Digital Twin Thermal Monitoring for AM Flaw Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Additive manufacturing processes, particularly LPBF, are prone to flaw formation due to thermal history variations, process drifts, and cyber security threats, which hinder their adoption in safety-critical industries like aerospace and biomedical.
Innovation Solution
A digital twin approach combining graph theory thermal simulation with in-line thermal measurements to detect flaws by integrating real-time sensor data with physics-based models for timely and accurate flaw detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If thermal simulations are performed using traditional methods to detect flaws in additive manufacturing, then measurement precision can be achieved, but processing time becomes excessively long
Solution Approach 1:
The patent creates a digital twin (virtual copy) of the physical additive manufacturing process that runs in parallel. This digital twin uses graph theory thermal simulations to replicate the thermal history and detect flaws in real-time, eliminating the need for slow traditional post-processing thermal analyses while maintaining detection accuracy.
Solution Approach 2:
The system performs thermal simulations and flaw detection during the additive manufacturing process itself, rather than after completion. By integrating graph theory simulations that run concurrently with printing, the system identifies flaws in real-time, transforming a post-process analysis task into an in-process monitoring function.
2Loss of information
If traditional thermal simulation methods are used to model the entire part, then comprehensive thermal history can be obtained, but computational complexity increases significantly
Solution Approach 1:
The patent segments the thermal simulation into two distinct components: graph theory simulations that model thermal diffusion throughout the entire part, and finite element simulations that provide detailed thermal history for specific regions of interest. This segmentation allows comprehensive thermal modeling while reducing overall computational complexity by distributing the modeling tasks appropriately.
Solution Approach 2:
The patent introduces an intermediary approach where graph theory simulations serve as a computationally efficient framework that captures overall thermal behavior, while finite element simulations are used selectively to supplement detailed thermal history where needed. This intermediary strategy balances computational efficiency with thermal history completeness.
3Productivity
If real-time sensor data is collected during additive manufacturing to detect flaws, then detection speed improves, but false alarms increase
Solution Approach 1:
The patent implements a feedback mechanism where the digital twin continuously compares predicted thermal history from graph theory simulations with actual sensor measurements from the physical process. When deviations occur, the system adjusts and refines the simulation, creating a closed-loop system that reduces false alarms by validating real-time detections against physics-based predictions.
Solution Approach 2:
The patent merges two independent detection approaches: real-time sensor data analysis and graph theory thermal simulations. By combining these methods, the system cross-validates flaw detections, reducing false alarms that would occur with sensor-only approaches while maintaining the speed benefits of real-time monitoring.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables real-time detection of flaws with a five to ten times reduction in processing time, allowing for immediate recognition of complex fault formations and reducing false alarms, thus ensuring flaw-free production.
Implementation Method 1
The simulation can utilize a configurable node-based architecture to approximate thermal diffusion and steady state within the part
Implementation Method 2
Fault detection can occur when significant deviations between the predicted history and actual history of thermal measurements are detected by thermal sensors
Data Source
AI summary
Described herein are systems and methods for detecting flaws during an additive manufacturing (AM) process. A method can include accessing, by a computer, simulation results of a computer-modelled part representing a physical part to be formed using the AM process. The simulation includes a thermal history model for the computer-modelled part. During run-time formation of the physical part, the method includes receiving, from sensor devices, real-time sensor data of temperature values for nodes within regions of the physical part as each region is formed. The method also includes determining, for each region as the region is formed in the physical part, a deviation between the real-time sensor data of temperature values for nodes within the region and temperature values of the thermal history model for the computer-modelled part, and identifying flaws in the physical part based on determining that the deviation satisfies criteria indicating a flaw.