Interactive demo
Coffin-Manson, in your browser
See how junction temperature swing, peak Tj and the fatigue exponent shape the cycles-to-failure of an IGBT power module — the core physics behind every InverterAI prediction.
Junction temperature swing (ΔT_j)60 K
Amplitude of each thermal cycle. Utility-scale PV inverters typically see 30–90 K, with daily and intra-day cycles.Maximum junction temperature (T_j_max)125 °C
Peak silicon temperature during operation. Si IGBTs are usually rated up to 150–175 °C; SiC modules can run hotter.Fatigue exponent (α)5.0
Higher α means larger swings hurt disproportionately more. Typical values are 4–6 for SAC305 solder joints in power modules.E_a = 0.5 eV · activation energy and Boltzmann constant kept fixed for clarity.
Cycles to failure (N_f)
7.14 × 10^3
thermal cycles
RUL (estimated)
19.5 years
@ 1 thermal cycle / day
vs. baseline
+0%
Change in N_f vs. the demo baseline (60 K, 125 °C, α=5).
The equation behind the slider
N_f = A · (ΔT_j)^(−α) · exp( E_a / (k_B · T_j_max) )Coffin-Manson with Arrhenius temperature acceleration. InverterAI extends this with rainflow counting, Foster/Cauer thermal networks and a PINN trained on field SCADA so each inverter gets its own RUL curve, not a textbook average.
This is one piece of the InverterAI engine.
Production deployments combine Coffin-Manson, Arrhenius capacitor decay, virtual Tj sensing and a Physics-Informed Neural Network trained on multi-year SCADA history.