(1) Panel Comment: On the basis of its potential scientific
importance, the panel ranked this proposal extremely highly. However,
the panel did not feel competent to judge the likelihood that the
research could or would be completed successfully. The reason for
this is that the proposal is mainly about a laboratory MHD experiment
and no one on the panel has any direct experience with such
experiments. The panel therefore recommends that the NSF program
director seek reviews from several researchers with the relevant
experimental expertise. If they conclude that the experiment is
feasible and likely to be completed successfully by the PI and Co-Is,
then the panel believes this proposal would be among the very best it
has reviewed and should be funded by the NSF.
Reply to (1): The PI has spent his earlier career as an
experimentalist, initially in nuclear weapons testing followed by a
large number of liquid sodium MHD experiments establishing MHD for
fusion plasma confinement. The observation of the tearing in a liquid
sodium fountain led to elucidation of the "tearing mode". Later, many
fusion plasma experiments demonstrated the lack of confinement of
tearing unstable pinches. (This plasma physics of "flux conversion" is
now central to the question of radio lobe evolution. ) However, at
the time there were "pleas" by many theorists to construct and test a
liquid sodium dynamo (particularly Johnny Wheeler). The minimum size
required appeared to be many 10's of meters requiring many thousands
of kW of power. Even for LLNL this was too much. However, in the late
80's with Greg Willette, we recognized the importance of the
rotationally coherent nature of axially displaced plumes in a rotating
frame. This source of coherent helicity has made this experiment
feasible.
(2) Reviewer #1 Concerns: On the "minus" side, I worry
about whether there is enough experimental "depth" on this team to
pull off the experiment. All in all, I would trust Colgate's
considerable expertise and his LANL collaborators to bring other
scientists from LANL into this as appropriate.
Reply to (2): The experimental tradition at NMIMT is
extensive, especially in observational and experimental atmospheric
physics. Please see the NMIMT web site at http://www.nmt.edu/. If it could
have been funded at LANL, it would have cost more than 10 times as
much and take many times longer.
(3) Reviewer #2 Concerns: There is no evidence that they
can get such a high flow speed, up to 100 m/s, thus high magnetic
Reynolds number product of 4000 to achieve self-excitation (critical
Rm of 1000). The power needed to generate flow speed in a turbulent
state should be proportional to cubic of flow speed, instead of speed
squared as used in the proposal. The nominal achievable speed is 10
m/s, a factor of 10 smaller than the estimate, therefore, there is no
evidence that the goal of self-excitation can be achieved for the
given driving power of 50 kW.
Reply to (3): There is an extensive engineering analysis in
the year 2000 proposal in response to this question. The Couette flow
is only very weakly turbulent, which is why it is chosen. The
pressure curves are now part of this proposal relative to the MRI
investigations, but were known initially for the original design. The
present apparatus is twice the dimension of the size considered in
this first proposal and was chosen on the basis of ease of
construction, lower technology, lower pressures for a given magnetic
Reynolds number.
(4) Reviewer #3 Comments: I am not competent to judge the
engineering aspects of the proposed experiment. Achieving the 1-2 x
104 cm/sec with the associated rotating plumes is the key to
success. It will not be easy, but the proposers have given
considerable thought to the problem, and there is no reason to think
that their ingenuity will not be successful.
Reply to (4): The size of the experiment has been doubled
for ease of construction, reduced stresses and technology.
(1) Panel Comment: The external expert referees raise
critical questions about its (the experiment's) safety and engineering
feasibility. We wonder whether this experiment can be realized
without a greatly increased investment in funds, technology and
engineering. Funds might be spent in engineering rather than a Co-I
and graduate student who are not hardware experts.
Reply to (1): The experiment has been designed by physicists
in conjunction with engineering advice. The physics considerations
have been paramount. The original proposal was funded at 100k.
Roughly the same has been spent in addition since then. The apparatus
to test Phase I, the Ω-flow, is nearing completion as shown in
Figs. 1 and 2.
(2) Reviewer #1 Comment: There is little past track record
for this type of experiment. Perhaps a scale-down version of this
experiment should first be used to demonstrate its feasibility and
test the diagnostics.
Reply to (2): As discussed in the proposal there are three
sodium dynamo experiments in this country, and two successful ones in
Riga, Latvia and in Karlsruhe, Germany. The later experiments are
up to 10 times the proposed experiment size.
