Simcenter Testlab Neo features a modal "dashboard" (Figure 1) that enables users to get high quality results more efficiently than before.
Figure 1: The Simcenter Testlab Neo Modal Analysis consists of a single dashboard with MAC (lower left), stabilization diagram (center), FRF synthesis (upper right), and mode shapes (lower right).
This single-view dashboard interface consolidates almost everything an engineer needs to validate frequency, damping values, and mode shapes.
Within the dashboard, users can simultaneously view:
A clear list of identified modes created by Artificial Intelligence (AI) assisted algorithm
Auto-MAC (Modal Assurance Criterion)
The stabilization diagram, often showing the summation of Frequency Response Functions (FRFs).
Individual FRFs and their synthesized counterparts
Animated mode shapes directly on the geometry
This article is a step-by-step guide for using the Simcenter Testlab Neo modal curvefitter.
1. Getting Started 2. FRF Validation 2.1 Heatmap FRF 2.2 Heatmap Coh 2.3 Phase DP 3. Modal Dashboard 3.1 Data Selection 3.2 Curvefit Option 3.3 Frequency Range and Stabilization 3.4 4. Validation 4.1 Synthesis 4.2 Modal Assurance Criterion (MAC) 5. Modal Ecosystem
1. Getting Started
Start Simcenter Testlab Neo and open a project that contains measured Frequency Response Functions (FRFs) with an associated geometry.
Under “File -> Add-ins” turn on the “Modal Analysis” and “Polymax” as shown in Figure 2.
Figure 2: Under “File -> Add-ins” turn on “Modal Anlaysis” (77 tokens) and “Polymax” (26 tokens).
A new tab is added to the bottom of the screen called “Modal” as shown in Figure 3.
Figure 3: The Modal tab in Simcenter Testlab Neo.
The “Modal” tab has two sub tabs called “FRF Validate” and “Modal Dashboard”.
Next the Frequency Response Function (FRF) data is selected to be analyzed. Right click on the FRF data in the project and choose “Add to Input Basket” as shown in Figure 4.
Figure 4: Right click on the FRF set and choose “Add to Input Basket”.
The FRF selection can be further refined in the “FRF Validate” tab described in the next section.
2. FRF Validation
The "FRF Validate" step is used to assess the quality of the measurement data before proceeding with modal curvefitting. It is an optional step. If the FRFs selected in the input basket need no further changes, this step can be skipped.
2.1 FRF Heatmap
Click on the “Heatmap FRF” icon in the ribbon. This view displays an FRF heat map, where columns represent references and rows represent responses as shown in Figure 5.
Figure 5: The FRF Heatmap makes it easy to view individual FRFs.
The color intensity indicates the RMS (Root Mean Square) value across the selected frequency range. Inspect individual FRFs by moving your cursor over the map.
Click on the checkboxes to identify and discard problematic references or responses by simply turning them off if they show poor data quality.
2.2 Coherence Heatmap
Click on the “Heatmap Coh” icon to switch to the coherence view to see the average coherence as a color over the frequency range of interest (Figure 6).
Figure 6: Coherence heat map makes it simple to find poorly measured FRFs.
Measurements with high coherence will be red in color. Low or poor coherence will be blue.
Low coherence indicates poor signal-to-noise ratio (accelerometer or cable may have gone bad) or non-linear behavior, suggesting that the FRF might need to be re-measured or excluded.
2.3 Driving Point Phase
This map helps verify the quality of the driving point measurements in the modal FRF data set. Click on the “Phasemap DP” icon to see the driving point measurements phase versus frequency (Figure 7):
Figure 7: To assess the phase of the driving points, click on the “Phasemap DP” icon.
For a true driving point, the imaginary part of the FRF should consistently move in one direction (always positive or always negative) away from zero.
Use this map to identify if a supposed driving point is not behaving as expected, which might indicate a measurement error (e.g., hitting a few inches away from the accelerometer instead of directly on it).
3. Modal Dashboard
Next the modes can be calculated from the FRF measurements in the “Modal Dashboard”.
3.1 Data Selection
In the "Modal Dashboard", select the FRF data source (Figure 8):
Figure 8: Choose the FRF data to use by clicking on the first icon in the “Modal” ribbon.
With the first icon in the “Modal Ribbon” select one of the following options:
"Input basket": If you want to use the FRFs directly from your input basket without prior validation.
"FRF validate": If you have used the "FRF Validate" step and wish to use the potentially refined dataset from there.
Next set the modal curvefit options and select the frequency range for the analysis.
