitude (mm, g, mm/sec)
     Test duration
     Or specification to be applied
- Random vibrations
     Spectral density (g2/Hz)
     Band width
     Duration
     Or standard to be applied
- Shocks
     Type of shock
     Amplitude (g)
     Duration (ms)
     Or spectrum (attach curve)
     Or standard to be applied
- Constant acceleration
     Force g
     Or standard to be applied

Environmental specification

- Temperature (operating, non operating)
- Contaminants
- Humidity
- Salt spray
- Ozone
- Radiation
- Fire resistance
- Protection (standard to be applied)
- Fungus

5.2 - Selection criteria

Selecting an appropriate suspension for a specific vibration problem.

5.2.1 - Natural frequency required from the mounting system when subjected to a sinusoidal excitation.

Some specifications give the response required from the suspended system when subjected to sinusoidal excitation. This makes it possible to define the natural frequency of the mounting system considering the largest possible given output / input ratio. The data can be supplied in the form of displacement, acceleration or velocity.

The dynamic stiffness at the specified amplitude and frequency can be calculated from the natural frequency of the mounting system using the following equation :

Kdyn = f 2. M 4π2 where :
- M is the mass of the suspended system in kg
- f is the natural frequency of the system in Hz
- Kdyn is the dynamic stiffness of the system in N / m.

The suspension can be selected by taking into account :
- the dynamic stiffness of the mounting (equal to the dynamic stiffness of the system divided by the number of mountings)
- static load per mounting
- environmental conditions
- compatibility between the amplitude of the sinu- soidal vibration and the capability of the moun- ting.

If the input excitation is given in the form of acceleration level, the peak displacement can be determined using the following equation :

x = g/f24π2 where :
- g is the acceleration in ms-2
- f is the natural frequency of the system Hz
- x is the displacement amplitude in m (zero to peak)
1g = 9.81 ms-2
The peak displacement is half the peak to peak displacement.

5.2.2 - Performance required under random excitation.


5.2.2.1 - Theory of random excitation.


(fig. 11)

Consider the signal shown here, which would usually be the output from an accelerator with units of g. Clearly a random vibration is not easy to define, nor will it be exactly predictable or repeated if subsequent measurements are made. Statistical methods have to be used to describe the vibration. This signal has a zero mean value. However, by multiplying the signal by itself, i.e, squaring it , we will obtain a signal which will always have a positive value, and hence the mean is no longer zero. The mean value is known as the mean square acceleration and will have units of g2. The square root of this mean value is known as the root-mean-square (r.m.s) acceleration and, of course, will have units of g.

It is usually denoted of the total or overall energy in the signal (similarly the mean square level is known as the overall mean square level).

It will often be the case that a vibration environment will have one or more dominant frequencies even though the overall vibration is random. It is possible to analyse a signal so that a measure of the vibration of energy content with frequency is obtained.

The origins of this technique lie in analysis of electrical signals and the term power spectral density is used to describe the variation of power with frequency.

For vibration analysis the variation of mean square acceleration (i.e., a measure of power) with frequency is used as a standard form of presentation and hence units of g2 / Hz are applied and the terms acceleration spectral density (A.S.D) or acceleration power spectral density are used.

5.2.2.2 - Analysis of the system.

The response of equipment on anti-vibration mounts to a random vibration environment often requires complex analysis.

However, for the simple case of equipment mounted so that the response at resonance is uncoupled, i.e., translation without significant rotation , the response to a random vibration input will be vibration at almost constant frequency (i.e. the resonance frequency) with randomly varying amplitude.

For an input ASD of S (g2 / Hz) that is constant, or nearly so, at the resonance frequency fn (Hz), and if the magnification factor of mounts at resonance is Q, then the peak acceleration response with r = 3 (statistical approach 3σ)will be, approximately :

A = 3 (g)

and the peak displacement will be, approx : X = ± 9.81 (m)

These equations can be used to estimate the sway space requirement and displacement capability required of the mounts. The output amplitude at the resonance stud is X.Q.

NB : Elastomer parts typically respond sinusoidally when subjected to a random vibration input of 2 σ. It is therefore advisable to use r = 2 (2 σ) when determining resonant frequencies rather than r = 3 (3 σs) which gives peak responses.

5.2.3 - Performance required under shock.

If the fragility of the equipment under shock is the critical factor specified, the natural frequency of the system will depend on the level of protection required.

1 - Calculating the transmissibility of the shock

T = g0/g where g0 is the level of g permitted
by the equipment
g is the input level of g

T should be less than 1.

2 - Calculating the natural frequency

The calculation depends on the shape of the shock
Half sine wave
Triangular
Pulse
T approx. 4 fn. τ
T approx. 3.1 fn. τ
T approx. 3.2 fn. τ

where τ is the duration of the shock.

The natural frequency calculated in these conditions is different from that calculated for a system subjected to these sinusoidal or random excitation.

3 - Calculating the displacement associated with the shock

X = g0/fn24π2

The mounting will be chosen using the same procedure as described above : natural frequency, displacement capability of the mounting are the main consideration.

When calculating for shock, a stiffness equal to 1,4 times the static stiffness of the mountings should be used.

5.3 - Different types of mounting.

Various types of elastic mountings are shown in this catalogue :
- bonded mountings
- non-bonded mountings
- all metal mountings

It is preferable to use bonded mountings for suspending equipment in more severe environmental conditions.

They are often matched for radar, optical or gyroscope applications.

The matching conditions vary from application to another.

Biconical mounting systems are frequently used for suspending equipment as they are isometric.