USING CONDUCTIVE INKS AND NONWOVEN TEXTILES FOR WEARABLE COMPUTING by byrnetown72

VIEWS: 74 PAGES: 15

									                                                                                   Karaguzel


USING CONDUCTIVE INKS AND NONWOVEN TEXTILES FOR WEARABLE
COMPUTING
B. Karaguzel, C. R. Merritt, T. Kang, J. Wilson, P. Franzon, H. T. Nagle, E. Grant , B.
Pourdeyhimi

ABSTRACT

This paper deals with the interaction between conductive inks and nonwoven substrates, to
produce circuits and embedded systems for vital signs monitoring, and wearable computers.
The performance metrics related to the circuits are impacted by the ink viscosity, contact
angle of the ink on the surface of the nonwoven structure that dictates the manner in which
the ink distributes into the substrate. The competition between in-plane versus
through-the-plane distribution of the ink is ultimately responsible for the quality of the
printed media. As a demonstration of principle, an ECG monitoring system has been
designed, fabricated, and integrated into a garment made from a nonwoven material suited
to the application. At this stage the ECG system is interfaced to a Palm Pilot computer for
data gathering, data interpretation, and data transmission. Trials are continuing.

1. INTRODUCTION

Flexible circuit boards continue to be a high-growth technology in the area of electrical
interconnectivity. Over the use of traditional rigid printed circuit boards (PCB’s) we
investigate using conductive inks for printing circuits onto nonwoven flexible substrates as a
low-cost alternative because they offer printability, lightweight, durability and washability.

Interaction between ink droplet and nonwoven substrate has been previously investigated.
The impacting and spreading of liquid drops on solid surfaces are scientifically and
practically important physical processes in many applications such as spray coating,
delivery of chemicals, as well as printing (Park et. al.). Understanding droplet impact
requires knowledge about the flow within the droplet and the surrounding gas, and about the
movement of the dynamic contact line. The impact process consists of three phases, first is
the initial impact phase where the droplet hits the substrate. In the second phase there is a
rapid fluid flow and in the third phase the fluid comes to rest in a process of rebound
followed by inertial oscillations, damped by viscous dissipation (Van Dam et. al.). One of
the objectives of this investigation is to provide information on droplet spreading in
nonwoven substrates.

Electrical characterization of the conductive traces was performed before washing and after
different wash cycles. Measurements of trace DC resistance and Time Domain
Reflectometry (TDR) measurements were performed to determine the impact of washing on
the printed nonwovens. These measurements show that printed trace on certain textiles can
fail after washing due to mechanical fatigue of the structure (e.g. creases in the fabric).




                                              1
                                                                                         Karaguzel


EXPERIMENTAL PROCEDURES

2.1 Test Materials

The substrates selected for the screen-printing were Fruedenberg’s Evolon, DuPont’s Tyvek,
and BBA FiberWeb’s Resolution Print Media (RPM). Two silver conductive inks were
tested with these fabrics. These were CMI 112-15 made by Creative Materials and
CSS-010A made by Precisia.

2.2 Screen Printing

The screen printer used for the experiments is a DeHaart EL-20 flatbed semi-automatic
screen printer with a dual squeegee print head. In these experiments, we used conductive
inks for printing circuits onto nonwoven flexible substrates. The screen included several
patterns consisting of different line widths along with several groups of straight lines of
different line widths. For example, group G4 is made up of the 6 cm lines in the middle
with widths, 0.6 mm, 1.0 mm, and 1.3 mm (Figure 1). Screen printing parameters were
altered in order to achieve an even coverage of ink over the entire image. These included
mesh count of the screen, squeegee durometer, snap-off distance, and print speed. Once
printing was concluded, the printed samples were cured at temperatures recommended by
the ink manufacturer and within the safe temperature range for the nonwovens. The
temperatures selected for curing were 110 0C for Tyvek, 140 0C for Evolon and 160 0C for
Resolution Print Media respectively.




Figure 1. (left) Screen layout of Antenna’s and Lines Used for Screen Printing (right) Conductive Ink
Results Obtained from Screen Printing Antenna’s and Lines onto Evolon.




                                                 2
                                                                                     Karaguzel


2.3. Textile Transmission Line Design

2.3.1. Coplanar Waveguides

The design of a coplanar waveguide transmission line consists of a center conductor that acts
as the signal (S) while being surrounded by two ground planes (G). Figure 2 shows the
cross-section of a coplanar waveguide ground-signal-ground (GSG) structure indicating the
dimensions that influence the characteristics of the line. In the figure a stands for the signal
width, W stands for signal-ground gap, h stands for the height of the substrate and t stands for
line thickness.




