Features and Limitations of 2D Active Magnetic Levitation Systems Modelingin COMSOL Multiphysics
Adam Piłat
AGH – University of Science and Technology,Department of Automatics,Mickiewicza 30 Ave, 30-059 Kraków, Poland
Presented at the COMSOL Conference 2010 Paris
Agenda
• Interdisciplinary design
• Active Magnetic Suspension
• Active Magnetic Bearing
• Automata for modelling
• Optimisation
• PDE+ODE
• Conclusions
The proposed interdisciplinary Design Approach where virtual prototype is being developed and studied
System specification
Mechanics
Electronics
Control system
Virtual prototype
Validation, Optimization
Prototype
Manufacturing
Knowledge base
Tests, Identification
Interdisciplinary dynamics modelling and simulation
0 0.02 0.04 0.06 0.08 0.1-0.01
-0.008
-0.006
-0.004
-0.002
0
Control ON Control OFF
Time [s]
Displacement [m]
Magnetic levitation system MLS1EM in action.
Cross-section model of the MLS1EM
This figure presents the magnetization and magnetic field in the formof streamlines and arrows respectively. One can find that the levitated object is self centered with respectto the actuator geometryand the iron based components are magnetized.
MLS2EM - upper electromagnet replaced with cylindrical one
Magnetic flux density represented in the wire frame mode.
Active Magnetic Bearing
Geometry generator
Geometry generator
FEM analysis
Manual Automatic (programmable)
With fem do
• select application mode
• assign geometry objects
• assign parameters and equations
• initialize mesh
• solve
• analyze results using post processing functions
Geometry generator
Optimization
Initial
parameters set-up
AMB geometry generator
Solver
Criterion Optimization
Lab algorithms
Linear and nonlinear solvers linear and nonlinear constraints
Optimization
Initial
parameters set-up
COMSOL Toolbox
AMB geometry generator
Solver
Criterion
Optimization, GADS
Toolboxes
MATLAB
Linear and nonlinear solversincluding Genetic Algoritms
linear and nonlinear constraints
Optimization
Initial parameters
set-up
MATLAB + COMSOL Toolbox
AMB geometry generator
Criterion
Optimization, GADS
Toolboxes
Multiphysics model
Embedded Multiphysics
model
Controller
Simulink model
ODE Solver
PDE + ODE Solver
Example
1 2 3 4
x 10-3
20
40
60
80
100
Pole radius [m]
Re
lativ
e F
lux
de
nsi
tya)
2 4 6 8
x 10-3
20
40
60
80
100
Pole width [m]
Re
lativ
e F
orc
e g
ain
b)
Position of the levitated object steeredby the PD controlled – both implemented in COMSOL Multiphysics.
0 0.5 10.009
0.0095
0.01
0.0105
0.011
0.0115
Time [s]
Dis
plac
emen
t [m
]
Conclusions
Design, modelling and simulation
Modelling as is – geometry, materials
PDE + ODE
Optimisation
Computational effort
Controller architecture
Real-time calculation
Simulink data exchange and link
To Do:
Rotor axial motion and rotation in the AMB plane
Thank You for Your Attention