Introduction
The following chapters show excerpts of a training session using the durability templates for solving common measurement and evaluation tasks.
Example: Trim measurement data
Introduction – background related to durability
Measurements are completely unpredictable: You never know when something happens, and you do not need it in your data. Thus, measurement data files have to be trimmed in order not to have irrelevant data in the measurement files – at least at the beginning (e.g., vehicle is not moving) and the end (e.g., vehicle was shut down and the power supply was not stable). This step is frequently referred as validation. Besides removing irrelevant data the validation consists also of removing spikes, drift, resampling the data, exporting in the target format to name only a few.
Goal of this exercise is to trim all provided datasets – with the prefix “original” – to shorten the data and to remove spikes or peaks at the end of the measurement which may be caused by unstable power supply. This will happen by hand using the jBEAM Durability tool “Visual Signal Editor”.
Step by step instruction:
Start the jBEAM template “trim_start_end.jbt”
Use the Import button to import one original data file from the example data
Use the Modify visually button to define the range for the trimming.
Pay attention to:
The region which has to be cut from the signal can be modified with the mouse:
Left click on the boundary of the region and drag it to the needed position
Several regions can be defined
First region should start at a time t <= 0 and end before relevant data is seen in force (WFT_LF_Fz) and velocity (vel)
Second region should end at a time t >= max(time of the measurement)
It does not matter if the first region starts in negative time domain
It does not matter if the second region ends where no data from the file is seen in the preview anymore
Accept & close the Visual Signal Editor with OK
Modify the channels for which you want to compare the original and the trimmed time series with the button Change viz data
Choose the vertical loads Fz for the left front wheel force transducer (WFT_LF_Fz):
Save new dataset to file with the Export button to the folder jBEAM trimmed example data
Repeat the previous steps for all original files
Further jBEAM Durability functionalities to validate the acquired data
In the following several further tools and functionalities are listed which can be usefull in the validation phase:
Offset and drift correction – a jBEAM Durability component
To account for offset and drift in jBEAM Durability several methods have been implemented as e.g., correction procedures based on fits, FFT, moving statistics in the tool offset and drift correction.
Manual value adjustment – a jBEAM Durability component
To adjust single values the manual value adjustment is the suitable tool.
Resampling - a jBEAM Durability component
After data acquisition frequently the data must be resampled. For that jBEAM Durability offers a special tool called Resampling.
Filtering - a jBEAM Durability component
Filtering is also a basic task after data acquisition is done. In jBEAM Durability there exist different tool to filter data. The most relevant are combined in the tool Universal Signal Filter.
In addition to that some users need customizable filter curves. For that need the Customizable Filter is the proper tool.
Statistics - a jBEAM Durability component
Statistical values can reveal invalid data. Thus jBEAM Durability offers several tools to calculate statistics for channels. Starting with extremal values and average values it is also possible to calculate more durability related values like the root mean square (RMS).
And others – a jBEAM Durability opportunity
There are always situations, in which the available tools are not sufficient to get the needed information out of the data. Therefore, jBEAM Durability offers tools like the advanced Formular Editor for numeric Objects, which enables the user to combine channels and define own function by very basic programming in one line.
For more advanced users with the scripting tool, nearby every problem can be solved efficiently. The language used out-of-the-box is Groovy. Groovy is a script language and its syntax can be compared with Java. In this training package there are some scripts included as examples.
Example: Visualization and comparison of data with a dashboard, export of pdf report
Introduction – background related to durability
In durability the time series are frequently visualized in load spectra to compare them to existing measurement files and to get a feeling for the intensity of the load of the acquired data. Often dashboards or standard report overviews are used for this purpose.
In this exercise we assume, that the dashboard for a durability related comparison of data is already existing. We are in a meeting with management in order to get a decision form them and we plan to use this existing dashboard.
On this dashboard you can see data visualizations which are characteristic for durability:
Time series
Level crossing load spectra (Force over number of load cycles)
Range pair over accumulated damage
Range pair over load cycles
Power Spectral Density
In the table you can see the following statistical values:
Maximum of the currently seen portion of the time series (loc. max)
Minimum of the currently seen portion of the time series (loc. min)
Average of the currently seen portion of the time series (loc. ave)
Damage (D)
Root mean square value (RMS)
Step by step instruction
Start the jBEAM template “report_validation_data.jbt” from the jBEAM templates folder.
Modify the data source to import different data from the example data or even better the trimmed data, which you created in the previous exercise.
First, the existing dummy data sources can be deleted with the button delete all. Second, load new files form the folder Proving Ground by clicking on the button Load file(s)…
Choose one dataset for each loading condition: empty, partially, fully and click the button Open.
Rearrange the loaded data according to the loading condition of the vehicle by shifting the data file for the fully loaded vehicle down and rename the import sources by order with the button rename by order:
On the first page (and only page) the data is displayed which is defined with the button Change data. Modify this data to display the wheel force transducer data in Fz for the left front and the left rear corner of the vehicle.
Describe what is seen on this page by modifying the text line through a double click on it.
Add a new page and modify the data to display the wheel force transducer data in Fz for the left front and right front corner with the button Change data.
Add a new page and modify the data to display the hub acceleration data in z-direction for the left front and left rear with the button Change data.
Adjust for both added pages the description line accordingly.
Export the report as pdf file and have open it in a pdf reader.
Example: Concatenate different measurement files with mixing towards a specific sequence
Introduction – background related to durability
In the context of durability, the acquired data from measurements on proving grounds is in general available as one file per maneuver – a specific section or ground on the proving ground, see also the figure of the proving ground in Lommel. To prepare the data files for testing of whole vehicles or components on test rigs it is needed to combine the relevant measurement files to one target test file.
The lifetime of a vehicle or a component does not only depend on the absolute amplitude and the number of load cycles, but also on the history of loads. Thus, it is important to consider a certain representative sequence, e.g., a measurement with high loads is combined with a measurement file with smaller loads first and this combination is then repeated ten times.
For this purpose, jBEAM Durability offers a special tool called Concatenate Channels with Transition.
If the goal is not to get a target measurement file but only the final damage number of a limited set of channels, this procedure can also be done based on rainflow matrices:
Calculate the rainflow matrices for all needed channels
Combine these rainflow matrices to one using the jBEAM Durability tool Rainflow Superposition
Calculate the damage of this combination.
Further the damage of a combination of measurements can also be calculated in the following way:
Calculate the damage for all needed channels
Combine the damage values according to the needed mixture.
All possibilities are part of this example.
Step by step instruction
Start the jBEAM template concatenate-measurement-files_calculate_damage.jbt from the jBEAM templates folder.
Modify the data source to import different data from the example data or even better the trimmed data, which you created in the first example by clicking on the button import data.
First, the existing dummy data can be deleted with the button delete all. Second, load new files form the folder Proving Ground by clicking on the button Load file(s)…
Choose one dataset for each loading condition: empty, partially, fully and click the button Open.
Rearrange the loaded data according to the loading condition of the vehicle by shifting the data file for the fully loaded vehicle down and rename the import sources by order with the button rename by order:
On the first page in the jBEAM Durability graphic window the data is now updated, and you can see all results directly. The results are based on the parameters defined in the template. The 3 measurement files are concatenated accordingly to the following formular:
2 * (File2 + 2*File3) + File1
That means: the measurement with partial loading is repeated two times in total, the file with the fully loaded vehicle four times and the measurement file with the empty vehicle only ones.Now modify this mixture formular by modifying the component Concatenate Channels with Transition:
