TB 26-08 Complex Modes in Lightly Damped Structural Systems:A Two-Mode Interaction Perspective

NASA's Chemical Equilibrium with Applications (CEA) code is a crucial tool for analyzing propulsion systems, providing equilibrium chemistry, rocket performance, shock, and detonation calculations. The National Engineering and Safety Center (NESC) Activity TI-22-01730 recently modernized the legacy CEA2 Fortran code into CEA v3, an object-oriented software package with enhanced interface support, updated thermochemical data, better maintainability, and improved integration with modern workflows.

This updated version retains compatibility with the original CEA input format while allowing direct use in contemporary analysis environments such as Python, C, MATLAB, and automated design studies.

Classical modal analysis typically assumes lightly damped structures behave like nearly proportional systems with real mode shapes. However, this interpretation can be inaccurate when two modes are closely spaced. In such cases, nonproportional damping introduces an off-diagonal coupling term ε. The crucial question then becomes whether the coupling is sufficiently large compared to the baseline half-frequency separation δ, which is calculated as (ω2 - ω1) / 2.

The reduced perturbation framework offers a method to analyze this situation. For a closely spaced pair of modes, the central eigenvalue λ0 is approximated as -ζω + iω. The coupled eigenvalues are then written as λ1,2 = λ0 ± s. Under light damping and close-mode assumptions, the local quartic eigenvalue problem simplifies to s² = ε²/4 - δ².

This relation determines the direction of eigenvalue separation in the complex plane. When 0 ≤ ε ≤ 2δ, the split is predominantly horizontal, indicating a predominantly decay rate separation. As ε increases further, β increases, leading to stronger separation of the decay rates while their frequencies remain approximately equal.

It is essential to note that complexity in a mode shape should not be immediately attributed to heavy damping, nonlinearity, or poor test data. Instead, the framework suggests examining the complex poles together with the mode-shape phase. Frequency convergence accompanied by near-quadrature modal components is consistent with the imaginary-s regime, while nearly equal frequencies accompanied by decay-rate separation indicates the real-s regime.

The framework also provides a practical link between test data and analysis, as the effective interaction strength can be estimated by calculating s = (λ1 - λ2) / 2 from the identified poles, while δ can be estimated from the uncoupled or pretest analytical modal spacing. These quantities can then be used to determine proximity to the transition boundary between different regimes of modal interaction.