Florida State University DigiNole Commons Electronic eses, Treatises and Dissertations e Graduate School July 2014 Shrink Tube Insulation Apparatus for Rebco Superconducting Tapes for Use in High Field Magnets Andrew Whitington Florida State University Follow this and additional works at: hp://diginole.lib.fsu.edu/etd is esis - Open Access is brought to you for free and open access by the e Graduate School at DigiNole Commons. It has been accepted for inclusion in Electronic eses, Treatises and Dissertations by an authorized administrator of DigiNole Commons. For more information, please contact [email protected]. Recommended Citation Whitington, Andrew, "Shrink Tube Insulation Apparatus for Rebco Superconducting Tapes for Use in High Field Magnets" (2014). Electronic eses, Treatises and Dissertations. Paper 9115.
63
Embed
Shrink Tube Insulation Apparatus for Rebco Superconducting ... · SHRINK TUBE INSULATION APPARATUS FOR REBCO SUPERCONDUCTING TAPES FOR USE IN HIGH FIELD MAGNETS By ANDREW D. WHITTINGTON
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Florida State UniversityDigiNole Commons
Electronic Theses, Treatises and Dissertations The Graduate School
July 2014
Shrink Tube Insulation Apparatus for RebcoSuperconducting Tapes for Use in High FieldMagnetsAndrew WhitttingtonFlorida State University
Follow this and additional works at: http://diginole.lib.fsu.edu/etd
This Thesis - Open Access is brought to you for free and open access by the The Graduate School at DigiNole Commons. It has been accepted forinclusion in Electronic Theses, Treatises and Dissertations by an authorized administrator of DigiNole Commons. For more information, please [email protected].
Recommended CitationWhitttington, Andrew, "Shrink Tube Insulation Apparatus for Rebco Superconducting Tapes for Use in High Field Magnets" (2014).Electronic Theses, Treatises and Dissertations. Paper 9115.
Andrew D. Whittington defended this thesis on July 3, 2014. The members of the supervisory committee were:
David Larbalestier Professor Directing Thesis
William Oates Committee Member
Jonathan Clark Committee Member
Ulf Trociewitz Committee Member
The Graduate School has verified and approved the above-named committee members, and certifies that the thesis has been approved in accordance with university requirements
iii
TABLE OF CONTENTS
LIST OF TABLES .......................................................................................................................... v
LIST OF FIGURES ....................................................................................................................... vi
ABSTRACT ................................................................................................................................. viii
Table 5: Total cost of materials for STIA. .................................................................................... 42
Table 6: Comparison of the cost per meter to insulate a 200 m length of YBCO with heat shrink tubing for Manual Insulation and STIA. ....................................................................................... 43
vi
LIST OF FIGURES
Figure 1: Critical surface plots for different types of superconductors [18] ................................... 3
Figure 2: Dependence of the critical current density (JC) on the strength of the applied magnetic field (B) [20]. .................................................................................................................................. 3
Figure 3: Different layers of the HTS conductor YBCO produced by Super Power Inc. .............. 4
Figure 5: Testing setup for the needle and magnet threading technique. ..................................... 10
Figure 6: Diagram of implementing a pulley to reduce the length of the manual insulation process........................................................................................................................................... 11
Figure 7: Conceptual diagram of a vertical pulley set up using guide tracks to move the tape up the pulleys in a snake like pattern. ................................................................................................ 12
Figure 8: Front view of the final prototype design. ...................................................................... 13
Figure 9: Heat gun stability system is shown mounted to the Prototype’s two heat gun guide rails made from T-slotted aluminum. ................................................................................................... 14
Figure 10: Heat gun and modified attachment description. .......................................................... 15
Figure 11: Example of threading Kevlar string through shrink tubing with a blunted needle and guide magnet. ................................................................................................................................ 17
Figure 12:YBCO tape moving into a length of shrink tubing held in place by a tube holder. ..... 17
Figure 13: The progression of moving an insulated level of tape from one guide track to the next....................................................................................................................................................... 18
Figure 14: Yates-Star data of local critical current over 120 m of YBCO. .................................. 20
Figure 15: Final design of STIA. .................................................................................................. 22
Figure 16: The support structure for STIA made of T-slotted extruded aluminum. .................... 23
Figure 17: The forces acting on the tape moving around a stationary pulley ............................... 24
Figure 18: Tension and stress on each guide track level for the maximum back tension before conductor failure. .......................................................................................................................... 25
Figure 25: Comparison of Super Power’s 4mm wide YBCO tape, on the left, and the Fujikora 5mm tape, on the right. ................................................................................................................. 32
Figure 26: Handheld scale connected to in-house made pressure clamp ...................................... 35
Figure 27: Tension test data with capstan equation predicted values. .......................................... 36
Figure 29: Delamination of Japanese 5mm tape due to soldering and bending. .......................... 39
Figure 30: Follower arm potentiometer used to monitor the change in radius of the Storage and Collection spools. .......................................................................................................................... 46
Figure 31: Simplified diagram of tape movement between two spools. ....................................... 47
Figure 32: Dancer roller and potentiometer assembly. ................................................................. 48
viii
ABSTRACT
An increasing number of applications require the use of high temperature superconductors (HTS)
such as (RE=Rare Earth) Ba2Cu3O7-x (REBCO) coated conductors [1]. HTS conductors show
particularly great potential for high field magnets applications [1] due to their high upper critical
fields [2], But several groups have shown that REBCO coated conductors are prone to
delamination failure [3] [4] [5]. Under relatively low transverse stress the HTS film separates
from the substrate and the conductor degrades [6]. This is problematic due to high transverse
stresses that occur in fully epoxy impregnated solenoids wound with this conductor. Application
of thin walled heat shrink tubing introduces a weak plane around the conductor, preventing
delamination degradation [7]. However, manual application of the shrink tubing is impractical,
requiring three operators limited to insulating 100 m lengths or less of REBCO conductor. The
high risk of damage to the conductor, also associated with this process, shows the need for a
mechanized process.
Strict guidelines for the capabilities of the mechanized process were set: a single operator must
have the capability to insulate continuous 200 m lengths of REBCO conductor with a minimum
insulation rate of 50 m per work day. This thesis presents the ideation, prototyping, and
construction of such a mechanized insulation process. Results of prototyping yielded an
insulation rate of 100 m per work day with two operators, and the capacity to insulate 130 m of
REBCO conductor [8]. Local critical current test indicated that no damage is caused to the tape
from the insulation process [9]. A final mechanized Shrink Tube Insulation Apparatus (STIA)
was developed, increasing the conductor length capacity to 450 m and reducing the number of
operators to one. Due to unforeseen motor and tape tension issues, modifications to STIA were
required, which increased the number of operators to two. Through a collaborative effort
between the Applied Superconductivity Center (ASC) and the Kyushu University in Japan, a
continuous 200 m length of Fujikora-produced REBCO conductor was provided, to be insulated
by STIA. An insulation rate of 100 m per work day was achieved with two operators. In the
latter of the paper, current work to rectify the issues causing the need for two operators is
presented.
1
INTRODUCTION
Motivation
In 2009 the most powerful nuclear magnetic resonance (NMR) spectrometer to date was
inaugurated into the arsenal of NMR devices at the European Nuclear Magnetic Resonance
Center (CRMN). Boasting a 1 GHz resonant field range with a 23.5 Tesla (T) superconducting
(SC) magnet, this spectrometer promises to assist in unlocking the mysteries associated with
cancer research, protein structures, and chemical property analysis [10] [11]. Despite this record
breaking spectrometer, the demand for high magnetic fields in the NMR and condensed matter
physics communities have pushed current commercial superconductors to their critical field
limit. Low temperature superconductors (LTS) have been exclusively used to produce
commercial superconducting magnets; which are used in NMR spectrometers [10] [11],
accelerator magnets for CERNS Large Hadron Collider (LCH) [12], and essentially all other
commercial applications for superconductors [1]. Now the next generation of high field
technology demands fields which are impossible for LTS materials to reach. Discovered in 1986
[13], high temperature superconductors (HTS) have garnered much research intrest for the past
28 years. Yet, commercial production and application of these materials has been sparse [1].
Difficulties in material and high field coil production with HTS materials, coupled with
commercially viable and easy to use LTS materials, led to the exclusive use of LTS in
commercial SC applications.
The highest field any commercial LTS conductor can reach before losing superconductivity is
around 25 T. HTS conductors are currently the only option to push the high field community
forward [14]. Coil manufacturing techniques are still being developed for HTS conductors due
to many inherent difficulties associated with these materials. The lab group at the Applied
Superconductivity Center (ASC) has developed high field HTS record breaking coils. Some of
these HTS materials, such as YBa2Cu3O7-x (YBCO), have been used in some record breaking
coils and employ polyester medical shrink tubing as an electrical insulator [15]. Continuous
lengths ranging from 50 m-130 m of YBCO conductor are used for these coils with projected
lengths of 1km in the next few years. Current insulation processes are impractical and hazardous
to the superconducting material, threatening to hinder or prevent coil production due to
insulation time and or conductor damage. The motivation for this thesis work is to create a safe
2
and efficient insulation process to insulate continuous lengths of ReBCO tapes greater than
130 m, with the potential to be applied to future 1 km long tapes.
Superconductivity
Discovery
In 1911 Heike Kamerlingh Onnes (1853-1926) discovered a phenomenon of physics that has
steadily impacted humanity for over a century, superconductivity [16]. Electrical resistance is an
inherent property of all materials describing how easily current is carried through a material.
While observing the electrical properties of mercury at low temperatures Kamerlingh Onnes
discovered a sudden drop in resistance below 4.1 K. The material had no observable or
measurable resistance which was the first major superconducting property observed. In 1933 it
was shown that these elements also expel all magnetic flux [17]. This expulsion of magnetic
fields is known as the Meissner effect, which along with no resistivity are the uniquely defining
properties of superconductors. In the following years many other elements as well as certain
compounds, such as the Nb-based conductors, were shown to be superconductors. Now known
as low temperature superconductors (LTS), these materials operate in temperatures below 20 K
and fields as high as 25 T [14]. Over 75 years passed since the discovery of superconductivity
before high temperature superconductors (HTS) were developed. These conductors could
operate in temperatures above 77 K and reach fields previously thought impossible to obtain
[13]. Though HTS materials sparked a new era of superconducting research LTS materials had
been around for decades. For this and other reasons LTS materials dominate the commercial
superconducting market, though recently the electrical capabilities of LTS materials have
reached their critical field limit.
Properties
Electrical capabilities of superconductors are restricted by three parameters: temperature, current,
and applied magnetic field. If the limit of any of these parameters is breached in the
superconducting state, the superconductor will transition to the resistive or normal state. These
limits are called the critical temperature TC (K), critical field BC2 (T), and the critical current
density JC (A/cm2). A critical surface plot of the critical current density, temperature, and field
of a particular superconductor defines the range of operation; above this surface the
superconductor will revert to the normal state [18].
