Integrating Materials and Manufacturing Innovation最新文献

Abstract

Achieving targeted product performance requires the integrated exploration of design spaces across multiple levels of decision-making in systems comprising products, materials, and manufacturing processes—product-material-manufacturing process (PMMP) systems. This demands the capability to co-design PMMP systems, that is, share ranged sets of design solutions among distributed product, material, and manufacturing process designers. PMMP systems are subject to uncertainties in processing, microstructure, and models employed. Facilitating co-design requires support for simultaneously exploring high-dimensional design spaces across multiple levels under uncertainty. In this paper, we present the Co-Design Exploration of Multilevel PMMP systems under Uncertainty (CoDE-MU) framework to facilitate the simultaneous exploration of high-dimensional design spaces across multiple levels under uncertainty. The CoDE-MU framework is a machine learning-enhanced, robust co-design exploration framework that integrates robust, coupled compromise Decision Support Problem (rc-cDSP) construct with interpretable Self-Organizing Maps (iSOM). The framework supports multidisciplinary designers to (i) understand the multilevel interactions, (ii) identify the process mechanisms that affect material and product responses, and (iii) provide decision support for problems involving many goals with different behaviors across multiple levels and uncertainty. We use an industry-inspired hot rod rolling (HRR) steel manufacturing process chain problem to showcase the CoDE-MU framework’s efficacy in facilitating the simultaneous exploration of the product, material, and manufacturing process design spaces across multiple levels under uncertainty. The framework is generic and facilitates the co-design of multilevel PMMP systems characterized by hierarchical product-material-manufacturing process relations and many goals with different behaviors that must be realized simultaneously at individual levels.

Abstract

We present a new computational approach for large-scale segmentation and spatially-resolved analysis of melt pools in complex 3D printed parts and qualification artifacts. Our hybrid segmentation includes human-in-the-loop image processing of a few representative optical images of melt pools that are then used for training machine learning models for automated segmentation of melt pool boundaries in large parts. Our approach specifically targets minimizing the need for manual annotation. Considering imperfect segmentation and errors unavoidable with most algorithms, we further propose chord length distribution as a statistical description of melt pool sizes relatively tolerant to segmentation errors. We first show and validate our new approach on optical images of melt pools in a simple 3D printed plate sample (IN718 alloy) as well as selected regions of a complex qualification artifact (AlSi10Mg alloy). We then demonstrate the application of our approach on an entire cross section of the artifact.

Microstructure plays a crucial role in predicting the properties of parts by additive manufacturing. Fluid flow and temperature gradient are always recognized as key factors influencing the final microstructure. However, the effects of flow field were often ignored during microstructure simulation inside the molten pool. In this study, the Marangoni flow is firstly calculated using the finite element method. Fluid flow increases the temperature gradient and the cooling rate at the solid front. Subsequently, the temperature field and flow field are input to phase-field model to simulate the microstructure inside the molten pool. This integrated model is then applied to study the solidification behavior of IN718 alloy during additive manufacturing. The microstructure evolutions are analyzed in detail under different processing parameters. The simulation results demonstrate that the Marangoni flow has great effects on both molten pool and solidification microstructure. The integrated model developed in this work can predict the molten pool and solidification microstructure more accurately by combining the thermal, flow and microstructure models together.

Abstract

Thermal distribution during the sapphire growth process determines to a great extent the thermal stresses and dislocation density in sapphire. In this work, thermal and defect simulations of sapphire growth in a simplified single-boule furnace are presented. The heat transfer in the furnace is modeled via ANSYS Fluent® by considering conduction, convection and radiation effects. A dislocation density-based crystal plasticity model is applied for the numerical simulation of dislocation evolution during the crystal growth of sapphire. The physical models are validated by using a temporal series of measurements in the real furnace geometry, which capture the crystal–melt interface position during the technological growth process. The growth rate and the shape of the crystal growth front are analyzed for different side and top heater powers which result in different thermal distributions in the furnace. It is found that the cooling flux at the crucible bottom wall determines to a great extent the growth profile in the first half of the growth stage. Only toward the end of the growth stage, different top and side power distributions induce different growth front shapes. The effect of the convexity of the growth surface on the generation of dislocation defects is investigated by the crystal plasticity model. The results of simulations show that the convexity of the growth surface has a significant effect on the generation of dislocations.

