Multi-sensor geometry measurement on large-scaled gears
Final Report Abstract
Demand for scalable measurement systems for comprehensive gear measurements is growing, especially for large gears. Therefore, this project introduced a scalable model-based multi-sensor system for large gear meas- urements. The mean base circle radius was to be determined with a measurement uncertainty < 5 µm (k = 1) as a fundamental shape parameter and correlates with the gear shape parameter profile slope deviation. Using a parametric involute geometry model, the base circle radius was iteratively evaluated from the gear measurement data. Two scalable multi-sensor approaches (static: n ≥ 4 sensors, dynamic: n ≥ 1 sensors combined with a rotary table) were designed for large gear measurements and investigated concerning the achievable measurement uncertainty. Theoretical investigations showed that the use of several sensors or the acquisition of many meas- uring points significantly reduces the achievable uncertainty. Regardless of the gear size, simulations verified an uncertainty < 5 µm. Experiments on a laboratory scale showed that sensors with small acceptance angles can be used by arranging the sensors tangentially to the nominal base circle. By using commercial triangulation and confocal-chromatic sensors, gear measurement is possible without retroreflectors. The interaction of the sensors with the surface topography significantly influenced the uncertainty. Confocal-chromatic sensors show great potential here and measure the tooth flank shape with single-digit µm uncertainty. Due to lacking knowledge of the exact sensor arrangement, the multi-sensor system was calibrated using a known gear to reduce systematic errors. Afterward, the mean and tooth-specific base circle radii were measured with an uncertainty < 5 µm. Based on the correlation between base circle radius and profile slope deviation, the profile slope deviation could also be quantified with an uncertainty < 2 µm. Large gear measurements showed that the mean base circle radius cannot currently be measured with the aimed uncertainty < 5 µm. Unknown systematic errors dominate the total uncertainty and could not be corrected with the designed calibration strategy. Although the scalability of the model-based multi-sensor system could not yet be confirmed experimentally, simulations and the random error < 3 µm experimentally achieved demonstrate the existing potential of the multi-sensor approach for large gear measurements. Compared to tactile measure- ments, the multi-sensor approach reduced the measurement time by a factor of 15 (for n = 1 sensor combined with the rotary table used) for a comprehensive measurement of all teeth. If the number of sensors n is increased, the measuring time can be further reduced.
Publications
- Multisensory measurement of the base circle radius as a fundamental shape parameter of large gears. International Conference on Gears, München, 18.- 20.9.2019, pp. 1207-1214
M. Pillarz, A. von Freyberg, A. Fischer
(See online at https://doi.org/10.51202/9783181023556-1207) - Combined partitioning and approximation for optimized gear inspection. 20th International Conference of the European Society for Precision Engineering and Nanotechnology (EUSPEN), E-conference, 8.-12.6.2020, pp. 289-290
A. von Freyberg, A. Fischer
- Determination of the mean base circle radius of gears by optical multi-distance measurements. Journal of Sensors and Sensor Systems 9(2):273–282, 2020
M. Pillarz, A. von Freyberg, A. Fischer
(See online at https://doi.org/10.5194/jsss-9-273-2020) - Gear shape parameter measurement using a model-based scanning multi-distance measurement approach. Sensors 20(14):3910 (16 pp.), 2020
M. Pillarz, A. von Freyberg, A. Fischer
(See online at https://doi.org/10.3390/s20143910) - Optical multi-distance measurements of spur gears. Sensor and Measurement Science International (SMSI 2020), Proceedings, 2020, No. C6.4, pp. 193-194
M. Pillarz, A. von Freyberg, A. Fischer
- Gear Shape Measurement Potential of Laser Triangulation and Confocal-Chromatic Distance Sensors. Sensors 21(3):937 (22 pp.), 2021
M. Pillarz, A. von Freyberg, D. Stöbener, A. Fischer
(See online at https://doi.org/10.3390/s21030937)
