Droplet Shape Estimation for Additive Manufacturing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Additive manufacturing processes face significant limitations in predicting the as-printed shape due to manufacturing imperfections, such as variability in droplet deposition location, frequency, and temperature, leading to uncertainties in the final product's mechanical properties and surface quality.
Innovation Solution
A computationally efficient reduced-order model using hybrid machine learning, specifically training an Artificial Neural Network (ANN) to predict the shape of solidified droplets on substrates of arbitrary geometry, significantly reducing computational time from hours to milliseconds, allowing for rapid estimation of printed part shapes and visualization of printing processes without actual printing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional computational models are used to predict droplet shape, then accuracy is maintained, but computational time increases from milliseconds to hours
Solution Approach 1:
The patent creates a simplified copy of the complex droplet formation physics by training a neural network on representative droplet shape data. The neural network learns to replicate the behavior of traditional computational models without requiring the full computational complexity, enabling rapid prediction of droplet shapes while maintaining acceptable accuracy for manufacturing parameter optimization
Solution Approach 2:
The neural network is pre-trained offline using comprehensive droplet shape data covering various substrates and parameters. This preliminary training phase captures the essential physics relationships, so that during actual manufacturing planning, the pre-trained network can rapidly predict droplet shapes without requiring time-consuming simulations
2Manufacturing precision
If comprehensive manufacturing parameter optimization is performed, then part quality improves, but computational cost and time increase significantly
Solution Approach 1:
The patent replaces traditional mechanical computational models (finite element analysis, fluid dynamics simulations) with a data-driven neural network system. This substitution enables rapid evaluation of multiple manufacturing parameters and their effects on droplet shapes, allowing comprehensive optimization studies to be performed in minutes rather than hours or days
3Measurement precision
If detailed droplet shape prediction is performed for every parameter variation, then accuracy is maintained, but computational resources are excessively consumed
Solution Approach 1:
The neural network performs partial computation by learning to predict only the essential droplet shape characteristics needed for manufacturing optimization, rather than performing complete physical simulations. This partial action approach provides sufficient accuracy for parameter optimization while consuming minimal computational energy
Data Source
AI summary
A geometry of a substrate surface is received at a neural network. The neural network is trained using one or more training sets. Each training set comprises a different type of substrate geometry and a collection of manufacturing process parameters. The substrate is configured to receive at least one liquid droplet. A shape of the at least one droplet after it has been deposited on the substrate is determined based on the received geometry. An output representing the determined shape of the at least one droplet is produced.