Logo image
An offset multilayered optic sensor for shear and pressure measurement
Conference paper

An offset multilayered optic sensor for shear and pressure measurement

Chao-Shih Liu, Gai-Wen Chou, Yi-Lang Lyu, Per G. Reinhall and Wei-Chih Wang
Proceedings of SPIE - The International Society for Optical Engineering, Vol.6935, 69351B
2008

Abstract

Bend loss sensor Fiber optic sensor Neural network Pressure sensor Shear sensor Electronic Optical and Magnetic Materials Condensed Matter Physics Computer Science Applications Applied Mathematics Electrical and Electronic Engineering
Simultaneous recording of shear and pressure is an important requirement for study the causes of foot ulceration. In order to obtain a more robust and meaningful picture of what is occurring on the plantar surface of the foot, we have developed a multi-layered optical bend loss sensor that can be accommodated for shear and pressure measurement of an extended area. The sensor is made of two layers of crisscross fiberoptic sensor array separated by an elastomeric layer. Each sensing layer has multiple fibers molded into a thin polydimethylsiloxane (PDMS) substrate to form a mesh array. The top layer uses 6 fibers to create a 3 by 3 mesh with 9 intersection points and the bottom layer uses 8 fibers to create a 4 by 4 mesh with 16 intersection points. The space between the adjacent fibers is 0.5cm. Measuring changes of light intensity transmitted through the fiber provides information about the force induced changes of the fiber's radius of curvature. Pressure is measured based on the force induced light loss from the two affected crossing fibers divided by each sensing area. Shear was measured based on the relative position changes on these pressure points between the two fiber mesh layers. The design is an offset layout because the intersection points of the top and bottom layer are offset by 0.25 cm which can increase the shear sensing sensitivity. For testing the sensor with various loading condition, a neural network algorithm is induced to identify the loading pattern and the shear direction. Three loading patterns with 5 different loading directions were tested and a >90% accuracy was obtained using an algorithm using 2 neural networks.

Metrics

1 Record Views

Details

Logo image