NA Digest Tuesday, December 6, 1994 Volume 94 : Issue 49

Today's Editor:
Cleve Moler
The MathWorks, Inc.

Submissions for NA Digest:

Mail to

Information about NA-NET:

Mail to


From: Cleve Moler <>
Date: Tue Dec 6 02:34:27 EST 1994
Subject: Please Keep NA Digest Articles Short

I would like to ask everyone to shorten their contributions
to the NA News Digest. I try to keep each week's edition under
1000 lines of text. Some of the items that are contributed are
longer than that by themselves.

Please try to limit your contribution to a few dozen lines.
If you have a longer announcement, summarize it and include an
e-mail or FTP address for the complete document. If it's a
meeting announcement, most of our readers are not immediately
interested in the times for the coffee breaks or the special
price at the local motel.

The text should be ASCII characters, with at most 80 characters
per line, no tabs and no Macintosh "special quotes" or other
nonstandard characters. A limited amount of Tex notation is
certainly OK, but don't assume everybody has a Tex viewer.

Contributions should be sent to:

Thanks for your support of the NA News Digest and for your
brevity in future contributions.

-- Cleve Moler
Editor, NA News Digest


From: Leonidas Xanthis <>
Date: Friday, 02 Dec 1994 18:40
Subject: Ivo Babuska Receives Honorary Doctorate

Professor Ivo Babuska of the University of Maryland was awarded
the degree of Doctor of Science `Honoris Causa' by the University
of Westminster at a ceremony that took place in London on Monday
28th of November 1994.

The citation: Professor Ivo Babuska is an internationally
distinguished applied mathematician, numerical
analyst and educator, who pioneered theoretical
and applied research in the field of reliability
of finite element methods, e.g. a posteriori
error estimation and other adaptive procedures.
He also laid the foundations and practice of the
hp-version of finite element methods. The well-known
Babuska-Brezzi condition (obtained independently by
Babuska and Brezzi) became the basic theoretical tool
for the analyses of the various finite element methods
(e.g. mixed methods) which profoundly influenced both
the theory and practice of computational mechanics.
He is a prime example of a fine mathematician also
serving engineering.

Leonidas Xanthis, Centre for Techno-Mathematics &
Scientific Computing, University of Westminster, London, England.


From: Jack Dongarra <>
Date: Fri, 2 Dec 1994 21:07:12 -0500
Subject: Second Edition of the LAPACK Users' Guide

We have just finished the 2nd edition of the LAPACK Users' Guide (LUG).
SIAM will be publishing this edition, which should be available in
hardcopy around February. The list price is $28.50 and the SIAM member
price is $22.80.

As with the first edition, the royalties from the sales of this
book are being placed in a fund to help students attend SIAM meetings
and other SIAM related activities. This fund is administered by SIAM
and qualified individuals are encouraged to write directly to SIAM
for guidelines.

SIAM has allowed us display the html version of the LUG on the Internet.
To access this through a mosaic interface use the url:


From: Donald Fredkin <>
Date: Fri, 2 Dec 1994 03:55:39 -0800 (PST)
Subject: Greengard Method

I was asked yesterday at a meeting whether multigrid (whose application to
an applied physics problem one of my graduate students had just explained)
was related to "the Greengard method". We have not heard of the Greengard
method. Can anyone answer the question and, perhaps, give us a pointer to
the Greengard method?


Donald R. Fredkin


From: Isabelle Flamant <>
Date: Tue, 29 Nov 94 14:28:14 +0100
Subject: Incomplete Bessel Function


I need to calculate the incomplete Bessel
AND Y. The form of this function is defined
by the formula,

K (x,y)= int exp(-(xt+(y/t))) * (t**(n-1)) dt
n 1

with x and and arbitrary n.
More particularly, I have to evaluate this for

Does anybody have any information on the availability
of algorithm(s) for this.

Thanks in advance

Isabelle Flamant


From: Cleve Moler <>
Date: Tue Dec 6 02:23:11 EST 1994
Subject: Pentium Division Difficulties

Status of a Hard/Software Pentium FDIV Workaround
and of a Pentium-aware MATLAB

A couple of days ago, Tim Coe, Terje Mathisen and I proposed a
workaround for the Pentium FDIV bug. This is a short report on
the status of that proposal.

