Beruflich Dokumente
Kultur Dokumente
EDITORIAL.
! FLUX STUDIO:
Dimensioning of an
electromagnet.
Commercial
Commercial Team
Team -- CEDRAT.
CEDRAT.
! Design of an actuator
based on Thomson effect.
! Magnetostatic and Thermal
Analysis of an Electromagnet.
>> Motors:
! Asynchronous machine
end winding characterisation...
! SRM: Drive modelling using
Flux to Simulink technology...
>> MLI:
! FLUX and SIMULINK:
Example of modelling...
>> Network:
! PSCAD and transmission
lines.
>> Software:
PSCAD
FLUX/Matlab
SIMULINK
FLUX and Simulink links (pages 89); SRM : Drive modelling... (pages
10-11).
FLUX
Studio
Piezo Actuators
APA towards new
horizons (page 13).
CEDRAT
FLUX
FLUX: Asynchronous machine end windings characterisation
(pages 3-4); Design of an actuator based on Thomson effect
(page 5); How to denes properties in FLUX2D using tabulated values (page 14).
SOFTWARE>>
-2-
ntroduction
The knowledge of the end
winding inductance is an
important parameter to determine
the voltage-current characteristics
in motors.
The 3D finite element model
allows to represent precisely the
three-dimensional geometry of
end windings with which it can
be obtained a complete analysis
of electromagnetic fields in the
extremities of machines, for
the computations of losses and
electromechanical efforts on end
windings.
Among the leakage inductance, one
distinguished:
The 2D leakage, located in
the straight part of the machine
where the magnetic field is bidimensionnal: Slot leakage,
differential leakage, harmonic
leakage
The 3D leakage at the extremities
of the machine where the looping of
the conductors induce "3D" leakage
ux: Two 3D leakage inductance are
dened, the end ring inductance
at the rotor and the end winding
inductance at the stator. This latter
is the parameter studied in the
following Finite Element Method.
The model
The model is based from a
classical type of asynchronous
machine (LS180). This machine is
voluntary simple in order to validate
the calculation method on clear
basis but the study can be applied
with more complicated machines
(with screen and mounting anges).
The geometry of the model is shown
in gure 1.
Results
The calculation of end winding
inductance has been performed
with four different nite element
methods by using energy and
ux computations. Whatever the
hypothesis, the four methods give
very close results with only 4% of
maximal variance.
The computation has first been
completed with hypothesis
inspired from the existing IEC
norms concerning the experimental
determination of end winding cyclic
inductance (L-M). The computations
have been executed for several
lengths of iron in order to show that
the value is independent of it. This
conrms that the computed value is
an extremity parameter. The results
founded with the four methods
with this reference simulation are
presented in the following graph:
The study has been completed with
other simulations:
Single phase (configuration
which gives the self and the mutual
inductance),
(continued on page 3)
SOFTWARE>>
-3-
Asynchronous machine end winding characterisation with FLUX3D. (continued from page 2)
Fabrice Foucher, Claire Blache - CEDRAT.
Connections with a
FLUX2D study
The computed value can be
introduced in the CIRFLU module
of FLUX2D in order to realise a
complete study on motors. With
the knowledge of the end winding
inductance, your study will take
into account the extremity leakage,
which are important in transient.
Studied device.
Table 1:
Computing times
SOFTWARE>>
-4-
with experience
in FLUX2D and
having already
good knowledge
of the products
should be obviously
much
shorter.
Studied device.
SOFTWARE>>
-5-
ntroduction
Energy balance
Figure 4 proposes a representation
of the energy balance for the study
of such electromechanical devices.
It gives an idea of how the energy
provided initially by the capacitor is
distributed in time. It can be noted
that the propeller based on the
Thomson effect has a very weak
efciency, of about a few %. This
is the price paid to obtain a quick
movement of the mobile part.
Multi-parametric study
Figure 1: Description of the device based
on the Thomson effect.
Conclusion
FLUX2D simulation
The propeller based on the Thomson
effect we study is axisymmetric
and therefore able for finite
element analysis with FLUX2D.
