Particle Simulation Rigid Body Force Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current molecular dynamics simulations, including the Fast Inertial Relaxation Engine (FIRE), face challenges in converging solutions efficiently when simulating elastic analysis objects as aggregates of particles, particularly due to issues with oscillations and setting attenuation constants, leading to delays in convergence and potential calculation failures.
Innovation Solution
A simulation method and apparatus that represent an elastic analysis object as an aggregate of virtual particles, apply FIRE to alleviate oscillations by calculating forces and velocities as a rigid body, and distribute additional forces to particles, allowing for faster convergence by solving the equation of motion with adjusted attenuation constants for each particle.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If FIRE is applied to each particle individually in molecular dynamics simulation, then oscillation convergence is attempted, but the calculation fails or convergence is delayed due to inappropriate attenuation constants
Solution Approach 1:
The patent merges the treatment of particles by applying FIRE to the entire analysis object (rigid body) rather than to each particle individually. The force calculation is performed at the rigid body level, and then distributed to all particles, ensuring consistent attenuation constant application and avoiding convergence failures
Solution Approach 2:
The patent changes the parameter application level from individual particle parameters to rigid body-level parameters. By calculating forces and velocities at the rigid body level and then distributing them to particles, the attenuation constants can be appropriately set for the entire object rather than struggling with individual particle parameters
2Measurement precision
If molecular dynamics simulation is performed on elastic analysis objects, then detailed particle-level analysis is achieved, but oscillations persist and equilibrium convergence is delayed
Solution Approach 1:
The patent segments the simulation approach into two levels: rigid body level for force calculation and particle level for detailed analysis. This segmentation allows efficient force computation at the rigid body level while maintaining particle-level detail for accurate stress and strain analysis
Solution Approach 2:
The patent introduces the rigid body as an intermediary between external forces and individual particles. The rigid body receives external forces, calculates resulting forces and velocities, and then distributes these to particles, acting as a mediator that accelerates convergence while preserving particle-level analysis capability
Data Source
AI summary
Provided is a simulation method that represents an elastic analysis object as an aggregate of a plurality of virtual particles, and the method includes, when solving the equation of motion for each particle, regarding the analysis object as a rigid body, and calculating a force acting on the analysis object and a velocity of the analysis object, based on the force acting on each particle obtained at each time step and the velocity and the position of each particle, applying FIRE to a motion of the analysis object to calculate a force to be applied to the analysis object, obtaining a force to be additionally applied to each particle, by distributing the force to be applied to the analysis object to the plurality of particles, and solving the equation of motion, by adding the force to be additionally applied, to the force acting on each particle.