We have now joined forces with Peter Tang, a computer scientist from
Argonne Laboratory who is currently visiting the Chinese University
of Hong Kong, and a team of computer scientists and engineers from
Intel Corporation. It is our intention to refine and implement the
workaround in such a way that it can be used for compilers, libraries,
and floating point intensive applications. The result will be an
efficient and reliable assembly language macro replacement for the
Pentium FDIV instruction. This macro can be inserted by software
developers into their source code or emitted by compilers.

At the MathWorks, MATLAB is providing a prototypical implementation
and test bed for the approach. It had been our intention to have
a Pentium aware version of MATLAB available by now, but the expanded
scope of the project will add another week or two to the release date.

The basis for the workaround is Coe's characterization, now confirmed
by Tang and Intel, of bit patterns that must be present in the
divisors of operand pairs that lead to the error. A quick test
of the divisor is done before each FDIV is attempted. The absence
of the bit pattern indicates that the FDIV can be done safely. The
presence of the pattern does not guarantee that the error will
occur, it is just a signal that it might. In this case, scaling
both operands by 15/16 takes the divisor out of the unsafe region
and insures that the subsequent FDIV will be fully accurate.

With this approach, it is not necessary to test the magnitude of
the residual resulting from a division. It is known a priori that
all divisions will produce fully accurate results. An additional
test can compare the result of scaled and unscaled divisions and
thus count the number of FDIV errors that would occur on an
unmodified Pentium. We will offer this test in MATLAB, but it
may be desirable to turn it off for maximum execution speed.

A simple characterization of the regions to be avoided expresses
floating point numbers in the form

(n + f)*2^e

where, using MATLAB notation, n is an integer in the range 16:31
and f is a fraction in the range [0,1-eps/2]. The at-risk
divisors occur in the five bands characterized by

n = 17, 20, 23, 26 or 29

If the test were based on this fact alone, than 5/16 of uniformly
distributed operands would be rescaled.

A more refined characterization reduces the width of the bands
and hence the number of rescalings required by checking for the
number of high order bits in f that are equal to one. A test
requiring the first six bits of f to be one reduces the frequency
of rescaling to 5/1024.

For example, the denominator in Coe's now famous ratio



3145727 = 3*2^20-1 = 23.99999237060547*2^17

In this case, n = 23 and f = 1-2^(-17). The 17 consecutive high
order ones in f make this example an instance of worst-case error.

Together with the team at Intel, we are now in the process of
refining, proving correct, implementing, testing, timing, validating
and documenting this approach. We will continue to report on our

(For background information on the Pentium FDIV problem, including
our previous postings, see World Wide Web site
or anonymous FTP site, directory

-- Cleve Moler


From: Luca Dieci <>
Date: Tue, 29 Nov 1994 11:15:13 -0500
Subject: Conference on Dynamical Systems/Numerical Analysis

First Announcement

Conference on Dynamical Systems & Numerical Analysis
Georgia Tech, Atlanta. December 14-16, 1995

This will be a three day conference on dynamical numerical
analysis. Specific topics represented include: convergence
of integration algorithms over long time intervals including
long time error bounds and convergence of invariant sets;
stability of integration algorithms over long time intervals
including preservation of dynamical structure and spurious
solutions; new approaches to error analysis, including
shadowing results and backward error analysis; design of
computational techniques for invariant sets including
algorithms to compute invariant torii, inertial manifolds,
heteroclinic orbits, homoclinic orbits, etc. etc. .

There will be no concurrent sessions. The following people
have agreed to give invited talks:

D. Aronson U. Ascher W. Beyn E. Doedel
C. Elliott C. Foias E. Hairer A. Iserles
P. Kloeden S. Larsson J. Lorenz K. Palmer
G. Raugel R. Russell J. Sanz-Serna R. Skeel
D. Stoffer A. Stuart E. Titi J. Yorke.

