# Simcenter Testing Solutions Shock Response Spectrum (SRS)

2019-08-29T16:35:43.000-0400
Simcenter Testlab

## Details

A Shock Response Spectrum (SRS) is a frequency based function that is used to indicate the magnitude of vibration due to a shock or transient event. A SRS can quantify transient vibration from a myriad of different events: from earthquakes to pyrotechnic events to ballistic shocks (Figure 1). Figure 1: Shock test during commissioning of naval ship.

Take the example of a battleship which experiences transient vibration from the shock of a nearby explosion. A Shock Response Spectrum can be measured and captured on key components of the battleship during the shock event.

Using the Shock Response Spectrum, laboratory tests can then be devised to recreate the transient shock vibration experienced by the components to ensure they would survive the event.

1. What is a Shock Response Spectrum?
2. How is an SRS measured?
3. Why Use an SRS?
3.1 Time Signal Uniqueness
3.2 Amplitude Differences

4. SRS Calculation parameters:
4.1 Q-factor
4.2 Points per Octave and Frequency Spacing
4.3 Instances and Amplitudes
4.4 Correction Methods
4.5 Dimensions

5. Calculating an SRS in Simcenter Testlab
6. History of the Shock Response Spectrum

1. What is a Shock Response Spectrum?

The amplitude of a Shock Response Spectrum is the peak vibration response versus frequency of a series of mass-spring-damper systems created during a shock or transient event as shown in Figure 2. Figure 2: The peak motion of the response of a series of mass-spring-damper systems to a transient event is captured and plotted in a Shock Response Spectrum.

Historically, actual mass-spring-damper systems were used in the measurements. Today, the response of the mass-spring-damper systems is calculated virtually with computer software from measured or predicted acceleration data.

The steps in a SRS calculation are as follows:

• Input: A system experiences an acceleration which is measured. The force that created the acceleration is not measured or known in the SRS calculation.
• System: A series of mass-spring-damper systems respond to the transient acceleration. The natural frequency (fn) and damping (Q-factor) of these mass-spring-damper systems are selected to cover the frequency range of interest. In the Figure 3, the number of mass-spring-damper systems are numbered 1 to i, where i is the number of systems. Figure 3: Base with a series of mass-spring-damper systems attached. Each mass-spring-damper system is tuned to a different frequency to cover the entire frequency range of interest for a shock measurement.

• Responses: The acceleration versus time response of each mass-spring-system is recorded. The peak levels of the signal are determined. There are different methods for calculating the peak level which are covered later in the article.
• Output: The peak level response of each mass-spring-damper system is plotted as a function versus the corresponding natural frequencies of the systems. This is the Shock Response Spectrum (Figure 4). Figure 4: A Shock Response Spectrum (SRS) is a plot of the maximum excursions of a series of mass-spring-damper systems.

For the Shock Response Spectrum calculation, vibration is measured only at one location. A base with mass-spring-damper systems distributed across it is only used to illustrate how the calculation works.

The mass-spring-damper transient analysis is performed on the acceleration time history output of the accelerometer using a software algorithm on a computer. The vibration response can be plotted in acceleration, velocity, or displacement.

2. How is an SRS measured?

An SRS is normally measured so that it captures the input into a component that is experiencing a transient event.

Imagine a product that consists of a base with a component mounted to it. The SRS measurement is not made directly on the component, but on the base adjacent to component attachment location as shown in Figure 5. Figure 5: An SRS function is typically measured at the base of a component (left) during a field test. This allows recreating the SRS in laboratory testing (right).

By measuring the SRS at the base, the input into the component is captured during field usage. Then, if a similar transient event needs to be recreated in a laboratory setting, it can be reproduced by placing the control accelerometer of a closed loop vibration control shaker test in a similar location on the shaker head.

If a new or different component will be mounted to the same base location, they can be tested using the same SRS profile as shown in Figure 6. Figure 6: Left – An SRS is calculated from a measurement on a base. Right – Different product designs can be tested using the same SRS profile.

Simcenter Testlab Shock Control can use a SRS function as the target for a shock test.

3. Why Use an SRS?

There are a few reasons why a Shock Response Spectrum (SRS) might be used rather than a Fourier Transform to capture transient vibration events.

3.1 Time Signal Uniqueness

For a transient vibration signal:

• The Fourier Transform of a time domain transient yields both magnitude and phase information.
• A Shock Response Spectrum is an envelope (magnitude) only, and no phase information is retained from the input signal.

