Carnegie Mellon University

Wright State University

Jack Lee Beuth, Jr

Jason Cho Fox

Nathan W Klingbeil

Joy Davis Gockel

Sneha Prabha Narra

A method for training a machine learning engine for modeling of a physical system includes receiving process data representing measurements of a physical system. The method includes applying a transform to values of the at least two variables of the process data to generate a dimensionless parameter having a parameter value corresponding to each measurement of the physical system for the at least two variables. The method includes training the machine learning engine using a set of generated training data including the non-dimensionalized parameter, to output a prediction of a value of a physical effect of the physical system for values of the variables that are not included in the process data. The method includes controlling an additive manufacturing process for the material by setting the at least one physical property to the value of the at least one process variable during fabrication of a part.

Non-dimensionalization of variables to enhance machine learning in additive manufacturing processes

In one aspect, a method includes conducting a plurality of tests on process variables of a thermal process, with a test of the plurality of tests being conducted on two or more process variables, the test comprising: locally heating a region of a structure, wherein the local heating results in formation of a thermal field in the structure; assessing one or more temperature integrals of the thermal field; and based on results of the plurality of tests, generating a process map of the one or more temperature integrals of the thermal field, with the one or more temperature integrals based on a function of the two or more process variables.

Process mapping of average temperatures and process sensitivity

A method includes conducting a plurality of tests on process variables of a manufacturing process, with a test of the plurality of tests being associated with two combinations of process variables, the test having first values for a first combination of process variables at a first time and second values for a second combination of process variables at a second time, the test comprising: locally heating a region of a structure, wherein the local heating results in formation of a thermal field in the structure; assessing one or more thermal characteristics of the thermal field during a transition between the first combination of process variables and the second combination of process variables; and based on results of the plurality of tests, generating a process map of a transient response of the one or more thermal characteristics of the thermal field.

Process mapping of transient thermal response due to value changes in a process variable

A method performed by one or more processing devices includes conducting a plurality of tests of a manufacturing process. Each test is conducted at a different combination of at least a first process variable and a second process variable, and each test comprises locally heating a region of a structure, where the local heating results in formation of a thermal field in the structure, and assessing a temperature derivative of the thermal field. Based on results of the plurality of tests, a process map of the temperature derivative of the thermal field is generated, with the temperature derivative based on a function of the first process variable and the second process variable.

Process mapping of cooling rates and thermal gradients

Conducting a plurality of tests of a manufacturing process, with each test conducted at a different combination of a first process variable and a second process variable. Each test includes locally heating a region of a part, wherein the local heating results in the formation of a thermal field in the part, and assessing a dimension of the thermal field. In some cases, based on the results of the plurality of tests, a process map of the dimension of the thermal field is generated as a function of the first process variable and the second process variable.

Growing community of inventors

Pittsburgh, PA, United States of America

Jack Lee Beuth, Jr