There will also be room for twenty-minute contributed talks.
The selection of these talks will be done so that the broadest
possible panorama of topics is represented.

We are seeking notification from all who are interested in
attending. We prefer to get responses by e-mail. If you
are interested in giving a contributed talk, please indicate
this in your response, and send a tentative title.

Sincerely, Luca Dieci and Don Estep.
School of Mathematics, Georgia Institute of Technology
Atlanta, GA, 30332 U.S.A.


From: Marcela Kraemer <>
Date: Wed, 30 Nov 94 11:17:49 +0100
Subject: Conference on Centenial Hurwitz on Stability Theory

May 21 - 26, 1995

With this conference ETH wants to honor Adolf Hurwitz who
was in Zuerich from 1892 - 1919. In 1895 his paper ``Ueber
die Bedingungen unter welchen eine Gleichung nur Wurzeln
mit negativen reelen Theilen besitzt'' appeared in print.
It represents a milestone in the development of the stability
of dynamical systems. The stability still plays a key role in
todays system theory. In the last few years there have been
major new developments especially in robust stability and
multidimensional systems.

The aim of the conference is to show in a few lectures the
historical development of the article by Hurwitz and its impact
in other fields. The major emphasis however will be on modern
development in stability theory and its application in theory
of control and numerics. In particular stability of the
following problems will be treated: linear, nonlinear,
timedependent systems, discretizations of ordinary and partial
differential equations, systems with mensional systems.
In addition rebust stability and problems related to
the stability radius will be treated.

Form of the conference

This conference will bring together about 60 people from the
various fields related to stability. There will be about
20 invited 40-minute lectures in which outstanding experts
will cover in survey lectures the history and the new
developments of stability theory and its application
especially in control theory and numerics. Other participants
may present short 20-minute talks.

Since the number of participants is limited, participation
is on invitation only. However, potential participants should
not hesitate to contact one of the organizers, specifying
their area of research and indicating whether they would like
to give a talk, and if so, on which subject.

The meeting takes place in a small conference centre on
the Monte Verita, near Locarno and Ascona, in the Italian
speaking part of Switzerland. Lodging and meals for the
participants are also provided in the conference centre,
which is run by ETH Zuerich.

Invited Speakers

B. Anderson (ANU Canberra)
B. Barmish (University of Visconsin-Madison, Madison)
R. W. Brockett(Harvard University)
G. Collins (University of Linz)
D. Hinrichsen (University of Bremen)
E. Jury (University of Miami)
R. Jeltsch (ETH Zuerich)
T. Kailath (University Stanford)
H. Kimura (University Osaka)
V. Kharitonov (State University St. Petersburg)
M. Mansour (ETH Zuerich)
R. Tempo (CENS-CNR Torino)
P. M. Van Dooren (University of Illinois, Urbana)
M. Vidyasagar (Center for Artificial Intelligence and Robotics Bangalore)
G. Zames (McGill University Montreal)


Prof. Dr. Rolf Jeltsch
Seminar fuer Angewandte Mathematik
ETH Zentrum, HG G57.3
CH-8092 Zuerich
Telephone +41 1 632 3452
FAX +41 1 632 1085

Prof. Dr. Mohamed Mansour
Institut fuer Automatik
ETH Zentrum
CH-8092 Zuerich
FAX +41 1 632 1211


From: Bette Byrne <>
Date: Wed, 30 Nov 1994 10:44:58
Subject: ICFD Conference

This is a reminder that the closing date for the ICFD Conference on
Numerical Methods for Fluid Dynamics is Friday, 9th December, 1994.

