Machine learning assisted development in additive manufacturing

1. A method for optimizing additive process parameters for an additive manufacturing process, comprising:

(a) receiving, at a computing device, initial additive process parameters;

(b) generating, using the computing device, an uninformed design of experiment utilizing a specified sampling protocol to vary the initial additive process parameters in a design space;

(c) generating, using the computing device based on execution of the uninformed design of experiment in a first process, first response data;

(d) generating, using the computing device and a machine learning model based on the first response data, a specified objective for the additive manufacturing process, the sampling protocol, and on previous design of experiment comprising at least one of the uninformed design of experiment and a prior informed design of experiment, an informed design of experiment;

(e) evaluating, using the computing device, the specified objective with respect to second response data produced in a second process execution of the informed design of experiment; and

(f) iteratively repeating steps (d)-(e) until the specified objective is satisfied.

2. The method of claim 1 , wherein the initial additive process parameters are associated with at least one of material microstructure, material chemistry and material properties.

3. The method of claim 2 , wherein the material properties are associated with microstructure of a material.

4. The method of claim 1 , wherein generating the first response data comprises obtaining the first response data from one of an additively manufactured part or a computer implemented model executing the first process.

5. The method of claim 1 , wherein the specified objective comprises multiple objectives and the informed design of experiment comprises optimization for the multiple objectives.

6. The method of claim 5 , wherein the multiple objectives of the informed design of experiment comprises optimizing for at least one specified objective and reducing model uncertainty.

7. The method of claim 1 , wherein the specified objective is associated with at least one of an attribute or a property of a material.

8. The method of claim 1 , wherein the machine learning model comprises a probabilistic machine learning process.

9. The method of claim 8 , wherein the probabilistic machine learning process is based on a Bayesian Hybrid Model (BHM) protocol.

10. The method of claim 1 , wherein the specified sampling protocol comprises intelligent design analysis of computer experiments (IDACE).

11. The method of claim 1 , wherein the specified sampling protocol comprises an expect improvement (EI) protocol.

12. A computer-implemented process for optimizing additive process parameters for an additive manufacturing process, comprising:

(a) receiving, at a computing device, initial additive process parameters;

(b) generating an uninformed design of experiment utilizing a specified sampling protocol to vary the initial additive process parameters in a design space;

(c) generating, based on execution of the uninformed design of experiment in a first process, first response data;

(d) generating, using a machine learning model based on the first response data, a specified objective for the additive manufacturing process, the sampling protocol, and on previous design of experiment comprising at least one of the uninformed design of experiment and a prior informed design of experiment, an informed design of experiment;

(e) evaluating, using the computing device, the specified objective with respect to second response data produced in a second process execution of the informed design of experiment; and

(f) iteratively repeating steps (d)-(e) until the specified objective is satisfied.

13. The computer-implemented process of claim 12 , wherein generating the first response data comprises obtaining the first response data from one of an additively manufactured part or a computer implemented model.

14. The computer-implemented process of claim 12 , wherein step (f) further comprises generating the informed design of experiment until multiple specified objectives are optimized.

15. The computer-implemented process of claim 14 , wherein one of the multiple specified objectives comprises reducing model uncertainty.

16. A system for optimizing additive machine additive process parameters of a material for use in an additive manufacturing process comprising:

a computer comprising a processor operably connected to a storage device and a communication device; and

a test device operably connected to the computer, the test device comprising a test platform for accommodating an additively manufactured part and a plurality of measurement devices;

wherein the storage device of the computer comprises instructions that cause the processor to:

(a) receive initial additive process parameters;

(b) generate an uninformed design of experiment utilizing a specified sampling protocol to vary the initial additive process parameters in a design space;

(c) generate, based on execution of the uninformed design of experiment in a first process, first response data;

(d) generate, using a machine learning model and the sampling protocol based on one of the first response data and first material data, and based on previous design of experiment comprising at least one of the uninformed design of experiment and a prior informed design of experiment, an informed design of experiment;

(e) receive, from the test device, second material data of a part additively manufactured based on the informed design of experiment;

(f) evaluate a specified objective with respect to the material data produced in a second process execution of the informed design of experiment; and

(g) repeat steps (d)-(f) until the specified objective is satisfied.

17. The system of claim 16 , wherein generating the first response data comprises obtaining the first response data from a computer implemented model.

18. The system of claim 16 , wherein repeating steps (c) and (d) further comprises generating the informed design of experiment until multiple specified objectives are optimized.

19. The system of claim 18 , wherein one of the multiple specified objectives comprises reducing model uncertainty.