3
Figure 1: Critical surface plots for different types of superconductors [18]
HTS and LTS materials have similar JC at low fields yet a sharp decrease in JC is seen in LTS
material as applied field increases as compared to HTS materials. Around 25 T no LTS material
can maintain the superconducting state. At high fields near 25 T LTS materials lose their ability
to be effective conductors due to their low critical field. In contrast the HTS’s high critical field
permits these conductors to effectively conduct current well past 100 T [19]. Shown in Figure 2
below is the JC versus applied field for LTS and HTS materials, which shows the field-
dependence for a variety of materials [20].
Figure 2: Dependence of the critical current density (JC) on the strength of the applied magnetic field (B) [20].
4
YBCO, Insulation and Coils
YBCO is one of the most promising HTS conductors due to its high JC in applied magnetic fields
compared to other superconductors. This is known as a coated conductor being comprised of
several layers of different materials in which a thin film of YBCO is coated onto. Buffer layers
provide a template for the YBCO structure to be grown as well as a diffusion barrier, during film
deposition, between the stainless steel and YBCO layers [21]. Stainless steel or nickel based
alloys, known as the substrate, provide a robust structure with high tensile strength which the
YBCO thin film can be grown on.
Figure 3: Different layers of the HTS conductor YBCO produced by Super Power Inc. Not clearly shown in the figure is the electroplated copper layer completely surrounding the surface on all sides of the tape [22].
Coating the conductor in copper provides structural integrity, high electrical conductivity, and
the ability to solder onto the tape. Silver allows for low resistance transfer of electricity to the
YBCO film during operation [23]. High stresses develop in these superconducting coils partially
due to the resultant force of electrons moving through magnetic fields. This force is known as
the Lorentz force, described by the relationship below.
( 1 )
Where q is the charge of a particle, v is velocity, and B is the magnetic field. The main
component of the Lorentz forces are directed in the radial direction in standard solenoid coils
requiring structural support to prevent the coil from being destroyed. Epoxy impregnation of
coils provides the needed structural support to withstand these high hoop stresses. YBCO coils
also require electrical insulation over the conductor to prevent over-voltages during a quench,
5
which is a sudden loss of superconductivity in the coil. There are many different methods of
electrical insulation which can survive 4.2 K environments [24] [25], though all have drawbacks.
Insulation can be provided by helically wrapping adhesive back polyimide tape, such as Kapton,
over the conductor. Overlaps must be made to insure there are no gaps and complete electrical
insulation is achieved. The thinnest wrappings available are thicker than desired [25], and
coupled with required overlaps increase coil size, cost, while decreasing coil-winding current
density. Another method of insulation is dip coating the conductor in various polymer varnishes,
epoxies, or acrylates [24]. Thin insulation coatings, less than 10 μm thick, can be obtained
through dip coating. Yet, coatings are usually non-uniform and do not adhere well to the thin
edge of the tape where cracking ultimately occurs and bare tape is exposed. Though these
insulation have deficiencies delamination of the YBCO thin film is the greatest concern [3] [4]
[5].
Conductor delamination is when the YBCO layer lifts off or pulls away from the surrounding
buffer or stainless steel layers. Epoxy impregnated coils are designed to prevent the conductor
from moving during cool down and operation. Lorentz forces and the low temperature
environment cause the conductor to want to move and shift inside the epoxy encapsulate. Shear
stresses develop on the surface of the tape from the epoxy constraining the tape. These shear
stresses are large enough to delaminate the conductor. Though YBCO has high tensile yield
strength, around 700 MPa, the transverse yield strength is magnitudes lower possibly around
5 MPa [6]. Solutions to delamination in high field YBCO coils are still a focus of research in the
high field community.
Heat Shrink Tubing
Insulation Options and Application Methods
In 2011, the lab group at the Applied Superconductivity Center (ASC) developed a solution to
prevent delamination through the application of medical heat shrink tubing over the length of the
conductor [7]. Unlike other electrical insulation, shrink tubing mechanically couples to the
conductor by the friction caused by the shrinkage instead of through the use of adhesives. Shrink
tubing decouples the conductor from the epoxy encapsulate. During operation, stress
concentrations build up on the interface of the conductor surface and epoxy. When the
conductor is insulated with shrink tubing these stresses are removed from the conductor surface
6
and are instead taken on by the shrink tubing. Shrink tubing provides a weak plane where stress
concentrations are allowed to build upon without damaging the conductor. . A properly sized
shrink tube, once shrunk, conforms to fit tightly over the outer perimeter of the tape. A uniform
profile is achieved, including the thin edges, as well has high packing density due to a wall
thickness of 12.5 μm when shrunk. Though overlaps are required between lengths of shrink
tubes, this minimal increase in thickness is significantly less in contrast to other insulations with
helical wrapping overlaps. No cracking or damage to the shrink tube is observed in low
temperatures down to 4.2 K. The shrink tubing is able to provide the needed electrical insulation
and epoxy decoupling [7].
Figure 4: Properly sized shrink tube profile on YBCO conductor. Properly sized shrink tubing will slide onto the conductor and tightly conform to the tape’s surface upon shrinking.
Choosing the proper shrink tube is based on the conductor’s cross-section shape and size.
Typical YBCO tape used by the ASC has a cross-section that is approximately 4 mm wide by
0.1 mm thick. Using the known properties of the shrink tubes and the dimensions of the
conductor an appropriate sized shrink tube can be chosen by using the inequality below [7].
( 2 )
Where w and t are the width and thickness of the tape respectively, is the fractional shrinkage
of the insulation, and is the inside diameter of the pre-shrunk heat shrink tubing. This
inequality ensures that the shrink tubing used will properly shrink onto the conductor to provide
a tight and uniform fit. Use of this relation prevents the purchasing of shrink tubing that is too
large to conform to the perimeter of the conductor when shrunk or too small to fit or slide onto
the conductor when un-shrunk.
4(a) 4(b) 4(c)
7
Manual Insulation
Manual insulation of REBCO conductor with shrink tubing can be broken down into three main
phases: setup, sheathing and heat shrinking, and spooling. This process was developed by
insulating a continuous 100 m length of YBCO to be used in a high field coil [13]. For the setup
phase, 2.4 m long tables were set end to end until a continuous length of approximately 50 m of
table space was achieved. As an unobstructed 50 m length of space was unavailable; a slight
bend around a wall was required to reach this length. A spool storing the length of YBCO to be
insulated is set at one end of the tables. YBCO is then unspooled down to the opposite end of
the table, leaving a 50 m long length of bare conductor on the tables. The final step in the setup
phase is connecting Kevlar string to the exposed end of the tape via a copper wire leader
soldered onto the end of the conductor. Thin copper sheeting is folded down so that it can fit
through a shrink tube and attached to the other end of the Kevlar string.
Concluding the setup phase, the sheathing and heat shrinking phase begins by moving the Kevlar
string through a length of shrink tubing. Compressed air is used to push the folded copper
sheeting through the length of shrink tubing, 1.3 m long, which in turn moves the Kevlar string
through the tube as well. This allows the tape to be pulled tight to be able to guide the shrink
tube onto the tape. After the tube is completely on the tape, an operator must walk the shrink
tube down the length of conductor to the spool. Once the tape reaches the spool, a
programmable heat gun with a nozzle attachment is set to 300ºF to shrink the tube onto the
conductor. An operator steadies the heat gun while the other two operators, one on either side of
the heat gun operator, introduce the tape and insulation into the stream of heated air. After the
entire length of shrink tubing has been shrunk onto the conductor, the sheathing and heating
process is repeated. For each repetition of sheathing and heating process each shrink tube is
moved down the conductor until an already shrunk tube is reached, a 15 mm-38 mm overlap is
created to ensure complete insulation coverage then the new tube is heated. This process is
repeated until the 50 m of conductor along the length of the table is insulated, using
approximately forty 1.3 m lengths of shrink tubing.
Having completely insulated the 50 m of conductor, the spooling phase begins by removing the
Kevlar string and copper leader from the conductor tip. This end of the tape is attached to an
empty spool and the 50 m of insulated conductor is spooled up onto this new spool. This spool is
8
now mounted at one end of the tables while the other spool unspools the 50 m of bare conductor
over the length of the tables. At this point, the setup phase begins at connecting the Kevlar string
to the exposed tip of the conductor. This process is then repeated for this 50 m of uninsulated
conductor until it is completely covered with shrink tubing leaving no area of conductor exposed.
Yet the deficiencies of this method, described below, reveal that this was an impractical
insulation process liable to cause damage to the conductor.
Impracticalities of Manual Insulation
Manual insulation was enough to help prove that shrink tubing was a viable electrical insulation
and decoupling mechanism for high field YBCO coils. However, longer subsequent coil projects
will require continuous lengths of YBCO greater than 120 m. Manual insulation requires an
operation space equal to half of the conductor’s length. There is no room available to expand
making it impractical for this process to insulate conductors over 100 m. Another issue which
prevails through all three phases is the need for three operators. In the setup phase all three are
needed to unspool the 50 m of tape and lay it across the tops of the tables. While one operator
moves the tape down the other two must make sure that the tape does not fall off the table.
Conductor falling off the tables is a significant issue, which is a constant concern throughout the
insulation process, and one which can irreparably damage the conductor.
The sheathing and heat shrinking phase of the insulation process is too slow to be a practical
option for insulation. Blowing the folded copper sheet through the shrink tube took multiple
attempts, rarely working on the first try. Moving the conductor required an operator to hold the
end of the tape stationary, another to move the tape down the conductor, and another to prevent
the tape from falling off the table. By walking the insulation down the conductor the operator
must walk over 4 km to insulate 100 m of tape. Heating the shrink tube onto the conductor was
cumbersome and hazardous to the tape. The heat gun method used to heat and guiding the tape
resulted in a slow and inconsistent process. For proper heating the tape must have the wide edge
of the tape parallel to the flow of heated air coming out of the heat gun. Often the tape was not
in this proper configuration which increased heating time. Occasionally the tape would touch the
metal nozzle attachment which would sometime burn away the insulation. A small piece of
insulation must then be cut and moved down to the burned area and shrunk on which further
increased insulation time.
9
The spooling phase saw the same issues as the setup in keeping the tape from falling off of the
length of tables. In each of the three phases of manual insulation, the conductor is at high risk of
damage. Currently 100 m batches of YBCO can have lead times upwards of three months and
cost about $80/m. Irreparable damage to the thin YBCO layer can be caused by minor bends or
kinks in the conductor. Having a thickness of 0.1 mm YBCO is easily bent and damaged. Any
process that has a high risk of damaging the conductor is undesirable. Insulation rate of this
process was about 4.2 m per hour, requiring three operators three full work days to insulate
100 m of YBCO conductor. Due to the hazards exposed to the conductor, insulation time, and
personnel required this insulation process was an unfeasible solution for sheathing YBCO with
shrink tubing.
Required Development for a New Insulation Process
Compared to other electrical insulations for high field REBCO coils medical heat shrink tubing
is thought to be the best options for various reasons such as delamination prevention. Currently a
practical means of sheathing long lengths of REBCO, greater than100 m, with shrink tubing does
not exist. Without an effective insulation process shrink tubing would become impractical to
use. Research time and money would then focus on finding another solution to electrical
insulation and delamination prevention instead of high field coil manufacturing and testing..
Development of a practical means of sheathing long lengths of REBCO conductor with medical
shrink tubing is needed if shrink tubing is to be a viable insulation for high field REBCO coils.
