Modeling of Infrared Gas Sensors Using a Ray Tracing Approach

Many gas molecules absorb electromagnetic radiation at characteristic wavelengths in the infrared region. This absorption can be used to identify defined substances like CO2, ammoniac, and so forth. A lot of different types of gas sensors are based on the principle of infrared absorption like photoacoustic sensors (e.g., Golay cells), dispersive infrared instruments (e.g., utilizing a diffraction grating), or Fourier transform infrared (FTIR) spectrometers. However, the most commonly used types of infrared absorbing gas sensors are nondispersive infrared (NDIR) sensors. Such a NDIR gas sensor consists of some basic function blocks, i.e., an IR-source, the sensor cell or optical path containing the sample gas, a gas specific filter, and an IR-detector. One of the central issues in the design of this kind of sensors is the geometry of the sensor cell. In this paper we investigate the use of statistic ray tracing to predict the efficiency of 3-D cell geometries for NDIR gas sensors.We demonstrate the feasibility of the method and show examples on how to apply it on given 3-D sensor models where we illustrate the agreement between simulated and measured gas response data.