The magnetic transient regime
characterises the operation of this
actuator. That is why it is advisable
to use a time-varying magnetic
simulation, with translating air-gap
feature and circuit equations taken
into account. Indeed, the induced
currents developed in the disk must
SOFTWARE>>
-6-
(continued on
page 7)
SOFTWARE>>
PSCAD and transmission lines.
-7-
Example 2 : Network
PSCAD simulates an entire electrical
network, from the source to the
load. With the following example,
we represent a transmission line
from the source to a first node
where the network is separated
into two branches, one simulated
by a three phase RC single line
load, and the other is simulated
with a second transmission line
at a different voltage level. From
there the network is again split into
two branches, both connected by a
short distribution line and a three
phases balanced load.
SOFTWARE>>
-8-
Results
The results of simulation with
Flux2D Simulink coupling have
been compared below with those
of a FLUX2D simulation where the
states of switches are dened by an
user sub-program.
(continued on page 9)
SOFTWARE>>
-9-
Figure 5: Simulink
model of the device.
Partner !!
Sbastien Cadeau-Belliard - CEDRAT.
use Simulink to
design, model,
and
simulate
systems, and to automatically
generate code for the embedded
algorithms. This interface enables
those engineers to incorporate
detailed models of electromagnetic
and electromechanical devices
easily in their Simulink simulations,
leading to better understanding of
system behaviour and faster testing
of designs.
We invite you to contact CEDRAT
to get full information about the
wide and new possibilities this
technology offers.
http://www.mathworks.com/products/connections/product_
main.shtml?prod_id=416
SOFTWARE>>
- 10 -
1 - Introduction
Switched reluctance machines
are relatively simple electric
machines but with a high degree of
nonlinearity. This yields a drive with
complicated algorithms to control
its nonlinear behaviour. SRM-drives
can be optimized for different
customer demands (maximum
efficiency, maximum torque,
minimum noise and vibration ).
To optimize its control parameters,
an accurate nonlinear model of
the SRM is required. The model,
presented in this article, uses the
Flux to Simulink technology. The
advantage of this coupled method
is the integration of a simple 2D FEmodel of the motor geometry into
a complex Matlab/Simulink model
of the drive with different control
optimizations.
2 - Basic principles of
SRMs
The concept of a switched
reluctance machine is very simple.
It consists of stator and rotor poles,
made of laminated steel with high
magnetic permeability. Only the
stator poles are excited by coils.
A typical conguration is the SRM
with 8 stator en 6 rotor poles, a socalled 8/6 SRM. With 8 coils on the
stator, 4 phases are created with
the corresponding coils in parallel.
A sequence of anti-clockwise
excitations of the different phases
results in a clockwise rotation of
the rotor due to a positive torque
generation. SRMs are characterized
by a ux-linkage and inductance
L, varying with position and
current. The position-dependent
behaviour is due to the geometry
of the overlapping stator and rotor
teeth. In most SRM-applications,
saturation occurs, resulting in a
nonlinear inductance and fluxlinkage. This nonlinear behaviour
explains the difculty in modelling
and controlling SRM-drives.
5 - Flux to Simulink
technology
The coupling between drive model
and FE-analysis is represented in
gure 4. The input parameters for
the FE-analysis are the applied
phase voltages, the switch
resistance values (RON or ROFF) and
the rotor speed.
3 - FE-model
SOFTWARE>>
- 11 -
The output parameters of the FEanalysis are the phase currents, the
coil voltages and the instantaneous
electromagnetic torque in the
airgap.
At every time-step (ts = 2.10-5 s),
data is exchanged between the
drive model and the FE-analysis.
