Figure 1: A Simcenter Testlab Neo process consists of multiple connected methods that automate a data analysis.
Processes can encompass multiple functions including:
Processes can start with raw time data and/or spectrums/orders/FRFs etc
Data can be conditioned including filters, detrending, segmenting
Processing options include spectrums, FRFs, orders, statistics.
Output options include Excel, Powerpoint, etc.
*** The processes described in the next sections are in a zip file attached to this article. ***
This article has example videos and highlights of several different processes: 1. Process Basics 2. Calculating and Comparing Orders and Overall Levels 3. Averaging Multiple Channels with Block Calculator 4. Modulation Metrics 5. Data Validation 6. Rosette and Critical Plane Analysis 7. Early Detection of Bearing Faults 8. Damage-based Time Compression 9. Anomaly Detection 10. Human Body Vibration 11. Exporting Processed Results to Excel
1. Process Basics
The Process ribbon in Simcenter Testlab Neo Process Designer contains icons for opening already created processes as well as creating new ones (Figure 2):
Figure 2: The ribbon in the Simcenter Testlab Neo Process designer allows processes to be created, opened, and saved.
The ribbon creates and opens process files with a *.process extension. These files contain all pertinent methods and connections and can be shared with other users.
In this process, raw time data with rpm data is processed into orders and overall levels. Averages are calculated over multiple runs. Key methods include Spectral Map, Orders Sections, Overall Level, and Run Average (Figure 3).
Figure 3: Simcenter Testlab Neo Process that calculates order, overall level and averages over multiple runs.
This process shows how the spectrums of multiple channels can be averaged together. Different means (selecting specific channels versus using wildcards *) for selecting the data are shown with the Block Calculator method.
In this process, the amount of modulation (periodic rising and falling in level of a signal) is evaluated using several different methods: "Envelope", "Modulation Spectrum", "Modulation Map", and "Modulation Depth". The human response to modulation is also calculated using "Fluctuation Strength" and "Roughness" methods.
In this process, an automotive braking data validation process is explained in detail. In this case, specific data is extracted by creating an "Index channel" (has a value of 0 or 1 depending on criteria applied versus time) using the "Calculate" method. The "Decompose" method is used to reassemble the extracted data for further processing. Statistics are then calculated from the extracted data.
Simcenter Testlab Neo Process Designer has multiple ways to analyze measured strain gauges. This process shows how to use the "Rosette" and "Critical Plane" methods.
To decrease the amount of time it takes to perform a durability test, it is possible to remove small amplitude fatigue cycles (that do not create large amounts of damage) from a time history. A new condensed time history that has close to the original amount of damage potential can then be used to perform the durability sign-off test in a condensed amount of time (Figure 5).
Figure 5: The original time history (top, red) is condensed (top, green) but still preserves 80% of the damage potential (green lower right) versus the original damage (red lower right).
This process also has settings that help preserve the frequency content as well.
During data collection many potential "anomalies" can occur. Sensors can come loose and fall off of their intended measurement locations. Electro-magnetic interference can induce electrical noise in the measured data.
Spikes and drift can also occur in data (Figure 6).
Figure 6: Top: Spikes/dropouts are one type of anomaly that can occur. Bottom: Offsets can occur in sensor reading over time creating drift.
It can be very tedious and time consuming to inspect hours of recorded data over hundreds of channels. The anomaly detection process helps flag potential measurement issues in an automated way.
Simcenter Testlab Neo Process Designer includes processes for:
ISO 2631: Whole Body Vibration
ISO 5349: Hand Arm Vibration
These processes are used to take vibration measured at human interface locations (seats, steering wheels, drill handles, etc) and predict either human comfort or exposure limits.
This process highlights how both measurement functions (spectrums, etc) and statistics can be exported to Excel in Simcenter Testlab Neo Process Designer (Figure 7).
Figure 7: Simcenter Testlab Neo process with "Report to Excel" method.
The "Report to Excel" method is used in these processes.