3 - 6 APRIL 1995

Organising Committee

Professor K W Morton (Oxford)
Dr M J Baines (Reading)
Dr M J P Cullen (Meteorolgical Office)
Dr C L Farmer (INTERA)
Dr M G Hall (DRA, Farnborough)

Invited Speakers include:

H Deconinck (VKI, Belgium)
J C R Hunt (Meteorological Office)
M A Leschziner (UMIST)
D Mavriplis (ICASE, NASA Langley)
K W Morton (Oxford)
K G Powell (Michigan)
A Quarteroni (Milan)
S Rill (Deutsche Aerospace Airbus GmbH, Germany)
A N Staniforth (Quebec)
B Stoufflet (Avions Marcel Dassault, France)
J Thomas (NASA Langley)
N P Weatherill (Swansea)
P Wesseling (Delft)

This is the fourth international conference on CFD
organised by the ICFD (Institute for Computational Fluid
Dynamics), a joint research organisation at the
Universities of Oxford and Reading set up in 1983 with
the support of the SERC to collaborate with industry in
this area. Previous conferences were held at the
University of Reading in 1985 and 1992 and at the
University of Oxford in 1988 in a series of conferences
on the same theme held on the two sites over the last
thirteen years.

Abstracts and enquiries regarding the conference should
be addressed to:

Mrs B Byrne
ICFD Secretary
Oxford University Computing Laboratory
Wolfson Building
Parks Road
Oxford OX1 3QD

Telephone: 0865 273883
Facsimile: 0865 273839


From: Graham Carey <>
Date: Wed, 30 Nov 1994 07:43:04 -0600
Subject: Advances and Trends in Computational and Applied Mathematics

The Texas Institute for Computational and Applied Mathematics (TICAM)
announces a
To be held at

April 20-22, 1995


A two-and-one-half day symposium on recent advances and trends in the broad
area of computational and applied mathematics (CAM) is to be held at The
University of Texas at Austin during the third week of April, 1995. The
meeting will bring together leading researchers and practitioners in
Computational and Applied Mathematics from a world-wide community of
scholars and researchers in this area. The purpose of the meeting will be
to bring to the forefront the most recent advances in mathematics applied
to computation, in numerical algorithms and computational mathematics, in
parallel processing and supercomputing, and in applications to key fields
in science and engineering, including in particular computational fluid
dynamics, semiconductor modeling, acoustics and scattering, electromagnetic
field theory, nonlinear dynamics, numerical relativity, nonlinear control
theory, nonlinear solid mechanics, fluid structure interactions,
environmental modeling, reservoir simulation, and other areas. The meeting
will feature invited lectures from key scientists together with organized
sessions of contributed papers.

CALL FOR PAPERS: Persons wishing to contribute a paper for consideration
at the meeting should send a one-page (single-spaced) abstract to the
Organizing Committee prior to January 1, 1995.

Send by email to

or US mail to:

Symposium on Computational and Applied Mathematics I
Texas Institute for Computational and Applied Mathematics
TAY 2.400
The University of Texas at Austin
Austin, TX 78712

Pat Bozeman/G. F. Carey
ASE/EM Dept., WRW 301
The University of Texas at Austin
Austin, TX 78712


From: Weimin Han <>
Date: Thu, 1 Dec 1994 16:11:08 -0600
Subject: FTP Site at University of Iowa

Reports on Computational Mathematics from the University of Iowa
are now available via anonymous ftp.

To get there type

Login as anonymous, give your e-mail address as password, and then
cd pub/comp_math_rep

Now you will find index files (index.tex, index.dvi and with
a list of the reports on Computational Mathematics, and a collection
of files . If you are interested in one of these
postscript files and have a postscript printer, you can download
the file, uncompress it using
and print.

Some of the reports are not available from the ftp site. If you
are interested in such a report, you can request one by sending
a message to one of us:
K. E. Atkinson, e-mail address:
W. Han, e-mail address:
F. A. Potra, e-mail address:
E. Venturino, e-mail address:

Best regards,

Weimin Han
Department of Mathematics
University of Iowa
Iowa City, IA 52242


From: Jerzy Wasniewski <>
Date: Fri, 2 Dec 1994 15:29:06 +0200 (METDST)
Subject: Copenhagen PARA94 Proceedings

The Copenhagen Workshop on Parallel Scientific Computing, PARA94
Proceedings has already been printed as a volume 879 of the Lecture
Notes in Computer Science series of Springer-Verlag. The list price
is DM 104,00. Every PARA94 speaker will receive one free copy.
Authors or editors of Springer books, as well as any contributing
authors, for example to LNCS proceedings (PARA94 proceedings), are
entitled to purchase this or any other book published by
Springer-Verlag for personal use at a discount of 33 1/3 % directly
from the Springer -Verlag company. Orders are to be sent to
Springer-Verlag, Vertrieb, Postfach 31 13 40, D-10643 Berlin, Fax:
+49 30 8207 301. A cheque covering two thirds of the list price plus
DM 5,00 for postage has to be enclosed to the order or a credit card
number has to be given; only Eurocard, Visa, Diners or American
Express are accepted.