This method has the great
advantage that a complex drive
model in Matlab/Simulink can be
used in combination with accurate
flux-linkage calculation, taking
into account the mutual coupling
between adjacent phases. The
disadvantage of this method is the
rather high calculation time, caused
by the high amount of elements in
Figure 4 : SRM drive model with Flux to Simulink technology.
the thin airgap. Figure 5 shows the
occurs, due to the excitation of adjacent
simulation results
phases D and B. Conform the ux distribution
of coil voltage,
of gure 3, this voltage is induced in phase A
phase current and
when phase B or D is activated together with
electromagnetic
phase C (see gure 5).
torque production
for a reference
current of 7.5 A
6 - Conclusion
and a rotor speed
The Flux to Simulink technology has proven
of 50 rad/s. The
to be an efcient tool to model complex
results clearly
motor drives in combination with accurate
show that, besides
flux-linkage calculation. Effects, such as
the normal ON,
mutual couplings and induced voltages, can
OFF and chopping
be analysed. Attention must be paid to the
voltage of phase
sample-time in Matlab/Simulink and the
A,
an
extra
number of elements in the FE-model in order
Figure 5 : Simulation results of Flux to Simulink technology.
induced voltage
to keep the computation time acceptable
Figure 9:
Fault on the
line.
Figure 10:
Instantaneous traces
for a Line to Line fault
on A and B phase,
30% of the line.
SOFTWARE>>
- 12 -
A DC electromagnet for a
directional control valve
is characterised by the
presence of a cylindrical guide for
the mobile core.
This electromagnet is used to drive
a directional control valve with a
maximum pressure of 75 bars.
A magnetostatic study has been
carried out with FLUX2D for the
optimisation of the force in the
nominal working conditions, and
a thermal study with FLUX3D to
obtain the optimised magnetic
circuit to reduce temperature
in order to increase, in ne, the
force.
The difculty of the design consists
in the necessity to obtain a curve
of sufcient force to gain the total
resistant force for cold or hot
device, with the respect of project
bonds.
2D Magnetostatic
analysis
This phase of study is fundamentally
useful to optimise the geometry of
the cylindrical guide and the core.
The magnetic circuit is of armored
type and can be approximated to
axial-symmetrical type in FLUX2D,
considering an equivalent thickness
for the external part that gives
the same useful section for the
ux ow. In gure 1 we can see
the description of all regions in a
2D cross section of the magnet
central section. Only the upper coil
is energised. The electromagnet is
bidirectional and both displacements
are 3,0 millimetres long.
The coils nominal power was set
initially to 50W and a nominal
voltage source of 12V.
Shape and dimensions of the bronze
soldering, diameters and lengths of
the cylindrical guide and the core
have been optimised.
3D Thermal analysis
Due to the geometry of the
electromagnet, a bi-dimensional
axis symmetrical approximation
is not enough to obtain denitive
results: this is due to the external
embedding in polymeric resin
containing the device. Considering
the correct right external surfaces
and thickness of such insulating
shield is fundamental to get the
right heat exchange and so precise
temperature results.
Volume regions are represented in
the gure 4:
We can see the external magnetic
circuit with the cylindrical guide.
For symmetry reasons, half of the
device is represented.
Physical properties
The coil region has been dened
like a heat source. To obtain results
under high temperature condition,
the power of the same coil has been
set up as a function of the state
variable TEMPERATURE (Celsius
degrees temperature). The coil
power function is showed below:
*
where the numerator is the power
at the ambient temperature (20C),
0.0039 is the temperature coefcient
of copper and 1.4326.10-5 is the
volume (m3) of the half part of the
designed coil. With this formulation
the heat power generated of the coil
depends on the temperature of the
same region.
After setting the geometric
parameters up and defining the
heat sources, the convection
coefcients have been assigned to
the air-contact surfaces, making
opportune distinctions between
Where:
T1: wall temperature [K]
T2: ambient temperature [K]
h: wall characteristic dimension [m]
The K coefcient is assigned as
function of the type of wall: for a
vertical wall it is 5.6, for a horizontal
wall under a device is a 20% less
(4.48), while for an upper vertical
wall it is 20% more (6.72).
Figure 2.
Figure 3.
In the same way h is the height of
the vertical wall, or in case of an
horizontal wall, half its width. In all
the formulae the rst term regards
convection, while the second
regards radiation.