Imminent coil projects will require 130 m of REBCO conductor with subsequent coil projects
projected to use multiple 200 m lengths of this conductor. It is then desired to know if it is
possible to practically sheath long lengths of REBCO conductor with medical shrink tubing. A
practical method of insulation would reduce all risk of damaging the conductor, have the ability
to insulate greater than 200 m lengths of conductor, and have a minimum insulation rate of
5 m/h.
10
PROTOTYPING OF THE INSULATION MACHINE
Conceptual Ideation
New Insulation Concept
Sheathing and heat shrink phase of the manual insulation process required the most innovative
revisions, which is why our modifications to the process began with this phase. Compressed air
used to move the folded copper sail was inconsistent, taking an unnecessary amount of time to
thread the shrink tube. A more consistent method was developed using a blunted steel needle
and a neodymium guide magnet. Using Kevlar string attached though the eye, the needle would
be pulled through the tube with the magnet. To prevent damage to the tube from the contact of
the needle with the magnet, a spacer was put between the needle and magnet. The set up to test
this needle and magnet method is shown below.
Figure 5: Testing setup for the needle and magnet threading technique. It should be noted that the shrink tube was held in place by hand as the needle was pulled through the tube by the magnet.
Holding the shrink tube in place during this procedure required a light pressing of the finger tips
on the shrink tube against the cardboard surface. The tip of the needle needs to be inserted into
the tube and at least 3.175 mm thick separating wall was required for the method to work
properly. A thicker spacer prevented the guide magnet from moving the needle where a thinner
spacer incurred damage to the tube from the needle movement. After many tests this process
was shown to be more consistent than the previous method.
Preventing the tape falling off the tables during insulation required a tape guidance structure. A
strip of aluminum, 6.35 cm x 50.8 cm x 2.4 m with a 5.08 mm wide and 1.5 mm deep channel
cut down the middle of its length, was used to prevent the tape from falling. Channel dimensions
11
were based on the 4 mm wide tape used for the coil projects and for the dimension of the
insulation, the details of which will be discussed later in this section. The strips of aluminum
would be mounted down the length of the tables with clamps for quick set up. Tables must be
aligned in order for the tape to lie comfortably in the channels as described before; the tables had
to curve around a corner. To keep the tables aligned, a pulley could be used to redirect the tape
in a snake-like pattern. Placing a pulley at the end of a 25 m length of tables would redirect the
tape 180 degrees moving the tape parallel but in the opposite direction. This requires
approximately eleven of the rented tables used previously in the manual insulation process to be
put end to end. An example is shown in Figure 6 below, where the eleven tables are shown to be
one complete 25 m long table.
Figure 6: Diagram of implementing a pulley to reduce the length of the manual insulation process. This is a bird’s eye view of how the concept of mounting guide tracks and a pulley on to the 11 rented tables (boarders between tables not shown) can reduce the length of the insulation procedure. 25 m is the combined length of the 11 tables.
With this set up, guide channels can be used to keep the tape from falling off the table while
reducing the length required for manual insulation from 50 m to 25 m. Due to the
aforementioned low transverse yield stress inherent to YBCO tapes, it is desired that the
conductor is not twisted along its length. To obviate damage to the conductor, the tape would
have to be oriented with the flat face perpendicular to the table to round the pulley without
twisting. This tape orientation would require the tape guide channel to be approximately the
same thickness as the tape. Machining a thin channel such as this would have been difficult and
expensive but there were also other problems with this set up as well. Some concerns included;
the tape not going back into the channels, when lifted out, while moving the shrink tubes over
the tape and difficulties with heating the shrink tubing in such a thin channel. Some of these
problems could be solved by restructuring this setup in a vertical configuration.
12
Orienting this machine vertically allows for the tape to lie in the 5 mm x 1.5 mm channel,
removes the need for tables, and reduces the space needed; yet moving the shrink tubes was still
not possible. If the shrink tubes could not be moved about the pulleys, perhaps the tape could be
moved through the tubes and around the pulleys instead. It was realized that a shrink tube could
be placed in a guide channel and held stationary by hand while the tape was pulled through the
tube by the string already threaded through the tube. Laying shrink tubes in the guide channels
end to end would allow all the tubes to be threaded together. This then requires the operator to
pull the string to sheath the ReBCO material instead of walking each individual tube down the
length of the tape. The length of the machine required to insulate the same amount of tape the
manual insulation process required would be dependent upon the number of guide channel
levels.
Minimum bending radius for REBCO conductors varies between different conductors being
strongly dependent on the materials and thickness chosen for each layer of the conductor [26].
Though the minimum bending radius can be as small as 4.5 mm-5 mm for certain conductors a
conservative approach was taken when designing the pulley diameter which was 152 mm [26].
For an operator to safely use the device, the first guide track should be at least 1 m above the
floor. Keeping the device around 1.5 m in height this would allow for seven levels of guide
channels with six pulleys between them. If constructed where the tape guide tracks are 7.6 m in
length, approximately 50 m of conductor could be insulated. This would allow 50 m of
conductor to be insulate in a ~8 m space as compared to the 50 m space required for hand
insulation.
Figure 7: Conceptual diagram of a vertical pulley set up using guide tracks to move the tape up the pulleys in a snake like pattern.
13
Prototype
Design Overview
To demonstrate the functionality of this insulation process a prototype, approximately 2.4 m with
seven guide track levels, was constructed. Though shorter than the desired full scale machine
length of 7.6 m the prototype was able to demonstrate whether or not the process is capable of
practically sheathing long lengths of REBCO conductor with shrink tubing. The final prototype
design is shown in Figure 8.
Figure 8: Front view of the final prototype design.
T-slotted extruded aluminum, or T-Slot Al, is used as the support structure. A variety of
fasteners and brackets, specifically designed to mount directly into the T-slot in the T-slotted Al,
are commercially available. As a result, no permanent fixtures such as welds, glue, or mounting
holes in the T-Slot Al are required to construct this prototype. Having no permanent fixtures
allows for quick assembly, adjustments to any fixture, and no specialized skill to construct. The
design was such that if the prototype was successful all, T-Slotted parts would be recycled into
the final device as well. Standard lengths of T-slotted Al, at 2.4 m, as well as the tape guide
channels were responsible for the designed length of the device.
Aluminum angle stock came in a standard length of 2.4 m, being the maximum length a
machinist could cut the desired guide channel. Angle stock could provide a thin enough
separating wall, as well as being non-magnetic, for the needle and magnet technique. Angle
stock was rigid enough to support the shrink tube holders without appreciable bending. Shrink
14
tube holders were constructed out of door hinges, 2.3 mm x 25.4 mm x 1.07 m strips of
aluminum, and 0.8 mm x 25.4 mm x 1.07 m strips of adhesive backed rubber. The door hinges
were mounted to the back of the tape guide tracks and to the aluminum strips. Strips of rubber
were glued to the lengths of the aluminum strips with their adhesive backing. During manual
insulation it was seen that the shrink tube had to be flattened in order for the tape to fit into the
tube. As previously mentioned the guide channels were designed to form the shrink tube into a
rectangular profile to allow the tape to be inserted. Different dimensions were tested but the
5 mm x 1.5 mm channel restricted tape movement while allowing the needle and tape to be
easily pulled through the insulation. Hinges on the tube holder, T-slotted extruded aluminum,
and the open design of this structure was necessary for the method heating the insulation. Using
a few extruded aluminum scrap pieces, available wheel attachments for T-slotted Al, a linear
bearing and a precision milled rod, a heat gun stability system was created. The setup of this
stability system is shown in Figure 9 below.
Figure 9: Heat gun stability system is shown mounted to the Prototype’s two heat gun guide rails made from T-slotted aluminum.
Height adjustments are made possible by the linear bearings ability to slide up or down the
vertical heater support, allowing the heat gun to reach any track level. The T-slotted wheel
attachments allow for a smooth linear movement down the full length of the device. As
15
discussed in the previous section, heating the shrink tubing required two operators, one to
operate the heat gun and the other to feed the tape through the nozzle attachment. Burns on the
insulation, and possibly the tape, from contact with the metal nozzle attachment were of major
concern. To solve this issue, a delrin sled was created, modifying the existing heat gun nozzle
attachment. The sled was designed to slide in the tape guide channel, lift the tape sheathed with
insulation vertically off the track into the optimum location in the nozzle attachment, and then
lay the tape back in the channel. A thermo couple was used to find the spot in the nozzle
attachment that was closest to the programed temperature. Orientation of the nozzle on the sled
allows for the 4mm side of the tape to be parallel to the flow of hot air. This orientation heats
both sides of the tape evenly preventing incomplete shrinkage on one side. The modified nozzle
attachment and operation of the heat gun with this attachment is shown below in Figure 10.
Figure 10: Heat gun and modified attachment description. (a) Modified heat gun attachment is shown with Insulated tape on it. (b) Side view of the modified attachment depicting hot the heated air stream heats and shrinks the insulation onto the YBCO tape.
The bottom of the sled fits into the guide channel keeping the sled from moving off the track.
Heating the shrink tube onto the tape with the sled now requires only one operator due to the sled
guiding the tape safely through the nozzle. T-Slotted Al has wheel attachments, which fit into
the slotted groove of the heat gun guide rails. This provides a smooth linear movement of the
heat gun allowing for a more controlled and consistent heat shrinking of the insulation. All of
these improvements to the heating system decrease the number of operators, heating time, and
improve the safety of the process. Hinges on the tube holders are now an obvious feature due to
the path of the delrin sled during heating. Though ReBCO conductors have a minimum bending
radius on the order of a few millimeters, before the conductor will incur irreparable damage, the
pulley diameter was conservatively designed [26]. To assure that no damage will occur to the
16
conductor and to allow room for the heat gun to fit on to the guide track the pulleys were
designed with a 178 mm diameter. A total of six delrin pulleys were made due to the seven tape
guide track. These pulleys were mounted on an aluminum shaft supported by ball bearings,
reducing the resistance of the tape moving from one track to the other. Delrin was chosen for its
light weight, yielding a low moment of inertia and low coefficient of friction to allow the tape to
slide over the pulley if the bearings seize.
As in our manual insulation process, YBCO material was taken directly from the storage spool.
This spool will be mounted at the end of the bottom guide track. Another spool is placed at the
end of the top guide track on the opposite side of the insulation device. After threading the
insulation tubes, the string will be connected to this Collection spool. Moving the tape through
the tube will be powered by the operator using a hand crank connected to this Collection spool.
This crank handle has a threaded coupling that screws onto the shaft of the Collection spool.
Both the Storage and Collection spools have crank handles that can be attached or disconnected
depending on the direction the tape is being moved on the machine. With seven guide tracks
spanning approximately 2.4 m long, a total of 17 m can be insulated on the machine at once. The
following section covers how this operation is executed to insulate the entire 130 m of
continuous ReBCO material
Operation
Shrink tube insulation is laid in the guide channels on all seven guide track levels. Standard
length for the shrink tube insulation is 1.3 m, meaning that two lengths can be placed on one
guide track with about 152 mm extra. However, this excess insulation interferes with pulley
operation, thus we cut off about 100 mm of one insulation tube. A blunt needle with Kevlar
sting attached to the eye is inserted into the shrink tube on the bottom guide track, closest to the
Storage spool. Pressing gently, by hand, on the shrink tube, the needle is moved through it by
moving a magnet underneath the guide track. Once the needle exits the tube, the tube holder is
engaged on this tube. After the next tube is threaded and tube holder engaged the string is
moved around the pulley at the end of the track, up to the next guide track, and inserted into the
next shrink tube. Each consecutive tube is threaded in this manner; Figure 11 below shows an
example of this threading process.
