1.

INTRODUCTION
Austenitic stainless steel capillary tubes are widely used in medical devices,
including catheter shafts and needles for injection or biopsy procedures. A smooth
interior tube surface is required to prevent contamination, but as the tube diameter
decreases, the more difficult the internal finishing becomes. In a magnetic field,
magnetic flux flows unimpeded through nonferrous work-piece material, and
ferrous materiala component of the magnetic toolis suspended by magnetic
force. It is possible to influence the magnetic tool motion by controlling the
magnetic field, thus enabling the finishing operation to be performed not only on
easily accessible surfaces but also on areas that are difficult to reach by
conventional mechanical techniques. A variety of Magnetic Field- assisted
Finishing processes using this principle have been developed for the internal
finishing of components. The potential of the process using magnetic abrasive for
the internal finishing of capillary tubes has been demonstrated for tubes with inner
diameters down to 0.4 mm.
In the magnetic abrasive nishing process, material is removed as a result of the
relative motion between the work-piece surface and magnetic abrasive under the
inuence of a magnetic eld. Control of the magnetic eld at the nishing area is
required in order to generate the magnetic force necessary for nishing. The
smaller the tube diameter (e.g., capillary tubes), the more critical is the control of
the magnetic eld. A simple method to adjust the magnetic eld without
modifying the nishing equipment is to change
(1) Magnetic properties or geometry of the magnetic abrasive or
(2) Magnetic properties, geometry, or concentration of magnetic tools (such as
ferrous particles, a ferrous rod, or a permanent magnet) inserted with the magnetic
abrasive or conventional abrasive slurry. The effectiveness of these techniques
have been demonstrated for capillary tubes, widely used in medical applications
and in precision analytical instruments. However, the default nished length in the
Department of Mechanical Engineering, GEC, Thrissur.
1

tube axial direction is limited by the pole-tip width. For example, a pole-tip width
of 4 mm has been considered adequate for capillary tubes with inner diameters
(ID) between 0.4 mm and 1 mm. By feeding the pole tip in the tube axial direction
theoretically allows the entire internal surface of along tube to be nished.
However, difculties associated with the insertion of the magnetic abrasive into
the nishing area limit the total nished length in practice to just a few times the
pole-tip width. To nish the entire surface of a long tube, several short nishing
steps are required, and this approach leads to a long nishing time. To overcome
this issue, an MAF process using a multiple pole-tip system has been proposed.
Magnetic properties of a metastable austenitic stainless steel tool can be changed
by heat treatment, and multiple partial heat-treatments on a single tool can provide
alternating magnetic and non-magnetic regions. The insertion of such a tool into a
capillary tube with the magnetic abrasive generates multiple nishing regions as
the magnetic abrasive is attracted to the border between the magnetic and nonmagnetic regions on the single tool. This
(1) Facilitates the delivery of the magnetic abrasive deep into the tube and
(2) Realizes the simultaneous nishing of multiple regions.
The feasibility of the concept was demonstrated at a tube revolution rate of 2500
rpm. For further improvement of the finishing efficiency, an increase of the tube
rotational speed is necessary. The seminar studies the application of a multiple
pole- tip system for high-speed finishing of capillary tubes. Very recently
finishing equipment with double pole-tip sets is developed, which enables a tube
to rotate up to 30000 rpm on which effects of tube revolution on abrasive motion
are investigated through the tube finishing experiments and the finishing
mechanisms of the high-speed finishing are discussed
For each machines discussed in the seminar, firstly, finishing equipment and its
characteristics are discussed. Secondly, the mechanism of metal removal from the
internal surface is discussed. Finally, the surface produced and its quality is
compared with reference to various tools and methods used along with special
mention to latest developments and results obtained.

Department of Mechanical Engineering, GEC, Thrissur.

2

2. MAGNETIC ABRASIVE MACHINING

The internal nishing of capillary tubes are usually carried out using a magnetic
abrasive nishing process. Such tubes are used with Nano-scale technologies and
meet the demands of the present age in medical and chemical equipment. Since
the working conditions undergone by such tubes are highly critical as in medical
applications for needle syringes, catheter shafts etc. and for precision instruments
employing capillarity property, the internal finish for the tubes are absolutely
essential.

