A comparison of calculations and measurements of the magnetic characteristics of the SSC (Superconducting Super Collider) design D dipole Page: 1 of 4
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to Digital Library by the UNT Libraries Government Documents Department.
The following text was automatically extracted from the image on this page using optical character recognition software:
esented at 1987 Particle Accelerator Conference, Washington, D.C., March 16-19, 1987.
A COMPARISON OF CALCULATIONS AND MEASUREMENTS OF TIlE
MAGNETIC CHARACTERISTICS OF THE SSC DESIGN D DIPOLE
R.C. Gupta, G.H. Morgan, and P. Wanderer
4rookhaven National Laboratory
The SSC design D dipolelNl has two, circular, supercon-
ducting, 10 mm thick coil layers with an inner diameter of 40
mm. Surrounding these coils and compressing them is a 15 mm
thick, stainless-steel collar. The collar has tabs at the poles
and on the midplane which fit into notches in the iron yoke,
the latter having an outside diameter of 267 mm. The coils
are composed of partially-keystoned cable, molded into radial
blocks separated by copper wedges and positioned in asimuth
by a protrusion inward of the Nitronfc 40 collar at the poles.
Between the protrusion and the coils are G-10 shims of variable
thickness. The nominal maximum working field is 6.6 tesla.
Eight, 4.5 m long magnets of this design (the first was no. 8)
were built and tested at BNL, and more recently, two, 16.6
m long magnets were built at BNL and tested at FNAL. The
present report compares measurements which do not include
the ends, on two of the 4.5 m magnets, no.s 8 and 9, with com-
puter calculations. These two magnets were tested at higher
fields (up to about 7.5 T) than the others by subcooling, so
data from them is best suited for comparison with calculations
showing the effects of saturated iron. Similar magnets have
been made and measured at LBLI2J.
Three computer programs were used in the design and
analysis for the present report. The first is a program which
optimises the coil positions to acheive high field quality subject
to the various constraints. This program, the current version
of which is called "PAR2DOPT', evaluates analytical expres-
sions for the harmonics due to a polygonal conductor inside
a. circular, infinitely-permeable iron aperture. The other two
programs are general-purpose, two dimensional, saturable-iron
programs. The first is a version of GFUN3I called MDP. It is a
finite element program which solves integral equations for the
field of both the conductors and the iron. No meshing of the
conductor or air regions is required, which greatly facilitates
use of the program, and also results in almost exact modeling
of the conductors. The third program is POISSON, the present
version of which incorporates a means of effectively extending
the outer boundary to infinity"1, and has improvements in the
mesh generator51 permitting more accurate modeling of the
The accuracy of modeling of the coil structure can be ex-
amined by comparing Fourier harmonics of the field computed
by the three programs. The two saturable-iron programs are
used with low coil current to get high permeability and with a
circular iron aperture. The results are shown in Table I. We
observe that all harmonics (6; = 10'B;/Bo at a radius of 10
mm) are in agreement to 0.2 units.
PROGRAM Bo/l T/kA 4y 44 44 45 4,0
PAR2DOPT 1.0361 -.1 0.00 0.12 0.67 -.03
POISSON 1.0350 -.19 0.24 0.19 0.89 -.03
MDP 1.0344 -.24 0.00 0.12 0.87 -.03
Research carried out under the auspices of t
Although the cable is thicker on one edge than the other,
the current per unit width of cable is no greater at the thick
edge than on the thin; the wires there are simply compressed
less. Since the programs assume constant current density, to
obtain a realistic current density distribution, two approaches
are used. In PAR2DOPT and in MDP, each cable is modified
from the physical trapezoidal shape to a rectangle having the
same base and radial width. It is this shape that is shown
in Figure 1(a). In POISSON, because of meshing limitations,
the cables are not modeled individually; each group of cables
between the wedges is modeled as a block and the block outlines
closely follow the cable outlines. Each block is divided into four
radial sub-blocks having the same radial width and carrying
the same current. Figure 1(b) shows this model.
Comparison of Calculations and Measurements
The effects of iron saturation on harmonic content in these
round iron dipoles has been discussed elsewherelel. The notches
at the poles in the present design, shown in Fig.2, modify
the effects of iron saturation in a distinctive way. Field lines
which would normally be almost perfectly vertical and uni-
formly spaced at the poles are diverted towards the near corner
of the notch, leading to premature iron saturation at this corner
at relatively low fields with an associated increase in sextupole,
b2. The path length'in air is also increased, resulting in both a
small decrease in transfer function (TF) and an increase in b2
at low field. Saturation of the iron on the midplane causes a de-
crease in b1 and an increased reluctance of the yoke, resulting
in a rapid decrease in transfer function at high excitation. A
notch in the aperture at the midplane aggravates these effects.
Both MDP and POISSON predict a decrease of 0.17% in
the low field transfer function due to all notches from the case
of a smooth circular iron aperture. With the notches the values
are 1.0318 T/kA by MDP and 1.0332 T/kA by POISSON.
The relative decrease with excitation is given in Table II. The
measurements are the average of the two magnets, except at
7.4 kA which is Magnet 8 alone.
I, kA 3.0 5.0 0.0 7.0 7.2 7.4
MDP 1.000 .995 .967 .973 .970 .966
POISSON 1.000 .995 .965 .98a 964 .960
Meaured 1.000 .996 .952 .969 .967 .962
The low field transfer function was measured accurately
using an NMR devicel7l; the value reported is the average of 6
magnets. Since the coil positioning shims differ from magnet to
magnet and the transfer function is affected by the shimming,
a correction to the shims used in Magnet 8 (the first of the 6)
was made to the transfer function of the other 5 magnets. The
value calculated by MDP and POISSON must be corrected to
the same shim sizes; the factor is 1.0018. The measurements
are made at a temperature of 4.5 K, and the calculations are
based on room temperature dimensions, so a correction to the
calculated values for the decrease in coil and iron sizes must
DISTRIBUTION OF THIS DOCUMENT IS iJNLlMITE
eontf- /S W) _50 t - -/ y)
Here’s what’s next.
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
Gupta, R.C.; Morgan, G.H. & Wanderer, P. A comparison of calculations and measurements of the magnetic characteristics of the SSC (Superconducting Super Collider) design D dipole, article, January 1, 1987; Upton, New York. (https://digital.library.unt.edu/ark:/67531/metadc1212388/m1/1/: accessed April 22, 2019), University of North Texas Libraries, Digital Library, https://digital.library.unt.edu; crediting UNT Libraries Government Documents Department.