Maxwell Linear Generator (Input Power < Output Power and losses?)

[2diamondm]

Hello,

I’m running a 2D transient FEM simulation of a linear generator using ANSYS Maxwell. The simulation runs without any error or warning, but I’m facing a problem related to energy conservation and inconsistencies between different software versions.

---

Main Issue:

In ANSYS Electronics R2022, I observe:

• Output power: 95.95 W

• Input power: 99.33 W

• Losses (Core + Solid + Coil): 6.129 W

• → Total (Output + Losses) = 102.08 W, which is more than the input, violating energy conservation.

In Maxwell 14.0 (same model):

• Output power: 96.05 W

• Input power: 100.28 W

• Loss values were not available, but overall energy behavior was more consistent.

---

Simulation Details:

• Software: Maxwell 14.0 and ANSYS Electronics R2022

• Model: 2D linear generator

• Motion: Translational, Global Z, ±5.5 mm stroke

• Speed: Constant at 1.727 m/s

• Time step: 0.1 ms

• Simulation time: 80 ms

• Excitation: 220-turn coil

• Circuit: R = 24.925 Ω, C = 226 µF

• Boundary: Vector Potential = 0

• Input power: Force_z × Speed

• Output power: Current × Voltage

• BH curve material: 35PN230 (nonlinear)

• Losses used: CoreLoss, SolidLoss, StrandedLoss

---

Additional Notes:

• Increasing the time step from 1 ms to 4 ms caused output power to drop (96 W → 84 W), showing time-step sensitivity.

• Mesh refinement had little effect on the output at 1 ms.

• In a second test, increasing the magnet thickness (3 mm → 3.8 mm) gave a more realistic result (Input > Output + Losses).

---

Questions:

1. Why does output + losses exceed input power?

2. Does R2022 compute losses differently compared to Maxwell 14?

 

Any help or insights would be appreciated.

Best regards,

Muhammad Mohsin