Showing papers by "Yung C. Shin published in 2019"

In-Process monitoring of porosity during laser additive manufacturing process

Abstract: This paper describes a deep-learning-based method for porosity monitoring in laser additive manufacturing process. A high-speed digital camera was mounted coaxially to the process laser beam for in-process sensing of melt-pool data, and convolutional neural network models were designed to learn melt-pool features to predict the porosity attributes in deposited specimens during laser additive manufacturing. With the image processing tools developed in this paper, the extraction of porosity information from raw quality inspection data, such as cross-section images and tomography data sets, can be automated. The CNN models with a compact architecture, part of whose hyperparameters were selected through cross-validation analysis, achieved a classification accuracy of 91.2% for porosity occurrence detection in the direct laser deposition of sponge Titanium powders and presented predictive capacity for micro pores below 100 μm. For local volume porosity prediction, the model also achieved a root mean square error of 1.32% and exhibited high fidelity for both high porosity and low porosity specimens.

Ultrafast Laser Applications in Manufacturing Processes: A State-of-the-Art Review

Abstract: With the invention of chirped pulse amplification for lasers in the mid-1980s, high power ultrafast lasers entered into the world as a disruptive tool, with potential impact on a broad range of application areas. Since then, ultrafast lasers have revolutionized laser-matter interaction and unleashed their potential applications in manufacturing processes. With unprecedented short pulse duration and high laser intensity, focused optical energy can be delivered to precisely defined material locations on a time scale much faster than thermal diffusion to the surrounding area. This unique characteristic has fundamentally changed the way laser interacts with matter and enabled numerous manufacturing innovations over the past few decades. In this paper, an overview of ultrafast laser technology with an emphasis on femtosecond laser is provided first, including its development, type, working principle, and characteristics. Then ultrafast laser applications in manufacturing processes are reviewed, with a focus on micro/nano machining, surface structuring, thin film scribing, machining in bulk of materials, additive manufacturing, bio manufacturing, super high resolution machining, and numerical simulation. Both fundamental studies and process development are covered in this review. Insights gained on ultrafast laser interaction with matter through both theoretical and numerical research are summarized. Manufacturing process innovations targeting various application areas are described. Industrial applications of ultrafast laser based manufacturing processes are illustrated. Finally, future research directions in ultrafast laser based manufacturing processes are discussed.

Self-Sufficient Modeling of Single Track Deposition of Ti–6Al–4V With the Prediction of Capture Efficiency

Abstract: Understanding the capture efficiency of powder during direct laser deposition (DLD) is critical when determining the overall manufacturing costs of additive manufacturing (AM) for comparison to traditional manufacturing methods. By developing a tool to predict the capture efficiency of a particular deposition process, parameter optimization can be achieved without the need to perform a costly and extensive experimental study. The focus of this work is to model the deposition process and acquire the final track geometry and temperature field of a single track deposition of Ti–6Al–4V powder on a Ti–6Al–4V substrate for a four-nozzle powder delivery system during direct laser deposition with a LENS™ system without the need for capture efficiency assumptions by using physical powder flow and laser irradiation profiles to predict capture efficiency. The model was able to predict the track height and width within 2 μm and 31 μm, respectively, or 3.3% error from experimentation. A maximum of 36 μm profile error was observed in the molten pool, and corresponds to errors of 11% and 4% in molten pool depth and width, respectively. Based on experimentation, the capture efficiency of a single track deposition of Ti–6Al–4V was found to be 12.0%, while that from simulation was calculated to be 11.7%, a 2.5% deviation.

Effects of Composition and Post Heat Treatment on Shape Memory Characteristics and Mechanical Properties for Laser Direct Deposited Nitinol

Abstract: Nitinol structures were synthesized in a fully dense form using a laser direct deposition method. The pure elemental metal powders of nickel and titanium were used and powder ratios were controlled to arrive at the prescribed final chemical compositions of Nitinol. The transformation temperatures of synthesized Nitinol samples with different chemical composition and post heat treatment conditions were systematically analyzed and compared with those of conventional Nitinol. Compared to Nitinol parts produced by other techniques, the laser engineered net shaping (LENS) created the least amount of secondary phase, indicating the possibility of high corrosion resistance. Two step post processing of solution heat treatment and aging heat treatment was carried out to improve the homogeneity of the microstructure and to investigate its effects on phase transformation temperatures. The resultant phase transformation temperatures could be controlled by the heat treatment parameters. Compression test results showed mechanical properties of synthesized Nitinol samples are largely affected by its post heat treatment history while the effect of initial chemical composition was negligible.

Thermo-fluid Topology Optimization and Experimental Study of Conformal Cooling Channels for 3D Printed Plastic Injection Molds

