Methods for fast setup and robust dimensional measurements with industrial x-ray computed tomography
Zusammenfassung der Projektergebnisse
Industrial CT for dimensional metrology is a highly innovative and versatile technology for the nondestructive testing and entire inspection of industrial parts with an outstandingly high measurement point density. In particular for the reduction of the time to market of new products CT can speed up the production part approval process. Current limitations concerning the application of industrial CT are mainly due to the reachable accuracy for dimensional measurements and the time-consuming process of setting up and performing the tomographic imaging. The measurement is affected by fix systeminherent factors, i.e. hard- and software, and by the parameter setup of the system defined by the system operator. Depending on his experience, the chosen parameter values may lead to higher measurement deviations for a given workpiece. The aim is to minimize the operator influence on the measurement and find the most suitable parameter set for a certain task, so that a reliable and objective decision based on the measurement result can be taken, e.g. if the production of a new part can be started or not. As CT is a complex process with a lot of influencing factors and many non-linear interdependencies, the finding of adequate parameter values for the task currently exceeds the operators capability since experts can often not explain why a certain setup has been chosen. Expertise is learned knowledge, which cannot easily be distributed or transferred to other operators. How can the operator optimize his setup in an efficient way without knowing the analytic function of the whole measurement chain? The approach is to utilize cognitive methods modelling operator’s expertise with the calculation performance of computers combined within a Knowledge-Based System (KBS). The hypothesis is that the KBS reduces the uncertainty in measurement for the CT measurement process using the available knowledge about the part and the tomographic imaging process. Using already existing and available knowledge about the workpiece, the tomography process, previous tomographies and the measurement task, improved parameter values are derived. The KBS uses a simulator which models relevant parts of the imaging process, on which a recommendation for workpiece parameter values can be done. A new task is compared to already existing parts in the data base to check if it is possible to adapt the parameter values to the actual case. Important to this issue is a sufficient similarity regarding the shape, size and material of the workpiece. In case of insufficient similarity, the simulator recommends workpiece parameter values. Within the project a KBS was designed and implemented. The required degree of detail for the process simulation was adapted from the project’s results. To find workpiece parameter values, different cognitive methods were compared and evaluated. The performance of the KBS to predict workpiece parameter values was validated by comparing the recommended KBS setups to setups done by experts and derived from elaborate experimental studies using quality criterions such as the uncertainty in measurement, the scatter variation, the Signal-to-Noise-Ratio (SNR) of the projection images and the imaging time. The part similarity was assessed quantitatively with a defined set of characteristics. Sensible tolerance values for the similarity were suggested. The quality of the tomographic imaging process as a base for an accurate measurement was also assessed by taskindependent characteristics. Apart from the tomographic imaging process the influence of the segmentation onto the workpiece geometry was analyzed by comparing two different surface extraction processes. The findings of this project improve both issues by developing a method for supporting the setup of CT measurements. This will lead to higher robustness and reliability of the obtained results whereby the number of possible applications for industrial CT can be increased.
Projektbezogene Publikationen (Auswahl)
- Einfluss der Bauteilorientierung auf die Messunsicherheit bei dimensionellen Computertomografie Messungen. In: Industrielle Computertomografie Tagung 2010 – Proceedings, Hrsg.: Kastner, J., 1. Aufl. , Shaker Verlag Aachen 2010, ISBN 978-3- 8322-9418-2, S. 221-226
Schmitt, R.; Niggemann, C.
- Reproducibility of dimensional Measurements performed by computed tomography. In: Proceedings of 10th International Symposium on Measurement and Quality Control 2010, September 5-9, Hrsg.: Takaya, Y., Osaka University Osaka, Japan 2010, ISBN 978-4-9905119-1-3, S. 1-4
Nardelli, V.; Donatelli, G.; Schneider, C.; Niggemann, C.; Schmitt, R.
- Uncertainty in measurement for x-ray-computed tomography using calibrated workpieces. In: Measurement Science Technology 21 (2010), 5, ISSN 1361-6501, S. 054008-1-054008-9
Schmitt, R.,Niggemann, C.
- Messunsicherheitsbestimmung für die industrielle Computertomografie anhand kalibrierter Bauteile. In: tm - Technisches Messen 78 (2011), 3, ISSN 0171-8096, S. 133-141
Schmitt, R.; Damm, B.; Niggemann, C.
- Method for efficient Identification of Similar Workpieces for X-Ray Computed Tomography. In: International Symposium on Digital Industrial Radiology and Computed Tomography (Proceedings): June 20 - 22, 2011, Berlin, Germany / DGZfP, Deutsche Gesellschaft für Zerstörungsfreie Prüfung e.V., ISBN 978-3-940283-34-4
Schmitt, R; Niggemann, C.
- Feature-based analysis for quality assessment of x-ray computed tomography measurements. In: Measurement Science and Technology 23 (2012), 10, ISSN 1361-6501, S. 105006
Nardelli, V.,Arenhart, F.,Donatelli, G.,Porath, M.,Niggemann., C.,Schmitt, R.
(Siehe online unter https://doi.org/10.1088/0957-0233/23/10/105006) - Knowledge-Based System to improve dimensional CT Measurements. In: Conference on Industrial Computed Tomography (ICT) 2012- Proccedings, 19. - 21. September 2012 University of Applied Sciences Upper Austria/ Wels Campus, Hrsg.: Kastner, J., Shaker Verlag Aachen 2012, ISBN 978-3-8440-1281-1, S. 363-372
Schmitt, R.; Isenberg, C.; Niggemann, C.
- Wissensbasiertes System zur Verbesserung dimensioneller Messungen mittels Röntgen-Computertomografie, Apprimus-Verlag Aachen 2012, ISBN 978-3-86359-093-2, Dissertation
Niggemann, C.
- Computertomografie für industrielle dimensionelle Messtechnik Computed tomography for industrial dimensional metrology. In: tm - Technisches Messen 81 (2014), 5, ISSN 0171-8096 ; 2196-7113 (Online), S. 201 – 208
Schmitt, R.; Isenberg, C.
(Siehe online unter https://doi.org/10.1515/teme-2014-0421) - Design of a multiwave standard to evaluate the frequency response of CT measuring systems. In: Proceedings of the 11th International Symposium on Measurement and Quality Control (ISMQC), 11-13 Sept. 2013, Cracow-Kielce, Poland, Hrsg.: Adamczak, S. et al. IMEKO: Abstracts of the ISMQC 2013 ID 198, Commitee of Mechanical Engineering of the Polish Academy of Sciences Kielce/ Poland 2014
Arenhart, F.; Nardelli, V.; Donatelli, G.; Porath, M.; Isenberg C. ;Schmitt, R.