                        Figure 2. Cross-section of GSG coplanar waveguide

2.3. SEM Imaging
A series of scanning electron microscope (SEM) images were taken to determine the height
of the printed lines and thickness of the substrates. CMI 112-15 ink printed samples were
selected for these measurements. From each of these samples the entire G4 section, Figure 1,
was removed to capture the cross-section of all three lines. The G4 samples were then
prepared for imaging by placing them into a liquid nitrogen bath. While still in the liquid
nitrogen bath, the lines were cut with a straight razor and the samples were removed from
the bath. Finally, the samples were placed inside the pressure chamber of the SEM for
imaging.

2.5. Ink Penetration Dynamics Analyzer

Ink penetration tests were performed using a Penetration Dynamics Analyzer Measuring
Device. The substrate attaches in the sample holder. There is an ultrasonic wave transmitter
and to the opposite side the receiver is arranged. The sample is moved towards the test liquid.
When the test liquid and sample contact, the transmitter expels a supersonic wave. When the
liquid permeates to the sample, the receiver draws the change of the signal with a curve. The
intensity of ultrasound increases when the air voids within the sample are filled with liquid or
the air solid interface converts to liquid solid interface (Figure 3).




                                               3
                                                                                    Karaguzel




                 Figure 3. Schematic representation of penetration dynamics analyzer measurement

3. RESULTS AND DISCUSSION

3.1. SEM Imaging for Ink Height and Substrate Thickness

Figure 4 shows the cross sectional SEM images of different substrates of Evolon, Tyvek and
RPM. The silver ink can be observed on the surface of each substrate.

Data summarizing various properties related to ink height and fabric thickness are given in
Table I. The relatively low ink height for RPM is the result of the penetration of the ink
inside the structure, which correlates well with higher resistance values due to the less
conductive medium for the flow of electrons. Ink spreads more in Evolon than Tyvek
because of the structure differences that exist between the two. Tyvek has very small and
very few capillaries available on the surface and consequently, the printing is a surface
phenomenon.




                                              4
                                                                     Karaguzel
                             Evolon 80




Conductive
   Ink




         Evolon 130                                Evolon 100




             Tyvek                          Resolution Print Media




      Figure 4. SEM cross-sectional images of printed substrates




                                    5
                                                                                                Karaguzel


Table I. Mean Fabric Thickness and Ink Height extracted from SEM Cross-sectional images

                            Evolon                                               Resolution
      Fabric Type           80                           100         130         Print        Tyvek
                            gsm                          gsm         gsm         Media
      Fabric Thickness (µm) 320.22                       391.22      416.91      398.72       241.9
      (StdDev)              (17.43)                      (16.84)     (25.53)     (13.15)      (9.07)
      Ink Height (µm)       36.27                        40.48       41.17       21.02        43.51
      (StdDev)              (12.05)                      (8.13)      (12.01)     (13.02)      (4.94)

3.2. DC Resistance

DC resistance measurements were performed on each continuous line using a digital
multi-meter. The results obtained show that the CMI 112-15 ink performed better than the
CSS-010A ink possibly because it has a higher viscosity and larger percentage of silver.
The CMI 112-15 may tend to stay on the surface more which will result in a more
continuous coverage of ink while the CSS-010A ink is broken up by absorbing into the
fabric. It is apparent that CMI 112-15 printed on Tyvek® produces the best results in terms of
DC resistance (Figure 5).

                                                                     6 cm CSS-010A
                                  8                                  6 cm CMI 112-15
                                                                     2 cm CSS-010A
                                  7                                  2 cm CMI 112-15
                                  6
                  DC Resistance




                                  5
                                  4
                                  3
                                  2
                                  1
                                  0
                                           Evolon            RPM             Tyvek


                                  Figure 5. Average DC resistance for 2 cm and 6 cm lines



3.3 Viscosity

Viscosity measurements were performed using a Brookfield Cone/Plate Viscometer. CMI
112-15 shows higher viscosity values, Figure 6, which has an impact on ink penetration
through the substrate as will be discussed later.