Abstract: With the advent of additive manufacturing, innovative design methods, such as network-based techniques, and structural topology optimization have been used to generate complex and highly efficient cooling systems in recent years. However, methods that incorporate coupled thermal and fluid analysis remain scarce. In this paper, a coupled thermal-fluid topology optimization algorithm is introduced for the design of conformal cooling channels. The problem is formulated based on a coupling of Navier- Stokes equations and convection-diffusion equation. The problem is solved by gradient-based optimization after analytical sensitivity derived using adjoint method. With this method, the channel position problem is replaced to a material distribution problem. The material distribution directly depends on the effect of flow resistance, heat conduction, natural and forced convection. The algorithm leads to a two-dimensional conceptual design having optimal heat transfer and balanced flow, which is further transformed into three-dimensional cooling channel design. Here, a comprehensive study is presented, starting from design, simulation, 3D printing process and experimental testing of an injection mold with conformal cooling channels in industrial production environment. A traditional mold model is provided by an industrial collaborator. To enhance the overall thermo-fluid performance of the mold and improve final product quality, a redesign of this mold core is done with conformal cooling channels inside. The final design is 3D printed in pre-alloyed tool-steel powder Maraging Steel using Truprint 3000 metal 3D printing machine. The printed core required some heat treatment and finishing processes and added features to be incorporated to make it production ready. Once all the preparation was complete, the core was tested experimentally in a multicavity injection molding machine in real industrial environment at our industrial partner’s production facility. This paper describes all the steps starting from design, analysis, die 3D printing and finally ending at final experimental testing, as well as recommendations for tool designer and injection molding industry to implement additive manufacturing for their benefit. This paper is not just focused on a specific aspect such as design, simulation or manufacturing, but rather a comprehensive paper presenting a case study on implementation of topology optimization and additive manufacturing in real life industrial production scenario.

Enhancement of weld strength of laser-welded joints of AA6061-T6 and TZM alloys via novel dual-laser warm laser shock peening

Abstract: In this paper, an experimental study is presented on an investigation to improve the weld strength of laser-welded joints via post-processing by warm laser shock peening (wLSP). A dual-laser setup was utilized to simultaneously heat the sample to a prescribed temperature and to perform the wLSP process on the laser-welded joints of AA6061-T6 and TZM alloys. Joints in overlap and bead-on-plate configurations were created by laser welding by a high-power fiber laser and post-processed with wLSP. The tensile tests carried out on wLSP-processed AA6061-T6 samples demonstrate an enhancement in the strength by about 20% over as-welded samples and the ductility of samples processed by wLSP improved by 30% over as-welded samples. The bead-on-plate (BOP) welds of TZM alloy processed with wLSP demonstrated an enhancement in strength by about 30% and the lap welds processed with wLSP demonstrated an increase in the joint strength by 22%. Finite element analysis revealed that the depth and magnitude of compressive stresses imparted by wLSP were greater than room temperature laser shock peening (rtLSP), which contributed to the enhancement of the joint strength for processed samples.

Investigation of the Machining Behavior of Ti6Al4V/TiC Composites During Conventional and Laser-Assisted Machining

Abstract: Conventional machining of Ti6Al4V/TiC composites is a very difficult process, which exhibits a peculiar cutting force pattern where the thrust forces are higher than the tangential forces. This behavior results in rapid tool wear and consequently very short tool life. This study is concerned with describing the reasons for the attendant behavior using experimentally validated 3D finite element simulations and alleviating this behavior via laser assisted machining (LAM). Simulations were conducted using an equivalent homogeneous model (EHM) and a multiscale heterogeneous model (MHM) of the Ti6Al4V/TiC composite. Results showed a good agreement between the tangential forces obtained from experiments, EHM, and MHM for conventional machining and LAM. However, only the MHM was able to successfully predict the unusual high thrust forces. The MHM simulation results showed that the tool/particle interaction along the tool nose region presented the highest resistance due to the high resistance against pushing the TiC particles by the tool into the machined surface. This resistance results from the efficient load transfer capability between the particles and the matrix below the machined surface. When using LAM, the stated resistance was decreased by the reduction in load transfer capability of the Ti6Al4V/TiC workpiece such that the thrust and tangential forces were reduced by 78% and 37%, respectively, according to the MHM simulation. The experimental results showed that the tool wear was improved by 68% by LAM. All the results demonstrated that the MHM successfully captured the underlying machining mechanism of the Ti6Al4V/TiC composites.

A Framework for Estimating Mold Performance Using Experimental and Numerical Analysis of Injection Mold Tooling Prototypes

Abstract: Additive Manufacturing (AM), 3D printing, rapid prototyping, or rapid tooling refer to a range of technologies that are capable of translating virtual CAD model data into physical model. It is executed in growing number of applications nowadays. A wide range of materials are currently being used to produce consumer products and production tools. AM has brought in revolutionary changes in traditional manufacturing practices. Yet, there are certain drawbacks that hinder its advancement at mass manufacturing. High cost associated with AM is one of them. Using 3D printed tooling can provide long-time cost effectiveness and better product quality. Additively manufactured injection molds can increase the cooling performance, reduce production cycle time, and improve surface finish and part quality of the final plastic product. Yet, manufacturers are still not using the printed molds for industrial mass production. Numerical analysis can provide approximation of such improved performance, but, factual experimental results are necessary to satisfy performance criteria of molds to justify the large investment into tooling for existing industries. In this research work, a desktop injection molding machine is used to evaluate performance of 3D printed molds to develop a cost and performance analysis tool. It serves as a baseline to predict the performance of molds in real-time mass manufacturing of consumer products. The analysis describes how appropriate the estimation can be from any simulation study of molds, how much the scaling down of tool and molding system can affect the prediction of actual performance, what correction factors can be used for better approximation of performance matrices. Several “scaled down” prototypes of injection molds have been used. They have design variations as: with or without cooling system, conformal or straight cooling channels, solid or lattice matrix, and metal or tough resin as the mold material. The molds are printed in in-house printing machines and can also be printed online with limited charges. This also provides an excellent demonstration of using inexpensive material and manufacturing process, such as resin to estimate the performance of highly expensive 3D printed stainless steel molds. The work encompasses a framework to reduce overall cost of implementing AM, by lowering time and monetary expenses during the research and development, and prototyping phases.