(3) Reviewer #1 Comment: Safety issue is only lightly
addressed.
Reply to (3): An extensive safety report is posted on our
project web site; please see http://physics.nmt.edu/~dynamo/. A panel
member of the 2001 proposal reviewed this with a positive comment; see
first reviewer's summary statement, AST-0098474.
The experiment will be performed by remote operation, at ~100 m
distance, in the open without building confinement. The experiment
will be performed at the NMIMT high-explosive testing grounds, the
Energetic Materials Research and Testing Center (EMRTC). More than
20,000 high explosive tests have been safely performed at EMRTC in the
last 55 years. This is a major national high explosive testing
facility. The safety practices and procedures are reviewed and
overseen by both the EMRTC engineers and the NMIMT Research Division's
Hazardous Materials and Safety Officer.
(4) Reviewer #2 Comment: The proposal makes little mention
of sodium handling and safety. I appreciate that one of the proposers
have worked with liquid sodium before (but at a time before much of
the present liquid metal technology had developed), and that the
liquid sodium volume is small. Nonetheless, the proposal leaves me
curious as to how they intend to proceed.
Reply to (4): This is discussed in the safety section of our
project's web site: Safety Practices, Procedures and Compliance.
Please see http://physics.nmt.edu/~dynamo/.
(5) Reviewer #2 Question: How the vessel is charged and
emptied of sodium in order to service the magnetic probe arrays?
Reply to (5): Neither of the two magnetic probe arrays
require emptying and recharging the vessel for service. There are two
magnetic probe arrays associated with the experiment: a 2-D matrix
array and a radially-aligned linear array. The matrix array
is mounted on the external surface of the driven end-wall of the
apparatus, and thus it is directly serviceable. The radially-aligned
linear array is inserted into, and removed from, a radial probe
housing that is itself sealed into the outer cylinder. The radial
probe housing is a hollow aerodynamic "vane", and remains in place
whether the sodium is liquefied or solid. Thus the linear array is
serviceable as required, independent of the sodium.
(6) Reviewer #2 Question: How cold spots are eliminated
during the charging given only the inner cylinder is heated?
Reply to (6): The thermal time constant of the sodium when
solid is about 5 minutes. When liquid, the convection time is much
shorter. In addition the external shield can be electrically heated
if necessary.
(7) Reviewer #2 Question: How catastrophic would
a vessel rupture be, even with the safety shields?
Reply to (7): 140 kg of liquid sodium sprayed into the
atmosphere at operating pressure and velocity will make a fine
particle cloud that may or may not be able to be ignited at a
temperature of ~110 C. If it burns, it will be about the same as
the equivalent mass of gasoline or jet fuel. This is why the
experiment is performed unconfined outside, not in a building. The
experiment is to be remotely operated.
(8) Reviewer #2 Question: What strategy is proposed for the
sliding seal through which the titanium piston is pushed. If this is
simple packing material, are the pressures there small enough to
eliminate a significant sodium leak?
Reply to (8): The titanium piston itself is guided on its
outer surface via the bore of the inner reservoir plenum cylinder and
internally by a linear bearing and high-temperature hydraulic seal
sliding on the volume-compensating shaft. There is no seal on the
piston's outer surface, but the by-pass flow should be negligibly
small (clearance area / piston area). The volume-compensating shaft
allows for the internal volume of the apparatus to remain constant
while the piston extends and retracts in the source and sink
reservoirs. The piston rod (hydraulic shaft), titanium also, is
guided through linear bearings and high-temperature hydraulic seals.
The hydraulic seals are capable of restraining pressures in excess of
3000 psi (~3x actuation and operating pressures) and
withstanding temperatures of 150 C.
The axial bearings and seals will be in an oil environment. A finite
amount, ~1 liter, of mineral oil will intentionally be left in
the apparatus when filled with sodium metal. Due to the different
specific gravity of mineral oil vs sodium metal, ~0.81 vs
~0.93 at 100 C, when the apparatus in spun up to the very low
rotation rate of 42 rpm, the mineral oil will centrifugally float to
the region of the rotational axis relative to the sodium. Thus the
seals and bearings internal to the apparatus will operate in a hot
mineral oil environment, not a liquid sodium environment. After an
experimental run, the apparatus will be cooled while rotating so that
the sodium will solidify with the mineral oil retained at the
axis. Thus the only time that the seals could come in contact with the
sodium is under the static configuration during the filling or
draining process. In these circumstances the pressure will be small
and would only be due to gravity acting on the liquid sodium. This
means that all seals are oil seals, NOT sodium seals. These
components are therefore standard high-temperature industrial
components.
