Adaptive multi-process additive manufacturing systems and methods

Patent metadata

Field	Value
Publication number	US-11203160-B2
Application number	US-201916370871-A
Country	US
Kind code	B2
Filing date	Mar 29, 2019
Priority date	Mar 29, 2018
Publication date	Dec 21, 2021
Grant date	Dec 21, 2021

Abstract

Apparatuses and methods are provided that provide adaptive multi-process additive manufacturing systems for monitoring, measuring, and controlling additive manufacturing processes. A first laser (e.g., a fiber laser) is used for melting and consolidating the powder, and a second laser is utilized for dual purpose: (a) for metrology to measure the surface roughness, dimensional accuracy, material properties, etc., and (b) based on the evaluated measurements to take corrective actions (laser ablation, etc.) to attain the desired surface finish and dimensional accuracy. Various elements provide defect detection, defect identification, and defect response actions which remove defect related material or address under print or missing material in a build object.

First claim

The invention claimed is:

1. An additive manufacturing (AM) system including an in-situ measurement and corrective action system comprising:
an AM system that produces a plurality of AM build layers based on an input AM build file;
a plurality of lasers that are operable to selectively orient on and ablate material from one or more of the AM build layers and measure the plurality of AM build layers during and after they have been produced by the AM system;
an orienting system for the plurality of lasers;
a sensing system that includes a sensor or imager system that selectively orients towards and generates sensor outputs or sensor data captures of each AM build layer during or after printing, the sensor outputs or sensor data captures include electromagnetic and/or thermal images, the sensing system further can include one or more of a group comprising a scanning electron microscope (SEM), a transmission electron microscope (TEM), thermal profile, and distortion/dimensional profile sensing systems;
a measuring system that receives inputs from at least one of the plurality of lasers and measures the AM build layer during and after the AM system prints or produced the AM build layers;
a data storage or hard drive storing a plurality of in-situ monitoring and corrective action system machine instructions to operate the system comprising:
a defect exploration system that operates the sensing system to generate the sensor outputs or sensor data capture of said AM build layers;
a defect identification system that classifies or identifies one or more defects in the AM build layers that correlate elements of the sensor outputs or sensor data with stored defect data, the defect identification system accesses a defect library including a plurality of detectable build layer defect patterns including comparison of the sensor system and defect exploration system outputs with entries in the defect library files by a feature or pattern recognition system to identify a respective one of the one or more defects, wherein the defect library includes pixel patterns associated with one or more defects which are compared by the feature recognition system, wherein the defect identification system comprises a feature or pattern recognition system which receives outputs of the sensing system including images of the build layers from a build object produced from the AM system, the feature or pattern recognition system comprises a neural network and/or other machine learning system;
a corrective action section that includes a corrective action library that includes a plurality of build layer defect corrective action processing sequences that control the AM system or at least one of the plurality of lasers, wherein the corrective action section or the in-situ monitoring system further includes machine instructions that selects one or more of the build layer defect correction action processing sequences in the corrective action library based on correlation between a detectable build layer defect pattern or patterns and an associated build layer defect correction action processing sequences stored in the corrective action library, wherein each of the build layer defect correction action processing sequences include predetermined corrective response actions associated with each of the defect library entries including an overprint and under print corrective response action, wherein the plurality of build layer defect correction action processing sequences includes machine instruction sequences for controlling at least one of the plurality of lasers to ablate some or all of a detected said build layer defects and machine instruction sequences for controlling the AM printer to fill in or adjust a geometry or section of a subsequently applied build layer adjacent to respective detected defects;
a defect library and corrective action library input section that provides an operator of the in-situ measurement and corrective action system a user interface to accept operator inputs of new said plurality of build layer defect corrective action processing sequences and new said plurality of detectable build layer defect patterns; and
an in-situ monitoring system section that receives inputs from the feature recognition system and interacts or controls the defect exploration system, the defect identification system, and the corrective action section as well as the AM system; and

at least one controller system that includes at least one processor which executes the machine instructions for controlling various elements including the AM system, the plurality of lasers, the measuring system, the in-situ monitoring and corrective response action system, and the orienting system.

2. The AM system of claim 1 , wherein the plurality of lasers is used to execute ablative removal of the one or more defects from the plurality of lasers that is used to perform AM layer build operations.

3. The AM system of claim 1 , wherein a first laser is used for melting and consolidating additive manufacturing material for a respective one of the AM build layers and a second laser is utilized for a dual purpose comprising metrology to measure surface roughness, dimensional accuracy, material properties of each AM build layer and, based on evaluated measurements, to execute one or more of the corrective actions comprising laser ablation to produce a desired surface finish and dimensional accuracy of each said AM build layer.

4. A method of operating an additive manufacturing (AM) system including an in-situ measurement and corrective action system comprising:
providing an additive manufacturing (AM) system including an in-situ measurement and corrective action system comprising:
an AM system that produces a plurality of AM build layers based on an input AM build file;
a plurality of lasers that are operable to selectively orient on and ablate material from one or more of the AM build layers and measure the plurality of AM build layers during and after the have been produced by the AM system;

an orienting system for the plurality of lasers;
a sensing system that includes a sensor or imager system that selectively orients towards and generates sensor outputs or sensor data captures of each AM build layer during or after printing, the sensor outputs or sensor data captures include electromagnetic and/or thermal images, the sensing system further can include one or more of a group comprising a scanning electron microscope (SEM), a transmission electron microscope (TEM), thermal profile, and distortion/dimensional profile sensing systems;
a measuring system that receives inputs from at least one of the plurality of lasers and measures the AM build layer during and after the AM system prints or produced the AM build layers;
a data storage or hard drive storing a plurality of in-situ monitoring and corrective action system machine instructions to operate the system comprising:
a defect exploration system that operates the sensing system to generate the sensor outputs or sensor data capture of said AM build layers;
a defect identification system that classifies or identifies one or more defects in the AM build layers that correlate elements of the sensor outputs or sensor data with stored defect data, the defect identification system accesses a defect library including a plurality of detectable build layer defect patterns including comparison of the sensor system and defect exploration system outputs with entries in the defect library files by a feature or pattern recognition system to identify a respective one of the one or more defects, wherein the defect library includes pixel patterns associated with one or more defects which are compared by the feature recognition system, wherein the defect identification system comprises a feature or pattern recognition system which receives outputs of the sensing syst

Assignees

• Us Navy

Inventors

• Vora Hitesh D
• Sanyal Subrata

Classifications

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/09

  Supervised learning · CPC title

• B33Y40/20

  Post-treatment, e.g. curing, coating or polishing · CPC title

• B33Y10/00

  Processes of additive manufacturing · CPC title

• B29C64/268

  using laser beams; using electron beams [EB] · CPC title