The 2021 3D Printing Industry Awards shortlists are open for voting, have your say now. Make sure to tune in to the live stream of the event on our YouTube channel.
A group of researchers from the University of Nebraska-Lincoln, Drexel University, Navajo Technical University, and SigmaLabs have developed a new process for detecting flaws in laser powder bed fusion (LPBF) 3D printed parts that use digital twins.
In a new paper, the team outlines a digital twin strategy that integrates physics and data to provide real-time detection of flaws as they form during the LPBF printing process. By combining in-situ meltpool temperature measurements with computational predictions, the researchers were able to detect and identify three different types of flaws in stainless steel impellers.
The goal of the study was to address concerns around the process’ tendency to create flaws in order to make it suitable for precision-driven industries like aerospace and biomedicine, while protecting against cybersecurity threats such as process tampering.
Flaw formation during LPBF
Despite the geometric freedoms and significant cost and time savings LPBF 3D printing can achieve, high-precision industries such as aerospace and medical have so far been hesitant in adopting the technology to make safety-critical parts, due to its tendency to create flaws.
Cybersecurity risks have also become another emerging concern not just within LPBF but in other 3D printing processes too, with malicious parties potentially able to tamper with the process parameters and plant flaws inside a part to compromise its performance.
Increasing research is being undertaken to address these issues and reduce the chance of defects within the LPBF process. The causes of microcracking in certain metals have been investigated in order to improve the process, as has the effects of beam-shaping.
Texas A&M, in particular, has carried out a lot of work in this area, having worked with Argonne National Laboratory to deploy machine learning to predict defects in 3D printed parts, and also establishing a LPBF “speed limit” at which defects such as J-shaped bubbles are less likely to form on 3D printed parts.
Just last month, Texas A&M scientists introduced a universal method of LPBF 3D printing flawless metal parts based on single-track printing data and machine learning. The team claims that its method is less expensive, time-consuming and simpler than existing parameter-optimizing methods, making it well-suited to aerospace, automotive and defense applications.
The digital twin approach
Flaws tend to be formed during LPBF processes as a result of thermal occurrences during the melting, cooling, solidification, and remelting of powder by the laser. At the micro-scale, the melting of the powder creates a wake of molten material, called the meltpool, within which the temperature distribution, flow, and spatter influence the part’s microstructure, porosity, and cracking.
At the macro-scale, the rapid scanning action of the laser and continuous melting of the material at high temperatures causes heating and cooling cycles which can lead to residual stress and part deformation.
To address this, the latest study aims to develop and apply a data and physics integrated strategy for online monitoring and detection of flaw formation in LPBF parts. To do this, the team has combined in-situ meltpool temperature measurements with a thermal simulation model that rapidly predicts the temperature distribution in a part.
According to the researchers, the novelty of their approach lies in the temperature distribution predictions provided by the model, which are updated layer-by-layer with the in-situ meltpool temperature measurements. As such, the scientists are calling their method the ‘digital twin’ approach to detect flaw formation.
The digital twin strategy is able to provide feedback for correcting anomalies in parts, thereby reducing waste from failed builds. The researchers are offering their strategy as an alternative to purely data-driven process monitoring techniques in order to overcome the drawbacks of these processes, namely delays in detection, poor generalizability of data-driven models to part shapes, and the expense and resource-intensive nature of acquiring data.
Additionally, because the digital twin incorporates both the macro-scale effect of part shape on thermal history and the micro-scale effect of laser-material interaction in the form of the meltpool temperature, it can incapsulate the effect of different processing parameters, such as the scanning pattern, hatch spacing, laser power, and velocity.
Testing the digital twin
To test their method, the team 3D printed four stainless steel impeller-shaped parts using an EOS M290 LPBF system that had different types of flaws covering process drifts, lens delamination, and cyber intrusions. To create the flaws, the researchers made changes in the processing parameters, prompted machine-related malfunctions, and deliberately tampered with the process to create flaws inside the part.
The team chose to print impeller parts to demonstrate their digital twin as it is divisible into three distinct regions along the build direction – the base, mid, and fin sections. Each of these sections include complex features that are challenging to print, such as a teardrop-shaped internal cooling channel, which resulted in varied cooling time between layers and subsequently a complex thermal history.
During the build, the process was monitored continuously using an array of three coaxial photodetectors integrated into the laser path. Signals obtained from the sensor array were processed to create two types of measurements, namely thermal energy planck (TEP) and thermal energy density (TED). The TEP signature was correlated to the meltpool temperature, while TED captured the broadband chamber radiation.
These signatures were then incorporated into the graph theory model to continually update it with the meltpool’s micro-scale activity throughout the process.
The digital twin was able to detect all three types of flaw in the 3D printed impeller parts during the LPBF process. According to the researchers, the results demonstrated that the method enabled precise and interpretable detection of flaw formation as opposed to the use of sensor data alone. To this end, the digital twin approach overcomes the need for transferring sensor signatures to a separate data analysis algorithm, and therefore prevents delays in detecting flaws.
Going forwards, the team will look to extend the capabilities of its digital twin to detect other types of flaws, such as distortion. They will also test the approach with different processing parameters, scanning strategies, and part shapes.
Further information on the study can be found in the paper titled: “Digitally twinned additive manufacturing: Detecting flaws in laser powder bed fusion by combining thermal simulations with in-situ meltpool sensor data,” published in the Materials and Design journal. The study is co-authored by R. Yavari, A. Riensche, E. Tekerek, L. Jacquemetton, H. Halliday, M. Vandever, A. Tenequer, V. Perumal, A. Kontsos, Z. Smoqi, K. Cole, and P. Rao.
Subscribe to the 3D Printing Industry newsletter for the latest news in additive manufacturing. You can also stay connected by following us on Twitter and liking us on Facebook.
Looking for a career in additive manufacturing? Visit 3D Printing Jobs for a selection of roles in the industry.
Subscribe to our YouTube channel for the latest 3D printing video shorts, reviews and webinar replays.
Featured image shows the researchers’ physics and data integrated digital twin strategy. Image via Materials & Design.