                                                            6
                                                                                                              Karaguzel



                        90.00
                        80.00                                                          CSS-010A
                        70.00                                                          CMI 112-15
    Viscosity (Pa.s)

                        60.00
                        50.00
                        40.00
                        30.00
                        20.00
                        10.00
                         0.00
                                 0.00        50.00        100.00     150.00   200.00     250.00      300.00
                                                               Shear Rate (1/s)
                                               Figure 6. Effect of shear rate on ink viscosity

3.4. Ink Penetration

Ink penetration is measured by the intensity of the sound waves reflecting back from the
substrates. There is a decrease in intensity at the start of the test due to the presenece of air
voids within the pores of the nonwoven substrate. When the voids start to fill with ink,
intensity values increase. Higher intensity values correspond to more ink penetration
through the substrate. As can be seen from the Figure 7, Resolution Print Media shows the
highest intensity values, corresponding to higher ink penetration with both inks used. Tyvek
has a more even and finer porosity whereas Evolon has larger but rougher pores yet they
show comparable intensity values as well as ink penetration values.


                                                                                                  Evolon 130g, CSS
                       150
                                                                                                  Evolon 80g, CSS
                                                                                                  6145 RPM, CSS
                                                                                                  6180 RPM, CSS
     Intensity [%]




                                                                                                  Evolon 100g, CSS
                                                                                                  RPM, CSS
                       100
                                                                                                  Tyvek, CSS
                                                                                                  Yellow RPM, CSS


                       50
                             0          10           20       30       40         50     60
                                                             t [s]

                                        Figure 7. Change in light intensity vs. time for CSS-010A




                                                                       7
                                                                                                                     Karaguzel


3.5. Drop Testing

Drop testing was performed using a high-speed camera to monitor how droplets form onto
the nonwoven materials surface and diffuse into the material itself. Figure 8 and Figure 9
show the dynamic contact angle values with CSS-010A and CMI 112-15 respectively.


                        100
  Contact Angle (degrees)




                                   80

                                   60

                                   40
                                                                                                              Evolon 80
                                   20                                                                         RPM
                                                                                                              Tyvek
                                          0
                                              0       20           40         60              80     100       120         140
                                                                                   Time (s)
                                                           Figure 8. Dynamic contact angle with CSS-010A



                                          90
                Contact Angle (degrees)




                                          80
                                          70
                                          60
                                          50
                                          40                                                                  Evolon 100
                                          30                                                                  Evolon 130
                                                                                                              Evolon 80
                                          20                                                                  RPM
                                          10                                                                  Tyvek
                                           0
                                                  0        50           100            150         200       250          300
                                                                                   Time (s)

                                                           Figure 9. Dynamic contact angle with CMI 112-15




                                                                                   8
                                                                                        Karaguzel


As seen from Figure 10, CSS-010A Ink penetrates more into the nonwoven substrate over
time while the CMI 112-15 ink remains more on the surface of the substrate due to its higher
viscosity.

                               t=0s                       t = 40 s


    CSS-010A Ink




   CMI 112-15 Ink


                                     Droplets on Evolon 80

                                t=0s                      t = 40

    CSS-010A Ink




   CMI 112-15 Ink


                                     Droplets on Tyvek

                               t=0s                      t = 40 s


     CSS-010A Ink



   CMI 112-15 Ink


                                      Droplets on RPM

                Figure 10. Droplets on Evolon, Tyvek and Resolution Print Media (RPM)




                                               9
                                                                                     Karaguzel


Durability Properties

Specimens were washed for 15 cycles under appropriate conditions of temperature,
detergent solution and abrasive action using a Launder-Ometer which is a laundering
machine for rotating closed canisters in a thermostatically controlled water bath. The
abrasive action is a result of the frictional effects of fabric against canister, the low liquor
ratio and the impact of the steel balls on the fabric. Additionally, the fabrics were exposed to
5 more wash cycles in an industrial laundry, for a total of 20 wash cycles (“15+5”). Figure
11 shows the changes in the fabric structure upon washing for Evolon, RPM, and Tyvek.

