A Smart, Data-Driven Approach to Qualify Additively Manufactured Steel Samples for Print-Parameter-Based Imperfections

Author:

Alaparthi Suresh12,Subadra Sharath P.12ORCID,Sheikhi Shahram12ORCID

Affiliation:

1. Institute of Materials Science and Joining Technology, University of Applied Science Hamburg, Berliner Tor 13, D-20099 Hamburg, Germany

2. Forschungs-und Transferzentrum 3i, University of Applied Science Hamburg, Berliner Tor 13, D-20099 Hamburg, Germany

Abstract

With additive manufacturing (AM) processes such as Wire Arc Additive Manufacturing (WAAM), components with complex shapes or with functional properties can be produced, with advantages in the areas of resource conservation, lightweight construction, and load-optimized production. However, proving component quality is a challenge because it is not possible to produce 100% defect-free components. In addition to this, statistically determined fluctuations in the wire quality, gas flow, and their interaction with process parameters result in a quality of the components that is not 100% reproducible. Complex testing procedures are therefore required to demonstrate the quality of the components, which are not cost-effective and lead to less efficiency. As part of the project “3DPrintFEM”, a sound emission analysis is used to evaluate the quality of AM components. Within the scope of the project, an approach was being developed to determine the quality of an AM part dependent not necessarily on its geometry. Samples were produced from WAAM, which were later cut and milled to precision. To determine the frequencies, the samples were put through a resonant frequency test (RFM). The unwanted modes were then removed from the spectrum produced by the experiments by comparing it with FEM simulations. Later, defects were introduced in experimental samples in compliance with the ISO 5817 guidelines. In order to create a database of frequencies related to the degree of the sample defect, they were subjected to RFM. The database was further augmented through frequencies from simulations performed on samples with similar geometries, and, hence, a training set was generated for an algorithm. A machine-learning algorithm based on regression modelling was trained based on the database to sort samples according to the degree of flaws in them. The algorithm’s detectability was evaluated using samples that had a known level of flaws which forms the test dataset. Based on the outcome, the algorithm will be integrated into an equipment developed in-house to monitor the quality of samples produced, thereby having an in-house quality assessment routine. The equipment shall be less expensive than conventional acoustic equipment, thus helping the industry cut costs when validating the quality of their components.

Funder

Bundesministerium für Wirtschaft und Klimaschutz (BMWK) Zentrales Innovationsprogramm Mittelstand

Publisher

MDPI AG

Reference29 articles.

1. Assessment of grain structure evolution with resonant ultrasound spectroscopy in additively manufactured nickel alloys;Rossin;Mater. Charact.,2020

2. Metal additive manufacturing: A review of mechanical properties;Lewandowski;Annu. Rev. Mater. Res.,2016

3. (2022). Additive Fertigung-Grundlagen-Terminologie/Additive Manufacturing-Generl Principles-Fundamentals and Vocabulary (Standard No. DIN EN ISO/ASTM 52900:2022-03).

4. Application of finite element, phase-field, and calphad-based methods to additive manufacturing of ni-based superalloys;Keller;Acta Mater.,2017

5. Wire-feed additive manufacturing of metal components: Technologies, developments and future interests;Ding;Int. J. Adv. Manuf. Technol.,2015

Cited by 1 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3