11(a) 11(b)
17
Figure 11: Example of threading Kevlar string through shrink tubing with a blunted needle and guide magnet. (a) Needle with attached Kevlar string is inserted into a shrink tube. (b) Guide magnet is shown moving the needle through the shrink tube. (c) Needle exiting the shrink tube, threading the shrink tube with Kevlar string. (d) The threaded string is connected to YBCO tape and is pulled to move the length of YBCO through the shrink tube, here the tape is about to enter the tube to be pulled through.
As the needle and string exit the shrink tube on the seventh and last guide track, the string is
connected to the Collection spool. To allow for easier tape insertion into the shrink tube, the tip
of the tape has both corners cut off, as shown in Figure 11(d). Here the end of the threaded
string is connected to the ReBCO conductor with thin copper wire. A loop is made in the copper
wire where the string is tied and the other end is soldered to the tape. In this manner a smooth
profile obtained, whereas punching a hole would create a rough profile due to burrs. The crank
handle is now attached to the shaft of the Collection spool and all tube holders are checked for
good contact. Slowly the operator turns the crank handle to take up the string and move the tape
through the shrink tubes. Though the tape can enter the tubes without assistance, it is safer to
guide the tip of the tape into the entrance of each tube.
Figure 12:YBCO tape moving into a length of shrink tubing held in place by a tube holder. (a) The string pulls the tape towards the shrink tube. (b) The tapered tip helps to guide the tape into the tube. (c) Tape fully enters the shrink tube.
11(c) 11(d)
12(a) 12(b) 12(c)
11(a) 11(b)
18
Once the tip of the tape reaches the exit of the last shrink tube on the seventh and final track, the
tape is stopped and put under tension. Tensioning the tape is done with a brake, which is a bolt
that is tightened down onto the shaft of the spool. Both the Collection and Storage spool have
such a bolt mounted into their respective bearing blocks. Once tensioned, the heat gun assembly
is mounted on to its T-slotted Al guide rails and the tube holders on the top guide tracks are
lifted. The delrin sled is placed at the end of the top guide track closest to the Collection spool
and the heat gun is attached and turned on. Once the first shrink tube is heated, by moving the
heat gun assembly down its length, the gun is switch off. The next shrink tube is moved to
overlap, about 38 mm, the newly shrunk shrink tube then the heat gun is turned on and moves to
shrink this tube. After both tubes on this top guide track have been heated and shrunk, the gun is
turned off, the brakes on the spools are released, and the tape is wound slowly onto the Storage
spool until the newly shrunk insulation reaches the next lower level guide track. The un-shrunk
tube on this track is overlapped, again about 38 mm, on the already heat-shrunk tube. Adjusting
the heat gun assembly to this guide track, the tube holders are removed from the tubes and the
heating process continues as before.
Figure 13: The progression of moving an insulated level of tape from one guide track to the next. (a) Heat gun reaches the end of a guide track level having heated all shrink tube on this level onto the tape. (b) Tape is spool back onto the storage spool moving the insulated tape from one guide track level, around the pulley, to the next guide track level. (c) The presently shrunk tube is overlapped with un-shrunk shrink tubing. (d) The un-shrunk tube is heated and the overlap, approximately 38 mm, has been created leaving no gap in the insulation over the tape.
13(a) 13(b)
13(c) 13(d)
19
Heating the shrink tube in this manner continues until all the shrink tubes on the machine have
been shrunk onto the ReBCO conductor. Once completed the ReBCO material is wound back
onto the ion spool, as well as the length of string. Approximately 17 m of this length of ReBCO
conductor was insulated during this process; we will refer to each 17 m of insulated material as a
single ‘pass’ from here on. Multiple passes will be needed to insulate longer lengths of
conductor. For example, if each pass insulates 17 m of conductor, 6 passes will have to be made
to insulate a 100 m length. This process is repeated for each pass, with one minor change if the
conductor has been partially insulated with shrink tubing. As before, the string pulls the
conductor through the un-shrunk tubes, stopping the tape when its tip reaches the exit of the tube
on the seventh and final guide track.
The string must now be disconnected from the ReBCO conductor and removed from the
Collection spool. Since the insulated tape will now be connected to the Collection spool, the
string had to be removed to prevent damage to the tape from spooling onto an uneven surface.
Now the Collection spool is again moved slowly with the tape, spooling onto the Collection
spool until bare conductor reaches the exit of the shrink tube on the seventh guide track. Here
the entire process of heat shrinking each level begins again. In this manner, multiple passes can
insulate long lengths of YBCO tape.
Prototype Results
A 50 m length of ReBCO conductor was used to test our prototype. Approximately four passes
through the machine were needed to insulate the entire 50 m length, which took 8 hours to
complete with two operators. Visual inspection of the tape during and after the insulation
process showed that no damage was imparted to the tape. No bends, deep scratches on the
surface, or burns in the insulation or tape were observed. As a result, a 130 m continuous length
of ReBCO conductor, purposed for a magnetic field homogeneity test coil, was insulated by this
prototype. In total, the process took a little over 9 hours to complete. Insulation rate of this
prototype was 12.5 m/h, doubling the rate of the 50 m test, an increase attributed to the learning
curve of the operator. Visual inspection of this 130 m length showed no defects, much like the
50 m test length. Further verification that the insulation process did not damage the ReBCO tape
came from testing the critical current over the entire 130 m length of the conductor. Critical
current can be used to indicate whether or not the YBCO film is damaged. The 130 m of
20
insulated tape was tested on the Yates-Star, which is a machine that takes local critical current
measurements at 77 K over 120 m of the conductor with a resolution around 1 cm [9]. Shown
below is the critical current measurement over the length of the conductor [9].
Figure 14: Yates-Star data of local critical current over 120 m of YBCO. Two different methods of obtaining the local critical current are displayed in this graph. Disagreements between the two critical current data sets are contributed to calibration issues in the hall sensor array. Periodic IC variations and large dropouts are attributed to manufacturing induced defects [9].
Two different methods are used to measure critical current in the Yates-Star, which is the cause
for the two different critical current measurements in the above graph. Variations in critical
current observed in this 130 m length of conductor are those typically produced by Super Powers
YBCO manufacturing process [9]. Consistent critical current degradation over the whole length
of the conductor would be expected if damage was caused due to the insulation process. Further
validation is produced from the experimental success of the many different tapes insulated on
this prototype. These tapes, including tape for the Center for Advanced Power Systems, were
insulated and implemented in their respective projects. No ill effects were seen from the
insulation process, a result which we consider as a positive verification for this device.
Though tapes have been successfully insulated, there were a few problems with the prototype
that needed to be addressed. When pulling the conductor through the tubes, the spool feeding off
the tape will rotate faster, due to inconsistent tape speed, than the operator is pulling the tape.
21
This creates slack, that causes a sudden increase in tension when the slack is taken up, potentially
damaging the tape. Correction for this requires a second operator to slow the spool with the
brake, keeping constant tension in the tape. Bearings used for the pulleys were not of high
quality and would resist motion or would seize, creating high tension in the tape. When a
bearing seized or resisted motion, the operator would have to spin the pulley by hand to reduce
the tension and move the tape. Tube holders, being mounted with an epoxy weld, were difficult
to properly mount yielding poor contact on the shrink tube in some areas. Damage to some tubes
was a result of this poor contact, due to the tape catching the tube and scrunching the tube
against another tube, pulley, or tube holder. An additional drawback to this insulation method is
that it requires two operators and in its present form only allows lengths up to 130 m. In
principle, we would like this to be a one man operation that requires little training and no
specialized skill to operate so that undergraduate assistants, visiting scientist, or scientists from
other departments can operate this process with little training. Insulating lengths over 130 m will
require more than 9 passes, at 17 m of tape insulated per pass. More passes will increase
insulation time and will require many repetitions of spooling and unspooling of the tape.
Yates-Star data was performed on a tape that required 9 passes to insulate, which is the greatest
number of passes performed to date. The resulting data along with the multiple tapes insulated
on the prototype confirm that 9 passes are safe for YBCO conductor. It is likely that a greater
number of passes are safe for this conductor; however, we have found no data on the cycling
limit of YBCO conductor around pulleys. Due to this uncertainty, 9 passes will be the maximum
allowable number of passes. The prototype proved this method of insulation to be safe and
useful, yet major changes must be made to rectify certain issues. In order to fix these problems a
larger Shrink Tube Insulation Apparatus or STIA was constructed based on the prototype; one
which can handle continuous lengths of ReBCO tape exceeding 200 m.
Designing STIA
As previously described the two main issues with the prototype, preventing the insulation of
200 m or greater of ReBCO tape was the number of passes and the need for two operators.
Solutions to these two issues as well as the other problems previously stated are covered in the
section below. However, design of machinery is based not only on the functionality and
purposes of the machine but also its location. At the conclusion of the prototype and the start of
the design of STIA, the room of operations for this project was to be under construction in a few
22
months’ time. Though this would play a major role during the construction of STIA, part of the
result of this renovation was to create a dedicated, new lab spaces including space for STIA. The
final design of STIA is shown in Figure 15. Subsequent sections will discuss the design and
individual operation of the various components of STIA.
Figure 15: Final design of STIA.
Length of the Machine
Decreasing the number of passes to insulate a length of conductor requires that more tape is
insulated per pass, which is equal to the length of shrink tubing on the guide tracks. A
continuous length of 200 m of ReBCO tape must be insulated in 9 passes or less. Based on the
prototype there are two options. Three guide-track levels and pulleys could be added, increasing
the height to ~2.4 m, or track extensions less than 1m long could be added. Extending the length
of the machine as compared to the prototype was the chosen solution. As stated before, a 7.6 m
long device would be able to insulated approximately 50 m/pass, insulating 200 m in ~4 passes.
Another benefit of this extended length is that it matches the standard lengths of the shrink tube
insulation. Standard shrink tubing lengths are 1.3 m or 2.4 m. With 7.6 m long guide track
levels both standard shrink tube sizes will cover nearly the entire guide track length without
having to waste any shrink tubing. For these reasons, it was requested that a 9.1 m long space be
23
set aside for STIA in the new renovation plans. This request was accepted and therefore the
length of the machine approved. The support structure for STIA is shown in Figure 16 below.
Figure 16: The support structure for STIA made of T-slotted extruded aluminum. Highlighted here are the 1.5m long guide tracks, which are connected and supported by their respective T-slotted Al support towers.
Each guide track level is comprised of five separate lengths of angle stock, placed end to end,
each measuring 1.5 m long. To obtain a straight guide track channel, the machine shop was
commissioned to cut the channel requested that the track lengths be no more than 1.5 m long.
For this reason, vertical supports are placed every 1.5 m along STIA, to connect the two guide
track lengths for each level.
