A Consistent Procedure Using Response Surface Methodology to Identify Stiffness Properties of Connections in Machine...

Area of Science:

• Mechanical Engineering
• Computational Mechanics
• Structural Dynamics

Background:

• Accurate finite element (FE) models are crucial for structural analysis during the design phase.
• Uncertainties in material properties, boundary conditions, and connections cause discrepancies between predicted and actual dynamic characteristics.
• Model updating is necessary to reconcile FE predictions with real-world system behavior.

Purpose of the Study:

• To present a systematic three-step procedure for updating FE models of machine tools with numerous connection uncertainties.
• To integrate statistical, numerical, and experimental techniques for robust model updating.
• To achieve improved agreement between FE model predictions and experimental measurements.

Main Methods:

• Gradual application of fractional factorial designs to identify significant variables affecting dynamic responses.
• Analysis of variance (ANOVA) to determine variable significance.
• Development of quadratic response surface meta-models relating design parameters to modal responses.
• Identification of updated design variables using regression equations and experimental modal data.

Main Results:

• The integrated procedure effectively updates FE models despite a large number of uncertainties.
• The updated FE models demonstrate dynamic responses that closely agree with experimental measurements.
• The method achieves this accuracy at an acceptable computational cost.

Conclusions:

• The proposed three-step procedure successfully refines FE models of machine tools.
• The integration of statistical, numerical, and experimental methods provides a powerful approach for model updating.
• Accurate dynamic predictions are achievable even with significant system uncertainties, offering practical benefits for design and analysis.