Fig 1.Schematic Diagram of Working Principle

Fig. 1 shows a schematic of the internal finishing process using a work rotation
system for stainless steel capillary tubes. Permanent magnets generate the
magnetic field needed for attracting the magnetic abrasive to the finishing area,
pressing it against the inner surface of the work-piece.
In a non-uniform magnetic field, the magnetic force F acts on the magnetic
abrasive and is shown in the following:
F = VH grad H
Where V is the volume of the magnetic abrasive, is the susceptibility, and H and
grad H are the intensity and gradient of the magnetic field, respectively.
Department of Mechanical Engineering, GEC, Thrissur.
3

If the tangential component of the magnetic force acting on the magnetic abrasive
is larger than the friction force between the magnetic abrasive and the inner
surface of the work-piece, the magnetic abrasive shows smooth relative motion
against the inner surface of the work-piece when the work-piece is rotated at high
speed. Material is removed from the surface by the magnetic abrasive as a result
of this relative motion, and the surface is smoothed. Moreover, manipulating the
poles along the work-piece axis causes the mixture to move in the axial direction
following the poles motion, thereby efficiently finishing a wide area of the inner
work-piece surface.
The nishing characteristics are inuenced by the magnetic abrasive behavior
against the inner surface of the capillary, which is controlled by the supplied
amount of magnetic abrasive and the magnetic force acting on it. The
development of the nishing unit identies the characteristics of the magnetic
eld, which controls the magnetic force, required for the necessary magnetic
abrasive behavior. The run-out of the capillary while rotating at high speed under
the cantilever tube support method causes instability of the magnetic abrasive
behavior. The run-out may be eliminated or reduced to a minimum by using
additional rigid supports in order to complement the already installed support
system.
To be effective in pressing the magnetic abrasive onto the tube surface, the iron
particles must be larger than the 80 m mean diameter magnetic abrasive 12 m.
On the other hand, reduction of the tube ID creates difficulty in introducing the
mixed-type magnetic abrasive into the tube, which establishes an upper limit to
the particle size. Iron particles with 150 m mean diameter were mixed with the
magnetic abrasive.
Also lubricants needed to be provided to the tube during machining either by
interruption of the process or by initial lubrication. Lubricants used include
Soluble-type barrel finishing compound and are added in volumes of 0.04mL.

Department of Mechanical Engineering, GEC, Thrissur.

4

3. SINGLE POLE-TIP SYSTEM

Fig.2 shows a schematic of a magnetic abrasive finishing process with a single
pair of pole tips (hereafter called a single pole-tip system) and a photograph of the
finishing machine. The desired magnetic field in the finishing area is generated by
two permanent magnets attached to a steel yoke. Magnetic abrasive in the tube is
magnetized, and it is pushed against the tube surface by magnetic force. As the
tube is rotated at high speed, the magnetic abrasive exhibits relative motion
against the tube surface and thus removes material from the tube surface. The
default finished length is determined by the pole-tip width in the tube axial
direction, and motion of the pole tip along the tube axis extends the finished area.

Fig 2.Schematic and Photograph of single pole-tip system

Department of Mechanical Engineering, GEC, Thrissur.
5

Reduction of the tube ID makes this approach difficult. For example, magnetic
abrasive tends to build up in the area corresponding to the entry side of the poletip edgesinstead of being distributed at the region corresponding to the pole tip
with following the line of magnetic forceand plugs the tube. Regardless of the
pole-tip motion, the magnetic abrasive remains at the initial position and finishing
is not performed. This is due to a lack of magnetic force acting on the magnetic
abrasive, and the magnetic force yields to friction between the magnetic abrasive
and tube inner surface. Since the magnetic force is a function of magnetic field
strength and its gradient, it can be controlled by the geometry and magnetic
property of the pole-tip or the magnetic tool inserted with magnetic abrasive.

Department of Mechanical Engineering, GEC, Thrissur.