                                        Evolon 130




                       Before wash                     After wash
                                Resolution Print Media




                          Before wash                  After wash
                                           Tyvek




                          Before wash                 After wash
                                               10
               Figure 11. Images of unwashed and washed nonwoven substrates
                                                                                       Karaguzel




                Unwashed                                          Washed
Figure 12. Images of unwashed and washed (“15+5”) coplanar wave lines of Resolution Print Media

DC Resistance measurements were done using a Fluke 77/BN Multimeter, each
measurement was repeated to insure consistent measurements (ohm meter cables = 0.1 Ω).
For the lines that were washed 15 times in the Launder-Ometer and 5 times in the industrial
washer the DC resistance measurement was made on a 5cm long section, due to the break
that bisected all but one of the CPW lines. Data summarizing measured properties related to
DC resistance under washed and unwashed conditions for various CPW signal line widths (a)
are given in Table II and Table III. These measurements show that washing does degrade the
traces on the textile. Traces exposed to a 10 wash cycle showed a 15% to 20% increase in
per unit length resistance. The textile subjected to the “15+5” wash cycle suffered from a
crease that caused a break in all but one of lines, and the per unit length resistance increased
significantly.

       Table II. DC Resistance measurement results for unwashed and 10 cycle wash
                      CPW signal      Resistance         Resistance
                     line width (a)    unwashed         10 wash cycle
                        1200 µm       0.14 Ω/cm          0.16 Ω/cm
                        1400 µm       0.12 Ω/cm          0.14 Ω/cm
                        1600 µm       0.10 Ω/cm          0.12 Ω/cm

 Table III. DC Resistance measurement results for unwashed, 10, and “15+5” cycle wash
       CPW signal           Resistance       Resistance     10        Resistance
      line width (a)        unwashed             wash cycle       “15+5” wash cycle
         1200 µm            0.14 Ω/cm            0.16 Ω/cm            1.86 Ω/cm

To further characterize the effects of washing on traces printed on textiles using conductive
inks, we conducted Time Domain Reflectometry (TDR) measurements using a Tektronix
11801 with a SD-24 Sampling Head. The TDR measurement is useful for extracting


                                                11
                                                                                      Karaguzel


information about an electrical network. The characteristic impedance can easily be
determined as well as the losses, both DC and high frequency. This measurement required
the design and fabrication of a custom test fixture to enable measurements free of artifacts
from cables and connectors. The traces on the textiles were measured directly by using DC
to 40GHz microwave probes with a ground-signal-ground (GSG) configuration. Figure 13
shows the TDR response for two CPW lines, unwashed and after 5 wash cycles. Figure 14
shows the TDR response for three CPW lines, unwashed and after 10 wash cycles. In Figure
13 & 14, the bold lines indicated the unwashed case and the light lines the washed.

             Unwashed vs 5 wash cycles: 10cm line on RPM b=2000um a=700,900um
       140
       130
       120
       110
Zo [Ω] 100
          90
          80
          70
          60
          50
               9.56    9.57    9.58    9.59       9.6    9.61   9.62   9.63    9.64
                                               Time [seconds]                         -8
                                                                                  x 10

               Figure 13. CPW line TDR response for unwashed vs 5 wash cycle (RPM)



       Unwashed vs 10 wash cycle: 10cm line on RPM b=2400um a=1200,1400,1600um
       120

       110
       100
          90
 Zo [Ω]
          80
          70
          60
          50
                9.56    9.57    9.58    9.59      9.6    9.61   9.62    9.63    9.64 -8
                                               Time [seconds]                      x 10


               Figure 14. CPW line TDR response for unwashed vs 10 wash cycle (RPM)




                                                  12
                                                                                      Karaguzel


The results of the TDR measurements for the case of 5 wash cycles show little, if any,
degradation in the quality of the CPW lines (Figure 13). However, for the lines that were
subjected to 10 wash cycles the effects are evident (Figure 14). The measurement of the
lines subjected to 10 wash cycles show an increase in impedance. This increase was also
evident in the DC resistance measurements of Table II.

The samples that were subjected to the “15+5” wash cycles suffered from a break in the
lines caused by a crease in the fabric (see Figure13). All but one of these lines completely
failed and the TDR measurement of this line is shown in Figure 15. The line shows a large
increase in impedance (from DC resistance) and also has a section of high characteristic
impedance. The section where the characteristic impedance spikes is at the region where the
crease occurred. It is likely that the CPW line was “effectively” reduced in width; thereby,
increasing its characteristic impedance. After the creased region, the line impedance comes
back to the trajectory that it was on before the crease. The slope of the line is proportional to
the per unit length resistance of the CPW line.


