Physics Procedia 00 (2010) 000 000

www.elsevier.com/locate/procedia

12th International Conference on Magnetic Fluids

Experimental Study on Micropump using Reciprocating Motion of

Magnetic Ball Covered with Magnetic Fluid

Hiroshige Kumamarua,*, Satoshi Okamoto

a, Koki Arimoto

a,

Kazuhiro Itoha and Yuji Shimogonya

a1

a University of Hyogo, 2167 Shosha, Himeji, Hyogo, 671-2280, Japan

Elsevier use only: Received 30 January 2010; revised 21 June 2010; accepted 11 April 2010

Abstract

A micropump combining reciprocating motion of a magnetic ball covered with magnetic fluid and diffusers working as valves

was investigated experimentally in this study. The reciprocating motion of the magnetic substance pushes water in order to

transport it. The diffusers with different divergence angles serve to pump water in net one direction. In the present micropump

experiments, the maximum flow rate achieved with minimum backpressure was 3.89 l/min and the maximum pressure head

achieved was 42.4 mm water.

Keywords: Micropump; Magnetic Ball; Magnetic Fluid; Maximum Flow Rate; Maximum Pressure Head

1. Introduction

In recent years, a variety of microfluidic devices are developed for a wide range of application, from chemical

analysis systems to actuating systems in medicine and biology. Many researchers have tried to develop micropumps

to be applied to microfluidic devices based on various principles and methods. Among them, diaphragm and

electroosmotic micropumps are at the stage of practical application in some degree. However, there still remains

problems in both micropumps; for examples, endurance of moving parts in check valves and relatively large pump

size for the former, and limitation of pumped fluid only to electrolyte solution and bubble generation in the pump for

the latter.

A few researchers have made efforts to carry experiments for micropumps using magnetic fluid or magnetic

materials. Hatch et al. proposed a micropump using rotating motion of magnetic fluid shown in Fig. 1 [1]. The

magnetic fluid acts as both a valve and a piston. The magnetic fluid valve, held in place by a stationary magnet, is

always present in the short section of channel between the inlet and outlet of the pumping loop. The magnetic fluid

plug serving as a piston is drawn around the loop by a rotating magnetic field. As the mobile plug is drawn around

the loop, water is drawn into the loop through the inlet, and forced out through the outlet. The maximum flow rate

* Corresponding author. Tel.: 81-79-267-4833 ; fax: 81-79-267-4833 .

E-mail address: kumamaru@eng.u-hyogo.ac.jp .

c© 2010 Published by Elsevier Ltd

Physics Procedia 9 (2010) 238–242

www.elsevier.com/locate/procedia

1875-3892 c© 2010 Published by Elsevier Ltddoi:10.1016/j.phpro.2010.11.053

Open access under CC BY-NC-ND license.

Open access under CC BY-NC-ND license.

Hiroshige Kumamaru/ Physics Procedia 00 (2010) 000 000

achieved was 45.8 l/min and the maximum pressure head achieved was 135 mm water. However, the pump could

not get rid of the disadvantage caused by the channel contamination from the magnetic fluid and effluence of the

magnetic fluid.

Kim et al. proposed a peristaltic micropump using magnetic fluid shown in Fig. 2 [2]. Diaphragm between upper

and lower channels is made of silicon rubber. Collected magnetic fluid lump in the lower channel deforms the

silicon rubber diaphragm and then the deformed diaphragm pushes water in the upper channel. The maximum flow

rate was 3.8 l/min, however, the maximum pressure head is not shown in their paper.

In this study, a micropump combining reciprocating motion of a magnetic ball covered with magnetic fluid and

diffusers working as valves has been proposed and investigated experimentally.

Fig. 1 Hatch et al. s Micropump

Fig. 2 Kim et al. s Micropump

H. Kumamaru et al. / Physics Procedia 9 (2010) 238–242 239

Hiroshige Kumamaru/ Physics Procedia 00 (2010) 000 000

2. Experiments

Figure 3 explains the working principle of the present micropump. The magnet ball covered with magnetic fluid

serves as a piston, preventing the effluence of magnetic fluid observed in Hatch et al s micropump. The magnetic

fluid covering the magnetic ball seals the space between magnetic ball and square channel wall. The diffuser

(convergence/divergence nozzle) works as a valve. For the divergence, the loss coefficient for a divergence angle of

~6 degree is minimum, 0.15, and the loss coefficient for angles greater than ~40 degree is ~1.0. For the convergence,

the loss coefficient is very small, ~0, for all angles.

