About this Abstract |
| Meeting |
2025 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2025)
|
| Symposium
|
2025 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2025)
|
| Presentation Title |
Neural Network Prediction of Printing Parameters and Solar Absorptivity via Powder Bed Fusion |
| Author(s) |
Mathew Farias, Santosh Rauniyar, Ben Xu |
| On-Site Speaker (Planned) |
Mathew Farias |
| Abstract Scope |
Concentrated Solar Power (CSP) utilizes heliostats to focus solar radiation on a receiver where a heat transfer fluid is heated to either store or used to heat a power cycle. Currently, Pyromark 2500, is the industry standard coating used to maximize absorptivity; however, the coating degrades rapidly in 3rd generation receivers, which operate in temperatures above 700 °C. To circumvent this problem, it is proposed that the high surface roughness inherent in the powder bed fusion (PBF) process be increased to increase solar absorptivity. To achieve this, it is necessary to understand how laser process parameters directly affect the surface roughness. A machine learning neural network program is proposed that can be trained to both predict the solar absorptivity spectra as well as the printing process used to create a given surface. Both models proved to predict both laser parameters, power, scanning speed, and solar absorption spectra with accuracy. |
| Proceedings Inclusion? |
Planned: Post-meeting proceedings |