Result file overwrite limits during retried Ansys batch solves

When designing retry logic for Ansys batch jobs, the central uncertainty is whether a repeated execution of the same input or journal will overwrite existing result files or produce duplicate outputs. Mechanical APDL batch runs are generally understood to overwrite prior result files deterministically when output names remain unchanged, whereas Fluent journal replays do not automatically resume an interrupted mid-case solve unless explicit checkpointing is configured. No general idempotency-key or retry-without-duplicate-writes mechanism is documented across the Ansys command-line interface or Workbench job submission.

Because result file overwrite behavior depends on output naming and solve settings, and because semantics differ across Mechanical, Fluent, CFX, and Electronics, the limits of safe retry remain product-specific and require verification against installed documentation before automated use.

• Does Ansys Mechanical APDL provide a documented guarantee that a retried batch solve will overwrite prior result files without manual cleanup, or is this behavior incidental to output naming?
• When an Ansys Fluent journal is replayed after a partial failure, do existing case and data files get overwritten or appended, and what configuration controls this?
• Is there a supported cross-product retry mechanism in Ansys Workbench or the CLI that prevents duplicate result writes without requiring product-specific checkpoint or restart settings?