Page 347 - Biomimetics : Biologically Inspired Technologies

Bar-Cohen : Biomimetics: Biologically Inspired Technologies DK3163_c012 Final Proof page 333 21.9.2005 11:55pm

Multifunctional Materials 333

12.2.4.1 Integrating Sensing into Composites

Here we focus on the development of a new type of composite with integrated high density of small,
advanced sensors that would enable sensing without compromising the structural integrity. The
volume of work in related areas is vast, and there have been a number of contributions aimed at
incorporating novel nonstructural elements into composite systems (Varadan and Varadan, 2000;
Lin and Fu-Kuo, 2002; Zhou and Sim, 2002). In this section we discuss several of the challenges
associated with sensorized composites, including electronics, mechanical integration, and data
management. We conclude by presenting results of some preliminary work.

12.2.4.2 Sensor Communications and Power

We seek to take advantage of advances in microelectronics, new capabilities in MEMS and sensor
development, data feature extraction algorithms, multi-drop networking protocols, and composite
fabrication technology to enable in situ sensing and damage detection at scalable, and potentially
high areal density. Data bandwidth limitations require local data handling and efficient processing
algorithms. Power management will be necessary to make high sensor densities possible. Embed-
ded sensors require both power and a means of communication. Ideally, this would be done with the
fewest number of conductors possible, and use a multi-drop network if feasible. The 1-Wire 1
network allows for distribution of power and two-way communications among many devices over a
single pair of conductors. 3
The 1-Wire network is specified for operation over a supply voltage range of 2.8 to 6 VDC. The
positive, or data, line is held high by a nominal 5 kV impedance source. Multiple devices may be
attached to the same network wire pair. Each device has a unique 64-bit address. The network
protocol uses half duplex communications in master–slave architecture, where devices respond
only upon command from the master. The protocol uses serial bit communications at 16.3 kbps. An
overdrive mode is available that increases the data rate to 142 kbps. The network protocol not only
allows individual device addressing, but also enables a novel search algorithm that allows a
network master to discover all unique 64-bit addresses on any particular network. Power is
drawn by the network device from the data line while it is high, and stored onboard the device in
a nominal 800 pF capacitor for use during the intervals when the line is low.
The presence of multiple devices on a single network can increase the capacitive load on the
network. Reflections at longer network distances and higher capacitance per unit length conductor
pairs can cause slew rate problems, and contribute to limitations in the total fanout, or number
devices that can be accommodated on a single network. Active pull-up drivers, careful attention to
line termination, and proper transmission line design can compensate some of these effects. A well-
designed system might accommodate several hundred devices, and has been demonstrated in a
laboratory setting (Dallas, 2004).

12.2.4.3 Mechanical Integration

Previous efforts to integrate sensors and MEMS devices into fiber-reinforced composites have often
required that such a device be placed between the fiber layers of a composite as it is being
fabricated. As a result, the device is usually surrounded by the matrix phase, typically a polymer,
which is generally the weak phase within the composite. Interlaminar regions are often the source
of failure within a composite due to delamination, since often there is no reinforcement in the
thickness direction (this problem may be minimized by through-the-thickness stitching). Depend-
ing on the size of an embedded device, this region is further weakened by the presence of sensors
which can serve as stress concentrators. These sensors act as discontinuities in transferring stress
within the material, leading to matrix cracking, debonding, delamination, and ultimately mechan-
ical failure. Fiber optic sensors do not present a similar problem, since the sensor itself is a fiber that