DEFECT IDENTIFICATION USING MACHINE LEARNING IN AN ADDITIVE MANUFACTURING SYSTEM
An additive manufacturing system comprises an apparatus arranged to distribute layer of metallic powder across a build plane and a power source arranged to emit a beam of energy at the build plane and fuse the metallic powder into a portion of a part. The system includes a processor configured to steer the beam of energy across the build plane and receive data generated by one or more sensors that detect electromagnetic energy emitted from the build plane when the beam of energy fuses the metallic powder. The received data is converted into one or more parameters that indicate one or more conditions at the build plane while the beam of energy fuses the metallic powder. The one or more parameters are used as input into a machine learning algorithm to detect one or more defects in the fused metallic powder.
1 . An additive manufacturing system comprising:
an energy source arranged to fuse metallic powder;
a sensor arranged to detect electromagnetic energy emitted during the fusing of the metallic powder; and
a processor that receives data from the sensor and employs a trained machine learning algorithm to detect a defect in the fused metallic powder.
2 . The additive manufacturing system of claim 1 wherein the trained machine learning algorithm determines a type of the defect.
3 . The additive manufacturing system of claim 2 wherein the type of defect includes at least one of a lack of fusion of the metallic powder defect, a porosity defect or an inclusion defect.
4 . The additive manufacturing system of claim 1 wherein the processor is configured to calculate one or more parameters based at least in part on the electromagnetic energy.
5 . The additive manufacturing system of claim 4 wherein the one or more parameters includes determining a thermal emission density (TED) that includes measuring an amount of energy radiated from the fused metallic powder during one or more scans of the energy source and determining area of a build plane traversed during the one or more scans.
6 . The additive manufacturing system of claim 4 wherein the one or more parameters includes identifying spectral peaks associated with material properties of a batch of the metallic powder and selecting a first wavelength and a second wavelength spaced apart from the first wavelength, and determining an amount of energy radiated from a build plane based upon a ratio of energy radiated at the first wavelength to energy radiated at the second wavelength.
7 . The additive manufacturing system of claim 1 wherein the trained machine learning algorithm includes one or more training parameters based on a known-defective part.
8 . The additive manufacturing system of claim 7 wherein the one or more training parameters are derived from a known-defective part having at least one region having lack of fusion of the metallic powder.
9 . The additive manufacturing system of claim 7 wherein the one or more training parameters are derived from a known-defective part having at least one inclusion.
10 . The additive manufacturing system of claim 7 wherein the one or more training parameters are derived from a known-defective part having at least one region having porosity.
11 . The additive manufacturing system of claim 1 wherein the one or more sensors includes an on-axis photodetector.