2 Checking operational consistency

Getting source characterization first-time right, can be a challenging task. To achieve good quality, it is essential to detect overloads, detached sensors, broken cables, and noise issues.

The operational consistency evaluates the consistency of the response data through the Virtual Point Transformation.

At low frequencies, all sensors linked to a Virtual Point should move consistently as a rigid body. When this is the case, the operational consistency is close to 100% and your data is good. If this value drops, it means that your structure has some flexibility or that the operational dataset has some problems that you can understand and fix immediately before moving on.

The operational consistency provides information at Virtual Point level. If the operational consistency of a VP is low, you need to use other means (ODS, graphs, etc.) to spot which sensor or channel is causing the issue.

It is essential that you check the operational consistency of each recorded operational condition. You must check it right after each measurement to ensure that problems (such as detached sensors or sensor overloads) do not propagate while you measure different operational conditions.

Displaying operational consistency

  1. Operational consistency is plotted in the Graphs card of the Operational Data module.
  2. The operational consistency is calculated by checking the movement of the sensors linked to each Virtual Point. For this reason, you need Virtual Points to be defined in SOURCE. Virtual Points are automatically created in SOURCE when importing a DIRAC file.
  3. Indicator channels and Virtual Point forces must be added as master channels and assigned to the corresponding Virtual Point group. If you import a DIRAC project and create master channels from it, this step is unnecessary, as SOURCE takes the VP assignment and master channel positions directly from DIRAC.
  4. Visualize the graph by right-clicking on the dataset and plotting the operational consistency.

How to: Check operational consistency

  1. Select the channels you want to visualize from the drop-down.
  2. Select the blocks you want to visualize from the drop-down.
  3. Set the frequency range of interest by typing minimum and maximum values or zooming with the mouse.
  4. Increase the minimum value of the consistency until you see white areas appearing.

The white areas correspond to channels with the lowest consistency. If the whole plot turns white at the same time while zooming in on the consistency axis, you have very similar data quality. If you’re satisfied with the minimum consistency, you can move on to the next quality checks.

General guidelines for operational consistency

Here are some general guidelines when checking operational consistency.

A good model has an overall operational consistency at low frequencies of 90% or more. However, this threshold varies according to the component, the chosen setup and the checked frequency range.

  • Symmetric VPs with similar instrumentation have a similar consistency. For example, the consistency of the sensors at the left and right wheel for symmetric excitations is similar.
  • Virtual Points placed at similar locations on a component have similar consistency. For example, the consistency of the sensors placed at the three connection points of an e-drive to the subframe is similar.
  • Operational consistency is highly dependent on the signal quality. If the signal-to-noise ratio is low, the operational consistency also drops. For example, sensors located on the active side (before bushings) have higher operational consistency than sensors on the passive side (after bushings), due to the difference in signal-to-noise values.

The more sensors are in a VP group, the lower the overall consistency will be.

How to: Spot a detached sensor

You want to make sure that all sensors stay glued in place during your measurements. Due to the vibrations, sensors can detach and fall. This has to be spotted immediately after the measurement. The operational consistency of the whole VP group will be reduced mainly in the higher frequency range as the cable acts like a low-pass filter on the sensor. In this case, you should check if all sensors in this VP group are still well attached.

You could also check the ODS, to see which sensor does not move consistently.

How to: Spot sensor overloads

When a sensor measures a voltage that is higher than its measurement range, it gets overloaded. When a sensor is overloaded, the operational consistency drops at the time block when the overload happened.

If the operational consistency drops when comparing different time blocks, there is a signal problem, e.g., an overload. Overloads are also visible in time and waterfall plots. If you have overloads, you need to repeat the operational measurement. If you keep getting overloads, you might want to use less sensitive sensors. If you replace sensors, remember to repeat artificial excitation measurements and change the sensor in DIRAC.

How to: Spot wrong channels mapping

To perform the calculations needed for a source characterization, you must correctly map the FRF and the operational channels. If something goes wrong with the mapping, the operational consistency of the sensors of that Virtual Point will decrease in the whole frequency range.

If you see low operational consistency of one Virtual Point over the entire frequency range, you might have wrongly mapped the channels. Open the channels card, check and correct the mapping. ODS can also help to spot inconsistent sensor movement.

How to: Spot a wrongly oriented sensor

If one sensor is rotated on the test setup compared to how it is oriented in the DIRAC file, the operational consistency of that Virtual Point is low.

If you see low operational consistency of one Virtual Point over the entire frequency range and at different time blocks, you might have a wrongly oriented sensor in your test setup. You can check the sensors movement by playing the ODS.

To fix it, you must open the DIRAC file, compare the design of experiment with the physical test setup, and rotate the sensor in DIRAC correctly. Then, you must correct the orientation in SOURCE. You can either manually type the new sensor orientation in the channels card or remove the master channels, import the corrected DIRAC file, create new master channels and redo the mapping.


    Contact us for more VIBES


    Contact our support team or call us on +31 85 822 50 49

    ×