From: D.F.Griffiths <>
Date: Fri, 2 Dec 94 18:33:12 GMT
Subject: Dundee NA Conference 1995


Tuesday 27 June - Friday 30 June, 1995

Invited Speakers

The 16th Biennial Conference on Numerical Analysis will be held at the
University of Dundee from Tuesday 27 June - Friday 30 June, 1995

The special invited lecture in honour of A. R. Mitchell will be
presented by

Professor K. W. Morton

The other Principal Speakers will include

G. H. Golub A. Griewank D. F. Griffiths
R. Jeltsch C. Johnson W. A. Light
B. N. Parlett L. Petzold R. D. Russell
J. M. Sanz-Serna D. J. Silvester E. S u li
M. H. Wright J. Zowe

Leslie Fox Prize

The conference will be preceded by a one day meeting on Monday 26 June
at which talks will be given by those shortlisted for the Leslie Fox
Prize. Please indicate on the registration form whether you plan to
attend. A charge of 7 pounds will be made to cover lunch.

Address for correspondence

Dr D. F. Griffiths
Numerical Analysis Conference
Department of Mathematics and Computer Science
The University
Dundee DD1 4HN
Scotland, UK

Telephone: +44 (1382) 344467/344471
FAX : +44 (1382) 345516


From: Gene Golub <golub@sccm.Stanford.EDU>
Date: Mon, 28 Nov 94 19:30:31 PST
Subject: Position at Stanford


Faculty Openings

Stanford University's Departments of Computer Science and Electrical
Engineering seek applicants for a tenure track faculty position in
software systems and tools. There is particular interest in applicants
who have designed, built, and/or experimented with large software
systems or software tools that support engineering tasks. Example
tasks include parallel computing, computational modeling and
prototyping, scientific computing, rapid prototyping, embedded system
design and implementation, and real-time system design and

Applicants should have a Ph.D. in a relevant field, and should have a
strong interest in both teaching and research. The appointment will
be made at the level of Assistant Professor. The successful candidate
will be expected to teach courses, both in the candidate's specialty
area and in related subjects, and to build and lead a team of graduate
students in Ph.D. research. Stanford University is an equal
opportunity employer and welcomes nomination of women and minority
group members and applications from them. Applications, including a
resume, a publications list and the names of five references, should
be sent by March 1, 1995, to:

Professor Hector Garcia-Molina
Search Committee Chairman
Computer Systems Laboratory, ERL 444
Stanford University
Stanford, CA 94305-4055


From: <>
Date: Wed, 30 Nov 1994 18:08:10 -0500 (EST)
Subject: Position at University of Maryland, Baltimore County

Department of Mathematics and Statistics

We invite applications for an anticipated visiting or tenure-track faculty
position in Numerical Analysis starting Fall 1995, subject to availability of
resources and administrative approval. The position is at the rank of assistant
professor; in exceptional cases, however, a more advanced appointment may be
considered. Excellence in both research and teaching is essential, and a
doctorate is required. Only candidates in the area of NUMERICAL ANALYSIS FOR

Applicants should arrange for a curriculum vitae, (p)reprints and three letters
of recommendation to be sent to:
Numerical Analysis Search Committee
Department of Mathematics and Statistics
University of Maryland Baltimore County
Baltimore, MD 21228-5398

Please submit applications by February 15, 1995. Women and minority candidates
are encouraged to apply. UMBC is an Affirmative Action/Equal Opportunity


From: Kazufumi Ito <>
Date: Thu, 1 Dec 1994 14:11:21 -0500
Subject: Postdoctoral Position at North Carolina State

