Simcenter Testing Solutions Bad Test Data ? Let's Fix it!
2020-07-10T10:19:22.000-0400
Simcenter Testlab
Summary
Details
It is inevitable that bad data will cross the path of the best test practitioner in their lifetime. If you are lucky, you might be able to fix it!
What might you have encountered? Pick from the bad data and fix list below!:
1. Wrong Calibration Factor
Direct YouTube link: https://youtu.be/V886GdRxqjg
Figure 1: Acceleration vs Time data of three different accelerometers
Figure 2: Two accelerometers similar in appearance with different sensitivities
Figure 4: Waterfall plot of run-up from 1200 RPM to 4000 RPM contaminated by 60 Hertz and 180 Hertz electrical noise
Figure 5: Comparison of three different digital filters (FIR, IIR and Kalman)
Figure 6: Left Side – ‘Harmonic Tracking’ Add-in, Right Side – Time Signal Calculator function selection
Figure 7: Simcenter Testlab – Time Signal Calculator
Figure 8: Waterfall plot of run-up from 1200 RPM to 4000 RPM with electrical noise removed
For more information on Time Signal Calculator and how to avoid ground loops, see the following links:
Direct YouTube link: https://youtu.be/XhnCHGaPtN0
The occasional spike/dropout in the RPM signal in Figure 9, are often related to imperfections in the Zebra tape or perhaps dirt on contamination on the reflecting surface.
Figure 9: RPM Time History with occasional RPM spikes
Figure 10: Time Signal Calculator – ‘Time Data Editor’ Toolbar
Next, highlight the appropriate time trace in the data set area on the left of the screen. Then in the Overview display, click and drag the segment around the spike. Then press the [R] button to edit the RPM trace and the spike is replaced by a line.
For more information on Removing RPM Spikes, see the following links:
Direct YouTube link: https://youtu.be/XhnCHGaPtN0
Slight imperfections in a particular stripe of zebra tape can cause an RPM anomaly to occur at the same angular position within each rotation as shown in Figure 11 below.
Figure 11: A RPM-time history with a spike shown in every other rotation.
Figure 12: Time Signal Calculator – TACHO_MOMENTS_SPIKEREMOVAL _TO_RPM Function
After applying the RPM spike removal, results in Figure 13 show comparison the corrected RPM time history and the original RPM time history.
Figure 13 – Overlay of original RPM-Time history (red) versus the original time history with spikes removed (green).
What might you have encountered? Pick from the bad data and fix list below!:
- Wrong Calibration factor
- 60 Hertz (or 50 Hertz) Contaminated Data
- RPM Occasional Spike
- Spikey RPM
- Incorrect Point Id or Directions on Frequency Data
- Geometry and Measurement Data Identifications Not Matching
1. Wrong Calibration Factor
Direct YouTube link: https://youtu.be/V886GdRxqjg
Wrong calibration factor entered when acquiring the data? No problem!
In Figure 1, it’s clear the general trend of all three channels is similar, but one of the three channels have a very different amplitude.
Figure 1: Acceleration vs Time data of three different accelerometers
This is because the transducer sensitivity was not correct, 100 mv/g was used instead of 10 mv/g as shown in Figure 2.
Figure 2: Two accelerometers similar in appearance with different sensitivities
The Time Signal Calculator (26 tokens) can be utilized to apply a scale factor to the time history and correct the amplitude of the data.
The Time Signal Calculator formula to apply the correction is shown in Figure 3 below.
The Time Signal Calculator formula to apply the correction is shown in Figure 3 below.
Figure 3: Simcenter Testlab Time Signal Calculator with formula for scaling a channel CH3*(100/10)
The Time Signal Calculator can also be used to change the channel name, change units, resample data, or apply filters as needed.
For detailed information on how to use the Time Signal Calculator to correct data, check the links below:
In Figure 4, notice that there is high amplitude vibration at 60 Hertz and 180 Hertz. This is caused by electrical interference in the instrumentation and is not actual vibration.
For detailed information on how to use the Time Signal Calculator to correct data, check the links below:
In Figure 4, notice that there is high amplitude vibration at 60 Hertz and 180 Hertz. This is caused by electrical interference in the instrumentation and is not actual vibration.
Figure 4: Waterfall plot of run-up from 1200 RPM to 4000 RPM contaminated by 60 Hertz and 180 Hertz electrical noise
Electrical contamination noise is often caused by ground loops or electromagnetic interference when a power cable is laying too close to a transducer cable. Also, fluorescent lights could induce electrostatic interference on poorly isolated wires.
A Kalman filter can be used to eliminate the 60 Hertz and 180 Hertz electrical noise. Depending on the countries electrical power grid, the electrical contamination could be at 50 Hertz and multiples, rather than 60 Hertz.
