Data Mining and Visualization of High-Dimensional ICME Data for Additive Manufacturing

Kannan, Rangasayee; Knapp, G.L.; Nandwana, Peeyush; Dehoff, Ryan R.; Plotkowski, Alex; Stump, Benjamin; Yang, Ying; Paquit, Vincent

doi:10.1007/s40192-021-00243-2

Cited by 12 publications

(7 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specific alloys were chosen based on their estimated resistance to hot cracking in the dilution region that is formed when the deposited alloy is mixed with the representative cast substrate. The hot cracking estimation was done using an empirical hot cracking parameter [7] determined by our recently developed high-throughput CALPHAD approach to predict the solidification pathways, detailed in our published work [8].…”

Section: Methodsmentioning

confidence: 99%

Local Modification of Cast Aluminum Alloys via the Cast-and-Print Process

Knapp,

Brown,

Paramanathan

et al. 2023

View full text Add to dashboard Cite

Casting of aluminum alloys is a cost-effective way to mass manufacture bulk aluminum parts for automotive applications. However, the casting process requires filling of a mold, which imposes some limitations on the geometries that can be successfully cast and limits parts to be a uniform material. Therefore, we proposed that a hybrid casting plus additive manufacturing approach, Cast-and-Print, could be used to locally modify the properties and geometry of cast parts to create difficult-to-cast, functional features. Here, we report the results of the properties of deposited materials on cast substrate, the properties of the interface of the deposited and cast material, and the application of the Cast-and-Print method to deposit rivet tabs on high-pressure die cast plates. Ultimately, the approach appears viable, but there are engineering challenges to reliable additive processing of aluminum wires that need to be overcome for successful implementation of the technology.

show abstract

Section: Methodsmentioning

confidence: 99%

Local Modification of Cast Aluminum Alloys via the Cast-and-Print Process

Knapp,

Brown,

Paramanathan

et al. 2023

View full text Add to dashboard Cite

show abstract

“…A variety of computer tools have been developed to query, visualize, and analyze high dimensional AM datasets [2,7,8]. Some platforms are designed for sharing scientific datasets, which wrap up data with a generic set of metadata [9,10].…”

Section: Prior Artmentioning

confidence: 99%

Enabling FAIR data in additive manufacturing to accelerate industrialization

Li¹

2023

View full text Add to dashboard Cite

Additive manufacturing (AM) is an important enabler of Industry 4.0 but there are several hurdles that need to be overcome to realize the full potential of AM. These challenges include the need for a data infrastructure that supports the sharing of data generated by the technologies as they mature, and data interoperability is critical to a sustainable data infrastructure that is widely accessible. This paper outlines the need for building a common data stack for an interoperable AM data infrastructure. At the foundation is a common data dictionary (CDD),which defines a primary technical vocabulary for the AM community in logical buckets. On top of the CDD is a common data model (CDM), which defines the hierarchy and relationships of the terms in the CDD and enables the data to be linkable in a complex data system. The CDM empowers information integration and sharing throughout the lifecycles and value chains for different AM technologies. The comprehensiveness of the properties defined in the CDM makes AM data reusable. To enable the exchange of data between different systems, common data exchange formats (CDEFs) transform the CDM into targeted data packages. The design philosophy of the CDD, CDM, and CDEF are described to accelerate the development and adoption of AM technologies across alliances.

show abstract

“…43 J/mm 3 71 J/mm 3 135 J/mm 3 phase evolution because Cr eq and Ni eq account for ferrite stabilizing and austenite stabilizing effects of the constituent elements. 14 Once the compositions were generated, the TC-Python module of ThermoCalc was used to generate phase diagrams and extract data such as equilibrium phases while Scheil simulations were conducted to determine the phase fractions during solidification. The relevant data was extracted from the TC-Python calculations, following which a Gaussian Copula machine learning model was used to learn from the input data (i.e., data from thermokinetic calculations on the 1,000 compositions).…”

Section: Calphad and Data Mining For Compositional Specificationsmentioning

confidence: 99%

“…They explored 5,000 compositions of 316SS-L, all within the bounds of ASTM specifications, and found that there can be significant variation in the -phase. 14 Later, Kannan and Nandwana extended this approach to generate 1 million synthetic compositions to enable new alloy discovery. 15 This report has combined the two approaches (i.e., data mining and synthetic data generation) to understand the effects of various elements on phase evolution for 316SS-L and 316SS-H by exploring 1 million compositions, all within the ASTM specifications, to help narrow down the compositional specifications of printed 316SS for nuclear applications.…”

Section: Introductionmentioning

confidence: 99%

Preliminary Report on Compositional Specification for Printed 316SS

Nandwana,

Kannan,

Nayir

et al. 2023

View full text Add to dashboard Cite

The Advanced Materials and Manufacturing Technologies Program is focused on accelerating the development and deployment of advanced materials and components fabricated via additive manufacturing (AM). A key challenge for the widespread adoption of AM is the variability in material properties, which can originate from powder feedstock, process parameters, geometry, and the machine. AM powder feedstock compositions are based on those developed for conventional manufacturing. However, the smaller melt pools combined with multiple melt cycles over the course of a build can have substantial effects on the local phase selection resulting from minor batch-to-batch feedstock variability. Therefore, this report focuses on coupling CALPHAD calculations with data mining and visualization to aid the development of compositional specifications for stainless steel 316, an important alloy for nuclear applications and suitable for AM. This work reports that δ-ferrite formation has a stronger dependence on the amount of nickel in the alloy despite nickel being an austenite stabilizer. These findings show the need for exercising a tighter control on the amount of nickel in the alloy compared with the amount of chromium, which is a ferrite stabilizer.

show abstract

Data Mining and Visualization of High-Dimensional ICME Data for Additive Manufacturing

Cited by 12 publications

References 38 publications

Local Modification of Cast Aluminum Alloys via the Cast-and-Print Process

Local Modification of Cast Aluminum Alloys via the Cast-and-Print Process

Enabling FAIR data in additive manufacturing to accelerate industrialization

Preliminary Report on Compositional Specification for Printed 316SS

Contact Info

Product

Resources

About