Reusable validated simulation procedures with enforced standards.
Method packs are formalized simulation procedures that capture your organization's best practices as repeatable, enforceable standards. They go beyond templates by including validation criteria, parameter constraints, and approval workflows.
What method packs include
A method pack is a complete simulation procedure with six components:
Creating a method pack
Method packs are typically created from successful simulation setups:
Capture from a simulation
After completing and validating a simulation, capture its setup as a method pack. The system extracts the configuration, parameters, and validation criteria automatically.
Define parameter ranges
Specify which parameters users can vary and their allowed ranges. Lock down parameters that should not be changed (e.g., numerical scheme choices that are known to work).
Add acceptance criteria
Define what "success" looks like: convergence requirements, metric thresholds, quality checks. These are evaluated automatically after each simulation using this method pack.
Include benchmark cases
Add one or more reference cases with validated results. These serve as regression tests for the method pack itself.
Approval workflow
Method packs follow a governed approval process:
Draft: The method pack is created and can be tested internally
Review: Submitted for review by the designated methods owner
Approved: Published and available for use across the organization
Retired: Superseded by a newer version but preserved for historical reference
Approved method packs are controlled
Once approved, a method pack can only be modified by creating a new version that goes through the review process again. This ensures that validated procedures remain stable.
Enforcement
Organization guardrails can require specific method packs for specific problem types:
Mandatory: All simulations of a given type (e.g., "external aerodynamics") must use the approved method pack
Recommended: The AI suggests the method pack but allows deviation with justification
Available: The method pack is listed as an option but not enforced
When a mandatory method pack is in effect, the AI applies its guidance, parameter constraints, and validation criteria automatically. Users cannot bypass the constraints without admin override.
Default acceptance criteria
Each method pack defines default acceptance criteria that are applied unless the user specifies overrides:
Criterion type
Example
Convergence
Residuals below 1e-4 for all fields
Mesh quality
Maximum non-orthogonality below 65 degrees
Result range
Drag coefficient between 0.01 and 2.0
Conservation
Mass imbalance below 0.1%
These defaults reduce setup time and ensure consistency -- engineers don't have to remember the right thresholds for each simulation type.