The piston that drives the flow for the pulsed plumes will be used in
Phase II. Phase I of the experiment, the first year of this proposal,
will not use the jets and hence will not use the piston.
(9) Reviewer #3 Concerns: Sodium safety; a) caustic smoke
ventilation in case of fire, b) fire suppression of metal fires, c)
sodium handling, d) safety training of personnel/students.
Reply to (9): These facets are all discussed in the safety
section of our project's web site: Safety Practices, Procedures and
Compliance. Please see physics.nmt.edu/~dynamo/
(10) Reviewer #3 Concerns: Stress analysis; The proposed
experiment would be pushed to an internal stress in the stainless
steel of 30 ksi. Although such a design is possible, one should not
undertake this without a more detailed (numerical) stress analysis of
the design. The presented back-of-the-envelope calculations are not
suitable design criterion.
Reply to (10): The size of the apparatus has been doubled
and the vessel material has become aluminum alloy. The maximum stress
currently expected is 10 ksi compared to a limiting strength of
40 ksi. The vessel will be tested hydrostatically and the apparatus
will have integrity verification measurements performed with mineral
oil under rotational operating conditions.
(11) Reviewer #3 Concerns: Seals; The seals which hold the
sodium into apparatus between differentially rotating parts are a
concern.
Reply to (11): See above for discussion of oil immersion of
all seals and bearings; Sec. 1.2, Reply to (8), paragraph 2.
(12) Reviewer #3 Concerns: Cooling; The proposed maximum
mechanical power needed is 40 kW. This mechanical energy is converted
to heat by the actions of viscosity and possibly Joule heating.
Reply to (12): Experimental runs or tests at or near high
speed will be of the order of a few minutes duration. The total time
at the near maximum speed will be no more than a few minutes also.
The change of temperature during a test will be small, a few degrees
at maximum speed (~1 degree per minute at 40kW). The oil
circulation can be used as cooling if necessary, but we expect windage
heat loss to be larger than internal friction heat gain and therefore
have provided the opportunity for heating in the surrounding safety
shield. We may require greater power and the current motor can be
used to double capacity for the short periods of the experiment.
(13) Reviewer #3 Concerns: Rotating electronics;
Substantial rotating analog and digital electronics appear to be
proposed. These electronics are needed to survive 100 Hz rotation
leading to a 2000g acceleration and a temperature of 100
deg. C. Standard laboratory electronic techniques are unsuitable.
Discussion is needed about what Mil. spec. techniques are
available.
Reply to (13): Due to the doubling of the apparatus size,
the the outer cylinder assembly, low-speed drive and printed circuit
board (PCB) mount rotate at 30 Hz. The acceleration at the periphery
of the ~30 cm radius apparatus is ~1000 g. The electronics
undergoing this acceleration are the analog devices consisting of Hall
effect detectors, pressure transducers and temperature sensors. The
specific examples of each that have been chosen and purchased can
readily withstand 1000 g in a sustained thermal environment of 150 C.
They are in essence ``off-the-shelf'' items, thus Mil. spec. techniques
were not required. Power and signal communication with these analog
electronics takes place via twisted pairs from the PCB. The PCB
houses the balance of the analog electronics and all the digital
electronics. These consist of amplifiers, multiplexers, A-to-D
converters, and a programmable logic device (PLD). The PCB is
~15 cm in radius, thus the acceleration at its periphery is
~500 g. All components, or component sockets are wave soldered
to the PCB. Being individually low mass, they can readily withstand
500 g. Wiring traces on the PCB are configured such that induced
currents from the magnetic field coils are minimized or nullified.