3.2 Curvefit Options
In the “Modal Dashboard” ribbon, there are several other options that can be set as shown in Figure 9.
Figure 9: The modal ribbon has several options for the modal curvefitter.
These options include:
"Switch Table" Option (for Roving Tests): If you performed a roving hammer test (e.g., 4 stationary accelerometers and 48 impact points), ensure the "Switch table" option is turned ON. This tells Neo that the impact points are roving references and the accelerometers are stationary responses, preventing errors and correctly building the FRF matrix.
Solver Selection: Choose between "Time MDOF" or "PolyMAX" solver. The “Polymax” solver is better, but requires an additional 23 tokens.
Mode Type: Calculate "Real" or "Complex" modes. In a “Real” mode shape (also called a normal mode), all points move exactly in and out of phase. In a “Complex” mode this may not be the case,
Reciprocity: "Force the modes to be reciprocal" if your test setup allows for it.
Access "Advanced Settings" to fine-tune the analysis:
Input Data: Choose to use input data, synthesized FRFs, or system FRFs.
State Space: Include state space model FRFs if applicable.
Stable Pole Defaults: Define what constitutes a "stable" pole (e.g., for vector, frequency, damping).
User-Defined Model Orders: Optionally, specify a limited model order range (e.g., 10 to 100 with a step of 2) to potentially speed up calculations.
Residuals: Turn on lower and/or upper residuals. Residuals are computational modes used to compensate for contributions of modes outside the bandwidth
Scaling Options: Set your preferred scaling method (e.g., unity mass scaling, unity component scaling). If using unity component scaling, you'll need to select the specific Degree of Freedom (DOF) for scaling later.
3.3. Frequency Range and Stabilization
Once the options are set, define the frequency range for your modal analysis using the cursors as shown in Figure 10.
Figure 10: Select the frequency range for analysis and click the “Run” button to initiate the modal curvefit.
Hit the "Run" button from the state control in the lower left. Simcenter Testlab Neo will automatically calculate the stabilization diagram, pick the poles, show synthesized FRFs, display the Auto-MAC, and start animating the first mode (Figure 11).
Figure 12: Press the “Run” button in the modal software, and modes will automatically be selected (upper left) using the AI assisted algorithm. Additionally, the following is also calculated: an Auto MAC (lower left), FRF Synthesis (upper right), and mode shapes (lower right).
In addition to the automatically selected modes, the following is calculated:
AutoMAC (Modal Assurance Criterion): Indicates the similarity between mode shapes.
FRF Synthesis: Compares calculated FRF with originally measured FRF.
Mode Shape Display: Shows mode shapes selected in the MAC matrix.
The stabilization diagram by default shows “clusters”. “Clusters” indicate how the algorithm views and identifies modes in the FRF data. Dots of the same color indicate the likely presence of a mode.
Clustering is explained in the next section.
3.4 AI Assisted Algorithm
Simcenter Testlab Neo's automatic mode selection process, which is turned on by default and requires no additional license, operates through a multi-step algorithm which includes machine learning components (Figure 13):
Figure 13: Algorithm used to identify modes from FRF data.
Step 1: Pole Clustering (DBSCAN):
The stabilization diagram is fed into a DBSCAN clustering algorithm. DBSCAN stands for Density-Based Spatial Clustering of Applications with Noise. It is an unsupervised machine learning algorithm that groups closely packed data points into clusters based on density and identifies outliers as noise.
This algorithm groups similar poles together based on their frequency, damping, and participation.
A "cluster diagram" (plotting poles against damping ratio instead of model order) aids visualization.
From each identified cluster, a representative pole (typically the one closest to the cluster's median value) is automatically selected.
Step 2: Avoiding Fragmentation and Verification:
The system then verifies that a single physical mode hasn't been inadvertently split across multiple clusters.
If two clusters are deemed too similar based on their frequency, damping value, and mode shape, they are automatically merged. This prevents over-fragmentation and ensures each mode is represented by a single, well-defined cluster.
Step 3: Filtering Non-Physical Modes:
In the final step, each computed mode's contribution to the overall modal model is evaluated.
Modes that have negligible influence on the modal synthesis (i.e., do not significantly affect the synthesized FRFs) are identified as noise and filtered out. This ensures that only physically meaningful modes are retained in the final result.
The effectiveness of Simcenter Testlab Neo's automatic mode selection has been validated through "jury tests" involving users of varying experience levels (novice, intermediate, expert). The goal is to achieve consistent, high-quality results regardless of the user's expertise, which is a major benefit of the AI-assisted approach.