Thermal analysis
results
Therefore the problem has been
solved to verify the evolution of
*
Figure 1.
ACTIVE MATERIALS>>
- 13 -
EYSTONE INTERNATIONAL
Co. Ltd signed an agreement
of promotion and distribution
of our standard piezo products in
JAPAN. Mister Tamioshi KUROSAWA,
Marketing Director of Keystone, and
Mister Norio TATE, Sales Engineer,
would like to express to the Flux
magazine readers their feelings
about this new collaboration:
We are very pleased to have
received our recent initial order
from CEDRAT TECHNOLOGIES.
Having received the order of two
APA S series from Tohoku University
became the first work for us
Keystone International for CEDRAT
TECHNOLOGIES.
We were impressed the quality of
product when it was unpacked, all
members at Keystone International
were pleased that the quality
verication would lead to generate
more customers. We think that
IWAG AG signed an
agreement of promotion and
distribution of our standard
piezo products in Switzerland.
Mister Jean Jacques WAGNER,
president of DIWAG, is also the
inventor of an innovating optical
encoder called Dicod. By
distributing Cedrat Technologies
piezo products, Mr Wagner is willing
to enlarge the DIWAG technological
offer in the micro-positioning
eld.
These two agreements will launch
for the rst time a full and exclusive
commercial activity for our piezo
products in Japan and Switzerland.
These agreements also supplements
our precedent agreements covering
the Benelux countries, Italy, Poland,
Germany, Austria, USA and India.
The search for partners in other
countries (UK, Spain, Scandinavia)
still continues.
Conclusion
The important results obtained
from the thermal analysis allowed
to observe the negative infuence of
the supercial resin (low thermal
conductivity coefcient) on thermal
exchange. The thermal analysis
has concurred to preview the
value of the magnetizing current
at high temperature condition and
therefore a new magnetostatic
analysis with FLUX2D has been
carried out in order to verify if the
force of under high temperature
condition was sufficient to gain
total resistant force.
Experiences have confirmed the
theoretical forecasts, with minimal
Figure 5.
Figure 4.
SOFTWARE>>
- 14 -
LUX software is controlled under quality assurance procedures according to EDF (French Historical Power Supplier)
standards. These Quality Assurance procedures have been effective since 1994 for 2D applications (version 7.12) 1998
for 3D applications (version 3.10).
A Software Quality Assurance Plan has been set up to dene software development procedures and ill functionnings. Several
cases types has been dened:
Analytical cases: those cases test unique features comparing FLUX results with analytical formulae,
Numerical cases: those cases are run for every version to test the software non regression (around 50 cases for the whole
FLUX software),
Physical cases: FLUX results are compared to measurements.
In the mean time, the documentation quantity has been largely increased.
Three types of documents are now available:
Users guide (700 pages in 2D, 1000 pages in 3D),
Tutorials, explaining every command to complete the case (mouse click,
text)
-2D : Magnetostatic, Thermal, Electrostatic, Brushless DC machine,
Induction machine, Induction heating, Translating motion
-3D : Magnetostatic.
Technical papers, explaining briey all the commands to complete the
case as well as main results.
-2D : Superconductors, Transformers, FLUX2D to Simulink technology,
-3D : DC machine, End windings characterisation.
QUAL
A SURANCE
A
CER
TE
CEDRAT
ance
The whole documentation represents around 6000 pages. All the les
are available on FLUX CDRom with the PDF format, as well as the FLUX
les corresponding to both tutorials and technical papers.
AGENDA
- 15 -
th
European Conference on
Power Electronics and Applications
2 - 4 September 2003, Centre Pierre Baudis, Toulouse, France.
