Blog
About

0
views
0
recommends
+1 Recommend
1 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found

      Prediction of Hardness after Industrialized Bainitization of 100Cr6 based on Process Parameters by Application of Machine Learning Methods∗ Translated title: Vorhersage der Härten nach industrialisiertem Bainitisieren von 100Cr6 basierend auf Prozessparametern durch Anwendung von Methoden des Maschinellen Lernens

      Read this article at

      ScienceOpenPublisher
      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          14,000 data sets from an industrialized bainitization process, consisting of process gas furnace, salt bath and circulating air furnace, were used to predict the resulting Vickers hardness of cylinder heads made of 100Cr6 based on process data such as temperature and pressure. For prediction, machine learning methods such as ANNs, CNNs, ensemble methods and support vector regressors were compared. Meta features such as the furnace number as well as features extracted from the recorded time series were used. Data preparation and feature extraction were performed according to the machine learning methods used. The random forest achieved the best predictions with an R 2 score of 0.406 and also allows the evaluation of the most important features.

          Kurzfassung

          14.000 Datensätze aus einem industrialisierten Bainitisierungsprozess, bestehend aus Prozessgasofen, Salzbad und Umluftofen, wurden verwendet, um basierend auf den Prozessdaten, wie Temperatur und Druck, die resultierende Vickers-Härte von Zylinderköpfen aus 100Cr6 vorherzusagen. Zur Vorhersage wurden Methoden des Maschinellen Lernens wie ANNs, CNNs, Ensemble-Methoden und Support-Vector-Regressoren miteinander verglichen. Es wurden sowohl Metafeature wie die Ofennummer als auch Feature benutzt, die aus den aufgezeichneten Zeitreihen extrahiert wurden. Die Datenaufbereitung sowie die Feature-Extraktion wurden entsprechend der eingesetzten Methoden des Maschinellen Lernens durchgeführt. Der Random Forest erzielte mit einem R 2-Score von 0,406 die besten Vorhersagen und ermöglicht zusätzlich die Auswertung der wichtigsten Feature.

          Related collections

          Most cited references 11

          • Record: found
          • Abstract: not found
          • Article: not found

          Artificial neural networks for modelling the mechanical properties of steels in various applications

            Bookmark
            • Record: found
            • Abstract: not found
            • Article: not found

            New approach for the bainite start temperature calculation in steels

              Bookmark
              • Record: found
              • Abstract: not found
              • Article: not found

              Predictions of toughness and hardness by using chemical composition and tensile properties in microalloyed line pipe steels

                Bookmark

                Author and article information

                Journal
                htme
                HTM Journal of Heat Treatment and Materials
                Carl Hanser Verlag
                1867-2493
                2194-1831
                13 August 2020
                : 75
                : 4
                : 212-224
                Affiliations
                1 Robert Bosch GmbH, Powertrain Solutions Division, Manufacturing Technology Management, Stuttgart, Deutschland
                2 Robert Bosch GmbH, Corporate Sector Research and Advance Engineering, Materials- and Process Engineering Metals, Renningen, Deutschland
                3 KIT Karlsruhe, Institute for Applied Materials – Material Science and Engineering (IAM-WK), Karlsruhe, Deutschland
                Author notes
                4 Yannick.Lingelbach@ 123456de.bosch.com (corresponding author/Kontakt)
                [∗]

                Lecture held at the HeatTreatingCongress, HK; October 22–24, 2019 in Cologne, Germany. The lecture was awarded the Paul Riebensam prize.

                Article
                HT110415
                10.3139/105.110415
                © 2020, Carl Hanser Verlag, München
                Page count
                References: 22, Pages: 13
                Product
                Self URI (journal page): http://www.hanser-elibrary.com/loi/htme
                Categories
                Scientific Contributions/Fachbeiträge

                Comments

                Comment on this article