When the magnet ball in the pumping channel moves toward this side by the permanent magnet, the water

(pumped fluid) is transported toward left side in the micro channels (Fig. 3(a)). When the magnet ball moves

toward opposite side, the water is also transported mainly toward left side (Fig. 3(b)).

The pumping channel is 3.4 mm in width, 3.5 mm in depth and approximately 40 mm in length. The micro

channels except for the pumping channel are 1 mm in width and 0.5 mm in depth. The diffusers are

approximately 6 mm in length having divergence angles 6 and 127 degrees. The upper and lower plates of the

micropump are made of acryl. The flow channels are fabricated on the lower plate by cutting with end mills. The

upper and lower plates are attached by burning in an electric furnace at 140 C during 2 hours. The magnet ball is a

neodymium magnet plated with nickel with a diameter of 3 mm and a magnetic flux density of 320 mT.

(a)

(b)

Fig. 3 Present Micropump

240 H. Kumamaru et al. / Physics Procedia 9 (2010) 238–242

Hiroshige Kumamaru/ Physics Procedia 00 (2010) 000 000

Figure 4 shows the experimental apparatus. Two cylindrical containers are connected to the inlet and outlet of the

micropump in order to measure the flow rate and pressure head. The permanent magnet is put into reciprocating

motion by an electromotive slider and a stepping motor. The permanent magnet is a neodymium magnet with a

diameter of 10 mm and a thickness of 5 mm, and a magnetic flux density of 350 mT

3. Experimental Results

The stroke of the reciprocating motion of permanent magnet is 30 mm and the frequency of it is 0.5 1/s (CASE1)

and 0.25 1/s (CASE2). Two kinds of magnetic fluids covering the magnetic ball are used, N-304 (medium=

isoparaffin, magnetization= 33 mT, viscosity= 10 mPas, specific gravity= 1.14) and N-504 (medium= isoparaffin,

magnetization= 55 mT, viscosity= 22 mPas, specific gravity= 1.40).

Figure 5 shows total pumped water (i.e. integrated water flow rate) versus elapsed time. The total pumped water

became considerably larger for the combination of CASE1 (0.5 1/s) and N504 (larger viscosity) than for the other

combinations.

Figure 6 shows pump pressure head versus flow rate. The pump pressure head and flow rate were also

considerably larger for the combination of CASE1 and N504 than for the other combinations. The maximum flow

rate achieved with minimum backpressure was 3.89 l/min and the maximum pressure head achieved was 42.4 mm

water.

Fig. 4 Experimental Apparatus

6 12 18 24 30

500

1000

1500

2000

2500

3000

0T (hour)

Qto

tal

:CASE1-N504

:CASE1-N304

:CASE2-N504

:CASE2-N304

Fig. 5 Total Pumped Water versus Time

H. Kumamaru et al. / Physics Procedia 9 (2010) 238–242 241

Hiroshige Kumamaru/ Physics Procedia 00 (2010) 000 000

4. Conclusions

In the present micropump experiments, the maximum flow rate achieved with minimum backpressure was 3.89

l/min and the maximum pressure head achieved was 42.4 mm water. In future, it is desirable to replace the

diffusers by passive valves in order to improve the pump performance.

References

[1] A. Hatch, A.E. Kamholz, G. Holman, P. Yager, K.F. Bohringer, A Ferrofluidic Magnetic Micropump, J. of

Microelectromechanical Systems, 10 (2) (2001) 215-221.

[2] E.G. Kim, J.-G. Oh, B. Choi, A Study on The Development of A Continuous Peristaltic Micropump using

Magnetic Fluids, Sensors and Actuators A, 128 (2006) 43-51.

1 2 3 4

10

20

30

40

50

0

Q in)

H (m

m)

:CASE1-N504

:CASE1-N304

:CASE2-N504

:CASE2-N304

Fig. 6 Pressure Head versus Flow Rate

242 H. Kumamaru et al. / Physics Procedia 9 (2010) 238–242