Inexpensive Optical Accelerometer Is Simple to Fabricate
Breck Hitz
The advantage of optical accelerometers over their electrical counterparts is their ability to function in harsh electrical environments. Recently, scientists at Tecnische Universität Braunschweig in Germany and at Centre Nacional de Microelectrònica de Barcelona in Spain developed what they believe is the first polymer optical accelerometer, an inexpensive device that requires only two photolithographic steps to fabricate. Because it is very small and lightweight, its developers expect that it could be attractive for military and aerospace applications.
Figure 1. The center section of the inexpensive polymer accelerometer becomes misaligned with Y-axis acceleration, and the greater optical loss of light transmitted through the device is a measure of the acceleration. Images ©2005 IEEE.
The device is a trisection waveguide, with the middle section supported by four L-shaped beams (Figure 1). The operating principle is simple: When the device is accelerated in the horizontal direction (the Y-direction indicated in the figure), the middle section of the waveguide becomes misaligned. The optical loss through the trisectional waveguide is a measure of the acceleration.
The accelerometer is fabricated from the polymer material SU-8 from MicroChem Corp. of Newton, Mass., and measures 10.5 × 3.8 mm and is 280 µm thick. Besides being inexpensive and allowing simple device fabrication, the polymer's mechanical flexibility enables the use of the fishbone structures at each end of the accelerometer that automatically align optical fibers inserted into them, eliminating the tedious and expensive manual alignment required for most integrated optical devices. Moreover, SU-8 assures high-quality optical surfaces without polishing.
To compensate for the expansion of the beam in the free-space region between waveguide sections, each waveguide is wider than the preceding one. The waveguide widths in the input, sensing and output sections are 50, 60 and 70 µm, respectively. Because the entire device is fabricated from the same material, waveguiding in the central sensing section is accomplished by fabricating a periodic system of airholes along the edge of the waveguide.
In a computer simulation of the accelerometer's optomechanical performance, the scientists predicted a sensitivity of 11 dB/g. They illuminated the input section with 633-nm light from an LED and measured a loss of 8.3 dB when no acceleration was present. They believe that this was the result of residual loss at the SU-8 surfaces and of the fact that the output waveguide was thicker than the optical fiber that collected the exiting light, so that some of the light was lost.
Figure 2. The accelerometer demonstrated a sensitivity as high as 11 dB/g -- 2.5 times that of similar silicon-based optical accelerometers. The zero-acceleration optical loss of 8.3 dB has been removed from the experimental data to facilitate comparison with the computer simulation.
To measure the performance up to 1 g, the researchers merely changed the orientation of the device. When it was straight up -- that is, when the X-direction in Figure 1 was vertical -- it experienced no acceleration. When it was on its side -- the X-direction becoming horizontal -- it experienced 1 g. The scientists suggest that an initial misalignment of the waveguides was responsible for the disparity between the simulation and the experimental results (Figure 2).
Although their accelerometer was designed to measure relatively low acceleration, in the vicinity of 1 g, the scientists note that it could be redesigned to measure acceleration up to and perhaps beyond 10 g by thickening the L-beams that support the sensing section.
LATEST NEWS