SOLIDWORKS SIMULATION

(Choosing the Right Solver in SOLIDWORKS Simulation)

When working with SOLIDWORKS Simulation, selecting the appropriate solver plays a critical role in balancing accuracy, speed, and computational efficiency. SOLIDWORKS offers four solver options: Auto, FFEPlus, Intel Direct Sparse, and Large Problem Direct Sparse (LPDS).

For users who are just getting started, the safest and most effective choice is the Auto option. It automatically selects the best solver based on the study conditions, helping you avoid unnecessary complexity while ensuring reliable results.In this blog we will discuss about how each solver works and when you should (or shouldn’t) use it.

Auto Solver: TheDefault

The Auto solver acts as an intelligent decision-maker. It typically begins with the FFEPlus solver (iterative) and switches to the Intel Direct Sparse solver when certain conditions arise. The Auto switches solver based on the following factors such as

• Available system memory before solving
• Memory overflow during computation
• Activation of large displacement settings
• Use of remote loads or connectors
• High percentage of shell elements
• Contact complexity (especially contacts)
• Mixed mesh interactions
• Symmetry conditions
• Frequency shift options

Why use Auto?

• Ideal for beginners and general workflows
• Reduces manual decision-making
• Balances performance and accuracy automatically

FFEPlus Solver: Efficient for Large Models

FFEPlus is an iterative solver that uses implicit numerical techniques. It works by assuming a solution, calculating errors, and refining the result until acceptable accuracy (Convergence) is reached. The following factors can be considered of the usage of this solver

• Large models with more than 100,000 degrees of freedom (DOF)
• Linear static analyses
• For every 2,000,000 degrees of freedom, 1GB of RAM is required. In general, the FFEPlus solver requires less RAM than other solver.

Limitations:

FFEPlus may struggle or lose accuracy in cases involving:

• Incompatible meshes or complex bonded contacts
• Large differences in material stiffness
• Imported loads from other simulations
• Cyclic symmetry conditions
• Nonlinear analysis
• Frictional contacts
• Pin or rigid connectors

Intel Direct Sparse Solver: High in Accuracy & Performance

The Intel Direct Sparse Solver is a modern, high-performance version of the traditional direct sparse method, optimized using Intel’s Math Kernel Library (MKL). It computes exact solutions using advanced numerical techniques and is designed to take advantage of multi-core processors and optimized memory handling.The following factors can be considered of the usage of this solver.

• Highly accurate results for complex simulations
• Efficient multi-threading for faster computation
• Improved numerical stability for challenging models
• Better handling of matrices compared to iterative solvers
• Small to medium-sized models
• Nonlinear analyses
• Simulations requiring high precision
• Cases where FFEPlus may produce unreliable results

Limitations:

• Memory usage increases rapidly with model size
• Performance depends heavily on available RAM
• Not ideal for extremely large models with millions of DOF

Large Problem Direct Sparse (LPDS): When Memory Is Limited

The LPDS solver extends the Intel Direct Sparse approach by enabling out-of-core solving and better utilization of multiple CPU cores. It allows simulations to run even when system RAM is insufficient by using disk space as supplementary memory.The following factors can be considered of the usage of this solver.

• Intel Direct Sparse is required
• System RAM is not enough to complete the simulation
• Slower due to disk access
• Should be considered a fallback option rather than a primary choice

Choosing the Right Solver: Quick Reference

Scenario	Recommended Solver
Beginner or unsure	Auto
Large models (>100k DOF)	FFEPlus
High accuracy required	Intel Direct Sparse
Nonlinear analysis	Intel Direct Sparse
Limited RAM with direct solving needs	LPDS

Each solver in SOLIDWORKS Simulation has its strengths:

• FFEPlus excels in handling large-scale problems efficiently
• Intel Direct Sparse delivers superior accuracy and stability
• Auto simplifies solver selection for most users
• LPDS provides a safety net when memory becomes a constraint

Understanding these differences allows you to optimize your simulations, reduce solving time, and improve result reliability.

Thanks for reading!!

Please reach us for detailed info and training plan, if required.