The advantage of a Kalman filter is how narrow the filter is (see Figure 5) in comparison to conventional FIR and IIR digital filters.
A Kalman filter can be used to eliminate the 60 Hertz and 180 Hertz electrical noise. Depending on the countries electrical power grid, the electrical contamination could be at 50 Hertz and multiples, rather than 60 Hertz.
The advantage of a Kalman filter is how narrow the filter is (see Figure 5) in comparison to conventional FIR and IIR digital filters.
Figure 5: Comparison of three different digital filters (FIR, IIR and Kalman)
The narrow filter shape helps ensure only the electrical component is removed from the signal while preserving the actual signal content.
To use the Kalman filter, turn on the following add-ins in Simcenter Testlab: Time Signal Calculator (26 tokens) and Harmonic Tracking (20 tokens). This is done from the main Simcenter Testlab menu under "Tools -> Options".
In the Time Data Selection worksheet, add the function KALMAN_FREQ_CUT to the Time Signal Calculator (Figure 6) using the fx button.
To use the Kalman filter, turn on the following add-ins in Simcenter Testlab: Time Signal Calculator (26 tokens) and Harmonic Tracking (20 tokens). This is done from the main Simcenter Testlab menu under "Tools -> Options".
In the Time Data Selection worksheet, add the function KALMAN_FREQ_CUT to the Time Signal Calculator (Figure 6) using the fx button.
Figure 6: Left Side – ‘Harmonic Tracking’ Add-in, Right Side – Time Signal Calculator function selection
Each KALMAN_FREQ_CUT creates a new time trace of containing the time frequency of the electrical noise. It must be subtracted from the original time signal as shown in Figure 7.
Figure 7: Simcenter Testlab – Time Signal Calculator
Note increasing the Harmonic confidence parameter will make Kalman filter sharper or narrower. This may need to be adjusted interactively.
With the Kalman filter applied, everything looks great! (Figure 8)
With the Kalman filter applied, everything looks great! (Figure 8)
Figure 8: Waterfall plot of run-up from 1200 RPM to 4000 RPM with electrical noise removed
For more information on Time Signal Calculator and how to avoid ground loops, see the following links:
3. RPM Occasional Spike
Direct YouTube link: https://youtu.be/XhnCHGaPtN0
The occasional spike/dropout in the RPM signal in Figure 9, are often related to imperfections in the Zebra tape or perhaps dirt on contamination on the reflecting surface.
Figure 9: RPM Time History with occasional RPM spikes
To remove an occasional spike/dropout, use “Time Data Editor”. After selecting “Time Data Editor” under “Tools -> Add-ins” an editing toolbar is added to the ‘Time Data Selection’ worksheet (Figure 10).
Figure 10: Time Signal Calculator – ‘Time Data Editor’ Toolbar
Next, highlight the appropriate time trace in the data set area on the left of the screen. Then in the Overview display, click and drag the segment around the spike. Then press the [R] button to edit the RPM trace and the spike is replaced by a line.
For more information on Removing RPM Spikes, see the following links:
4. Spikey RPM
Direct YouTube link: https://youtu.be/XhnCHGaPtN0
Slight imperfections in a particular stripe of zebra tape can cause an RPM anomaly to occur at the same angular position within each rotation as shown in Figure 11 below.
Figure 11: A RPM-time history with a spike shown in every other rotation.
The "Time Signal Calculator" has a dedicated function called ‘TACHO_MOMENTS_SPIKEREMOVAL_TO_RPM’ to eliminate spikes in RPM on a per revolution basis. Add it to the time signal calculator as shown in Figure 12.
Figure 12: Time Signal Calculator – TACHO_MOMENTS_SPIKEREMOVAL _TO_RPM Function
The TACHO_MOMENTS_SPIKEREMOVAL_TO_RPM command is found
- Click on the "fx" button in Time Signal Calculator
- Highlight the "Tacho" group
- Select the TACHO_MOMENTS_SPIKEREMOVAL_TO_RPM
- Press "OK"
After applying the RPM spike removal, results in Figure 13 show comparison the corrected RPM time history and the original RPM time history.
Figure 13 – Overlay of original RPM-Time history (red) versus the original time history with spikes removed (green).
For more information on Removing RPM Spikes, see the following links:
Direct YouTube link: https://youtu.be/X6Kn2NDP4bw
Performed a modal test and entered some Point ID or directions incorrectly?
The point ID, direction or other properties of the data can be edited individually as well as on a group basis. This is done by right clicking on the data in the Simcenter Testlab Navigator worksheet and selecting “Edit Properties” as shown in Figure 14 below.