Pulleys, Motors, and Electronics
To guarantee that no damage comes to the insulation or tape, the process requires that the
operator inserts the tip of the tape into every shrink tube. It is necessary to know where the tip of
the tape is on the machine in order for the operator to slow the tape to an almost or complete stop
to insert the tip into the shrink tube. In order to create a process that requires only one operator,
motors will have to be employed to move the tape through the tubes. These motors will be
controlled by wireless remote so that the operator is free to move up and down the machine,
following the tip of the tape, to insert it into each shrink tube. A drive motor was connected to
the Collection spool to provide the driving force to pull the tape through the shrink tubes. To
address tension oscillation issues observed with the prototype, a torque motor was connected to
24
the shaft of the Storage spool to provide constant tension when moving the tape. Constant
tension will prevent the development of slack in the tape which causes sudden spikes in tension
when the tape is pulled tight. Back tension, however, was not used in the prototype and will
assuredly increase the tension in the tape when implemented in STIA. Due to the back tension
and elongation of the guide tracks, knowing how much tension the tape will experience is
essential. To understand how the tension will propagate through the guide tracks and over the
pulleys, the Capstan equation, shown below, was used.
Figure 17: The forces acting on the tape moving around a stationary pulley
Where T+dT is the tension in the direction of the desired tape movement, T is the tension in the
opposite direction of the tape movement, μ is the coefficient of friction, and is the angle
subtended by the tape. Substitution of the simplified forces for the y direction into the reduced
summation of forces in the x removes the normal force from the equations. With some equation
manipulation, the capstan equation is derived, shown below.
( 3 )
Where THold is the back tension applied by the torque motor and TLoad is the tension after being
pulled around a capstan. The Capstan equation applies to any material being pulled in tension
around a stationary object. Assuming a worst case scenario where the pulleys stop rotating, the
capstan equation can be used to find the maximum allowable back tension before failure.
ReBCO tapes tend to plastically deform above 700 MPa of tensile stress [6] which translates to
about 294 N of tension. Using this ultimate allowable tension the capstan equation was used to
find the tension in the tape after every pulley for the maximum allowable back tension, shown in
Figure 18 below. It should be noted that the guide track level 1 tension is the back tension
provided by the torque motor.
25
Figure 18: Tension and stress on each guide track level for the maximum back tension before conductor failure. Designation of guide track levels is shown alongside the maximum allowable stress for each level.
According the capstan equation approximately 6 N, or 14 MPa, of back tension will cause tape
failure if the pulley motors stop rotating while the conductor is pulled by the Collection spool. It
is not desired to operate near this failure condition, yet 6 N may not provide enough back tension
to regulate the tension oscillations during tape movement. Ideally, tension in the tape would be
as low as possible, at least half the yield stress, to lower the risk of damage to the tape.
Obviating the high tension, to allow for more back tension to be applied, drive motors and timing
belt assemblies were used to rotate the pulleys. When moving tape around the pulleys and
through STIA, the three pulleys on one side rotate in the same direction but opposite to the three
pulleys on the opposing side. Therefore, two motors are used, one to drive three pulleys on one
side, and one for the other three on the opposite side. Each respective drive motor is connected
to the center pulley through a gear coupling and to the other two pulleys through timing belts.
Figure 19 shows the drive motor and timing belt assembly setup used on either side of the
machine.
Figure 19: Motorized pulley assembly. (a) Front view of the pulley, motor and timing belt assembly. (b) Back view of the assembly.
19(a) 19(b)
26
Assuming the tape does not slip and the rotational speed of the pulley is equal to the tape
velocity, no additional tension will be imparted on the tape from rounding the pulleys. In this
condition the maximum tension in the tape will be the addition of the back tension plus the
tension due to the friction force of the tape moving through the shrink tube. Friction force
produced by the tape can be calculated by using the weight of the tape on the shrink tube over
the length of one guide track. Though YBCO is made of many layers, two materials dominate
the tensile strength: stainless steel and copper. It is assumed that the thickness of the YBCO is
made of only copper and stainless steel their respective thicknesses are 0.40 μm and 0.60 μm.
Using a tape width of 4.2 mm and a length of 7.62 m, for the length of the track, the friction
force from one guide track length is 5.4 x10-4 N, which is negligible. If the pulleys rotate at the
same speed as the linear tape speed then the increase of tension in the tape should be minimal, on
the order of a few Newtons. When the tape is being taken up on the Collection spool, the
velocity of the tape will increase with the increasing radius of the tape on the spool. In order to
have the pulleys rotating at the same speed as the tape, an encoder with a pinch idler roller will
directly measure the tape speed. This tape speed measurement system is shown in Figure 20
below.
Figure 20: Tape speed measurement system. Here YBCO tape is pulled off the storage spool through the tape speed measurement system, which is comprised of an encoder wheel and a spring-loaded pinch roller.
Contact between the tape and the encoder wheel is kept by the spring-loaded pinch roller. The
location of this tape speed measurement system is between the first guide track and the bottom
Storage spool. A size 20 series QDH20 optical encoder was chosen for its high resolution for the
low tape velocities during start up. With the tape speed measurement system designed, the
27
motors to control the pulleys and spools could be chosen. As stated before, the ideal operating
tension would be low, realistically less than 100 MPa, one seventh the maximum yield tension.
The drive motor would then have the minimum requirement of being able to apply
approximately 47 N with a spool radius ranging from 57 mm to 133 mm. Standard spools that
store the ReBCO tape have an inner diameter of 57 mm and will increase to 133 mm when
completely spooled with tape. These will be the spools used on the machine to house and store
the tape. Back tension applied by the torque motor could range anywhere below the desired
100 MPa limit due to the minimal increase in tension through the machine. A torque motor that
would be able to apply a minimum back tension of 25 N would be satisfactory. During
operation, the pulley motors would not have to overcome much tension, as long as the pulleys
are rotating. Aside from the resistance to rotation due to the inertia of the pulleys and shafts, the
main resistive force to rotation will be from the change in tension around each pulley. The
tension that the motors would have to overcome is the combined force resisting each pulley
connected to the motor. The force resisting the rotation of a pulley is the change in tension as the
tape rounds the pulley. For the motor to rotate the three pulleys, it must overcome the sum force
of resistance acting on the pulleys. Assuming pulley rotation and tape speed are equal and no
slip occurs, synchronous pulleys motors would share the back tension load. If a range of
6 N-25 N for back tension is considered the pulley motors will have to overcome a maximum of
25 N split between the two motors. As motors were about to be purchased, it was discovered
that a retired process for winding coils had put a winding rig out of commission. Three motors
and motor controllers from this winding machine were donated to this project by the magnet
design and technology department at the National High Magnetic Field Laboratory (NHMFL).
Two different drive motors and a torque motor were repurposed for STIA. Table 1 below shows
the requirements for each motor and the capabilities of the donated motors.
Table 1: Motor description, specification, and requirements
Donated Motors Motor type Motor Function Output Torque Desired Output Torque
From Table 1 it can be seen that all motors are over qualified for their respective motor jobs. For
these reasons, the three motors described above were used. A matching pulley drive motor was
purchased to allow for a simpler design of a control box that will regulate the motors, encoder,
and provide wireless control for the operator.
Control Box and Electrical Team
Electrical design and construction of the control box was done by the electronics workshop at the
NHMFL. It was uncertain how fast the tape could proceed through the tubes, how much back
tension needed to be applied, and what motor accelerations were possible without harming the
tape. For this reason, the control box was designed with the ability to manipulate the back
tension, acceleration of the drive motors, and the speed at which the motors rotate. Shown below
are the wireless remote and the front panel of the control box that runs off of a standard 120 v
wall outlet.
Figure 21: Handheld wireless remote and front panel of the control box. (a) Handheld wireless remote able to Run, Jog, and Stop the tape. (b) Front panel of the control box which regulates the motors, encoder, and wireless remote.
21(a) 21(b)
29
Some features on the front panel of the control box are obvious in function; emergency stop,
power on/off, acceleration potentiometer, torque potentiometer, and wireless on/off are self-
explanatory. Three modes exist for this device: Run, Jog, and Stop. Run is the mode which is
used to move the tape through shrink tubes or wind onto spools; this is the average operating
speed for the tape. Jog provides a much slower speed and acceleration and is used to move the
tape forward in a controlled manner for moments when the tip of the tape needs to be inserted
into a shrink tube or other situation where a small movement is needed. Stop is another self-
explanatory function, mentioned only to note that unlike emergency stop this only stops the
motors but does not cut off power to the system. The potentiometers labeled Run and Jog set the
speed at which these two modes move the tape. Direction of the tape is control by the switch
labeled Direction with the two options forward and reverse; which move the tape towards the
Collection spool or the Storage spool respectively. Switches labeled Brake and Motors allow for
the operation of just the torque motor or just the Collection spool motor. Brake when switched
to ON will prevent the Collection spool motor from moving whereas if the Motor switch is
flipped to take all motors besides the Collection spool motor are stopped. These modes allow for
individual spool adjustments, as needed for various situations such as when the tape is to be
connected or disconnected from a spool. The switch labeled torque is to allow the torque motor
to be on at all times, in manual mode, even when stop is pressed. When flipped to Auto the
torque motor is turned on and off with the drive motors. Manual mode allows the tension to be
held when the shrink tube is being heated onto the tape. Finally, the wireless remote has an
on/off switch, a Run, Jog, Stop, and trigger button. To use the Run and Jog functions, the trigger
button must be pressed at the same time, which is to prevent any accidental movement that could
possibly damage the tape.
Tube Holders and Heat Gun Assembly
The same concept for restricting the motion of the shrink tube was used for STIA; however,
some changes were made to improve their performance. Identical adhesive backed rubber strips
as those used in the prototype were mounted to 1.4 m long 3.2 mm thick aluminum angle stock.
Instead of connecting this to a hinge, the angle stock was connected to two 6.35 mm thick
aluminum strips mounted to the back of the guide track levels. This 6.35 mm thick aluminum
strip is known as the tube holder guide strip, which is perpendicular to the guide tracks. In
Figure 22 below the tube holder guide strip set up is shown.
30
Figure 22: Tube holder operation. (a) Tube holder in the top locked position. (b) & (c) Show the movement and engagement of the tube holder, which is operated by grasping the handle and sliding the holder down.
Vertical movement of the tube holder allows for better contact between the rubber strip and the
shrink tubing, partially due to the precise mounting as compared to the prototype tube holders. It
should be noted that when the tube holders are lifted and locked at the top there is enough room
to move the heat gun through. Only slight modifications were made to the heat gun mounting
assembly, the most noticeable of which is the addition of the extension cord shelf. If the heat
gun assembly is suddenly stopped and jerked while shrinking the insulation onto the tape,
damage could be caused to the tape. To prevent the extension cord from getting caught and
causing such a scenario a plastic catch shelf was installed.
Figure 23: Collection shelf for heat gun power cord. (a) Heat gun cable wrapped around the support rod to allow the heat gun to move vertically without having the cable interfere with heat treatment. (b) Collection shelf placement on STIA.