EPE Association is delighted to invite you to submit a synopsis for presentation at
the EPE 2003 Conference. More information regarding the submittance of synopses
(procedures, requirements, topics and subtopics) can be found on the EPE 2003
Conference internet site: http://epe2003.inp-toulouse.fr
Intending authors should note following deadlines:
Notication of provisional acceptance: 10th of February 2003
Receipt of full typescript for nal review: 15th of April 2003
Liste of topics
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Devices
Power electronic system design & Packaging
Power electronics converter circuits
Modulation strategies and specic
control methods for static converter
Application of control methods to electrical systems
Motion control and robotics
Measurements and sensors
Power supplies
Electrical machines and adjustable speed drives
Power electronics and drives in space and aerospace
Power electronics and drives in surface transport
Power Electronics in generation, transmission
and distribution of electrical energy
Systems engineering
Education
Dates to
remember
Training courses
February 11-13, 2003
Active Materials and Actuators
March 5-7, 2003
May 5-7, 2003
Magnetism for Electrotechnics
March 11-13, 2003
September 16-18, 2003
ATILA Introductory course
March 26-28, 2003
November 19-21, 2003
FLUX (2D) Introductory course
April 2-4, 2003
November 26-28 2003
FLUX (3D) Introductory course
April 9-10, 2003
FLUX (2D) Advanced course
A full list of training programs can be obtained from Jolle Lescne at CEDRAT, or
from our Internet site http://cedrat.com. The dates are already fixed up until the end of
year 2003, allowing you to prepare your training course calendar.
Email : training@cedrat.com.
ISEM 2003
Versailles, France.
11th International Symposium
on Applied Electromagnetics and
Mechanics: 12-14 May, 2003
For all your projects in electrical engineering, the CEDRAT Group provides a
fully integrated line of tools, products and
services.
CEDRAT and its partners form a worldwide
consortium with a reputation built on the
success of FLUX software, APA piezo
actuators, and R&D activity.
FLUX, quality-certified through EDF, has
more than 600 companies as clients,
active in research and production of every
type of electric machine or process, from
small appliances to large machinery, in
automotive and aerospace industries,
energy transfer and conversion,
telecommunications, household
appliance and medicine Flux supports
high-precision design of electrical and
electromechanical devices. Coupled with
other programs, Flux solutions expand into
drive technology, mechanics, acoustics and
thermal analysis.
- europA:
Austria - Switzerland: APPLIED MAGNETICS.
Rue de la Gare 15A, CH - 1110 Morges, Switzerland.
Phone: +41 21 803 58 68
magnetics@bluewin.ch
Fax: +359 2 9 88 51 84
CAE
Piezo
CAE
CAE
Phone: +49 89 65 12 20 0
anders@maccon.de
CAE
CAE - Piezo
CAE-Piezo
CAE
Kurosawa Building, 3F
13-27 Sakasai Kashiwa Chiba, 277-0042 Japan.
Phone: +81 45 586 0800
key@keystone-intl.co.jp
Piezo
Piezo
Piezo
CAE
CAE
Phone: +33(0)4 72 18 60 96
Fax: +33(0)4 78 43 39 62
alain.nicolas@eea.ec-lyon.fr http://cegely.ec-lyon.fr
Phone: +33(0)4 76 82 62 99
Fax: +33(0)4 76 82 63 00
jean-louis.coulomb@leg.ensieg.inpg.fr
http://www.leg.ensieg.inpg.fr
Fax: +33(0)1 39 01 01 64
http://prismeca.com
(InductionHeating).
(Motors).
software).
(Software).
http://rpi.edu
CAE
CAE
Poland: TermagSoft.
CEDRAT
Phone: +33(0)3 20 30 40 50
jean-claude.debus@isen.fr
Fax: +49 89 65 52 17
http://www.maccon.de
- AsIA:
India: Maxsoft.
Piezo
CAE
CAE
- AMERICA:
Brazil: ELECTROMAGNETICS TECNOLOGIA.
CAE
Phone: +359 2 9 88 51 84
vdla@vmai.acad.bg
Spain: INDIELEC.
Phone: +34 607 330 112
indielec@indielec.es
CAE
Partners :
UT
CEDRAT
ABOUT CEDRAT
CAE: Distribution of CAE software for Electrical Engineering - Piezo: Distribution of Piezo Products