Figure 14 –Right Click on data to select “Edit Properties”
6. Geometry and Measurement Data Identifications Not Matching
Direct YouTube link: https://youtu.be/lRxHeaQzyVs
Fill in the table to match the geometry and measurement names. The geometry names are selected on the left, and the corresponding measurement names are on the right.
Hope this helps!
- Removing Spikes from RPM Signals
- Problems with Laser Tachometers
- Zebra Tape Butt Joint Correction for Torsional Vibrations
5. Incorrect Point ID or Direction on Frequency Data
Direct YouTube link: https://youtu.be/X6Kn2NDP4bw
Performed a modal test and entered some Point ID or directions incorrectly?
The point ID, direction or other properties of the data can be edited individually as well as on a group basis. This is done by right clicking on the data in the Simcenter Testlab Navigator worksheet and selecting “Edit Properties” as shown in Figure 14 below.
Figure 14 –Right Click on data to select “Edit Properties”
Note: Data must be in active project, not under “My Computer” to edit the properties. Before doing this, make sure a backup copy of the project is made. This operation will permanently alter the data headers.
Choose the property “Point Direction” and make sure it is the only property selected. Choose the correct direction from pulldown list (Figure 15).
Choose the property “Point Direction” and make sure it is the only property selected. Choose the correct direction from pulldown list (Figure 15).
Figure 15 – “Edit Properties” menu allows data properties to be edited and changed, including the Point Direction
Choose either “Apply to Current” to change one frequency function, or “Apply to All” to update multiple functions at once.
For directions on creating a geometry, see the following article:
For directions on creating a geometry, see the following article:
6. Geometry and Measurement Data Identifications Not Matching
Direct YouTube link: https://youtu.be/lRxHeaQzyVs
Problems animating your mode shape or operational deflection shape?
An error message “No matching nodes found” (Figure 16) indicates that the measurements and geometry are not named the same.
An error message “No matching nodes found” (Figure 16) indicates that the measurements and geometry are not named the same.
Figure 16: Non-animating Golf Club Mode Shape Error Message
In Figure 17, the geometry and measurement names did not match. The measurements have labels “club” while the geometry is called “golfclub”.
Figure 17 – Left side – Frequency Response Function measurements have the name "club", Right side –Geometry with nodes named "golfclub"
As of Simcenter Testlab Version 17 and higher, an alias table (similar to a lookup table) has been introduced so that mapping can be done between the geometry and the modal data. This makes it possible to animate shapes without editing individual measurement names one at a time.
In the Navigator worksheet of Simcenter Testlab, open the alias mapping table by right clicking on the project name in the project tree and selecting "Add/Edit Alias Mapping" (Figure 18 -Left).
In the Navigator worksheet of Simcenter Testlab, open the alias mapping table by right clicking on the project name in the project tree and selecting "Add/Edit Alias Mapping" (Figure 18 -Left).
Figure 18: Left Side – Add/Edit Alias Map, Right Side – Alias Mapping Table
Fill in the table to match the geometry and measurement names. The geometry names are selected on the left, and the corresponding measurement names are on the right.
Figure 19: Properly animating golf club
For more information on Removing RPM Spikes, see the following article:
Hope this helps!
Questions? Email charles.rice@siemens.com or contact Siemens Support Center.
More Simcenter Testlab Acquisition Tips:
- Index of Testing Knowledge Articles
- Simcenter Testlab Signature
- Hands-Free Acquisition (Almost)
- Capturing Transient Events in Simcenter Testlab
- Cool Channel Setup tricks for Triaxial Accelerometers
- Copy Channel Setup from One Project to Another?
- View Channel Setup from Desktop Navigator!
- Data Acquisition Based on GPS Location
- How to calculate a VECTOR SUM
- Tips and Tricks for Acquiring Torsional Orders
- CAN Bus Measurements
- No more tach! Use OBDII to get Engine and Vehicle Speed...
- Record Raw CANBUS Data and Decode Offline!
- Single Plane Balancing in Simcenter Testlab
- Measuring Strain Gauges in Simcenter Testlab
- Simcenter Testlab: Working with Charge Accelerometers
- Simcenter Testlab: Measuring Sound Intensity
- Correcting a Bad Calibration Factor
- Simcenter Testlab Calibration
- Simcenter Testlab: Using Transducer Electronic Data Sheets (TEDS)
Associated Components
Simcenter Testlab Digital Image Correlation
Testlab Environmental
Testlab Acoustics
Testlab Data Management
Testlab Desktop
Testlab Durability
Testlab General Acquisition
Testlab General Processing & Reporting
Testlab Rotating Machinery & Engine
Testlab Sound Designer
Testlab Structural Dynamics
Testlab Turbine