Power for the electronics are brought on board via conventional carbon
brush slip rings. Bidirectional digital communication with the
electronics takes place via capacitative (air gap) slip rings. The
PCB mounting flange thermally isolates it from the bulk of the
convective and radiative heat such that it should have a saturation
temperature of less than 55 C. The standoff vanes to which the PCB is
physically mounted are configured to pump air behind it when the
apparatus is rotating, and thus force cool it also. When static
heating is taking place, such when charging/emptying of sodium, a fan
and duct will be used to cool the electronics. (The pressure and
temperature electronics will be used to monitor the apparatus during
charging and emptying.)
(1) Panel Comment: The experimental details and the specific
way in which the experiment works were not made optimally clear in the
proposal. For example the schematic diagram was not accompanied by a
step by step outline of how the experiments are performed.
Reply to (1): Additional information is on our web site and
added to this proposal. Please see physics.nmt.edu/~dynamo/
(2) Panel Question: Under what circumstances would Beckley
continue on the project given that his participation seems essential?
Reply to (2): Howard Beckley is desiring to stay on the
project through its completion. It is intended that he will receive
his Ph.D. at the completion of Phase I, the Ω-Phase of the
experiment. He hopes to continue working on the project as a postdoc
and guide the efforts of other graduate students pursuing their
advanced degrees. He has been responsible for the day-to-day research
guidance of the physics and electrical engineering undergraduate
student's efforts on the dynamo project's instrumentation development
and looks forward to continuing this aspect. All of the requested
funding for this proposal is for Beckley's postdoc salary.
(3) Panel Recommendation: It seems that the experiment is
well on its way, however the panel recommends having an experienced
outside experimentalist review the current status and the likelihood
of success and safety of the experiment.
Reply to (3): We will continue to present the experiment at
meetings. A paper has been accepted. We are posting more details on
our web site. We welcome visitors. The PI obtained his degree in
experimental physics with Bob Wilson of Cornell. As an experimentalist
he was responsible for the fast diagnostics (neutrons and gammas) for
the major thermonuclear weapons tests of LANL and LLNL during the
Castle series of 1954, including the 12 megaton equivalent Bravo test.
He published more than 50 SRD and open literature reports of plasma
physics experiments of the "fusion pinch" program at LLNL. He
published more than 40 reports and publications of atmospheric,
geophysics, and automated astronomy experiments, 1964 through 1975,
while at NMIMT. He continued with a major experimental program at
LANL in high explosives and plasma physics.
Howard Beckley has had comparable experience as a machinist for 23
years, including foreman of a machine shop for 3 years. He has done
considerable experimental work in astronomy, astrophysics, fluid
dynamics, and thermodynamics at NMIMT for 7 years and at Sacramento
Peak and Cerro Tololo for one summer each. We feel we have the
credentials and experience to make this experiment work.
We continue to have reviews within the institution on safety. For
instance, we question the safety of any high speed liquid sodium
experiments within a confined building without the exclusion of
oxygen. The safety experience and tradition of 55 years of more than
20,000 high explosive tests makes the NMIMT's EMRTC high explosive
proving grounds one of the major HE test facilities of the country.
(4) Panel Recommendation: We recommend that it be clarified
what a smaller level of funding would still be viable to ensure the
experiment's success.
Reply to (4): We are asking for a more modest 80k per year
for this proposal, primarily to support Howard Beckley as a postdoc.
This level of funding would allow us to complete both phases, I & II,
of the experiment in three years, or just Phase I of the experiment if
funding is for only one year. We expect to have roughly equal funding
available from other funding sources, including DOE Fusion, LANL and a
Frontier Center proposal, and in addition, separate support of the
graduate and undergraduate students.
(5) Reviewer #1 Question: I do think the proposer could
have mentioned the magneto-shearing instability. Why are Rossby waves
the mode of angular momentum transport rather than magneto-shearing
instability?
Reply to (5): The magneto rotational instability (MRI) was
mentioned in a foot note as an obvious first experiment during the
Phase I experiment in the year 1999 and year 2000 proposals. This was
not emphasized because we had not performed the stability analysis in
the presence of the Eckman layer induced turbulence. This analysis is
now a part of this proposal. There were just too few of us spread too
thin. The Rossby vortex mechanism of torque transmission in a
Keplerian disk has the possibility of torque transmission in an
insulating disk and therefore before magnetic fields are generated.
This is not a problem in stellar size disks around compact objects
because the companion star can always supply the flux to the Roche
lobe overflow. The scientific question is whether the MRI can act as
a dynamo and create its own flux. We hope to address this question in
the Phase I & II experiments.