The original input time signal cannot be regenerated from an SRS.

Because there is no unique relationship between an SRS and the time signal (unlike a Fourier Transform), there can be an infinite number of transient time signals that can generate the same SRS (Figure 7). Figure 7: Top Graphs – There are an infinite number of time series that can produce a given SRS. Bottom Graphs – For any Fourier Transform (and vice versa) there is only one unique corresponding time series.

This can be an advantage when recreating a shock transient event on a shaker table. For example, a time signal for a given SRS can be selected that is within the performance constraints (maximum displacement, maximum acceleration, etc) of the shaker system, so the test can be run without issues.

3.2 Amplitude Differences

Another advantage of using an SRS over the Fourier Transform is that the amplitude is not dependent on the time window (assuming the whole event is captured) or the frequency resolution used in the Fourier analysis.

Because the transient event is so short, the Fourier Transform has difficulty getting a consistent amplitude due to the tradeoff between frequency resolution and the time analysis window as shown in Figure 8. This is true whether the Fourier Transform is done with a Fast Fourier Transform or a Power Spectral Density. Figure 8: The frequency domain amplitude of both a FFT (left) and a PSD (right) of the same transient event changes depending on the frequency resolution used.

Functions like a Power Spectral Density (PSD) are better suited for quantifying random vibration.

The SRS processing bypasses these issues and provides a consistent amplitude for the transient event. The consistency of the SRS amplitude will be illustrated further in the upcoming points per octave section.

4. SRS Calculations

There are several parameters that can be set by the user during the calculation of an SRS:

4.1 SRS Calculations: Q-Factor

The Q-factor parameter sets the damping in the mass-spring-damper systems. When calculating an SRS, the Q-factor for the mass-spring-damper systems is set by the user.

The Q-factor determines with how much amplitude the mass-spring-damper system responds to a given input. By increasing the Q-factor, a mass-spring-damper system responds with higher amplitude to the same input as shown in Figure 9. Figure 9: System response increases with higher Q-factor.

The Q-factor is calculated by measuring the width of the response of the mass-spring-system as shown in Figure 10. Figure 10: The Q-factor of a mass-spring-damper system is calculated using the 3 dB down lower and upper frequency intersection values (f1 and f2) and the frequency of the peak (f0).

The higher the Q-factor, the lower the damping. The lower the damping, the sharper the peak of the system response, and the higher the amplitude of the system response.

Besides Q-factor, the damping of a mass-spring-damper system can also be expressed as percent critical (%Cr) damping as shown in Equation 1. Equation 1: Relationship between Q-Factor (Q) and Percent Critical Damping (%Cr).

The Q-factor, from a numerical point of view, is inversely related to damping. Percent critical damping is directly proportional to damping.

Depending on the software implementation, damping may be entered in either format. A Q-factor of 10, which corresponds to a critical damping of 5%, is often used in Shock Response Spectrum calculations.

Because the Q-factor affects the amplitude of the mass-spring-damper response, it is important to compare Shock Response Spectrums (SRS) functions that use the same Q-factor.

4.2 SRS Calculation: Points per Octave and Frequency Spacing

The Points per Octave parameter determines the number and natural frequencies of mass-spring-damper systems used in the calculation of the SRS. The higher the points per octave, the:

• Higher the number of mass-spring-damper systems that are used in the SRS calculation.
• Finer or narrower the frequency spacing between points in the SRS.

Together with the Q-factor, the points per octave should be selected so the response of the mass-spring-damper systems have an overlap as shown in Figure 11. Figure 11: Damping and points per octave should be selected so the responses of the mass-spring-damper systems overlap each other and cover the frequency range completely.

This overlap should ensure the entire frequency range of interest for the Shock Response Spectrum measurement (Figure 11) is covered, and no energy of the shock event is missed.

Like octave bands, fewer data points are calculated at higher frequencies. The spacing between the natural frequencies of the mass-spring-damper systems gets wider at higher frequencies.

Unlike the Q-factor, the points per octave does not affect the calculated amplitude for a given mass-spring-damper system. A comparison of 6, 12, and 24 points per octave on the same transient event are shown in Figure 12. Figure 12: Consistent amplitudes for 6, 12, and 24 points per octave SRS performed on the same transient event.

With more points, the SRS can contain more details as a function of frequency, but the amplitudes of any common natural frequencies will be the same. In Simcenter Testlab, possible values of the points per octave range from 1 to 48.

