A theoretical analysis of mass scaling techniques

Area of Science:

• Computational Mechanics
• Finite Element Analysis
• Structural Dynamics

Background:

• Mass scaling is a common technique in finite element models for structural dynamics.
• It aims to increase the critical time step in explicit time integration methods.
• The field currently lacks a robust theoretical basis, relying heavily on numerical experiments.

Purpose of the Study:

• To provide a rigorous theoretical foundation for mass scaling techniques.
• To connect existing mass scaling methods to established linear algebra results.
• To derive eigenvalue bounds and condition number estimates for improved assessment.

Main Methods:

• Comprehensive review of existing mass scaling methods.
• Application of established linear algebra results.
• Derivation of rigorous eigenvalue bounds and condition number estimates.

Main Results:

• Established rigorous eigenvalue bounds for successful mass scaling techniques.
• Derived condition number estimates providing theoretical insights.
• Unraveled the underlying mathematical reasons for well-known numerical observations in mass scaling.

Conclusions:

• This work establishes a strong theoretical basis for mass scaling in explicit finite element methods.
• The derived bounds and estimates offer a rigorous framework for assessing and understanding mass scaling.
• Provides a foundation for developing more theoretically sound and effective mass scaling strategies.