For the FeelMath solver, three different iterative methods are available: Fast (Conjugate Gradient), Intermediate (Memory Efficient Conjugate Gradient) and Memory Efficient (Neumann Series).

Fast: here the CG method is used (Conjugate gradient method - Wikipedia), which is the fastest available approach in ElastoDict, especially for strongly varying material parameters and nonlinear material laws as, e.g., plasticity or damage. The convergence rate is proportional to the square root of the stiffness contrast. However, it needs up to 4 times more memory compared to the Memory Efficient method. We recommend using the Fast method whenever possible.

Intermediate: this is a memory efficient CG method, which needs approximately 40 % less memory than the Fast method but may be 20-30 % slower.

Memory Efficient: here the Neumann series is used. It is the most memory efficient method, but the convergence rate is proportional to the stiffness contrast in the materials and can therefore be low. This method should only be used for applications with a low stiffness contrast, or if the RAM is not sufficient.

The Estimate memory button, can be used to decide which method is applicable for the current structure on the available computer.