Fatigue Life Prediction for Additive Manufactured Components
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for predicting the fatigue life of additive manufactured components are hindered by the complexity of multiple influencing parameters and the inability to account for localized material properties and variations, particularly surface roughness and porosities, leading to inconsistent and inaccurate predictions.
Innovation Solution
A machine learning approach using a Gaussian Progress Regression with Squared Exponential covariance function is employed to model fatigue life performance, allowing for localized parameter consideration and zone-specific calculations, enabling accurate prediction without prior assumptions on parameter interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional empirical rules and global property approaches are used for fatigue prediction, then the methodology is simple and easy to implement, but the prediction accuracy is poor and cannot account for localized material properties
Solution Approach 1:
The patent divides the component into multiple zones based on local material properties such as surface roughness, porosity, and print orientation. Each zone is assigned specific material properties that reflect the local conditions, allowing the fatigue life prediction to account for spatial variations in the additive manufactured component.
Solution Approach 2:
The patent implements zone-specific material properties where each region of the component has tailored mechanical properties based on local printing parameters and defects. This allows the model to capture localized phenomena such as varying surface roughness and porosity distributions, improving prediction accuracy without requiring a completely complex global model.
2Reliability
If multiple parameters including surface roughness and porosities are considered, then the prediction comprehensiveness improves, but the difficulty of developing and calibrating the mathematical model increases
Solution Approach 1:
The patent incorporates multiple printing parameters (laser power, scanning speed, hatch spacing), surface roughness, porosity, and print orientation as input parameters that influence material properties. By systematically varying these parameters across different zones, the model comprehensively captures the effects of multiple factors on fatigue life while maintaining a structured approach to model calibration.
Solution Approach 2:
The patent transitions from global property-based prediction to a spatially-resolved approach by introducing zone-specific material properties. This dimensional transformation allows the model to consider multiple parameters simultaneously across different locations in the component, achieving comprehensive prediction without overwhelming model complexity through systematic zonation.
3Adaptability or versatility
If global material properties are used for the entire component, then the model is simple to calibrate, but it cannot predict fatigue lives for complex components where parameters vary over the part
Solution Approach 1:
The patent segments the component into zones with representative material properties, allowing the model to adapt to complex geometries and varying local conditions. This segmentation enables the prediction methodology to handle complex components efficiently by focusing calculations on discrete zones rather than requiring exhaustive analysis of every point in the component.
Solution Approach 2:
By assigning zone-specific material properties that reflect local printing conditions and defects, the patent enables the model to adapt to complex components with varying characteristics. This local quality approach provides the versatility needed for complex geometries while maintaining computational efficiency through a finite number of representative zones rather than continuous spatial variation.
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a method and a system for Fatigue life prediction of additive manufactured components accounting for localized material properties. The method and the system is employed for prediction of Fatigue life properties of an Additive manufactured element, with a data collection step (1a, 1b) in which several data points for maximum stress vs. cycles to failure for different given processing steps of the element are collected, with a training step (2) in which a Machine Learning system is trained with the collected data, and with an evaluation step (5, 6) in which the trained Machine Learning system is confronted with actual processing steps and used to predict the Fatigue life properties of the element.