(6) Reviewer #1 Question: Why are the disks not turbulent
in AGN?
Reply to (6): So far there is no known accepted way to
produce turbulence in a Keplerian disk except the MRI. There are very
many papers that assume it, but none that prove it without the MRI.
This is why the MRI is considered so important. This requires flux
and conductivity. The Rossby vortex mechanism circumvents this
difficulty. (astro-ph/0012479)
(7) Reviewer #1 Question: Also, the Rossby waves should be
studied in the presence of magnetic fields. Are they present in the
dynamo region?
Reply to (7): The "turbulence " induced by the Eckman layer
flow stress will indeed be an ensemble of "Taylor columns" . However
the magnetic Reynolds number of these eddies, Rmag / (Rey1/2),
presumably following the law of the walls, or logarithmic profile
distribution, is too low to be significantly affected by the
turbulence.
(8) Reviewer #1 Comments: (Referring to 5,6,7) These should
be addressed in the associated theoretical work, but this is not the
main use of funding requested. We do not have experimental evidence
for magneto-shearing instabilities, only numerical evidence - perhaps
the experiment itself could test for it.
Reply to (8): Now a major thrust of this proposal.
(9) Reviewer #2 Concerns: The proposal also suggests that
the experiment is directly relevant to dynamos in astrophysical
settings, particularly accretion disks onto compact objects. I have
several misgivings about this conclusion, which bear on the
astrophysical relevance of the experiment (at least for compact object
accretion; for stellar or galactic dynamos the analogy may be more
appropriate).
Reply to (9): We apologize for implying that the α-Ω dynamo might
occur in the accretion disk around compact objects. This may or may
not occur depending upon the effectiveness of the MRI or Rossby vortex
mechanism. To apply the α-Ω dynamo to stars requires a deeper
understanding of convective plumes at the base of the convective zone
than is presently available. One might expect that the diffusive heat
flow, radiation transport transition to convection would truncate the
smaller scales leaving only the large scale convective elements, of
order H to carry the heat. At this scale such plumes would make
effective helicity-generating plumes for the α-Ω dynamo in nearly all
rotating stars.
(10) Reviewer #2 Comments: In addition, the Couette flow proposed in
the present experiment has an Ω~R-2 rotation profile. This is
a rather special rotation law for which the radial epicyclic frequency
of the system vanishes. It appears from numerical simulations that
dynamo action in an Ω~R-2 system can be rather different from that
in a Keplerian disk with Ω~R-1.5. The feasibility of varying Ω(R)
should be considered and would be an interesting addition to the
proposed experiment.
Reply to (10): We agree that other profiles are of interest.
The profile with maximum stable shear has the best possibility of
making a dynamo with positive gain as demonstrated by the
calculations. Lower shear will create a lower toroidal field (lower
magnetic Reynolds number) and hence less gain for a given helicity.
We will find out in Phase I of the experiment. One simply disconnects
the high-speed drive clutch and monitors the magnetic effects as the
Couette flow profile decays to the solid body profile.
(11) Reviewer #2 Statement: The proposed experiment would
be an interesting and scientifically important proof-of-principle for
dynamo amplification of magnetic fields; its astrophysical relevance
for the compact object accretion problem motivating the proposal is,
however, probably limited.
Reply to (11): Separately we have been demonstrating
theoretically and by observations (astro-ph/0106281) that the giant
radio lobes are an indication that the major free energy of BH
formation must be accessed to satisfy the minimum energy requirements,
1061 erg. In addition this must be a coherent, large-scale
production mechanism of magnetic flux. This magnetic energy is 107
greater than the galactic magnetic energy, and 103 greater than the
galactic binding energy. The flux is 104 greater than the galactic
magnetic flux. Only an α-Ω dynamo in the BH accretion
disk has the possibility of accessing this amount of energy and
creating this flux. The helicity generation must then be
non-axisymmetric and episodic by Cowling's theorem.
This document was generated using the
LaTeX2HTML translator Version 97.1 (release) (July 13th, 1997)
Copyright © 1993, 1994, 1995, 1996, 1997,
Nikos Drakos,
Computer Based Learning Unit, University of Leeds.
The translation was initiated by Howard Beckley on 15 November 2001
|