6

4. MACHINING QUALITY OF SINGLE POLE-TIP SYSTEM

Fig 3.SEM micrographs of unfinished surface and surfaces finished with 150 m
mean diameter iron particles
Fig.3 shows scanning electron micrographs of an unfinished surface and surfaces
finished for 416 min for an experiment conducted using single pole-tip system.
Deep, uneven depressions were observed on the unfinished surface. After 4 and 8
min, some of the depressions from the initial surface still seemed to be present,
and this indicated a lack of material removal. In the cases of 12 and 16 min, the
number and size of the depressions and pits on the surface diminished with
finishing time. Even if the measured value of the surface roughness is small, there
exists a certain finishing time required to remove material to obtain a smoother
surface without irregular features. The SEM observation suggested the finishing
period of 12 min was required to achieve successful finishing.

Department of Mechanical Engineering, GEC, Thrissur.

7

5. MULTIPLE POLE-TIP SYSTEM

Fig. 4 shows a schematic and a photograph of a method using a multiple pole-tip
system. In this machine, a pair of pole tips is added to a single pole-tip system and
yoked together to generate magnetic fields in two finishing areas. The finishing
area is doubled as magnetic abrasive is introduced into two regions of the tube and
pushes against the tube surface. Analogous to the single pole-tip system, rotating
the tube enables the relative motion between the magnetic abrasive and tube inner
surface, resulting in an efficient surface finishing. The number of pole-tip sets can
be increased if needed. For a constant pole-tip width, the finishing area will be a
function of the total number of pole-tip sets.

Fig 4.Schematic and Photograph of multiple pole-tip system

Department of Mechanical Engineering, GEC, Thrissur.
8

6. TOOL AND ITS GEOMETRY FOR MULTIPLE POLE-TIP

SYSTEM

Fig 5.Pole-tip geometry for multiple pole-tip system

In Multiple-Pole tip system, we introduce a tool into the tube which is to be
machined. The length and the magnetic properties of the tool are designed
previously according to the requirement. Fig. 5 shows the pole-tip geometry of a
multiple pole tip system.
For the multiple pole-tip system, the magnetic field is concentrated at the pole tips
as shown in Fig. 6.

Fig 6.Variation of magnetic flux density with change in pole-tip distance

Department of Mechanical Engineering, GEC, Thrissur.
9

The magnetic properties of the tool play an important role in realizing the
proposed method. Their effects on the tool behavior in the multiple pole-tip
system using tools with three kinds of magnetic properties are given below. Fig.7
below shows a photograph of three magnetized 54 mm long tools with iron
particles: Tool A is made of carbon steel, Tool B is 304 stainless steel, and Tool C
is partially heat-treated 304 stainless steel. In the case of Tool A, the iron particles
are more attracted by the tool ends because of the residual magnetism and
magnetic shape anisotropy. Tool B exhibits magnetic anisotropy, which was
generated during cold work (deformation-induced martensite transformation), and
the iron particles are attracted to the tool ends only. In the case of Tool C, the iron
particles are attracted to both ends of magnetic regions following the residual
magnetism. The center region (18 mm long) of the stainless steel Tool C was heattreated, and it has a face-centered-cubic structure. This crystal transformation
diminishes the magnetism and divides the magnetic section into two regions.

Fig 7.Geometry of various tools employed

Department of Mechanical Engineering, GEC, Thrissur.

10

7. TYPES OF TOOLS USED

Fig 8.Schematic of machining processes using various tools

Fig. 8 schematically shows the behavior of the three magnetic tools with mixedtype magnetic abrasive while the tube is rotated. Fig. 8(a) is the case with Tool A.
Initially, the mixed-type magnetic abrasive was supplied into two regions
corresponding to the pole tips. Once the pole-tip set started moving, some mixedtype magnetic abrasive stuck to the carbon steel tool surface regardless of the
pole-tip motion because of the ferromagnetism of the tool. The carbon steel tool
pushes the mixed-type magnetic abrasive against the inner tube surface with a
strong magnetic force. This increases the friction force between the tool and inner
surface feed in the experiment.

Department of Mechanical Engineering, GEC, Thrissur.

