Delphi Selects ALGOR FEA to Optimize Oxygen Sensor

Wide-Range Oxygen Sensors to Make Automotive Engines Cleaner and More Efficient

	The wide-range oxygen sensors under development at Delphi will help the next generation of automobiles to run more cleanly and efficiently.

Increasing fuel efficiency and reducing environmentally unfriendly exhaust emissions are two major, ongoing goals of the automotive industry ?goals which can be achieved by attaining the proper mix of oxygen and fuel. The next generation of automotive engines will use wide-range oxygen sensors, which can determine how rich or lean the air/fuel (A/F) ratio is and help to regulate clean, efficient motor operation. Delphi Corporation, a world leader in mobile electronics and transportation components located in Troy, Michigan, chose ALGOR FEA software to thermally optimize a wide-range oxygen sensor. 揑 selected ALGOR because it is a complete package with CAD support, meshing and easy-to-use analysis tools,?said Senior Project Engineer C. Scott Nelson, who optimized the sensor design. 

Better Sensors for Better Fuel Efficiency

Oxygen sensors have been used in automotive exhaust systems for over 25 years. To date, the type of sensor used, called a switching oxygen sensor, can only determine whether the A/F ratio is rich (excess fuel) or lean (excess oxygen). Replacing switching oxygen sensors with wide-range oxygen sensors is one of the technologies that is contributing to the development of lean burn engines, which lets the engine burn less fuel under low pressure demand, but increases intake to admit more fuel when needed, such as during acceleration. Burning less fuel contributes to fuel efficiency while lower emissions result from the fuel combusting more completely. 

In designing any exhaust sensor, thermal optimization is critical due to the extreme operating conditions from ?0癈 to over 1000癈. Durable, cost-effective materials and maintaining a short overall sensor length are also important design considerations. Important sensor components such as the terminal and seal are typically made of materials that can break down over time if the temperature of the sensor is not controlled. Although increasing the size of a part is an easy way to reduce the temperature, shorter sensors experience less potentially destructive vibration than longer ones. In addition, auto manufacturers prefer shorter sensors because they are easier to integrate into exhaust designs and easier to install. 

The illustration above shows the initial (left) and final (right) designs for the wide-range oxygen sensor. 

FEA-Based Thermal Optimization

To thermally optimize the sensor, Nelson worked with a 2-D axisymmetric cross-section of the sensor. The thermal loads and constraints represented worst case conditions of 1000癈 exhaust temperature, 150癈 ambient temperature and free convection (no air current). These conditions were simulated using a combination of convection, conduction and radiation loads. 

Over the course of dozens of analyses, Nelson optimized the geometry and material properties of the sensor components. The strategies behind the design changes were to restrict vertical heat flow, promote radial heat flow and increase heat flow through the components. In some cases, he even experimented with different materials having different thermal conductivity properties without having a particular material in mind and then researched materials that had similar properties. 

揢sing ALGOR, I was able to reduce the temperature at two critical locations in the sensor by 20% which kept the peak temperatures below the material抯 maximum threshold; this greatly improved sensor durability and robustness,?said Nelson. 

The wide-range oxygen sensor was modeled using a 2-D axisymmetric cross-section (lower right). The heat transfer analysis results of the final design are shown above (upper left).   

Laboratory tests using a dynamometer correlated well with the FEA results. 揗y results correlated to laboratory results within 4%,?said Nelson. 揑抦 very satisfied with this correlation, especially since the variables of an experiment can never be controlled as well in the laboratory as they can be with an FEA model. Still air is especially difficult to replicate experimentally; even a small amount of air flow can significantly affect the results.?

	C. Scott Nelson of Delphi Corporation used ALGOR software to optimize a wide-range oxygen sensor that will help automobiles to run more cleanly and efficiently.

This thermal optimization was not only Nelson抯 first project using FEA, but a departure from the way he has designed products in the past. 揑teratively analyzing designs and optimizing both the geometry and the materials used helped me to develop a much better design than I could have achieved without those virtual 慼ands-on?results,?said Nelson. 揂s a product designer, I find that performing my own analyses leads to a much more interactive and informed design process. It saves tens of thousands of dollars over iterative prototype testing and, with a simple model like this one, I can do far more iterations than would be logistically feasible if I turned the analysis work over to someone else.?p> The completed sensor design is currently being integrated into the next generation of automotive engines, which are expected to be used in 2004.

C. Scott Nelson is a Senior Project Engineer for Delphi Corporation. He holds a BSME from Lawrence Technological University and a MSE from Purdue University. Delphi Corporation is a world leader in mobile electronics and transportation components and systems. Headquartered in Troy, Michigan, Delphi supplies major automotive manufacturers as well as providing aftermarket automotive parts.