No products in the cart.
SOLIDWORKS Simulation – Choosing a Solver
July 19
Analysis Solvers
In general, all solvers give comparable results if the required solver options are supported. While all solvers are efficient for small problems (25,000 degrees of freedom or less), significant differences in performance (speed and memory usage) occur when solving large problems.
In finite element analysis, a problem is represented by a set of algebraic equations that must be solved simultaneously. There are two classes of solution methods:
- Direct Methods: Solve the equations using exact numerical techniques.
- Iterative Methods: Solve the equations using approximate techniques. In each iteration, a solution is assumed, and the associated errors are evaluated. The iterations continue until the errors become acceptable.
The Automatic Solver Selection is the default option and finds the best solver by considering the number of equations, load cases, mesh type, geometric features, contact and connector features, and available system memory. For instances where a specific solver is known to be more effective, use the Manual Solver Selection process.
FFEPlus (Iterative)
The FFEPlus solver works by assuming the solution and then evaluating the errors present, SOLIDWORKS uses the errors to estimate a new solution and repeats the processes until the errors are acceptable. The FFEPlus solver employs advanced matrix reordering techniques, which are particularly efficient for large problems. In general, FFEPlus is faster at solving large problems and becomes more efficient as the problem size increases, up to the maximum available memory.
However, it is important to be aware that FFEPlus has some limitations. It can fail when there is a significant difference in stiffness between two components or when temperatures and pressures are imported from another study. Additionally, FFEPlus may not always be the most efficient solver when dealing with contact interactions, virtual walls, connectors, or soft springs. In these scenarios, using one of the direct sparse solvers might be more effective.
Intel Direct Sparse / Direct Sparse
Intel Direct sparse solvers are known for their accuracy and stability advantages over FFEPlus in certain types of finite element analyses. They perform particularly well in small to medium-sized problems, especially when dealing with contact interactions, materials of vastly different stiffnesses, or connectors.
However, their efficiency in terms of RAM usage is a significant drawback, especially in larger studies. Intel Direct sparse solvers typically require much more RAM compared to FFEPlus. For the same number of degrees of freedom, a direct sparse solver might require up to ten times more RAM. This makes them less suitable for larger problems, where RAM limitations can become a critical issue.
As a guideline, it’s recommended to have at least 16GB of RAM to effectively use a direct sparse solver, with 32GB being the preferred option for regular use. Beyond a certain size, Intel Direct Sparse solvers can struggle or even fail due to their high RAM requirements. In such cases, simplifying the study or switching to alternative solvers like Large Problem Direct Sparse or FFEPlus may be necessary to successfully complete the analysis.
In summary, while direct sparse solvers offer superior accuracy and stability in specific scenarios, their inefficiency in RAM usage limits their applicability to smaller to medium-sized problems, unless adequate RAM resources are available.
**SOLIDWORKS Simulation 2023 is the last release to support the Direct Sparse solver.**
** For most cases, the Intel Direct Sparse solver is faster than the Direct Sparse.**
Large Problem Direct Sparse
The Large Problem Direct Sparse solver offers a significant advantage in handling simulation problems that exceed the physical memory (RAM) of a computer. This capability is achieved through enhanced memory-allocation algorithms, which allow the solver to efficiently manage and process data even when it cannot fit entirely within the available RAM.
When the Direct Sparse solver reaches memory resources limits, it prompts users to switch to the Large Problem Direct Sparse solver, which is specifically designed to handle out-of-core solutions.
Moreover, both the Large Problem Direct Sparse solver and the Intel Direct Sparse solver are optimized to take advantage of multiple processor cores. This parallel processing capability enhances their efficiency compared to solvers like FFEPlus.
Below is a summary of the solver information:
| Solver | Advantages | Limitations | Usage Recommendations |
| Automatic | Automatically selects the solver based on conditions | Might not always choose the optimal solver for specific cases | Default option for most study types |
| FFEPlus | Efficient for large problems, requires less RAM | May fail with large stiffness differences or imported data | Best for problems with >100,000 DOFs |
| Large Problem Direct Sparse | Handles problems exceeding physical memory, uses multiple cores | Requires significant memory, less efficient for small problems | Ideal for very large problems with enough RAM |
| Intel Direct Sparse | Efficient with multiple cores, Ideal for multi-area contact problems, handles complex models well | High memory usage | Best for static, thermal, and nonlinear studies |
Should you have any questions or need assistance, please do not hesitate to contact our support team.
Telephone : (65) 6226 3784
Email : support@seacadtech.com
WhatsApp : (65) 9372 1461
