| CPC B29C 64/118 (2017.08) [B29C 64/209 (2017.08); B29C 64/245 (2017.08); B29C 64/314 (2017.08); B29C 64/321 (2017.08); B29C 64/393 (2017.08); B33Y 10/00 (2014.12); B33Y 30/00 (2014.12); B33Y 40/10 (2020.01); B33Y 50/02 (2014.12); B33Y 70/00 (2014.12); B29K 2067/046 (2013.01)] | 9 Claims |
|
1. An optimization system for real-time control of adhesion dynamics during filament deposition, droplet formation, or both of an additive manufacturing process, the optimization system comprising:
a nozzle in communication with an additive manufacturing device including an amount of filament, the nozzle configured to receive the amount of filament from the additive manufacturing device for extrusion therefrom;
a print bed disposed beneath the nozzle, the print bed including a section of a thermally transparent thin film, such that the nozzle is aligned perpendicular to the section of the thermally transparent thin film of the print bed;
an optical capture device disposed above at least a portion of the print bed, below at least a portion of the nozzle, or both, the optical capture device being oriented orthogonally to the thermally transparent thin film;
a thermal capture device disposed beneath the print bed, the thermal capture device oriented within a sagittal plane of the print bed, such that the thermal capture device is aligned perpendicular to the section of the thermally transparent thin film of the print bed;
a processor communicatively coupled to the optical capture device and the thermal capture device;
wherein subsequent to receiving an electrical signal from the optical capture device, the processor is configured to activate a convolutional neural network comprising a plurality of characterization scheme datasets, the convolutional neural network being configured to identify, in real-time, a predetermined pass/fail criteria of thermoplastic adhesion along an interface between the print bed and the nozzle;
wherein the processor is configured to transmit an electrical signal to the optical capture device, whereby the optical capture device is configured to capture image-based feedback of the extrusion of the amount of filament from the nozzle to the print bed;
wherein, subsequent to the optical capture device capturing image-based feedback, the processor is configured to transmit an electrical signal to the thermal capture device, whereby the thermal capture device is configured to capture thermodynamic feedback of the extrusion of the amount of filament from the nozzle to the print bed; and
wherein based on a comparison of the captured image-based feedback and the captured thermodynamic feedback and the identified predetermined pass/fail criteria, the processor, via the convolution neural network, is configured to determine, in real-time, adhesion dynamics during deposition of the filament, whereby subsequent to determining suboptimal adhesion dynamics based on the real-time comparison to the pass/fail criteria, the processor is configured to thermocouple calibrate the thermal capture device, in real-time, thereby offsetting emissivity fluctuations of the thermal capture device from the thin film, the nozzle, or both, and
wherein the processor is configured to calculate a characteristic length of adhesion and adjust the distance between the nozzle and the print bed based on the calculated characteristic length of adhesion.
|