Magnetohydrodynamic Printer Drop Inspection via Temporal Averaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Magnetohydrodynamic (MHD) printers face challenges in maintaining precise drop placement and consistent jetting quality due to factors like dross build-up and contamination, which can lead to unstable or systematic degradation in jetting quality, compromising part quality.
Innovation Solution
An inspection system that captures images of drops after ejection, creates temporally averaged images, and uses a pretrained convolutional neural network to classify drop behavior, indicating unstable jetting and allowing for real-time interventions to maintain part quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a high-speed camera system running at 10,000 frames per second is used to record jetted drops, then the accuracy of drop observation is improved, but the cost and processing power requirements become prohibitively expensive
Solution Approach 1:
The system performs preliminary actions by capturing multiple images of drops at different positions along their trajectory before analysis. These pre-captured images are then processed to create a composite view that shows the complete drop path, eliminating the need for extremely high-speed continuous recording while maintaining accurate observation of drop behavior throughout flight.
Solution Approach 2:
The system creates a composite image that copies and combines information from multiple individual drop images taken at different times and positions. This composite representation reproduces the complete drop trajectory and behavior without requiring a single ultra-high-speed capture, thereby reducing processing requirements while maintaining measurement accuracy.
2Ease of operation
If the jetting quality is monitored periodically using a hand-held strobe light, then the operational simplicity is maintained, but the real-time detection capability and objectivity are compromised
Solution Approach 1:
The system enables self-service monitoring where the imaging system automatically captures and processes drop images without requiring manual intervention. The system self-evaluates jetting quality by analyzing captured images against predefined criteria, eliminating the need for operators to manually interpret strobe light observations while maintaining continuous monitoring capability.
Solution Approach 2:
The system implements automated feedback by continuously capturing images of ejected drops, analyzing their characteristics, and providing real-time assessment of jetting quality. This feedback loop enables objective, data-driven monitoring that replaces subjective manual observation while maintaining ease of operation through automated decision-making algorithms.
3Measurement precision
If multiple images of drops are captured and processed to create temporally averaged images, then the jetting quality evaluation accuracy is improved, but the processing time and computational requirements increase
Solution Approach 1:
The system extracts only the essential features and regions of interest from captured drop images, isolating the critical diagnostic information needed for jetting quality assessment. By extracting only relevant data elements rather than processing complete images, the system maintains evaluation accuracy while significantly reducing computational burden and processing time.
Data Source
AI summary
A printing system and method of inspecting drop ejection in a printing system is disclosed. The method includes capturing an image of each of a plurality of drops of a print material after ejection from an ejector of a printing system, creating a temporally averaged image from each image of the plurality of drops of print material, and classifying one of the plurality of drops of print material based on the temporally averaged image that was created. The use of a pretrained convolutional neural network for classifying one of the plurality of drops and comparing the temporally averaged image to another temporally averaged image to classify one of the plurality of drops may be employed. The printing system also includes a camera with a high-speed shutter where the shutter is synchronized to an ejector pulse, and a video analytic framework coupled to the ejector and the camera configured to generate a jetting result for each of the one or more drops of liquid print material.