North Carolina State University
Center for Research in Scientific Computation

The Center for Research in Scientific Computation at N.C. State University
in collaboration with the Lord Corporation's Thomas Lord Research Center,
Cary, N.C., expects to make a University/Industry Cooperative Postdoctoral
Research appointment starting August 16, 1995 (availability of the position
is contingent upon funding). The appointment will be in the area of applied
mathematics and scientific computation. The successful candidate for this
position is expected to participate in a collaborative multidisciplinary
team carrying out fundamental research investigations to provide a better
understanding and predictive capability of the dynamic mechanical behavior
of rubber-like polymer composites (elastomers). The research efforts will
involve the modeling of nonlinear dynamics for viscoelastic composite
materials subject to large deformations, development of computational
methods for both forward and inverse problem analytic studies, and design
and performance of experiments to aid in model verification and validation.
Since the research efforts will require physical modeling, theoretical
analysis and computational skills, candidates who are outstanding in at
least one of these areas and willing and able to learn quickly in the
others will be given highest priority. This position offers a unique
opportunity for multidisciplinary mentored post-doctoral research on a
mathematical project arising in an industrial/university collaborative
Applicants should send a vita and brief description of research interests
and have three letters of recommendation sent to Professor H.T. Banks,
Center for Research in Scientific Computation/Dept of Mathematics, Box
8205, NC State University, Raleigh, NC 27695-8205. Applications will be
considered anytime after Feburary 15, 1995, as funding becomes available.
NCSU is an AA/EOE. In its commitment to diversity and equity, NCSU and the
CRSC seeks applications especially from women, minorities, and the


From: Dave Dougherty <>
Date: Thu, 1 Dec 1994 15:31:53 -0500 (EST)
Subject: Position at University of Vermont


A non-tenure-track Research Assistant Professor position is available,
contingent upon funding, within the Department of Civil & Environmental
Engineering at the University of Vermont. A Ph.D. is required.
Candidates must demonstrate expertise in all of the following areas:
groundwater hydrology, subsurface characterization, remediation design,
artificial neural networks for characterization and inverse problems,
parallel computing, and optimization. Experience with a variety of
computing environments is needed. Highly developed written and
verbal communication skills are essential. Job requirements include
mentoring of graduate students, developing and sustaining an
outstanding research program, and interaction with other tenure-stream
and research faculty. Submit letter of interest, c.v.,
and the names of three references to CEE Research Assistant Professor
Search Committee, c/o Dawn Densmore, 109 Votey Building, College
of Engineering & Mathematics, University of Vermont, Burlington,
VT 05405. UVM is an equal opportunity/affirmative action
employer. Women and minorities are strongly encouraged to apply.
Applications received after December 15, 1994 may not receive
full consideration.


From: Richard Brent <>
Date: Fri, 2 Dec 94 17:13:07 EST
Subject: Research Fellowship at Australian National University

Australian National University

Cooperative Research Centre for
Advanced Computational Systems

Research Fellow or Fellow
(Level B or C)

REF: ACS 23.11.1.

The Cooperative Research Centre for Advanced Computational Systems (ACSys)
has been established under the Australian Government's CRC Program with the
mission of establishing an internationally competitive research capability
in the area of high performance advanced computational systems.
Facilities include access to a 128-cell Fujitsu AP1000, a 32-processor
Connection Machine (CM5), and a Fujitsu VP2200/10 vector processor.

The Numerical Algorithms Project is one of the research projects supported
by ACSys. Its aim is to develop state-of-the-art vector and parallel
algorithms for the solution of scientific and engineering problems,
and to implement the algorithms in high-quality transportable code.
The Project Leaders are Professors R. P. Brent and M. R. Osborne.

Applications are invited for a fulltime research position as Research Fellow
(academic level B) or Fellow (academic level C) in the Numerical Algorithms
Project of ACSys. Applicants should have a strong research record and interests
in one or more of the following areas:

vector/parallel algorithms for the solution of large
linear systems, eigenvalue problems, fast Fourier and related transforms,
differential equations, or optimisation;
vector/parallel algorithms for exact numerical computation;
parallel algorithms for combinatorial problems.

