Understanding Virtual Memory In Red Hat Enterprise Linux 4

Tobbe

Description

• mémoire - matière potentielle : management
• mémoire - matière potentielle : objects
• mémoire - matière potentielle : addresses from 16 mb to 1 gb2
• mémoire - matière potentielle : zones
• mémoire - matière potentielle : pools
• mémoire - matière potentielle : management unit
• mémoire - matière potentielle : in red hat enterprise linux
• mémoire - matière potentielle : translation
• mémoire - matière potentielle : by aliased addresses
• mémoire - matière potentielle : access architecture
• mémoire - matière potentielle : manager
• mémoire - matière potentielle : state machine
• mémoire - matière potentielle : control
• mémoire
• mémoire - matière potentielle : operations
• mémoire - matière potentielle : region
• mémoire - matière potentielle : above 1 gb
• mémoire - matière potentielle : management code
• revision
• mémoire - matière potentielle : at safe levels

Understanding Virtual Memory In Red Hat Enterprise Linux 4 Neil Horman Version 0.1 - DRAFT EDIT December 13, 2005 1

• virtual file system
• virtual memory manager
• physical address
• vm
• virtual address
• pages
• kernel
• memory
• page
• system

Sujets

• Management
• Object
• Translation
• Manager
• Finite-state machine
• Control

Informations

• Publié par: onne
• Nombre de lectures: 39
• Langue: English
• Poids de l'ouvrage: 9 Mo

Extrait

TRITA-ETS-2005-02
ISSN-1650-674X
ISRN KTH/R-0504-SE
ISBN 91-7178-032-7





Design and Analysis of a

Novel Low Loss Homopolar
Electrodynamic Bearing
Torbjörn A. Lembke
Doctoral Dissertation
Stockholm 2005
ELECTRICAL MACHINES AND POWER ELECTRONICS
DEPARTMENT OF ELECTRICAL ENGINEERING
ROYAL INSTITUTE OF TECHNOLOGY
SWEDEN
Submitted to the School of Electrical Engineering, KTH,
in partial fulfilment of the requirements for the degree of
Doctor of Philosophy









Printed in Stockholm, Sweden
Universitetsservice US AB, 2005


TRITA-ETS-2005-02
ISSN-1650-674X
ISRN KTH/R-0504-SE
ISBN 91-7178-032-7






















To my wife
Ann-Sofie
and my sons
Johannes & Fredrik



Abstract

A novel homopolar electrodynamic bearing, together with a suitable
permanent magnet drive, have been developed for high-speed applications
where low losses and high reliability are essential and exclude the use of ball
bearings, and yet where active magnetic bearings offer a too complex system
solution. Considered applications are small turbomolecular vacuum pumps,
and maintenance free flywheels for energy storage in remote telecom and
satellite systems. Other upcoming areas where these bearings offer interesting
technical and economic solutions are compressors for fuel cells and heat
pumps, applications which normally suffer from short bearing lifetime.

Unlike active magnetic bearings, forces are produced in electrodynamic
bearings without any control electronics, thanks to stabilizing eddy currents
induced by permanent magnets. In the novel homopolar concept eddy current
losses are reduced to a minimum using a homopolar design with ring magnets
instead of multipole or Halbach arrays.

Currents and forces are simulated using steady state 3D-FEM analysis, which
can take velocity into account using an implemented Minkowski transform.
From these results an analytical model has been developed, and the results
are compared. The results are converted into useful rotordynamic data that is
easily understood by machine engineers.

The bearing has been experimentally tested in a rebuilt turbomolecular
vacuum pump up to 90,000 rpm. Bearing forces have been accurately
measured on a specially designed spring suspended scales, in which the
bearing rotor is powered with the permanent magnet drive. Comparison of
measured data with results from the 3D-FEM analysis shows excellent
agreement.


Keywords: Homopolar bearing, Electrodynamic bearing, Induction
bearing, Eddy current bearing, High-speed drive, Permanent
magnet drive, Halbach rotor, Flywheel for energy storage

I

II

Preface

This work was initiated in 1997 in order to investigate an invention made by
the author two years earlier in the field of electrodynamic levitation. Having
finished his M.Sc thesis [5] in 1990 on the design of magnetic induction
bearings, the author continued to develop these bearings, also known as
electrodynamic bearings, for SKF and later also for the German vacuum
industry until 1994. However, though the bearings met the specifications set
by the application engineers, the losses generated by these bearings were still
too high to meet the requirements from a broader market, and it was also felt
that they could not be reduced to acceptable levels solely by optimizing the
bearing parameters. A completely new approach was needed; a development
task that the author initially carried out from his newly founded company
Magnetal.

