logo

Circular Waveform: Interactive graph to asset failure

August 5, 2021
Circular Waveform: Interactive graph to  asset failure

Waveform analysis is a fundamental technique for identifying failures in assets.

The DynaPredict Web analysis platform can now call upon a new visual alternative for this technique: the circular waveform.

The technique, already commonly used by vibration analysts, can now be applied to all temporal signals available on the Dynamox Web Platform, with interactive tools available according to the necessity of the user. 

The main objective of the circular waveform is to complement the current waveform analysis, converting the view to a polar coordinate system and making the identification of periodic characteristics that are related to the rotational velocity of the component more intuitive.

The graph is interactive with the manipulation of three main pieces of information: the rotation value registered to the spot, the initial time value from which the user wants to generate the view, and the number of axis rotations from that initial point.

The resulting graphic is an overlay of small portions of the original signal, each corresponding to the time of one rotation.

The combination of these inputs allows users to evaluate whether specific portions of the original signal have amplitudes that overlap in successive rotations, forming clear visual patterns.   

The circular waveform has direct interaction with the Cartesian waveform, by illustrating in real time the start and finish of the rotations defined by the user.

By defining the RPM value, it is also interacting directly with other resources that depend on this parameter, such as the bearing frequency markers.

Likewise, the different metrics, filter parameters and other signal transformations that are applied to the Cartesian view are immediately reflected on the circular waveform.

An example of this technique in use is the identification of failures in the outer race of bearings. In this case, the waveform is made up of a succession of equally spaced peaks on the time domain, which represent the rolling elements passing over defects formed during the bearing’s service life.

The autocorrelation technique, in particular, makes the patterns of interest even more evident during analysis when applied to the signal and viewed in the polar system. 

We will soon publish a text covering the autocorrelation technique in more detail.

However, as a spoiler and aiming to show the integration of the two techniques, the image below illustrates the autocorrelation of a gear transmission waveform signal with a cracked tooth. In this case, the peaks on the autocorrelation suggest, on the circular waveform, that the defect repeats itself once per rotation.

To facilitate the understanding of the available information, the tool has two main restrictions: prevention of viewing more than one sensor axis simultaneously, and allowing a visualization with up to ten simultaneous revolutions.

Development

Access

The circular waveform visualization is available on the Spectral Viewer and Spectral Comparison pages, and can be activated by clicking the Toggle button in the top-left corner of the Cartesian waveform.

On the Spectral Viewer page, when the circular waveform is activated, it is displayed alongside the Cartesian view. 

On the Spectral Comparison page, the circular waveform is positioned above the Cartesian waveform.

Parameterization

To the side of the circular waveform, users can find the parameters that define the start and the number of revolutions to be plotted on the graph.

The start value represents the point from which successive revolutions will be considered.

The revolutions field defines the quantity of axis revolutions that will be illustrated.

The RPM value registered to the spot can also be modified and the alterations will immediately be reflected on the two graphs.

This way of changing the RPM directly in the information bar was modified to simplify this interaction, since variations in the components actual rotation compared to its registered rotation can sometimes be expected.

This function can be useful for making adjustments to the registered value during waveform analysis, with immediate visual feedback of the calculated values.

Axis Selection

When a circular waveform is activated, the data that is viewed corresponds to the axis indicated on the legend below the Cartesian waveform graph. The axes can be changed by clicking on the legend, with the active axis being displayed in an expanded form.

Statistics

The signal statistics, usually available in the top-left corner of the Cartesian waveform, are also available in the top-left corner of the circular waveform graph when it is activated.

The data from the left column refers to the statistics of the active axis, for the complete original signal, while the data from the right column and the values below the polar graphic refer to information relevant to the circular waveform.

These statistics include: the number of rotations per second, the duration of a revolution, the number of revolutions and the start and end time interval of the selected revolutions. 

Incomplete Revolutions

When the combination of values selected by the user results in a time interval greater than the original signal duration, users will be alerted and some fields will be blocked to avoid incorrect analyses. 

In this article, we'll show you the Circular Waveform, a new fault analysis tool from the DynaPredict Web Platform. Check it out!

Metrics, filters and manipulating the signal

The other features of signal processing and alteration of metrics are immediately reflected on the circular waveform, when changed on the Cartesian waveform.

Also learn about the waterfall spectrum, for failure evolution analysis of components monitored by the DynaPredict Solution.


We excel at what we do

  • isso-27001
  • gptw

Are you in doubt? We will contact you