Effects of inspiratory waveform, inspiratory time, and pause on compartmental energy partitioning in a heterogeneous two-compartment respiratory model

In mechanically heterogeneous respiratory systems, the temporal pattern of inspiratory flow may influence how airway-opening work is partitioned between mechanically distinct compartments, independently of global mechanical power (MP). Using a simplified time-resolved two-compartment model with common airway-opening pressure, we investigated whether inspiratory waveform, inspiratory time, and end-inspiratory pause affect model-derived compartmental airway-level energy routing, including cases matched for global MP. Mechanical heterogeneity was introduced by varying compliance and resistance ratios across a 61 × 61 grid. Volume-controlled simulations included three waveforms, three inspiratory times, and two pause conditions, yielding 66,978 simulations. Global MP, compartment-specific inspiratory energy, the Partition Index (PI), and the Energy Inequality Index (EII) were calculated. MP ranged from 1.75 to 13.55 J/min, PI from 0.13 to 0.80, and EII from 0.00 to 0.74. Increasing asymmetry shifted the model-derived fraction of airway-level energy routed between compartments. Waveform, inspiratory time, and pause also altered compartmental energy routing when mechanical properties were unchanged. Across pairs matched for global MP, sensitivity analyses using 1–10 % tolerances yielded median absolute PI differences of 0.072–0.073, 95th-percentile differences of 0.239–0.241, and maximum differences of 0.488–0.519. Similar global MP may therefore coexist with substantially different model-derived compartmental energy-routing patterns. Given the simplified two-compartment structure of the model, PI should be interpreted solely as a model-derived descriptor of relative airway-level energy routing, not as a direct tissue-level metric or surrogate of stress, strain, energy dissipation, or regional lung injury risk. These mechanistic, hypothesis-generating findings warrant testing in higher-fidelity, experimental, and patient-specific models.