3.4 Display Options
By right clicking in the stabilization diagram and choosing “Stable Only”, the view can be switched between cluster representation and the identified poles as shown in Figure 14.
Figure 14: Right click on the stabilization diagram and choose “Stable Only” to toggle between the cluster view and underlying pole estimates.
The underlying poles have the following letters:
o – New: This solution was not present on the previous row of the stabilization diagram..
f – Frequency: If the frequency of the potential mode is stabilized with respect to the previous row, than the letter f is used.
d – Damping: Both frequency and damping values are stabilized relative to previous row.
v – Vector: Frequency and modal participation vector are both stabilized.
s – Stable: Modal solution is completely stabilized in frequency, damping, and modal participation vector.
There are other display options in the “Modal” ribbon (Figure 15).
Figure 15: Display options in the “Modal” ribbon of Simcenter Testlab Neo Modal Analysis.
The display options include:
Single versus Quad: Instead of a single view of the mode shape, it can be seen from four different views simultaneously.
Preview: Turn on "Preview" to see individual FRFs in the stabilization display.
Summation: Show the complex sum or imaginary sum FRF overlaid on the stabilization diagram.
Mode Indicator Functions (MIFs): Select how many MIFs to display based on your number of references (e.g., two MIFs for four references).
4. Validation
After the stabilization process, the selected modes can be validated using the FRF synthesis and Modal Assurance Criteria (MAC) capabilities contained in the Modal Dashboard.
4.1 Synthesis
Synthesized FRFs allow you to visually assess how well the identified modal model represents the measured data as shown in Figure 16.
Figure 16: Synthesis FRF (green) overlaid with measured FRF (red) in the FRF Synthesis display.
Synthesized FRF Display:
The dashboard displays the synthesized FRF alongside the measured FRF.
You can step through different responses and references (pulldown menus above the display area) to visually inspect the correlation.
The display also shows the correlation percentage and error for the currently selected FRF.
Ideally, the correlation percent should be as close to 100% as possible and the error percent should be 0%. This would be a perfect match between the synthesized and measured FRF.
In the “Modal” ribbon, click the "Synthesis results" button to open a comprehensive table summarizing the global synthesis quality as shown in Figure 17:
Figure 17: Click on the “Synthesis results” button in the “Modal” ribbon to get a list of correlation and error percentages for all measurement locations.
The synthesis results table shows:
Global correlation error for all measurement locations
Maximum and minimum correlation percentages
Maximum and minimum error values
Sort and filter this table to quickly identify FRFs with good or poor correlation, helping pinpoint areas where the modal model might need refinement.
4.2 Modal Assurance Criterion (MAC)
The Auto MAC table (Figure 18) provides a quantitative measure of the orthogonality and uniqueness of the identified mode shapes.
Figure 19: Auto-MAC diagram from Simcenter Testlab Neo Modal Analysis. Typically no two modes at different frequencies have the same shape so the off-diagonal terms should be close to zero.
Auto-MAC Table:
The modal dashboard automatically displays the Auto-MAC table, showing the correlation between the identified mode shapes to each other and themselves.
High diagonal values (close to 100) are typical when comparing mode shapes to themselves. Low off-diagonal values (close to 0) indicate that modes are orthogonal and not similar in shape.
Off-diagonal values that are not close to zero should be investigated. More measurements at additional locations may be necessary.
Mode Shape Animation:
Animate mode shapes directly from the MAC table by clicking on a cell, or from the list of identified modes.
This interactive visualization helps to understand the deformation pattern associated with each natural frequencies.
Simcenter Testlab Neo is not a standalone tool but an integral part of the broader Simcenter structural dynamics ecosystem. This helps support the complete process of modelling and updating structural dynamic models (Figure 20).
Figure 20: The broader Simcenter structural dynamics ecosystem consists of pre-test analysis (left), modal acquisition and analysis (middle), and correlation with updating (right).
Components include:
Pre-Test Analysis (Simcenter 3D): Engineers can leverage Simcenter 3D for pre-test analysis, guiding optimal sensor and excitation locations, and generating initial geometry for Testlab.
Simcenter Testlab (Measurement & Analysis): Testlab serves as the hub for instrumenting, measuring, analyzing, and validating experimental data, providing initial correlation (MAC) between test and CAE results.
Post-Test & Updating (Simcenter 3D): The validated test modes can then be brought back into Simcenter 3D for model updating, refining FEA models by adjusting material properties or shell thicknesses to improve correlation (e.g., FRAC) between measured and synthesized FRFs.
This seamless integration ensures a continuous digital thread from simulation to test and back, accelerating the development cycle.