11

Fig. 8(b) shows the case of Tool B, which exhibits anisotropic magnetism and
lower magnetic susceptibility than the mixed-type magnetic abrasive. The mixedtype magnetic abrasive separated into two finishing sections because it is attracted
to the pole tips more than to Tool B. Tool B simply lies on the mass of magnetic
abrasive. While the pole tips are moving, the mixed-type magnetic abrasive
follows the pole-tip motion. Both ends of Tool B are attracted to the magnetic
field according to its anisotropic magnetism; Tool B thereby stays in its initial
position regardless of the pole-tip and magnetic abrasive motion.
Fig. 8(c) shows the case with Tool C. The borders of magnetic regions of the tool
correspond to the pole-tip edges and are attracted to the magnetic field. The
mixed-type magnetic abrasive is attracted by the pole tips, and once the pole tips
are fed along the tube axis, both the mixed-type magnetic abrasives and tool
follow the pole-tip motion
Thus it can be concluded that, for effective movement of tool and abrasive during
pole-tip feed, two conditions must be satisfied.
(1)

The tool must have alternating magnetic and non-magnetic regions, and

(2)

The borders of the magnetic region of tool must correspond to the pole tip

edges
In turn a new tool satisfying above conditions were developed, namely Tool D as
shown in Fig. 9.

Fig 9.Tool geometry and Schematic of machining process for Tool D

Department of Mechanical Engineering, GEC, Thrissur.
12

8. MECHANISM OF METAL REMOVAL AND SURFACE FINISH

OBTAINED FOR VARIOUS TOOLS

Fig 10.Variation of surface roughness at the inner surface of the tube with various
tools
Fig. 10 shows the relationship between surface roughness (the average of ten
measurements) and distance X in the cases of the multiple pole-tip system using
Tool C, multiple pole-tip system using Tool D, and single-pole tip system. No
magnetic tool is used in the single pole-tip system. Above the figure is a
schematic of a sample work-piece tube that is machined using Tool C, Tool D and
single pole-tip system The roughness at X = 0 and 74 mm are measurements of
unfinished surfaces. In the MAF process, the finishing operation is performed by
masses of mixed-type magnetic abrasive aligned with the lines of magnetic force.
Due to the unstable mixed-type magnetic abrasive motion, the finishing capability
Department of Mechanical Engineering, GEC, Thrissur.
13

generally diminishes at the borders corresponding to the pole tip edges. As a

result, the areas around X = 0, 36, and 72 mm were less finished, which is more
clearly shown by the results from the single pole-tip system (Refer tool/pole-tip
Geometry).
The material removal realized by the multiple pole-tip system using Tool C was
about one third of that per finishing step of the single pole tip system. The number
of cutting edges active at each finishing area in the Tool C case is estimated to be
about half the number in the single pole-tip system. The single pole-tip system
does not use a tool, and the mixed-type magnetic abrasive forms chains by
magnetic force conforming to the shape of the tube. In turn, the mixed-type
magnetic abrasive removes the material not only from the peaks but also the
slopes of the surface asperities over the finishing area, as long as the mixed-type
magnetic abrasive enters the valley of the surface asperities. In contrast, the
multiple pole-tip system uses a tool (a solid rod), and the mixed-type magnetic
abrasive sandwiched between the tube surface and tool at the regions
corresponding to the pole-tip edges plays a major role in removing the material.
In the case using Tool C, some mixed type magnetic abrasive adheres to the
surface of the tube because the friction between the mixed-type magnetic abrasive
and tube surface exceeds the magnetic force acting on the magnetic abrasive. The
region corresponding to the chuck-end edge of the pole tip must have the least
amount of mixed-type magnetic abrasive regardless of the back-and-forth motion
of the pole-tip. This lack of abrasive also discouraged the finishing operation.
Accordingly, less mixed-type magnetic abrasive is involved in the cutting
performance in the multiple pole-tip system using Tool C than that in the single
pole-tip system case. The difference in the material removal rates and surface
improvement is attributed to these different mechanisms. The multiple pole-tip
system using Tool D smoothly finished the surface in an area deep inside the tube,
corresponding to the pole tip located closest to the chuck. A roughness was
achieved similar to that in the single pole-tip system case. This resulted from the
relative motion of the sufficiently distributed mixed-type magnetic abrasive
guided by the tool against the tube surface. After finishing with Tool D, the
material removal is approximately twice that obtained in the Tool C case, but the
Department of Mechanical Engineering, GEC, Thrissur.
14