Enquiries to: Professor R. P. Brent
phone 61-6-2493329
FAX 61-6-2494948

Closing date: 16 December 1994 (may be extended)

Salary: In the range Research Fellow (level B) $A41,574 - $A49,370 p.a.
or Fellow (level C) $A50,928 - $A58,724, depending on qualifications and
experience. A market loading may be applicable.

Appointment: Appointment will be for a fixed period of up to three (3) years.
Applicants wishing to undertake a period of secondment from another position
will be considered.

Applications: Applications should quote the reference number, include names
and addresses (including FAX or e-mail) of at least three referees, and be
addressed to:

The Secretary
The Australian National University
Canberra, ACT 0200, Australia

Assessment will be against selection criteria. Applicants are advised to
obtain a copy of the selection documentation and to frame their application
accordingly. Receipt of applications will not be acknowledged unless requested.
Applications should be in duplicate and include curriculum vitae and a list of

The selection criteria are available by ftp from


From: Mario Fafard <>
Date: Mon, 5 Dec 94 15:35:43 EST
Subject: Position at Laval University, Quebec


Applications are invited for a professional research position in the area
computational mechanics with object oriented programming paradigms.

The candidate will join a team of researchers and graduate students in the
GIREF (Groupe Interdisciplinaire de Recherche en Elements Finis)
research group in finite element. The project consists in developping a
finite element shell based on Object Oriented Programming in C++ in the
following field: Fluid and Solid Mechanics and Heat Transfer. The main
tasks are:

*Definition of objects, classes and sub-classes, inheritances, data
encapsulation and polymorphism with the researchers of the GIREF
*Definition of the data structure and the object manager
*Defining and programming the finite element shell
*Writting all documentation relative to the development procedures:
proceeding of the meetings, technical documentation, development
schedules, meeting convocation, research funds applications , etc.

The candidat must hold a bachelor degree in computer sciences or a Ph.D.
or a master's degree in enginering or in applied mathematics with
experience in engineering softwares and should be knowledgeable in the
area of Finite Elements, Object Oriented Programming and computer
programming in C++.

Starting date:
February 1995

The initial appointment will be for one year with a possibility of renewal
for two more years.

Commensurate with qualifications and experience.

Send complete curriculum vitae along with research experience, list of
publications and three references by Jannuary 31 1995, to:

Mario Fafard, Ph.D., ing.
CCivil Eng. dept.
Laval University
Quebec, Canada
G1K 7P4
Fax: 418-656-2928


From: Weizhong Dai <>
Date: Mon, 5 Dec 1994 15:23:05 -0600 (CST)
Subject: Position at Louisiana Tech University

Louisiana Tech University has recently received approval to establish a
Ph. D. Program in Applied Computational Analysis and Modeling. This
program is a broadly based interdisciplinary Ph. D. program for students
desiring to study mathematics, computer science and a companion science
or engineering discipline so that they can apply their computational
skills to solve scientific problems.

In association with this program, applications are invited for a senior
level position in the Department of Mathematics and Statistics. We are
seeking an applied mathematician with a national reputation and a proven
grant record whose research interests are in numerical analysis (with
emphasis on numerical solutions for differential equations, modeling and
computational science). It is preferred that the candidate has
experience in applications related to fields other than mathematics, such
as scientific or industrial problems. To apply send a resume and three
letters of recommendation to: R.J. Greechie, Head, Department of
Mathematics and Statistics, Louisiana Tech University, Ruston, LA 71272.
Nominations and suggestions are welcome.

The screening of applications has begun and will continue until the
position is filled. Louisiana Tech University is an Equal
Opportunity/Affirmative Action Employer. We are interested in receiving
applications from qualified women and minorities.


From: Ken McKinnon <>
Date: Mon, 5 Dec 94 20:54:37 GMT
Subject: Positions at University of Edinburgh

Three positions at University of Edinburgh

Lecturership in Mathematics (ref 490564)
Lecturer/Computing Officer (ref 490565)
Research position in Parallel Linear Programming (ref 490572)

Applications are invited for a new Lectureship in Mathematics tenable
from 1 September 1995 or earlier if this can be agreed. Preference will
be given to candidates with research interests in analysis (including
stochastic analysis), geometry, numerical analysis, analytical or
numerical aspects of partial differential equations, applied
differential equations or related areas. The position is tenable for
five years, but with the possibility of a permanent contract for an
exceptionally well qualified candidate.