31
As seen in Figure 23 above the black plastic shelf provides an area where the cord can move
freely without catching or causing a trip hazard. The cord is also wound around the height
adjustment shaft of the heat gun assembly to prevent the cord from interfering with the heating
process. This prevents the power cable from causing trip hazards or damage to the tape during
the heat shrinking process.
Spool Holder Design and Functionality
Longer lengths of tape must be handled safely and efficiently on this device. Part of the problem
with the prototype was spooling the tape. The problem came in with the string. The string had
to pull the tape through the tube which meant that the string had to be mounted on to the same
motor shaft as the spool. But the tape cannot be wound upon the string since the spools are not
large enough for the string and the tape. Also, there is a concern that the un-level winding of the
string would cause damage to the tape as it was wound on top of the string. Depending on the
tension when the tape is being spooled up, the stress could be enough to distort the tape and
create localized damage. To decrease the time to insulate the tape, a spool hub was designed to
allow for the string to remain connected to the tape. In Figure 24 below, the newly designed
spool hub, tape guide, and string guide set up and operation is shown.
Figure 24: Spool hub operation. (a) Spool is put onto the hub with removed front flange. (b) The string is pulled to the spool’s front nub. (c) String wrapped around front nub with needle secured in key way slot. (d) String is spooled around nub. (e) Tape is put into tape channel cut into the spool. (f) Front flange put back on spool to collect tape.
24(a) 24(b) 24(c)
24(e) 24(d)
24(d)
32
When the string exits the last shrink tube, it is threaded through the roller and eye hook, finally
connecting to the little nub on the spool hub. As seen in Figure 24 above, the spool is already
mounted to the spool hub without the front flange. The string is wound up onto the front nub
until the tape is right before the eye hook. Using the Jog function the tape is pulled by hand till it
reaches the spool. At this point any slack in the string is wound up onto the nub, the tape is
placed in the groove in the spool, and the front cover is mounted back onto the spool. Placing
the spool’s front flange over the string prevents the string from unraveling and allows for the
tape to be spooled up on to the spool. Using this new spool hub prevents the need for the string
to be disconnected and re-soldered onto the tape every time decreasing the overall insulation
time.
Kyushu University Collaboration
Towards the completion of STIA’s construction, collaboration between the ASC and the Kyushu
University in Japan to build a layer wound YBCO coil was arranged. The University wanted use
of our unique coil winding process and large resistive magnets in the user facility at the NHMFL
for a 200 m length of Fujikora-produced YBCO. Part of the coil winding process would be the
use of STIA to insulate the continuous 200 m length of YBCO. However the insulation device
was designed for a 4 mm wide tape and their tape is 5 mm wide. Guide track channels are
exactly 5 mm in width and will restrict the movement of the tape, possibly damage incurring
damage to the tape or shrink tube. Desiring to move forward with this collaborative arrangement
would include changes to the guide tracks and other various parts of the insulation device.
Figure 25: Comparison of Super Power’s 4mm wide YBCO tape, on the left, and the Fujikora 5mm tape, on the right.
33
Shown above is the comparison between the YBCO made by Super Power Inc. that the ASC
uses for their layer wound coils and the Kyushu University YBCO to be used for the
collaborative coil. To accommodate this new 5 mm tape, the guide tracks were sent back to
Exotic Machining to have the guide channels widened to 6.36 mm. Being the same width as the
guide channels, the delrin sled on the heat gun nozzle had to be widened as well, which meant
building a new sled altogether.
34
EXPERIMENTAL PROCEDURES AND TESTING
Testing the Operation of STIA
During this time the renovations commenced on the current staging area, which required STIA to
be relocated to a new staging area. Clearing a space on the first floor of the ASC provided room
for a new staging area, which once the modifications to STIA started the rest of the machine was
broken down and moved. Once all modifications had been completed STIA was erected as
before and was ready for testing. Currently the longest continuous length of YBCO available is
approximately 200 m. Though, this is quite difficult to produce and can take many months and
many batches of YBCO till a 200m length is acceptable. If this 200 m length of conductor is
damaged on STIA, there will be no way to replace the conductor and the project will have to be
terminated. Before attempting to insulate the Kyushu YBCO, multiple tests had to be performed
with another tape. Proper motor movement, electrical connections, and a test insulation run
through were needed to discover any operational issues with the system. After the device was
tested, the Kyushu 200 m long conductor will be the first live conductor insulated on STIA.
Electrical Testing
Due to the relocation of the device, all the electrical components were checked for proper
connections, loses fittings, or any changes due to the move. Electrical connections were
observed to be functioning properly. As a result, the functionality of the individual electrical
components was tested as well. Motors were tested for proper rotation, both in forward and
reverse. Changes to Acceleration, Jog speed, and Run speed were also tested in the forward and
reverse mode for the drive motor, with proper operation observed. Torque adjustments tested in
these modes were operating as desired. Encoder signals were responsive in both modes, as
evident by pulley rotation, increasing and decreasing rotational speed with the linear tape
movement. Switching from the “All” mode under the Motor switch to “Take” caused all motors
to stop functioning except the Collection Spool motor as desired. Likewise the brake switch,
when engaged, ceases the function of all drive motors leaving the torque motor operational.
Emergency stop, on/off, and the wireless remote function were all successfully tested. All
functions of the control box as well as the wireless remote were working individually. In order
to make sure that all the functions of the spool motors, pulley motors, and encoder would work
together a length of tape would have to be put on to the machine.
35
Substrate Testing
Due to the possibility of damaging the ReBCO, approximately $80/m, another material was used
to test the insulation process. ReBCO material left over from concluded projects can find
purpose as test coils, current leads, practice or test material for coil joints, and many other
miscellaneous but necessary uses. Long lengths of bare Substrate can also be used to practice
winding coils since it has approximately the same tensile properties and dimensions as YBCO.
A 70 m length of this bare substrate is owned by the ASC lab group for this very purpose.
Testing the motors, encoder, and the overall machine operation with this substrate would provide
near identical results to a test with YBCO tape. This length of substrate was therefore used to
test the functionality of STIA. First, to verify the motor design a series of tension test were
performed on the device to discover the actual tension seen in the conductor during operation. A
handheld scale, shown in Figure 26 below, measured the force required to move the tape from a
static position. Taking tension measurements from the static position should show the maximum
stress seen when attempting to move the tape. A clamp made in-house was used to connect the
tape to the handheld scale, as shown in Figure 26 below.
Figure 26: Handheld scale connected to in-house made pressure clamp
A spool of YBCO stainless steel substrate was mounted onto the torque motor and the handheld
scale and pressure clamp were connected to the end of the substrate. Tension data was taken at
the beginning and end of every guide track, to measure the tension before and after each pulley.
Four different tension tests were performed with the stainless steel substrate, where all test were
performed with stationary pulleys. Back tension was not applied to Test 1, nor did the tape move
through shrink tubing. Test 2 was the same as Test 1 with the addition of the substrate moving
Handheld Scale
Pressure Clamp
36
through shrink tubing. For Test 3 and 4 the previous two tests were repeated, respectively, with
the addition of applied back tension from the torque motor at a setting of 0. A list of these test
results are displayed in Table 2 below.
Table 2: Result from four tension test with different testing conditions.
Figure 27: Tension test data with capstan equation predicted values. The capstan equation was used to develop the predicted values of stress with respect to the location on STIA. Experimental values displayed are those of test 4 with the torque motor setting of 0 and with shrink tube insulation on the tracks.
37
Using the back tension of Test 4, which is 4 N, the capstan predicted tensions higher than those
measured. Tension differences are most likely due to lower than expected coefficient of friction
between the tape and the pulleys. A possible cause could be due to dust acting as a lubricant
between the tape and the pulleys. Figure 27 confirms that tension throughout STIA is as
expected in case of a scenario where the pulleys seize and stop rotating. Confidence for proper
motor operation comes from the results from a tension test performed with proper pulley motion.
Stainless steel substrate was pulled through the full length of STIA with a back tension of 6.6 N,
a torque motor setting of 1. The force required to move the tape with pulley assistance is about
9.8 N with an applied back tension of 6.6 N. A minor increase in tension is expected, which is
due to the friction of the tape against the shrink tubes. This confirms that pulley assistance will
allow for low tension operation with a larger range of back tension settings for the torque motor.
Having good agreement with the experimental and predicted results, the substrate can now be put
on STIA to find the best operational settings. All setting were put to zero for no tape movement
and then incrementally increased to see how the tape reacted to each setting. Table 3 below
shows the order of the increments tested on the tape, which were all performed in the forward
direction towards the Collection spool.
Table 3: Control box settings; Run, Acceleration, and Torque settings for different trial runs
For trials 1-3, the tape only ran along the bottom tape guide track, rounding the pulley at the end,
and then being directly connected to the drive motor. This was to reduce the tension in the line
to reduce the risk of damaging the tape during testing. Optimum operational settings were those
of test 3. During this test, the tape flowed smoothly from the torque motor around the pulley and
onto the drive motor at a reasonable tape speed. Other operating conditions with higher
accelerations or lower back tension settings saw tension oscillations in the tape. After a period
38
of time, the oscillations would stop and the tape would move along smoothly. By changing the
torque setting and the acceleration of the drive motor, the system reduced or eliminated the
oscillations. Operational settings were varied in this manner as the tape was move through the
next guide track level, bypassing the other levels, and connecting directly to the Collection spool.
In this way, the operational setting would be close to the optimal settings once the tape was put
through all the guide tracks and rounded every pulley. After testing the tape on five guide tracks
and 4 pulleys, the best operational setting was trial 7. At this setting, minor oscillations were
observed in the tape but would decay within a few seconds. Finally the tape was put around all
the pulleys and over all the guide tracks and tested with the previous settings to make final
adjustments to the operational settings. Yet when the Run button was pushed, the tape
experienced tensile failure after just a few seconds, exposing a flaw in the machine design.
When the drive motor engaged and applied force to the tape, the tape did not move due to the
high tension around the pulleys. After the tension built up, static friction was overcome and the
tape rapidly moved forward. This spike in tape speed was measured by the encoder which
caused the pulleys rotate to match the speed. High friction on the pulleys caused the pulleys to
move the tape forward faster than the drive motor was taking up tape, which created slack in the
line. Because of this slack, the pulleys did not rotate until the slack was taken in. Once the slack
was taken in then tension increased to where the pulley motors were not able to rotate, causing
the tension to rapidly build up as predicted by capstan equation. Tape failure was due to the lack
of pulley movement caused by the encoder wheel not viewing tape movement upstream of the
process, allowing tension to build up to catastrophic levels. A scientist from the Kyushu
University was to visit for about a month and a half, in which time the coil would have to be
insulated, wound, and tested. Manual insulation was no longer an option due to the renovation
and the prototype did not have wide enough guide channel groove. STIA is only method
available to insulate this conductor; this forced a quick but safe solution to be developed to get
STIA operational. The encoder wheel, from the tape speed measurement system, was removed
from its mounting and replaced with an idle roller. A handle was mounted to the encoder wheel
so the pulleys could be turned by hand.
The spool of substrate, mended by spot welding, was mounted to the Storage spool torque motor.
String was again pulled through the length of STIA and attached to the drive motor, which was
39
operated via control box. Another operator used the hand crank to rotate the pulleys to match the
tape speed. With this method, the tape was successfully moved through the full length of STIA,
under low tension and experiencing no damage. After a few attempts, it was discovered that no
tension oscillations would occur in the tape if an operator created slack at the torque motor end.