In a typical SRS calculation, the values are:

• Q-factor: A value of 10, which corresponds to 5% critical damping.
• Points per Octave: A value of 6 points per octave.

There are other considerations for how the amplitude of the shock event is calculated.

4.3 SRS Calculation: Instances and Amplitudes

When calculating an SRS, the transient response of the mass-spring-damper system is not treated as a single event. It is composed of two different instances.

The amplitude of the system response can be calculated from either the primary instance, the residual instance, or over both instances combined (Figure 13):

• The primary instance is the response while the transient excitation is applied to the system.
• The residual instance is response after the excitation is no longer being applied. Figure 13: Example of a half sine excitation input (light blue) creating a transient response consisting of primary and residual instances (both in green).

When determining the amplitude of the response from the primary instance:

• Absolute (1): The highest absolute amplitude of the primary instance of the time waveform is recorded.
• Positive (2): The highest positive amplitude value of the primary instance time waveform.
• Negative (3): The highest negative amplitude value of the primary instance time waveform.

When determining the amplitude of the response from the residual instance:

• Absolute (4): The highest absolute amplitude of the residual instance of the time waveform is recorded.
• Positive (5): The highest positive amplitude value of the residual instance time waveform.
• Negative (6): The highest negative amplitude value of the residual instance time waveform.

These values are calculated for the time history of each mass-spring-damper system as shown in Figure 14. Figure 14: Example of positive (top left), negative (top right), and absolute (bottom) shock response spectrum for the primary instance of a transient event.

In Figure 15, all six methods of calculating the amplitude of the Shock Response Spectrum from the same transient event are overlaid. In addition, the Maximax SRS is also shown. Figure 15: Overlay of the positive (red), negative (green), and absolute (blue) shock response spectrums from the primary (solid line) and residual (dashed line) of the same transient event. Also shown is the Maximax SRS (black) which is the absolute maximum of both the primary and residual instances.

The Maximax SRS is the absolute value over both the primary and residual instance. The Maximax SRS is commonly used. Some reasons why:

• When running a shock vibration control test, the exact time that the excitation is applied is known so the primary and residual instance are easily identified. For field data, the excitation is not always measured, so it is not easy to determine the primary and residual parts of the response.
• The absolute maximum value could occur either in the primary or residual response (especially possible for very low frequencies that take time to ramp up). The absolute maximum value may be of greatest interest, regardless where it occurs.

In Simcenter Testlab, the Maximax spectrum is calculated when the Instance is set to ‘Maximum’ and the Amplitude is set to ‘Absolute’ as shown in Figure 16. Figure 16: Simcenter Testlab SRS menu set to calculate Maximax spectrum.

At the end of the article, there is a more detailed explanation of how to calculate SRS functions in Simcenter Testlab.

4.4 SRS Calculation: Correction Methods

If the time signal being used in the SRS calculation is not centered about zero, but instead has offsets, the resulting amplitudes of the SRS function can be higher than expected. This can be the case in measured acceleration transient signals.

Offsets can be created in unexpected ways. For example, playing back data from a tape might introduce an offset, or an offset can be created by external signal conditioning equipment used with the accelerometers.

Even though the DC offset is at zero Hertz, the amplitude of the SRS calculation at every natural frequency can be affected. A correction is needed to compensate for this increased amplitude as shown in Figure 17. Figure 17: Top Left – Time history with no offset, Top Right – SRS of time signal with no offset does not benefit from offset correction, Bottom Left – Time history with positive DC offset, Bottom Right – SRS amplitudes are higher than expected without offset correction.

Offsets on data can be constant or vary slowly over time. Depending on the type of offset, there are different types of corrections available in the Simcenter Testlab software:

• DC Offset – Removes a constant acceleration offset from the time data.
• Velocity – Removes a linear 1st order velocity trend from the time data.
• Displacement – Removes a linear 1st order displacement trend from the time data.

For more information about recording acceleration data with offsets see the knowledge base article AC versus DC coupling.

4.5 SRS Calculation: Dimension

The SRS can be calculated with units (i.e., dimensions) of acceleration, velocity, or displacement. Sometimes there are advantages to viewing the SRS in different formats, based on the application.

Acceleration

A SRS can be shown in units of acceleration as shown in Figure 18. Figure 18: Acceleration SRS

In Simcenter Testlab, there are two ways of calculating the acceleration: absolute and equivalent static.