This work reports the measurement of laser powder bed fusion (PBF) process input signals, output signals, and structural data for a set of eight IN 718 samples. Data from multiple samples imparts statistical replicability to the measurements. The input signals are the real-time PBF laser position commands, power commands, and the beam radius set point. The output signals are thermographic videos from coaxial and off-axis infrared cameras, and temperature measurements from thermocouples embedded in the samples. The structural data are optical micrographs of all built surfaces. Data are collected for three testing regimes: First, the laser rasters over the samples under conditions that do not induce melting. Second, the laser rasters over the samples with conditions that induce melting. Lastly, five layers of IN 718 are built atop the samples. The main result is an open and comprehensive data set, comprising both raw and processed signal data, for validating PBF process and structure models.

Phase transformations in materials systems can be tracked using atomic force microscopy (AFM), enabling the examination of surface properties and macroscale morphologies. In situ measurements investigating phase transformations generate large datasets of time-lapse image sequences. The interpretation of the resulting image sequences, guided by domain-knowledge, requires manual image processing using handcrafted masks. This approach is time-consuming and restricts the number of images that can be processed. In this study, we developed an automated image processing pipeline which integrates image detection and segmentation methods. We examine five time-series AFM videos of various fluoroelastomer phase transformations. The number of image sequences per video ranges from a hundred to a thousand image sequences. The resulting image processing pipeline aims to automatically classify and analyze images to enable batch processing. Using this pipeline, the growth of each individual fluoroelastomer crystallite can be tracked through time. We incorporated statistical analysis into the pipeline to investigate trends in phase transformations between different fluoroelastomer batches. Understanding these phase transformations is crucial, as it can provide valuable insights into manufacturing processes, improve product quality, and possibly lead to the development of more advanced fluoroelastomer formulations.

Nickel-aluminum bronze (NAB) alloys are commonly used for marine applications such as propellers, piping, valves, bearings, and fasteners. These NAB components are conventionally manufactured using both casting techniques and rolling and heat treatment techniques. However, limited information is available regarding the high temperature properties of NAB. The following data descriptor article documents the thermo-physical and thermo-mechanical results for a C63200 wrought plate material. These results will help empower Integrated Computational Materials Engineering efforts through the integration with commercial software packages. The raw data, in machine-readable form, are available at the University of Michigan’s Materials Commons data repository: https://materialscommons.org/.

Non-metallic inclusions (NMIs) in continuous casting slabs will significantly reduce the performance of final steel products and lead to other defects in steel products. The traditional detection methods of NMIs in continuous casting slabs have the problem of low efficiency, and it is complicated to establish a prediction model of NMIs based on physics and chemistry. Therefore, we tried to use the machine learning method by integrating Transformer and Random Forest and established an RF-1DViT model to predict NMIs in continuous casting slabs. To predict the occurrence and the location of NMIs as accurately as possible, the whole process data of steelmaking, refining and continuous casting were used, and the continuous casting slab was processed in slices. The experimental results show that the proposed RF-1DViT model has an F1 score of 0.8991, surpassing Logical Regression, K-Nearest Neighbor, Support Vector Machine, Random Forest, AdaBoost, GradientBoost, Multi-Layer Perceptron and 1DViT model, and has good interpretability and strong feature extraction ability. By means of the Random Forest and histogram, the process importance can be analyzed and rules of inclusions generation can be given. The t-SNE manifold learning method can further assist researchers to accurately locate the defect.

To optimize and extend the service life of polymeric materials in outdoor environments, a domain knowledge-based and data-driven approach was utilized to quantitatively investigate the temporal evolution of degradation modes, mechanisms, and rates under various stepwise accelerated exposure conditions. Six formulations of poly(methyl methacrylate) (PMMA) with different combinations of stabilizing additives, including one unstabilized formulation, were exposed in three accelerated weathering conditions. Degradation was dependent on wavelength as samples in UV light at 340 nm (UVA) exposure showed the most yellowing. The unstabilized PMMA formulation showed much higher yellowness index values (59.5) than stabilized PMMA formulations (2–12). Urbach edge analysis shows a shift toward longer wavelength from 285 to 500 nm with increasing exposure time and an increased absorbance around 400 nm of visible region as the unstabilized samples increase in yellowing. The degradation mechanisms of PMMA were tracked using induced absorbance to dose at specific wavelengths that correspond to known degradation mechanisms. The degradation pathway of PMMA was modeled in a <Stressor | Mechanism | Response> framework using network structural equation modeling (netSEM). netSEM showed changes in degradation pathway as PMMA transition stages of degradation.