Mode Selection

After identifying poles on the stabilization diagram, the selected modes are listed in the Mode Selection table. This step allows you to review, refine, and finalize the set of modal parameters before proceeding to validation and mode shape animation.

Selected Modes Table

Each selected pole appears as a row in the table with the following columns:

Column	Unit	Description
Mode #	—	Sequential mode index, automatically assigned in order of increasing frequency.
Frequency	Hz	Natural frequency of the mode (damped frequency).
Damping	%	Viscous damping ratio ζ. Physical values typically range from 0.05% to 10%.
MPC	0–1	Modal Phase Collinearity. Values above 0.9 indicate a real normal mode; lower values suggest complex mode behavior or poor data quality.
Quality Score	0–100	Composite quality metric combining pole stability, MPC, and cross-validation results (PolyProMax only).
Model Order	—	The model order at which this pole was extracted.
Source	—	Algorithm that identified the pole (LSCF, LSCE, or Cross-Validated).

Quality Score Interpretation

The Quality Score provides a single-number assessment of each mode's reliability:

Score Range	Rating	Interpretation
90–100	Excellent	Highly stable pole with strong cross-validation. High confidence in frequency and damping values.
70–89	Good	Stable pole with minor variations across model orders. Suitable for most applications.
50–69	Fair	Pole shows some instability or moderate MPC. Verify with independent methods (e.g., FRF synthesis).
Below 50	Poor	Questionable pole. May be computational or result from poor data quality. Consider excluding.

Tip: Sort the table by Quality Score (click the column header) to quickly identify modes that may need further review or removal.

Editing Mode Parameters

You can manually adjust modal parameters after extraction. This is useful when the analyst has additional knowledge about the structure (e.g., from FE analysis or previous tests).

Editable Fields

Field	When to Edit
Frequency	Fine-tune if the pole is between two nearby frequency lines and you have a more precise reference value.
Damping	Adjust if the damping estimate is clearly inconsistent with physical expectations (e.g., negative damping from numerical artifacts).

How to Edit

Double-click the value cell in the table.
Enter the new value.
Press Enter to confirm or Escape to cancel.
The mode shape is automatically recomputed using the updated pole location.

Caution: Manual parameter editing overrides the algorithm's estimates. Use this feature sparingly and only when you have strong justification. Edited values are marked with an asterisk (*) in the table.

Deleting Modes

Remove modes that are identified as computational, duplicate, or otherwise not of interest:

Single mode: Select the row and press Delete, or right-click > Delete Mode.
Multiple modes: Hold Ctrl or Shift to select multiple rows, then press Delete.
By quality threshold: Right-click > Delete Modes Below Score... and enter a minimum quality score. All modes with lower scores are removed.

Deleted modes are removed from the selected modes table and their markers revert to the original stability color on the stabilization diagram. Deletion can be undone with Ctrl+Z.

Sensitivity Analysis

The sensitivity analysis tool evaluates how robust the selected modes are with respect to changes in analysis parameters. This helps identify modes that may be artifacts of a particular parameter setting.

Running Sensitivity Analysis

Select one or more modes in the table.
Click Sensitivity in the toolbar, or right-click > Sensitivity Analysis.
Choose the parameter to vary:

Parameter	Variation Range	What It Tests
Model Order	±10 orders around selected	Stability of pole location across model orders. Stable modes show minimal frequency/damping drift.
Frequency Range	±10% of band edges	Sensitivity to the analysis frequency band. Physical modes should persist when the band is slightly widened or narrowed.
Number of FRFs	Leave-one-out	Impact of individual FRFs on the extracted parameters. A mode that disappears when one FRF is excluded may be an artifact of that particular measurement.

Results Display

The sensitivity analysis results are shown as:

Frequency variation plot: Natural frequency vs. varied parameter. Flat curves indicate robust modes.
Damping variation plot: Damping ratio vs. varied parameter. Large variations indicate sensitivity.
Confidence interval: The 95% confidence interval for each parameter is displayed in the modes table.

Tip: Run sensitivity analysis on modes with Quality Scores between 50 and 70 to decide whether to keep or remove them. Modes that show less than 2% frequency variation across the parameter sweep are generally trustworthy.

Mode Sorting and Filtering

The modes table supports interactive sorting and filtering:

Sort: Click any column header to sort ascending; click again for descending.
Filter by frequency range: Enter min/max frequency in the filter row above the table.
Filter by quality: Use the Quality Score slider to hide modes below a threshold.
Highlight on diagram: Selecting a mode in the table highlights the corresponding pole on the stabilization diagram.

Mode Summary Statistics

The status bar below the table displays summary statistics for the current selection:

Total number of selected modes.
Frequency range of selected modes (lowest to highest).
Average damping ratio.
Average Quality Score.
Number of modes with MPC > 0.9.

Common Mode Selection Pitfalls

Pitfall	Symptom	Remedy
Selecting computational poles	High damping (>10%), low MPC, no corresponding FRF peak.	Remove the mode. Verify against FRF Sum or CMIF overlay.
Missing closely spaced modes	FRF synthesis shows residual at a frequency with a broad peak.	Return to stabilization diagram and check for two overlapping pole columns. Enable CMIF overlay.
Duplicate mode selection	Two modes at nearly identical frequency with high mutual MAC.	Keep the mode with the higher Quality Score; delete the other.
Selecting harmonic poles (OMA)	Extremely low damping (<0.05%), frequency exactly at machine order.	Exclude the pole. Check for harmonic warning markers on the diagram.
Incomplete frequency coverage	Synthesized FRF deviates from measured FRF in certain bands.	Expand the model order range and re-run the analysis for the missing frequency region.

Tip: A good rule of thumb is to select approximately 1.5 to 2 times the number of visible peaks in the FRF Sum. Extra poles from higher-order modes outside the visible range may still contribute to the overall response.

Proceeding to Validation

Once you are satisfied with the selected modes:

Review the summary statistics to confirm the mode set is complete.
Click Next or navigate to the Validation step.
The selected modes and their mode shapes are passed to the validation tools for animation, MAC computation, and FRF synthesis.