A few weeks ago, I asked for responses from labs that have
performed magnet shielding. From our perspective, it is
important to note that we are shielding a room below us;
this is much less common than shielding for a room above
magnets.
In this email, I give a short summary, and then attach
responding emails and some notes from a phone conversation
(minus names; ask if you want more details).
Thanks to all those that responded!
Charlie Fry
Original email:
>AMMRLs:
>
>I am experiencing moving pains, and am hoping you can help
>me once again.
>
>The department has decided to place a TA office directly
>underneath our facility. I have been asked to gather
>specific information as to the viability of shielding
>our magnets to provide a safe working environment for the
>TAs and the large number of students that will be visiting
>them. I have been in contact with Oxford and Bruker, but
>need to get information about current or past installations
>where shielding was placed directly on the floor beneath
>the magnet.
>
>I have real concerns about vibrations potentially
>recoupling back into the magnets, about the quality of
>water suppression, and about the downtime and cost of such
>a venture (we have four spectrometer that would have to be
>shielded), just to name a few.
>
>If any of you have direct experiences, or know of someone
>that does, with shielding high-resolution, high-field
>systems, I would very much appreciate your advise/comments/etc.
>
>Thanks,
>Charlie
SUMMARY: Construction of shielding without outside expert
help has been done in a number of laboratories. These labs
were always shielding rooms _above_ the magnet, and thus
dealing with much smaller stray fields than we are (except
for a couple cases of 750/800s). These facilities typically
were working to avoid extensive computer monitor distortions.
All these labs got imperfect shielding: stray fields were
reduced at the center of the metal shield plate, but were
also substantially increased at the edges of the plate.
Shielding by knowledgeable contractors might work, but would
come at a significant price tag (but considerably less than
replacing the magnets with new actively-shielded magnets).
Even then, it is not clear that significant performance
issues might not interfere with the use of the equipment.
Different labs seems to have different opinions; this fact
lessens confidence considerably. Temperature changes and
magnet movements (e.g., by weight of cryogens during fills
and slower during boil off) were mentioned as real problems
in some labs where shielding involved larger stray fields
similar to our configuration.
We have yet to make any decisions as to how to proceed for our
facility.
******************************************************************
ATTACHED RESPONSES:
******************************************************************
Charlie,
We have had no adverse effects on the magnets that we know of.
Let me describe our situation, without 800. Our ceiling is 123" concrete
to concrete. All systems except the 750 are on the floor, the 750 is in
a 4 ft deep pit. Below is dirt, above are labs and offices. Shielding was
put on the floor above with the objective of keeping the field below 2 Gs
at desk height to eliminate color distortion on monitors. This was not
fully achieved, a number of monitors sit in individual shielding boxes.
The shielding consists of low grade transformer steel, ARMCO M-43
or M-45. It comes in .025" rolls 3 ft wide. The supplier cut it to 8 ft
length, put it on pallets and shipped it by truck. (about 16000 lb total)
Above the 500's there are 6 layers. Above the 600 there are 8. Above
the 750 we have 16 layers. The direction of the strips was changed 90 deg.
with each layer.
The reason for all this came from the following, semi empirical guesses
and calculations:
1. Ideally, we would like to have a circular shield with the radius of the
2 Gs line around the magnet axis at the height of the shield. Since
circular is impractical, we settle for square or close to it.
2. We estimate the total field in Gs x m from the location of the shield
to infinity.
3. We pick the thickness of the steel to comfortably conduct that
field without going into saturation.
The problems we ran into was that the 2Gs line was outside the the
building or work area. So we picked a higher field line and kept the shield
as symmetrical as possible with respect to the magnet axis. As a result,
the stray field above the shield is higher in some areas, especially above
the 750, and near the edge of the shield.
The results are as follows: Above the 500's there are numerous PC
and Unix monitors. More than half of them were degaussed once
and are now ok. A few had to be put into additional shielding boxes.
Above the 600, more than half of them are in shielding boxes.
Above the 750, all of them are in additional shielding boxes, sometimes
with additional shims to correct for color distortion.