The homopolar induction bearing saw the crack of dawn in 1995. It has a
simple geometric design that promises very low losses, but it soon turned out
that the bearing was remarkably difficult to analyze. Nothing similar had ever
been published, and the computational tools required to simulate the bearing
had barely been developed.

Professor Dave Rogers at the University of Bath was contacted, who at the
time was developing a 3D-FEM software. He and his staff delivered a
version of the software, which they had optimized especially for this
application. Thanks to the software it was, for the first time ever, possible to
visualize how the currents are induced in the bearing, and from these results
the first steps towards an analytical model could be taken.

At that time also Prof. Chandur Sadarangani at the Royal Institute of
Technology in Stockholm was involved, and the current research project
started at his department, the department of Electrical Machines and Power
Electronics, in 1997. Professor Sadarangani has a long experience in eddy
current calculations regarding different types of electrical machines, and he
and his staff has contributed to a better analytical understanding of the
bearing.

Two test rigs were made to verify the analytical and computer simulated
results. The conformity of the results is striking. However, during the
experimental phase of the project, it was discovered that the existing
rotordynamic theories available did not correctly take eddy current losses and
its effect on stability into consideration, which caused difficulties to interpret
some of the dynamic effects of the first test rig. Nevertheless, several test
runs up to 90 000 rpm were performed with this rig. The second rig was
made to enable very accurate measurements at lower speeds. Low speed
involves much interesting physics, but is difficult to simulate in the
III

computer, since it causes instability in the Newton-Raphson equation solver.
Thus low speed measurements bring additional scientific values.

The original contributions of this work to the public knowledge on magnetic
bearings are believed to be:

• The development of a low loss homopolar electrodynamic bearing
concept with cylindrical conductors for high-speed operation.

• Parametric optimization using 3D-FEM simulations of the bearing.

• Development of a simplified analytical model of the bearing, which
explains the fundamental physics, and which is accurate enough to
comply with FEM data.

• Measurements of bearing forces and losses for various magnet
configurations.

• Development of a high-speed airgap BLDC motor with low radial
forces.

• Better understanding of rotordynamic influence from eddy currents
taking both resistance and inductance into consideration.



IV


Acknowledgement

First of all I would like to express my deepest gratitude to my supervisor
Prof. Chandur Sadarangani at the department of Electrical Machines and
Power Electronics at the Royal Institute of Technology in Stockholm, who
has guided me through this long and challenging project.

I gratefully acknowledge the financial support by the Swedish Business
Development Agency in Stockholm (NUTEK) and by Magnetal AB in
Uppsala.

Without the help and kindness from Prof. Dave Roger and Prof. Roger Hill-
Cottingham from the University of Bath regarding the 3D-FEM simulations,
I doubt it would have been possible to find ways to simulate this novel
bearing concept.

Prof. Hans-Peter Nee has been of great help in cross checking my theories.
He also helped me to bring understanding to some of the complicated 3D
effects, which I gladly, and without comment, will neglect in this report.

Many thanks to Mats Leksell, my roommate who helped me to interpret the
induction bearing in terms of induction generator terminology. He has also
encouraged me during the project, and has taught me the way of life in his
introduction movie "Dante Doktorand" for new employees.

Cheerful and encouraging comments were provided en masse by Peter
Bennich, who as a power quality researcher immediately saw the possibilities
with magnetically levitated flywheels.

Marcus Granström at Magnetal has been of great help in making the 3D-
CAD generated pictures of the bearings. Per Uselius at Magnetal and Jan
Olov Brännvall at the institutions laboratory, have done great jobs in finding
mechanical solutions and putting the machines together.

A new motor was developed for the high speed test spindle, and I especially
want to thank Louis Lefevre for his thorough analysis of that motor, and to
Dag Bergkvist at Magnetal and Matthias Milde who did a great job in trying
to understand the hand written winding diagrams from the author.

Without Göte Bergh, practical things would simply not work at the
department, and it will be a great loss for all of us when he soon retires.

V