Refining Finite Element Model of Integrally Bladed Disk
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current digital modeling techniques for integrally bladed disks (IBDs) in gas turbine engines fail to accurately predict blade mistuning frequencies, leading to inaccuracies in vibrational behavior predictions due to manufacturing tolerances and material property variations, which can result in high cycle fatigue damage.
Innovation Solution
A reduced order model (ROM) is used to determine the mistuning residual for each blade, allowing for residual mistuning corrections to be applied to the finite element model, improving its accuracy in predicting vibrational behavior and enabling more confident analysis of IBD modifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a finite element model is constructed to reflect measured dimensions of an IBD, then the model accuracy for predicting vibrational behavior is improved, but the model still fails to accurately predict blade mistuning frequencies due to residual dimensional measurement errors and material property variations
Solution Approach 1:
The patent uses experimental vibration data as feedback to iteratively refine the finite element model. The process involves comparing model predictions with actual measurements, identifying discrepancies in mistuning frequencies, and adjusting model parameters (such as blade stiffness, mass, or geometric dimensions) to reduce the gap between predicted and measured values. This closed-loop feedback mechanism continuously improves prediction accuracy.
Solution Approach 2:
The patent systematically varies key model parameters including blade dimensional tolerances, material property variations, and joint stiffness to match experimental observations. By adjusting these parameters within measured ranges and observing the impact on vibrational response, the model captures the actual mistuning behavior of the IBD more accurately.
2Reliability
If manufacturing tolerances and material variations are accounted for in the model, then the prediction of vibrational behavior becomes more realistic, but the complexity of the modeling process increases
Solution Approach 1:
The patent applies local quality by focusing computational resources on the most critical sources of mistuning. Rather than uniformly modeling all possible variations throughout the entire IBD, the approach identifies and prioritizes the blades and regions with the greatest dimensional and material property variations. This selective detailed modeling reduces overall model complexity while maintaining prediction accuracy for the dominant mistuning effects.
Solution Approach 2:
The patent manages complexity by systematically varying a limited set of key parameters (such as blade tip clearance, root geometry, and material density) that have the most significant impact on vibrational behavior. This parameter-based approach is more efficient than full geometric modeling of all tolerances, as it captures the essential mistuning mechanisms with fewer degrees of freedom.
3Manufacturing precision
If optical scanning is used to capture geometric data for model updating, then the dimensional accuracy of the model is improved, but small residual measurement errors remain that affect frequency prediction
Solution Approach 1:
The patent introduces an intermediate refinement step between optical scanning and final model usage. The process involves using the optical scan data to create an initial geometric model, then applying additional adjustments based on vibration test data. This intermediary calibration phase corrects for residual measurement errors and unmodeled effects, bridging the gap between geometric accuracy and dynamic behavior prediction.
Solution Approach 2:
The patent employs feedback from vibration measurements to correct residual errors in the optically scanned model. By comparing predicted and measured frequencies, the system identifies and compensates for small measurement inaccuracies and unmodeled physical effects, thereby improving frequency prediction accuracy beyond what optical scanning alone can achieve.
Data Source
AI summary
A reduced order model of an integrally bladed turbine disk (IBD) is used with experimental vibrational test data to modify a finite element model (FEM) of the IBD so that the FEM more accurately predicts the vibrational mistuning of the disk. The refined FEM can be used to evaluate a proposed modification of the IBD before the hardware is actually modified, and to evaluate the actual modification if there is a difference between the proposed and actual modifications.