ODS Analysis for Structural Dynamics

ODS analysis is about the determination and visualization of the vibration patterns of structures under operating conditions.

The vibration patterns – aka deflection shapes – are shown at different frequencies, orders, or time instances as an animated geometry representing the structural deflections, as well as listed in shape tables for the various points and directions (DOFs) on the geometry. The vibration values are either shown as acceleration, velocity, or displacement, with peak, peak-to-peak, or RMS scaling and in SI or imperial units.

Operating conditions can be defined by, for example, the rotational speed of various parts of a machine, and the load, pressure and flow the machine or structure is exposed to. Changing conditions will result in different vibration patterns.

The observed vibration signals contain a combination of the external forcing function acting on the structure, internally generated forces, and the dynamic properties of the structure defined by its modal parameters. At the resonances, the forces are significantly amplified, leading to large vibration levels. This can lead to everything from discomfort driving a car to structural damage with disastrous consequences such as aircraft and bridge collapses. Consequently, identification of the structural resonance frequencies and how the forces excite them is a very typical use scenario for ODS analysis.

Measuring the vibration and visualizing the deflection shapes as a function of time or at specific frequencies or orders gives a much better understanding of a potential problem or design consideration than just looking at measured vibration levels. And as such, it helps engineers to come up with more optimal solutions. Consequently, ODS analysis is often the first step into structural dynamics for many users.

As you are only measuring the output of the structure, ODS analysis can be performed on any structure (linear or non-linear), excited with any type of signal (for example, stationery, quasi-stationary or non-stationary), and have any type of boundary conditions (from free-free to fixed).

This makes ODS analysis very easy to use. In contrast to modal testing, no pre-test analysis is required to select the right excitation signals and boundary conditions, and during the measurement, no artificial input excitation is applied using impact hammers or modal exciters. However, as no model is created as in modal testing, no predictions of vibration responses at other conditions can be made. So, if the conditions change, you must measure again.

The use of ODS analysis is diverse. From design verification, target setting and benchmarking, to troubleshooting, quality control and machine health monitoring. Typical use scenarios include:

• Validation of FEM predictions from forced response simulations by comparing them to measured ODS data
• Comparing measured noise and vibration values to reference values
• Ensuring that structural resonances are not being excited and critical speeds do not occur
• Identifying and separating rotational and structural noise and vibration phenomena originating from rotating and fixed parts, respectively
• Relating noise radiation to structural vibration
• Reducing excessive noise and vibration
• Investigating machine wear resulting in, for example, unbalance, shaft misalignment, and rattling
• Analyzing transient phenomena such as door slams, explosions, launches and firing, crash and drop impacts
• Analyzing machinery running at constant or slightly varying speed
• Analyzing non-linear systems and performing frequency-variant analysis such as engine run-up/down
• Monitoring of civil engineering structures that are loaded by ambient forces, such as waves on offshore structures, wind load on buildings, and traffic load effects on bridges

With BK Connect, you can perform all three types of ODS analysis: Time ODS, Spectral ODS, and Non-stationary ODS.

• Time ODS Time ODS is used to investigate the vibration pattern of a structure as a function of time. Time ODS includes all frequencies in the analyzed frequency range and is very useful for showing an overall ODS at a given point in time for either stationary or non-stationary signals, such as transient signals.


• Spectral ODS Spectral ODS is used to investigate the vibration pattern of a structure for specific frequency or order components. For frequency component investigations, FFT analysis is used and the conditions must be stationary. For order component investigations, order tracking is used, and the conditions can be slightly quasi-stationary, for example, a slightly varying engine speed. The ODS of different spectral components is subsequently extracted, shown in a shape table, and animated.


• Non-stationary ODS Non-stationary ODS is used to investigate the vibration pattern of a structure for a specific frequency or order components as a function of rotational speed or time. Results are based on FFT analysis or order tracking. Non-stationary ODS is very useful for identifying which noise and vibration behavior is related to rotational and which to fixed parts of running machinery. Run-up/down ODS, where an engine is run up and down is a frequently used variant of non-stationary ODS.

As only the output of the structure/machinery is measured, the instrumentation for ODS analysis is relatively simple and inexpensive. The required instruments are a data acquisition system with hardware, measurement, and analysis software, response transducers (typically accelerometers), and potentially a tacho probe. The tacho probe is required if you perform ODS analysis on machines with slightly varying speeds, where order tracking must be performed to avoid frequency smearing. Or you might need a tacho probe to annotate run-up/down tests with RPM tags, even if order tracking is not performed.

Systems for ODS analysis range from simple 2-channel systems using roving accelerometers to systems with hundreds of accelerometers for measuring on complex structures with all DOFs measured simultaneously.

As the measured vibrations in an ODS test are often unknown, it is important to pay special attention to avoiding overloads and under-range situations. Much can be done by constantly doing trial runs and changing input attenuators accordingly. But for higher efficiency and data quality, the best solution is to use a data acquisition system with a high dynamic range matching the transducers used.

Also, when performing ODS tests on large structures such as large machinery, ships, trains, bridges, and buildings the cabling task can be immense. To reduce cabling cost, make the test setup simpler, and to eliminate the risk of errors, it can be very advantageous to use distributable data acquisition hardware that can be located close to the measurement points and connected either through simple standard LAN cables or wireless.

When performing ODS tests on large structures, the first vibration patterns of interest might be at quite low frequencies, close to 0 Hz. Consequently, both the data acquisition hardware and accelerometers must support measurements down to DC in these cases.

Regarding accelerometers, many of the requirements for other structural tests such as modal testing apply here as well. Parameters to consider range from dynamic range, frequency range, and sensitivity, to low weight to avoid mass loading and easy mounting using accessories such as clips and bases. Specific requirements such as high-temperature range, hermetic sealing, robustness, etc., might apply as well.

HBK offers complete solutions for all three types of ODS analysis – from accelerometers and tacho probes to data acquisition systems with hardware, and measurement and analysis software. Our data acquisition systems are based on our LAN-XI hardware, which can be configured as either a single front-end or as distributed systems, and on our user-centric BK Connect software.

Learn more about and watch our webinar recordings on ODS analysis and other structural dynamics applications:

• ODS analysis (applications and solutions)
• Structural Dynamics (applications and solutions)

GO BACK: THE DATA ACQUISITION AND ANALYSIS BLOG