The 800 is a separate chapter. It is in a different building, the remodeling
and installation of shielding was contracted out. We have had some
complaints and installed 3 shielding boxes on PC monitors. The worst
is a 27" color TV in a classroom used for training. The only practical
solution was to take the TV to the back of the room and have the
trainees turn around during the movie.
More details on request
Cheers Rolf
--------------------------------------------------------------------------
Rolf Tschudin Tel. 1 301 496 2692
National Institutes of Health Fax 1 301 496 0825
Bldg. 5, Room B2-29
Bethesda MD 20892-0505
USA e-mail: tschudin@nih.gov
--------------------------------------------------------------------------
********************************************************************
********************************************************************
Dear Dr. Fry,
Answers/comments follow.
>We have serious concerns about the ability to retain optimal
>performance with our spectrometers, which currently perform
>state-of-the-art experiments including long-term runs with
>water suppression.
>
>Some questions of particular interest to us are:
>
> What was the effort involved in terms of cost and downtime
> (I have not been able to obtain this information from any
> vendor yet, after multiple requests)?
Clint Gouveia at Oxford did a finite element analysis of our space to
determine
the shielding required to meet our specifications which were very tight,
less than
3 G outside the space. This calculation was included in our purchase. I
would imagine that your particular problem would be straightforward and
might cost on the order of 5K or so [cgfry: seems this is a guess at the
cost of a finite element analysis that precedes actual quotations; the
actual install is much more expensive]. You need to check with Oxford of
course.
The design was by Oxford and the shield was installed by Lindgren of
Chicago. The shield itself is simply cheap pig iron plates stapled to the
floor/walls/ceiling and welded together. Weight can be an issue but for
what you describe it shouldn't be.
> Were there particular problems with the installation that had
> to be resolved, or should be of special concern in evaluating
> shielding solutions with liquids spectrometers?
weight is the big issue for construction. as for performance, we have a 500
and 600 sitting on 5 mm iron plate with no apparent effect on performance.
> Do you know of other high-resolution liquids NMR users that
> have similar shielding installations?
no but andy byrd is in the middle of designing and constructing one.
> Are you sufficiently satisfied with the results of the
> shielding that you would recommend this as a general solution
> to the liquids NMR community?
yes. but actively shielded magnets are probably better.
>Shielding is an important issue for our department; we would be
>most appreciative of whatever advise you can provide.
>
if i understand correctly, you have new construction under existing magnets
that you wish to shield. the FDA rule is better than 5 G stray field in
unmarked areas. if that is all you are going for then I think the solution
will be simple: attach 5 mm plate to the ceiling. your best bet is to have
oxford comment.
A. Joshua Wand, Ph.D.
Professor of Biochemistry & Biophysics
University of Pennsylvania School of Medicine
Philadelphia, PA 19104-6059
Telephone: 215-573-7288
Facsimile: 215-573-7290
E-mail: wand@mail.med.upenn.edu
********************************************************************
********************************************************************
Dear Charles,
I have done shielding for two of our spectrometers here (500 MHz) and
the cost was about $5000 each case.
The rooms above us are classrooms ( 26 x 40 ) and traffic is heavy. Luck
enough that the class rooms use wooden chairs.
We used 1/16 inch thick "hot-roll" iron sheets, they came in 4x8 feet in
size, we have these sheets on the floor above the NMR lab in two layers
, sandwiched with 1/8 plywood, then a hard 3/4 plywood on top of them,
then, vinyl tiles. The sheets were painted to prevent rusting and glued
down with heavy duty construction glue but no nails. No gaps between the
sheets. ( The center above the magnet, we used three layers on the
floor just on the save side).
The whole room was covered as the fringe field tends to wrap around at
the edges.
We measured the strayed field, before shielding it was 10G 5 feet above
the floor (in the classroom above the magnet), after shielding it was
about 3 G at the floor (vinyl tiles). But at the edge of the shielding,
the field rises back up to 15-20 Gauss, but as it get into the cylinder
blocks and out to the window, it is ok.
We, Bruker, energized the magnet before the shielding was on, then,
check the cyro shim and field after the shielding was put up, there is
no noticeable drift. The magnet field did drift slowly down over the
years as I guess it is coupling to the soft iron sheet. We had the
magnet up for more than four years and so far the drift didn't bothered
us. We can run short term 1-D experiment without locking.
