A Partnership That Really Works
Work on a five-year, $11 million cost-shared program directed to enhancing the air flow passages of automotive components has recently been completed. A joint venture was formed for the collaborative effort, partnering Extrude Hone, Ford Motor Company, General Motors, University of Pittsburgh, University of Nebraska at Lincoln, and Extrude Hone (Irwin, Pennsylvania).
Work on a five-year, $11 million cost-shared program directed to enhancing the air flow passages of automotive components has recently been completed. A joint venture was formed for the collaborative effort, partnering Extrude Hone, Ford Motor Company, General Motors, University of Pittsburgh, University of Nebraska at Lincoln, and Extrude Hone (Irwin, Pennsylvania). Funded by the National Institute of Standards and Technology’s competitive Advanced Technology Program, the objective of the research was to develop sensing techniques and abrasive flow machining to enable automotive flow passages to be processed to a target airflow that is measured while the component is being machined. Abrasive flow machining (AFM), invented by Extrude Hone, is a process used to deburr and smooth the internal surfaces of a wide variety of flow passages. The central concept of flow control machining is one of machining to function rather than to geometry.
The focus of the ATP-funded project was the engine passages that control the volume of air ingested. These passages are too complex to be economically machined by conventional machining or grinding and are typically cast. The cast cavities vary in shape and position and have rough, irregular surfaces. These variations have significant impact on the performance, fuel efficiency and emissions of automotive engines. The ability to control these variations would enable building higher performance, more fuel efficient and cleaner burning engines at lower cost and in lower economical production volumes than currently possible.
Extensive data collection, before and after air flows of specific component test pieces, coupled with continuous process flow data, provided a large information base that, with appropriate analysis techniques, resulted in a derivation of a correlation between air and AFM media. Variations in the machining medium, especially in viscosity and abrasive grain condition, proved to have significant impact on the control of the flow machining process. The developed correlations rely on the characteristics of the media being held constant. This is next to impossible since the media will naturally change during the process with abrasive wear, loading of the media with particulates from the component machining, and the realities of a shopfloor environment. For these reasons, the control system was limited in its predictions of target flow, but it was instrumental in both process and tooling development.
Monitoring the flow rate of the machining medium in real time to the required resolution proved a significant task. Acoustic emission sensing capabilities showed the most promise for this application. Extensive experimentation was performed to correlate changes in flow resistance to characteristics of the AE signal. The research showed that an analysis of the outputs from the AE signal did indicate when to stop machining for flow control processing of intake manifolds.
The resultant flow control machining cell is comprised of multiple stations for tracking, inspecting airflow, processing and cleaning. By monitoring the results of the inspection stations and correlating these with the information from the processing station, it is possible to determine what action needs to be taken to maintain control of the machining process.
A manifold is comprised of twelve runners, or six pairs consisting of a long runner and a short runner. The initial goal of the effort was to achieve increased airflow on the overall component. One way of approaching flow control machining is to process a given passage until it achieves the target flow rate. At this point, this passage would be blocked and processing on it would stop. Although this would appear to be a very efficient approach, it is not always feasible for every component. Another method is to sequentially open blocked passages so that the passages requiring the most machining are open longer than those that do not. By choosing the appropriate point to open each passage, at the end of the cycle the passages are evenly processed. Not only does this produce improved airflow on the overall component, but it also balances airflow among the pairs of runners, resulting in a longer lasting product.
In the absence of a solid relationship between machining medium flow and airflow, the volume, scope and accessibility of the data allowed for an efficient process to be developed empirically. Acoustic emission data collected from different runners can be good indicators of the volume of machining medium necessary to produce a properly machined passage in the component. Using this information, engineers developed a processing scheme in which the opening times are easily determined. In practice, this proved to be an efficient method and worked well in high production.