• Absolute acceleration is derived from the maximum response acceleration of the SDOF mass relative to the base acceleration.
• Equivalent static acceleration is calculated by multiplying the relative displacement by the natural frequency twice.

Acceleration based SRS functions are commonly used in the aerospace industry.

Velocity

SRS can be calculated in units of velocity. In Simcenter Testlab, both relative velocity and pseudo velocity can be calculated (Figure 19). Figure 19: A pseudo velocity SRS has a left asymptote that is governed by peak displacement, a middle plateau governed by peak velocity, and a right asymptote governed by peak acceleration. A relative velocity SRS does not.

A pseudo velocity shock spectrum (pvss) has some useful properties. The shock response spectrum shape will have a hill like shape for a zero mean acceleration transient event:

• Maximum Displacement: There will be a left asymptote which is governed by the maximum displacement in the transient event.
• Maximum Velocity: The plateau of the hill is the maximum velocity reached. It is proportional to the damage on the object.
• Maximum Acceleration: The right asymptote is governed by the peak acceleration in the test.

It is helpful if this is plotted on what is called four co-ordinate paper (4cp). The acceleration, velocity, and displacements can be read directly from a 4cp graph.

The hill shape of pseudo velocity SRS is only possible if the acceleration signal has a zero mean. Depending on the segment of time that is being analyzed, even from the same acceleration recording, the SRS could change due to the mean of the segment selected as shown in Figure 20. Figure 20: The hill shape of the pseudo velocity SRS is only possible when analyzing a zero mean acceleration signal.

For a zero mean acceleration signal, the velocity versus time must be zero at the beginning and end of the recorded time signal. If a segment of time is analyzed where the velocity is not zero at the beginning and end, then the classic hill shape of the pseudo velocity does not occur.

The pseudo velocity SRS is often used in Naval applications. The pseudo velocity is obtained by multiplying the relative displacement by the natural frequency.

Displacement

The relative displacement model for shock response spectrum calculation is used when internal stresses or movements in the structure are of more interest (Figure 21). Figure 21: Relative displacement SRS

Stress is proportional to strain which in turn is related to the relative deflection (displacement) of the response mass. The relative displacement spectrum is derived therefore from the maximum relative displacement of the SDOF mass relative to the base. The input to the base is the acceleration provided by the time pulse to be analyzed.

The pseudo velocity SRS is derived directly from the relative displacement SRS by multiplying each value by its corresponding natural frequency.

Looking at a relative displacement SRS can be useful to indicate if internal parts of the test component could hit or interfere with each other. By knowing the clearances between parts in the component, the chances of interference can be assessed. The damping value, or Q-factor, must be indicative of material of the object under test.

5. Calculating a SRS in Simcenter Testlab

An SRS spectrum can be calculated either interactively in the Simcenter Testlab Desktop, or can be calculated in batch mode in Simcenter Testlab Signature Throughput Processing.

Simcenter Testlab Desktop

In the Simcenter Testlab Desktop, in the Navigator worksheet, there is a SRS calculation icon as part of the conditioning toolbar (Figure 22). Figure 22: Conditioning toolbar in Simcenter Testlab Navigator has a SRS calculation.

Pressing on the SRS calculation button while a time history selected in the display yields a menu where SRS calculation parameters can be set (Figure 23). Figure 23: Simcenter Testlab settings for SRS calculation.

After selecting the desired settings and pressing OK, the SRS spectrum will be placed in folder called “Conditioning” in the current section of the open project.

Simcenter Testlab Throughput Processing

In Signature Throughput processing, an SRS can be calculated on multiple channels and multiple files at one time as shown in Figure 24. Figure 24: Shock Response Spectrum settings in Signature Throughput Processing.

For more details on processing SRS functions in Simcenter Testlab, see the forum post on processing a SRS: How to Calculate a Shock Response Spectrum with Testlab?

6. History of the Shock Response Spectrum

1933 – Maurice Anthony Biot (a graduate of the Catholic University of Leuven, Belgium) describes the Shock Response Spectrum in his Ph.D. thesis from the California Institute of Technology. This is the first published reference to the Shock Response Spectrum. It was initially developed to understand earthquake vibrations.

1962 - The first edition of MIL-STD-810 for environmental test tailoring is published by the United States Department of Defense. Dr. Irwin Vigness of the U.S. Navy Research Lab did much to establish the method as an indicator of mechanical shock severity.

Questions? Post a reply or email nicholas.divincenzo@siemens.com.