Applications are also invited for a position as Lecturer/Computing
Officer commencing as soon as can be agreed. The position will entail
joint responsibilities for teaching and research and (with other
Computing Officers) the management and development of the Department's
computer installations. Candidates should have research interests in an
area of mathematics or statistics which makes significant use of
computation, and have practical experience in the running of computer
systems. The position is tenable for five years in the first instance.

Salary for both positions will normally be on the Lecturer Grade A scale
(14,756-19,326 pounds) or possibly on the Lecturer Grade B scale
(20,133-25,735 pounds)

Informal enquiries for the above posts may be addressed to Dr J Martin,
Department of Mathematics and Statistics, The King's Buildings,
Edinburgh EH9 3JZ, tel (44) 131-650-5062, email

A research fellow is required to start work as soon as possible within a
group developing methods for linear programming on parallel computers.
The work of the group is focused on the revised simplex method and
decomposition methods together with solution techniques for unsymmetric
sparse linear systems. The group uses computers, in particular the
Cray T3D, in the Edinburgh Parallel Computing Centre, which is the UK
national centre. Applicants should have a strong background, preferably
to PhD level, in mathematics, operational research, computer science or
some other numerate discipline. Salary will be on the scale 13941 -
16191 pounds per annum according to age and experience and the contract
will be for between 26 and 23 months.

Informal enquiries may be addressed to Dr K McKinnon, email

Three copies of applications (one copy from overseas candidates),
including curriculum vitae and the names and addresses of three academic
referees, quoting the appropriate reference number should be addressed
to the Personnel Office, University of Edinburgh, 1 Roxburgh Street,
Edinburgh EH8 9TB. Further particulars may be obtained from this
address or from the Departmental WWW home page

Closing date: 20 January 1995.


From: Iain Duff <>
Date: Thu, 1 Dec 94 09:59:43 GMT
Subject: Contents, IMA Journal of Numerical Analysis

Volume 15, Number 1. January 1995.

Janovsky V and Werner B
Constructive analysis of Takens-Bogdanov points with $Z_2$-symmetry.

Schecter S
Rate of convergence of numerical approximations to homoclinic
bifurcation points.

Jodar L and Ponsoda E
Non-autonomous Riccati-type matrix differential equations: existence
interval, construction of continuous numerical solutions and error bounds.

Fabiano R H, Knobel R, and Lowe B D
A finite-difference algorithm for an inverse Sturm-Liouville problem.

Miller J J H, O'Riordan E, and Shishkin G I
On piecewise-uniform meshes for upwind- and central-difference
operators for solving singularly perturbed problems.

Garcia-Archilla B and Mackenzie J A
Analysis of a supraconvergent cell vertex finite-volume method for
one-dimensional convection-diffusion problems.

Sun G and Stynes M
Finite-element methods for singularly perturbed high-order elliptic
two-point boundary value problems. I: reaction-diffusion-type problems.

Chu M T and Wright J W
The educational testing problem revisited.


From: Eduardo Sontag <>
Date: Thu, 1 Dec 94 16:00:31 EST
Subject: Contents, Math of Control, Signals and Systems

TABLE OF CONTENTS, Math of Control, Signals, and Systems
Volume 7, Number 1

B. d'Andrea Novel, F. Boustany, F. Conrad, and B.P. Rao,
Feedback stabilization of a hybrid PDE-ODE system: Application to an overhead
pp. 1-22.

George Weiss,
Regular linear systems with feedback,
pp. 23-57.

Richard M. Murray,
Nilpotent bases for a class of non-integrable distributions with applications
to trajectory generation for nonholonomic systems
pp. 58-75.

Takao Nambu,
Approximation algorithm for the infinite-dimensional Lyapunov equation,
pp. 76-93.


End of NA Digest