Hope this help, if you want to find out more about the materials,let me
know.
Regards,
Yiu-fai
-- Yiu-fai Lam NMR Lab. University of Maryland, Bldg #91 Chemistry & Biochemistry Department, College Park, MD 20742. Phone: 301-405-1869 (Voice-mail), Labs: 405-1870 or 405-8559 Fax : 301-314-9121 E-mail: YL2@UMAIL.UMD.EDU WEB: http://www.chem.umd.edu/facility/******************************************************************** ******************************************************************** Charlie,
We have a 400 MHz magnet on the 13th floor of a 17-floor building. The first 7 floors of the building are hospital parking garage. The magnet is sitting on anti vibration legs. Both my lab and the lab above mine have 1/2 inch steel plates put in for shielding purpose. Water suppression seems fine and there are no vibrational problems. The real problem is to control the floating height of the magnet relative to the floor. When the steel plates are magnetized, the shimming will change if the height of the magnet is changed (e.g, when the weight of the magnet changes as LN2 boils off). If you can get by without putting in the shield, don't.
Yan
+++++++++++++++++++++++++++++++++++++++= Yan Xu, PhD Associate Professor Anesthesiology and Pharmacology University of Pittsburgh School of Medicine Pittsburgh, PA 15261 (412) 648-9922 Fax (412) 648-9587 +++++++++++++++++++++++++++++++++++++++=
******************************************************************** ******************************************************************** Dear Charlie, I had the reverse problem to you with an office above the axis of a narrow bore 500 Mhz magnet. The occupant was wading through the 5G line at Approx knee height (I facetiously told him not to bother picking up his credit card if he dropped it). We laid #24 gauge galvanized iron sheets over the floor, which reduced the 5G line to below the floor level, but shot up to 25G at the edges next to the wall. The GI sheets were kept in place by gravity and the downwards attractive pull of the magnetic field. I am a little worried about placing such sheeting directly under the magnet footings, attaching the sheets to the ceiling below would be much safer. Make sure the placement of the sheets is symmetrically placed about the magnet centre. I experienced no adverse affect on the magnet homogeneity,etc but the distance from the coil to the sheeting on the above floor would be greater than in your situation.
Regards Ian
"I live with fear, death, and evil ... but sometimes I turn it off and use a Mac." Author Unknown
Ian Willing CSIRO Molecular Science I.Willing@molsci.csiro.au Ian Wark Laboratories Tel: +61 3 9545 2550 Bayview Avenue, Clayton Fax: +61 3 9545 2446 Postal. Private Bag 10 URL http://www.molsci.csiro.au Clayton South, VICTORIA 3169. AUSTRALIA.
******************************************************************** ******************************************************************** Charles:
In our case, we have a lab above a 500 magnet. I our case was simple because the LAB above us was being rebuilt at the same time that we were installing the magnet. The shielding was placed directly on the floor above us.
Below the magnet we have a machine shop, which have a computer that had to be moved to other place, because the shielding in this case had to installed at the ceiling. Due to security reasons we decided to do not install any type of shielding.
Other comments: the shielding installed above the magnet is not perfect, therefore the people in the LAB above us had to move some computers from their original place. Due to the shielding the lab next to us had to move a computer too.
Have a nice day.
JOE 8-) ___________________________________________________________
JOSE R. MARTINEZ ORTIZ joenmr@att.net UNIVERSITY OF PUERTO RICO joenmr@yahoo.com DEPT. OF CHEMISTRY joenmr@latino.com PO Box 21352 UPR Station San Juan PR 00931-1352 ____________________________________________________________
******************************************************************** ******************************************************************** Notes from phone conversation: an experienced engineer was involved in shielding a 750 from a room above. Said temp changes and building movements caused real degradation of performance in cancellation and suppression experiments. Highly recommends against attempting shielding in our case, where the shields would be so close to the magnets. Believes "our setup will cause real problems for the NMRs."
******************************************************************** ********************************************************************
---------------------------------------------------------- Charlie Fry Tel: (608)262-3182 Director, MR Facility Fax: (608)262-0381 Chem. Dept., Univ. Wisconsin Madison, WI 53706 USA email: fry@chem.wisc.edu ----------------------------------------------------------