material removal per length is calculated about half of the single pole-tip system.
The use of a solid tool encourages the amount of mixed-type magnetic abrasive
participating in the finishing performance and pushes it against the tube surface
more strongly than Tool C. However, since the material is predominantly removed
from the peaks of the surface asperities, the material removal rate must be lower
than the case with the single pole-tip.

Fig 11.Three Dimensional Surface images obtained using SEM at point X=36 mm
for various pole-tip stroke lengths
High values of surface roughness, similar to the unfinished surface, are measured
at the center of the finished area. They should be eliminated by enhancing the
mixed-type magnetic abrasive to contact with the tube surface, which can be
achieved by extending the pole-tip feed. The roughness peak exhibited around X =
36 mm in the case of the single pole-tip system was eliminated by lengthening the
pole-tip feed to 24 mm in the first phase and 18 mm in the second phase. The 6

Department of Mechanical Engineering, GEC, Thrissur.

15

mm overlap of the finished area between the first and second phases facilitated the
material removal from the center area and improved the uniformity of the finished
surface. Analogously, extension of the pole-tip feed length should diminish the
unevenness of the roughness in the finished area in the multiple pole-tip system.

Department of Mechanical Engineering, GEC, Thrissur.

16

9. HIGH SPEED FINISHING MACHINE EXPERIMENT

The experimental conditions are as given. Austenitic stainless steel tubes (304
stainless steel, 1.27x1.06x100 mm; 23 m Rz initial surface roughness) were
prepared as work-pieces for this study. A 304 stainless steel tool with three heattreated regions was used as a magnetic tool (see Fig. 12)

Fig 12.Geometry of tool used in high speed magnetic abrasive finishing

The heat-treated regions became non-magnetic by reverting from BCC crystalline
structure to FCC crystalline structure as a result of the normalization while the
four un-treated sections remained magnetic. The mixed-type magnetic abrasive
separates as it is attracted to the ends of the four magnetic sections. The four
magnetic regions correspond to each edge of the two magnetic pole-tips due to its
stronger magnetic force.
The pole tip feed length was set to 12.7 mm, the pole-tip width also being 12.7
and the feed rate was set to 0.59 mm/s. The tube rotational speeds were varied
between 5000, 10000, 20000, and 30000 rpm. To encourage uniform internal
surface coverage with the mixed-type magnetic abrasive prior to high-speed
finishing, the tube was rotated at 500 rpm with a single pole-tip stroke before
Department of Mechanical Engineering, GEC, Thrissur.
17

finishing. During the finishing process, the presence of lubricant is crucial to

encourage the smooth relative motion between the mixed-type magnetic abrasive
and the tube surface that facilitates finishing performance of the abrasive.
Accordingly, in the cases of 20000 and 30000 rpm tube revolution, the finishing
experiments were interrupted to inject lubricant every 5 min. For the cases of 5000
and 10000 rpm tube revolution, the finishing experiments were performed
continuously for 10 min without additional lubricant injection. Before and after
the finishing experiments, the tube was rinsed with ethanol in an ultrasonic
cleaner for 1 hr.

Department of Mechanical Engineering, GEC, Thrissur.

18

10. MECHANISM OF METAL REMOVAL AND SURFACE FINISH

OBTAINED
Figures 13 and 14 show intensity maps and oblique plotsmeasured by an optical
profiler at X=13 mmof the unfinished surface and surfaces finished for 10 and
20 min, respectively.

Fig 13.Intensity Profile and

Oblique Plots measured by
optical profiler at X=13 mm
for 10 min machining

Fig 14.Intensity Profile and

Oblique Plots measured by
optical profiler at X=13 mm for
20 min machining

Department of Mechanical Engineering, GEC, Thrissur.