           Unwashed vs 15+5 wash cycle: 10cm line on RPM b=2000um a=1000um
      150
      140
      130
      120
Zo [Ω]110
      100
        90
        80
        70
        60
        50
              9.56    9.57     9.58    9.59      9.6    9.61   9.62    9.63    9.64
                                                                                      -8
                                              Time [seconds]                      x 10

           Figure 15. CPW line TDR response for unwashed vs “15+5” wash cycle (RPM)



Figure 16 shows the TDR response for a line on the textiles that was subjected to the “15+5”
wash cycles. This CPW line had a complete break in the traces and therefore the
measurement of the line shows a length of approximately 5cm, instead of the physical length
of 10cm. This is due to the break near the middle of the CPW line. This failure in the
conductive traces due to a crease in the fabric indicates that RPM is not a good choice for
garments.




                                                 13
                                                                                        Karaguzel


              Unwashed vs “15+5” wash cycle: 10cm line on RPM b=2000um a=1200um
       140
       130
       120
       110

Zo [Ω] 100
         90
         80
         70
         60
         50
                9.56   9.57    9.58    9.59      9.6    9.61   9.62    9.63    9.64
                                              Time [seconds]                          -8
                                                                                  x 10


             Figure 16. CPW line TDR response for unwashed vs “15+5” wash cycle (RPM)

4. CONCLUSIONS

SEM imaging showed that the smallest ink height of RPM was due to the high penetration of
ink through the substrate which also correlated with high DC resistance values. Tyvek
having a surface coating showed the highest ink height values, correspondingly smallest DC
resistance values. It was observed that the ink tended to spread more on Evolon than Tyvek
substrates.

Drop testing results showed that CMI 112-15 ink tended to stay on the surface of the
substrate while the CSS-010A ink penetrated more through the structure. The CMI 112-15
performed better due to its higher viscosity and high tendency to stay on the surface which
resulted in a more continuous coverage of ink. Penetration Dynamics Analyzer
measurement results revealed that the structural parameters of the substrates such as porosity,
surface roughness had an impact on ink penetration values.

The results of the electrical characterization show that the printed inks do begin to degrade
(i.e. show an increase in resistance) after numerous Launder-Ometer washing cycles. When
printed traces on RPM undergo washing in home laundry equipment, they show severe signs
of failure due to creasing of the semi-rigid RPM substrate. Tyvek also shows this creasing
and would most likely suffer from the same type of failure with trace printed on its surface.
However, Evolon does not suffer the same type of mechanical failure due to its flexibility.
Future research to explore the effect of washing Evolon substrates with traces printed on
their surface.




                                                 14
                                                                                 Karaguzel


5. REFERENCES

Van Dam, D. B., Le Clerc C., “Experimental Study of the Impact of an Ink-jet Printed
Droplet on a Solid Substrate”, Physics of Fluids, 16 (9), 3403-3414, September 2004

Park, H., Carr, W. W., Zhu, J., Morris, J. F., “Single Drop Impaction on a Solid Surface”,
AIChE Journal, 49 (10), 2461- 2471, October 2003.

Toivakka M., “Numerical Investigation of Droplet Impact Spreading Spray Coating of
Paper”, Spring Advanced Coating Fundamentals Symposium, Spring 2003.

Cottet D., Grzyb J., Kirstein, T., and Tröster, G., “Electrical Characterization of Textile
Transmission Lines”, IEEE Transactions on Advanced Packaging, Vol. 26, No. 2, May
2003.

Tröster, G., Kirstein, T., and Lukowicz, P., “Wearable Computing: Packing in Textiles and
Clothes”, 14th European Microelectronics and Packaging Conference & Exhibition,
Friedrichshafen, Germany, 23-25 June 2003.

De Rossi, D., Carpi F., Lorussi F., Mazzoldi A., Paradiso R., Scilingo, E., P., Tognetti A.,
“Electroactive Fabrics and Wearable Biomonitoring Devices”, AUTEX Research Journal,
3(4), 180-185, December 2003.

S. B. Hoff, Screen Printing: A Contemporary Approach, Albany: Delmar, 1997.

ACKNOWLEDGEMENTS

The authors would like to thank NTC for funding this research. We would also like to thank
Freudenberg, BBA Fiberweb, DuPont, Creative Materials, Precisia for supplying the
materials, Goulston Technologies for their help in the contact angle measurements, and the
Electronics Research Laboratory at NCSU for the use of their TDR equipment.

RESPONDENCE ADDRESS:
Behnam Pourdeyhimi
College of Textiles
North Carolina State University,
Raleigh, NC 27695-8301 USA




                                             15

								
To top