With the method of one operator creating slack and manually rotating the pulleys while a second
operator regulated the drive motor, STIA was found to safely insulate tape without damage.
STIA Modifications
Upon the arrival of the scientist from Kyushu University, two lengths of their 5 mm wide YBCO
were provided, a 200 m length and a 25 m length. For the actual coil, the 200 m length would be
used, while the 25 m length would be for test winding and insulation practice. During insulation
trial runs with the 25 m tape, a few problems were discovered. Connecting the string with solder
did not work with this tape due to the different manufacturing process for this particular type of
YBCO. This YBCO has a copper layer soldered to the tape instead of electroplated Cu
surrounding the tape like the standard tape used by the ASC. Since this copper layer does not
wrap around the tape solder is used to adhere the copper layer to the tape. This meant that
whenever the copper lead was soldered onto the tape delamination would occur. To solve this
issue, a punch with a 1.6 mm hole was used so the string could be directly tied to the YBCO.
Figure 28: Delamination of Japanese 5mm tape due to soldering and bending.
About 5 mm from the tip of the tape, the hole was punched to prevent the tip from being able to
scratch the shrink tube by pivoting around the string in the hole. The tip of the tape was also
rounded to decrease the risk of damage and allow for the tape to enter into the shrink tube more
40
easily. Punching a hole in the tape caused burrs, which were removed with a fresh scalpel blade.
However, once the string was finally connected to the tape pulling it through the shrink tubing
proved impossible. Being a wider and thicker tape, the standard shrink tubing used for ASC
conductors would not work for this Japanese YBCO. As mentioned in the introduction, the
equation to determine the appropriate shrink tube size was once again consulted. Upon receiving
the properly sized shrink tubing, the modified insulation process was tested. Having successful
trial runs with all the new modifications to the procedure, it was decided to finally put the 200 m
length of Kyushu YBCO onto the machine to be insulated.
41
RESULTS
Results from the Kyushu Collaboration
Results from this test are presented in this section along with the total cost of the machine.
Performance specifications, machine features, and their implications will be discussed, in the
following discussion section. Insulating the 200 m of ReBCO received though the Japanese
collaboration was the first insulation run on STIA with live conductor. Insulating this length of
conductor took a total of 24 hours to complete, which was spread over a week due to end of
semester test conflicts between the two operators. Set up and clean up procedures are included
in this 24 hours insulation period. However, completing one pass takes approximately 4 hours to
complete. Using this information, a summary of the STIA performance specifications are listed
in Table 4 below.
Table 4: STIA performance specifications.
Insulation rate 100m/8hrs
Tape Width Range 4mm-5mm
Tape Length Range 1m-450m
Tape Insulated/Pass 53m
Substrate testing along with the insulation of the Japanese YBCO conductor showed that 4 mm-
5 mm wide tapes can be insulated on this device with properly chosen shrink tubing.
Continuous lengths from 1 m-450 m, based on the number of passes, (nine) were shown to be
safe on the prototype. Nine passes showed no degradation in critical current over the length of
conductor. Approximately 53 m can be insulated per pass on STIA, which when multiplied by
nine is the number of safe passes, gives an upper limit of 450 m of continuous conductor. The
final cost of STIA including the bill of material, electronics, and machining cost are shown in
Table 5 below.
42
Table 5: Total cost of materials for STIA.
Category Description Cost
Hardware & Software Structure $4,544
Spools and Pulleys $3,695
Electronics $1,104
Screws, Bolts, and Fasteners $488
Miscellaneous $368
Machining Angle Stock 1.2m length $817
Pulleys $780
Guide Planks $1,313
Track $1,225
Total $14,333
43
DISCUSSION
Operational Performance
STIA, as originally designed, failed to insulate any length of ReBCO material, succeeding only
in breaking a 70 m length of substrate. However, with modifications, back tension and pulley
speed were able to be manually regulated allowing for a 200 m length of YBCO to be insulated.
No visible damage was observed over the 200 m length, by the operator or the visiting Kyushu
scientist, as a result of the insulating procedure. Shrink tubing insulation provided complete
coverage over the entire length with no tears, holes, or gaps. Insulation of the continuous 200 m
length of the Kyushu University YBCO was declared successful, both by the operator and the
visiting Kyushu scientist. This process was completed in the desired time frame for the
collaborative effort. Increasing the length of the device not only increased the amount of tape
insulated per pass by ~36 m, but also eliminated shrink tube waste. Approximately 50.8 mm had
to be cut off from each length of shrink tubing for the prototype, at $9 for a 1.3 m length of
shrink tube $45 dollars is wasted to insulated 200 m of conductor. Longer lengths of shrink
tubing can also be used, such as the 2.4 m lengths instead of the 1.3 m lengths. Shrink tubes
need to be overlapped to guarantee complete coverage, longer shrink tubes decrease the number
of overlaps. Decreasing the number of overlaps increases the packing density of the conductor
when wound into a coil, an important coil design parameter. Longer shrink tubes also decrease
insulation time by reducing the number of times the tape must be stopped to be inserted into a
shrink tube. Based off of the Kyushu 200 m insulation an insulation rate of 100 m/8 hrs was
reported. Using this rate, hourly wage for graduate and undergraduate assistants, and shrink tube
insulation prices the cost to insulate a 200 m length of ReBCO is shown in Table 6 below. In
contrast the cost to insulate 200 m by the manual insulation process is shown as well.
Table 6: Comparison of the cost per meter to insulate a 200 m length of YBCO with heat shrink tubing for Manual Insulation and STIA.
Insulation Method Work Days Labor Cost Insulation Cost Total Cost Cost/m
STIA 2 $436 $1,524 $1,960 $9.80
Manual Insulation 6 $1,718 $1,524 $3,242 $16.21
A reduction of $1,280.00 in labor cost is seen when insulating 200 m of conductor with STIA as
compared to the Manual insulation procedure. After insulating another 2 km of YBCO, the
44
savings in labor cost will pay for the total cost of the machine. STIA takes a third of the time to
insulate 100 m as compared to manual insulation, translating to four work days of difference for
a 200 m length of ReBCO conductor. ReBCO manufacturers are being pressured by the
superconducting community to produce longer continuous lengths. It is the hope that within a
few years’ time lengths greater than 1km will be available. Manual insulation would need six
work weeks to insulate 1 km, which is impractical, whereas STIA requires only two work weeks.
Though it has not been tested, it is estimated that STIA can handle lengths in excess of 1 km. If
able to insulate 1 km, STIA would be able to operate for years to come with the ever increasing
lengths of ReBCO conductors.
Features
STIA has the ability to be customized and is applicable for many different needs. As shown
through the collaboration with the Kyushu University, STIA is able to insulate different widths
of conductor, limited only by the guide track channel and pulley width. YBCO tapes are
produced in a variety of widths from 3 mm-12 mm, for which different versions of STIA can be
designed and built. Length of the guide tracks can also be varied for different applications, such
as labs with limited space. Compact version of STIA could be manufactured for 10 m-50 m
REBCO insulation needs by reducing the guide track level lengths. Customization gives STIA a
flexibility to conform to research needs if other laboratories find shrink tube insulation as
necessary as the ASC for YBCO coil production [7]. Due to this flexibility of STIA and the
possibility of other research institutions desiring such a device, such as the Kyushu University, a
provisional patent has been applied for STIA [8].
Issues
STIA successfully insulated 200 m of YBCO, yet there are major issues that must be addressed.
Currently the safety of the tape relies on the technique and experience of the operators. Manual
operation, such as pulley speed and back tension regulation, increases the risk of damaging or
breaking the tape. For example, one operator has to control both the back tension and the pulley
speed. With divided attention, it is easy to focus on one task more than the other. If the pulley
speed drops due to this lack of attention, the tension in the tape will rapidly increase, possibly
causing damage to the tape. In the reverse situation too little slack could result in damage or tape
failure and too much slack could cause the tape to touch the floor. If the tape touches the floor,
45
dust and dirt can attach to the tape and possibly damage the tape or shrink tube through
scratches; or the tape could catch on the machine supports ruining a section of tape. Another
issue related to manual operation is the need for an extra operator. Having two operators
increases the cost of insulation, as well as the risk to the tape from human error. The current
design will not allow STIA to operate outside this manual regulation. Correcting this issue will
require a new control method for moving the tape at either a constant velocity or in constant
tension.
46
CURRENT WORK
Tape Movement
There are two main design techniques for moving web material, such as tape, through a machine;
constant velocity or constant tension of the web. STIA was designed to be a constant tension
device, relying on properly tuned motor acceleration speeds and back tension to prevent tension
oscillations. However it was shown that this design was inadequate due to the downstream tape
measurement system that could not see tape movement in time to rotate the pulleys. Tension
exponentially increases around stationary pulleys in accordance with the capstan equation. This
created a failure condition due to the tension applied by both the Collection and Storage spool
motors. To correct this issue and move towards a single operator process, a constant velocity
design is currently being pursued.
To create this constant velocity STIA, the torque motor will be removed and replaced with a
drive motor identical to the top Collection spool drive motor. Having identical motors will make
wiring and electrical logic simpler and easier to handle. It is desired that the operator can vary
the speed at which the tape moves through STIA with a potentiometer on the control box.
Moving the tape at a certain speed requires the motors to feed off or take up at the same rate.
Motor speed is not constant in this case due to the ever increasing or decreasing radius from
spooling or unspooling tape. To regulate the differences in speed, a potentiometer with a
follower arm will monitor each spool’s change in diameter.
Figure 29: Follower arm potentiometer used to monitor the change in radius of the Storage and Collection spools.
When the operator sets a tape speed, the motors will rotate at a certain angular velocity
corresponding to their respective potentiometer position, which will continue to regulate their
respective motor speed throughout the insulation process. Actual tape speed will be measured by
47
the encoder wheel used previously on STIA which will be moved up to the middle or fourth
guide track. This will provide a reference speed for the constant velocity control loop. Constant
velocity for the tape will be achieved with such a setup but tension is another matter and is of
great importance. Tension developed in a web between two rollers can be understood using the
mass conservation law for a control volume.
∙ ∮ ∙ ( 4 )
Figure 30: Simplified diagram of tape movement between two spools. Illustrated is an assumed control volume which the tape passes through, which is view as the only mass transfer through the volume.
Where V1 and V2 are the respective velocities of the Storage and Collection spool and ρ is the
density of the YBCO tape. If the control volume is taken as shown in Figure 30 above then the
mass flowing in and out of this volume must be conserved. Conceptually the mass can be
considered to flow through the cross sectional area of the tape entering the volume and the cross
sectional area exiting the volume. If the system is assumed to be in a steady state, all time
dependence is removed and the equation is simplified as shown below.
∆ ∆ ( 5 )
48
From the equation above, it can be seen that the ratio of spool velocities directly affect the strain
in the tape. Considering that typical stainless steel, a common substrate for YBCO, yield strains
are quite low; the ratio of spool velocities must remain close to unity to prevent damage or
failure. Constant velocity machine designs therefore require well regulated motor velocities. To
alleviate this strong dependence upon the spool velocities, two dancer rollers will be added to
STIA, one before each spool. Dancer rollers are used for controlling tension in the tape as well
as absorbing tension oscillations. The set up that will be used for STIA is shown in Figure 31
below.