19

The surface finished for 10 min at a tube revolution of 5000 rpm (Figure 13(b))
has a roughness of 0.15 m Rz, but multiple irregular asperities from the initial
surface remained. However, an extension of the finishing time for another 10 min
allowed the abrasive to remove those irregular asperities (Figure 14(a)). The
material removal after 20 min was more than twice the removal after 10 min.
Although the magnetic abrasive was not exchanged during the interruptions to
inject lubricant, the reconfiguration of the magnetic abrasive during these breaks
encouraged the relocation of the sharp abrasive cutting edges. The sharp cutting
edges and newly added lubricant seemed to refresh the finishing performance after
each intermission.

At a tube revolution rate of 10000 rpm, the surface was

smoothly finished (0.1 m Rz) after 10 min, and the roughness value remained
constant after 10 min of extra finishing time despite the additional material
removal (Figure 14(b)). Analogous to the case of finishing for 20 min at 5000
rpm, the pauses to add lubricant after 10 min aided the finishing performance. It
was confirmed that the increase in the tube revolution (i.e., cutting speed)
improves the material removal rate and finishing efficiency. However, due to the
high centrifugal force, the high-speed tube rotation creates more opportunities for
the mixed-type magnetic abrasive and magnetic tool to lapse into unstable
conditions. The lack of a uniform magnetic abrasive distribution under an unstable
rotating magnetic tool may lead to the deep, irregular scratches. This trend was
observed in the case of 20000 rpm (Figure 13(d)), the finished surface has deep
scratches and surface distortions. Extending the finishing time slightly increased
the material removal due to the longer duration contact of the magnetic abrasive
cutting edges against the tube surface and removed the relatively short wave
length surface asperities (Figure 14(c)); however, the deep scratches produced by
the irregular motion of the magnetic tool and mixed-type magnetic abrasive
remained on the surface. Further increase in the tube revolution, such as the 30000
rpm test, further enhanced the irregularity of the motion of the magnetic tool and
abrasive. The non-uniformly distributed mixed-type magnetic abrasive must be
pressed against the surface by a rotating magnetic tool that is unstable due to the
centrifugal force. This led to increased material removal and resulted in a surface
consisting of deep scratches over long-wavelength asperities, as shown in Figure
13(e).
Department of Mechanical Engineering, GEC, Thrissur.
20

11. CONCLUSION
Use of the previously developed multiple pole-tip finishing system produced a
uniformly finished surface (from 23 m Rz to ~0.2 m Rz) 72 mm longfour
times the pole-tip widthin 180 min. The high speed finishing system proposed
in this paper can produce a finished surface 50.8 mm long (four times the pole-tip
width) in 10 min with a roughness of about 0.1 m Rz. Thus the proposed highspeed finishing system is twelve times more efficient than its predecessor.
Thus it can be confirmed that the area of internal finishing of capillary tubes are
facing challenges regarding different aspects of machining. The field also
promises to be one, in which, there exist requirement for researches regarding
methods of machining that are absolutely novel and far different from the
conventional system. Hence the researches carried out to date provides firm
foundation for future works in the area.

Department of Mechanical Engineering, GEC, Thrissur.

21

12. BIBLIOGRAPHY

Junmo Kanga, Andrew Georgea, Hitomi Yamaguchia. High-speed

Internal Finishing of Capillary Tubes by Magnetic Abrasive Finishing.
5th CIRP Conf. on High Performance Cutting., 2012;(1):414-418

Junmo Kang, Hitomi Yamaguchi. Internal finishing of capillary

tubes by magnetic abrasive finishing using a multiple pole-tip
system. J. Precision Eng., 2012; (36):510 516.

Yamaguchi H, Shinmura T, Ikeda R. Study of Internal Finishing of

Austenitic Stainless Steel Capillary Tubes by Magnetic Abrasive
Finishing. J. Manuf. Sci. Eng., 2007; 129(5):885892.

Yamaguchi H, Kang J, Hashimoto F. Metastable Austenitic Stainless

Steel Tool for Magnetic Abrasive Finishing, CIRP Ann.,
2011;60(1):339-342.

Department of Mechanical Engineering, GEC, Thrissur.

22