Figure 31: Dancer roller and potentiometer assembly.
Pulling the tape through STIA will require a certain amount of tension as discussion in previous
sections. The dancer will compress under this desired tension providing a passive open loop
tension regulation system. As the tape is pulled through STIA, any minor tension oscillations
will be absorbed by the dancer rollers. However passive dancer roller systems can only help
absorb tension oscillations, these rollers provide no tension regulation. This is why a
potentiometer swing arm is connected to the dancer roller system. With the potentiometer swing
arm, the dancer roller displacement will be tracked allowing for a PI loop tension control as well
as direct measurement of tension. At the desired operating tension, the dancer roller will be
designed to be compressed to around 50% of its travel range. The potentiometer swing arm will
denote this position as the desired dancer roller height. If tension increases or decreases during
operation, the dancer roller will be displaced accordingly moving the swing arm away from the
desired position. While moving the tape, one of the dancer rollers regulates a drive motor while
49
the other only measures tension. Drive motor regulation is only needed when the drive motor is
feeding tape into the system. This will prevent the buildup of back tension in the system to
obviate tape damage or failure. Direct tension measurements will be performed by the dancer
roller nearest the drive motor taking tape in. Tension in the tape is calculated using simple
mechanics equations relating the tension required to compress the spring a certain distance. This
distance is monitored by the potentiometer swing arm and displayed on an LCD screen. Alarms
will be connected to this tension measurement to alert the user if the tension is dangerously high,
as well as an emergency shut off if the tension increases past a certain threshold. To regulate all
of these potentiometers, motors, LCD screens, alarms, shut offs, calculations, and encoders an A
Tmega644P microcontroller will be used. This microcontroller will be used to regulate STIA as
a whole, processing all signals and calculations needed. A USB port will be available as a type
of com port where a live read out of all the values the microcontroller is reading can be viewed.
Once installed and programed the microcontroller will be able to be reprogramed through this
port. This will also allow changes to the PI feedback loop gains. Certain gains and changes can
be made through this port. Eventually a Graphic User Interface (GUI) will be developed to
allow a continual display of these values.
50
SUMMARY
REBCO coated conductors, used in high field coils, are prone to delamination of the HTS film
from the conductor substrate. Medical heat shrink tubing prevents degradation from
delamination by providing weak planes around the conductor, in which the conductor and epoxy
encapsulate are decoupled. Manual insulation methods proved to be impractical due to required
operation space, personnel, and insulation time. To show that shrink tubing could be practically
applied to long lengths of REBCO conductor a new method of insulation was pursued. A novel
concept for insulating long lengths of REBCO conductor was developed and prototyped.
Continuous lengths of REBCO tapes, up to 130m long, were successfully insulated on the
prototype at a rate of 12.5 m/h. Development of this mechanized process to accommodate
lengths of REBCO up to 200 m in length was pursued. Collaboration between the Kyushu
University and the ASC provided a 200m length of Fujikora-produced YBCO to be insulated on
this newly developed process. Though issues with the mechanized process required two
operators, the Fujikora tape was insulated without damage. At a rate of 12.5 m/h, the 200 m of
YBCO was insulated in approximately 3 work days including set up. Current work on this
insulation device is being carried out to correct the issues with STIA and move to a single
operator process.
51
REFERENCES
[1] Noe, M., R. Heller, W.H. Fietz, W. Goldacker, and Th. Schneider. “HTS Applications”. Proc. of WAMSDO Proceedings, Forschungszentrum, Karlsruhe, Germany. 94-97. Web.
[2] YBCO: Tape, ∥Tape-plane, SuperPower "Turbo" Double layer (tested NHMFL 2009). Source: Aixia Xu and Jan Jaroszynski, June 2009. 20 T depression due to He bubble, dashed line estimates true performance
[3] Yanagisawa, Y., H. Nakagome, T. Takematsu, T. Takao, N. Sato, M. Takahashi, and H. Maeda. "Remarkable Weakness against Cleavage Stress for YBCO-coated Conductors and Its Effect on the YBCO Coil Performance." Physica C: Superconductivity 471.15-16 (2011): 480-85. Web.
[4] Takematsu, T., R. Hu, T. Takao, Y. Yanagisawa, H. Nakagome, D. Uglietti, T. Kiyoshi, M. Takahashi, and H. Maeda. "Degradation of the Performance of a YBCO-coated Conductor Double Pancake Coil Due to Epoxy Impregnation." Physica C: Superconductivity 470.17-18 (2010): 674-77. Web.
[5] Gupta, Ramesh, Mike Anerella, John Cozzolino, George Ganetis, Arup Ghosh, George Greene, William Sampson, Yuko Shiroyanagi, Peter Wanderer, and Al Zeller. "Second Generation HTS Quadrupole for FRIB." IEEE Transactions on Applied Superconductivity 21.3 (2011): 1888-891. Web.
[6] Walsh, R. P., D. Mcrae, W. D. Markiewicz, J. Lu, and V. J. Toplosky. "The 77-K Stress and Strain Dependence of the Critical Current of YBCO Coated Conductors and Lap Joints." IEEE Transactions on Applied Superconductivity 22.1 (2012): 8400406. Web.
[7] Hilton D. K., M. Dalban-Canassy, H. W. Weijers, U. P. Trociewitz, and D. C. Larbalestier. Encapsulating a Superconductor Tape, Comprises Providing a Heat-shrink Tubing Made of an Insulating Material, Placing the Heat-shrink Tubing around the Superconductor Tape, and Applying an Encapsulate over the Heat Shrink Tubing. Hilton D. K., assignee. Patent US2012142539-A1 ; US8530390-B2. 06 Dec. 2011. Print.
[9] A. Stangl. In Situ Comparison Between Direct and Magnetization Reel-to-Reel Critical Current Measurements in REBCO Coated Conductors. Technische Universität Wien, Masters thesis 2013
[10] CNRS International Magazine. Physics. Most Powerful NMR Spectrometer Now Operational. CNRS. N.p., n.d. Web.
[11] Davies, Antony N. "GHz NMR." Online posting. NIR News. International Council for Near Infrared Spectroscopy, 2013. Web.
52
[12] Zlobin, A.V., G. Ambrosio, N. Andreev, E. Barzi, B. Bordini, R. Bossert, R. Carcagno, Dr. Chichili, J. Dimarco, L. Elementi, S. Feher, V.s. Kashikhin, V.v. Kashikhin, R. Kephart, M. Lamm, P.J. Limon, I. Novitski, D. Orris, Y. Pischalnikov, P. Schlabach, R. Stanek, J. Strait, C. Sylvester, M. Tartaglia, J.C. Tompkins, D. Turrioni, G. Velev, R. Yamada, and V. Yarba. "R&D of Nb3Sn Accelerator Magnets at Fermilab." IEEE Transactions on Applied Superconductivity 15.2 (2005): 1113-118. Web.
[13] J. G. Bednorz and K. A. Müller. Possible high Tc superconductivity in the Ba-La-Cu-O system. Zeitschrift für Physik B. 64: 189-193, 1986
[14] Parrell, J.A., Youzhu Zhang, M.B. Field, P. Cisek, and Seung Hong. "High Field Nb/sub 3/ Sn Conductor Development at Oxford Superconducting Technology." IEEE Transactions on Applied Superconductivity 13.2 (2003): 3470-473. Web.
[15] Trociewitz, Ulf P., Matthieu Dalban-Canassy, Muriel Hannion, David K. Hilton, Jan Jaroszynski, Patrick Noyes, Youri Viouchkov, Hubertus W. Weijers, and David C. Larbalestier. "35.4 T Field Generated Using a Layer-wound Superconducting Coil Made of (RE)Ba2Cu3O7−x (RE = Rare Earth) Coated Conductor." Applied Physics Letters 99.20 (2011): 202506. Web.
[16] H. Kamerlingh Onnes. Leiden Comm. 120b, 122b, 124c (1911)
[17] W. Meissner and R. Ochsenfeld. Ein neuer Effekt bei Eintritt der Supraleitfahigkeit. Naturwissenschaften, 21(44):787-788, 1933
[18] Lei, Hechang, Kefeng Wang, Rongwei Hu, Hyejin Ryu, Milinda Abeykoon, Emile S. Bozin, and Cedomir Petrovic. "Iron Chalcogenide Superconductors at High Magnetic Fields." Science and Technology of Advanced Materials13.5 (2012): 054305. Web.
[19] Larbalestier, D. C., J. Jiang, U. P. Trociewitz, F. Kametani, C. Scheuerlein, M. Dalban-Canassy, M. Matras, P. Chen, N. C. Craig, P. J. Lee, and E. E. Hellstrom. "Isotropic Round-wire Multifilament Cuprate Superconductor for Generation of Magnetic Fields above 30 T." Nature Materials 13.4 (2014): 375-81. Web.
[21] 2014Savvides, N., and S. Gnanarajan. "YSZ Buffer Layers and YBCO Superconducting Tapes with Enhanced Biaxial Alignment and Properties."Physica C: Superconductivity 387.3-4 (2003): 328-40. Web.
[22] "More Flux." National High Magnetic Field Laboratory. Web. 18 June
[23] Duckworth, Robert C. Contact Resistance and Normal Zone Formation in Coated Yttrium Barium Copper Oxide Superconductors. Thesis. Thesis (Ph. D.)--University of Wisconsin--Madison, 2001. N.p.: n.p., n.d. Print.
[24] Mutlu, I.H., E. Celik, and Y.S. Hascicek. "High Temperature Insulation Coatings and Their Electrical Properties for HTS/LTS Conductors." Physica C: Superconductivity 370.2 (2002): 113-24. Web.
53
[25] "DuPont Kapton Polyimide Film General Specifications." DuPont Kapton, Web.
[26] Polikarpova, M. V., P. A. Lukyanov, I. M. Abdyukhanov, V. I. Pantsyrny, A. E. Vorobyeva, N. E. Khlebova, S. V. Sudyev, A. K. Shikov, and V. V. Guryev. "Bending Strain Effects on the Critical Current in Cu and Cu–Nb–Stabilized YBCO-Coated Conductor Tape." IEEE Transactions on Applied Superconductivity 24.3 (2014): 1-4
54
BIOGRAPHICAL SKETCH
Andrew Whittington enrolled at the Florida State University for his undergraduate degree in
mechanical engineering in 2007. Throughout his undergraduate career Andrew was active in
multiple organizations such as the Marching Chiefs, Rugby club team, Ultimate Frisbee club
team, Reverb Acapella choir, and the Chi Phi Social Fraternity. In 2011, Andrew was employed
by the Applied Superconductivity Center as an undergraduate research assistant. After
graduating with a B.S. in Mechanical Engineering Andrew applied for and was accepted into the
Graduate School for Mechanical Engineering to pursue his master’s degree. While working
towards his degree Andrew worked as a Graduate Research Assistant for the Applied
Superconductivity Center. After his master’s graduation in 2014 he will pursue a career in
mechanical engineer while preparing to be married to the love of his life in mid-2015.