Your engineering workflow¶
A Sindri project describes the mission, spacecraft and study using three editable
inputs: mission.yaml, spacecraft.yaml and study.yaml. Keep their intended
relationships intact when editing a project.
flowchart TD
accTitle: From authored project to engineering decision
accDescr: Mission, spacecraft and study inputs are validated and resolved, then executed into an indexed run bundle. Review the bundle before revising inputs and comparing designs.
INPUT["Mission, spacecraft and study"] --> VALIDATE["Validate inputs and resolve resources"]
VALIDATE --> RUN["Run the sizing study"]
RUN --> BUNDLE["Indexed results, issues and artifacts"]
BUNDLE --> REVIEW["Review budgets, convergence and constraints"]
REVIEW --> DECIDE["Compare designs or revise assumptions"]
DECIDE -. "Revise the project" .-> INPUT
The bundle is the record of what was calculated. Review its failures and missing results before using it to support the next design decision.
In the Console¶
Follow your first Console run once, then use this sequence:
- Load or open a project in Configure and make a deliberate input change.
- Save the project and Validate its documents and resource references.
- Run it and follow stage progress.
- Read the result's Issues, budgets and engineering checks before its score.
- Compare recorded designs, then revise the project for the next calculation. Previous results remain unchanged.
Validate from the command line¶
With a runtime available and a project in my-project, start with:
sindri validate-project my-project/study.yaml
sindri run my-project/study.yaml --output-root my-runs
my-project and my-runs are paths you choose. Inspect the command output for the
actual run directory and diagnostics. These commands describe the supported
interface, not a complete example configuration.
Review the evidence¶
Open the run's manifest.json and follow its result and artifact indexes.
Do not infer result paths from filenames belonging to a different run or product
version. Record the input assumptions, units and applicable catalog/model
choices before comparing designs.
Check the execution outcome, engineering issues, convergence and feasibility separately. If a result is missing or unavailable, investigate the recorded reason before using the remaining budgets. A comparison is meaningful only when the inputs being held constant really are unchanged.
See capabilities and availability for the current runtime and qualification boundary.
Continue with project authoring, run control and bundle inspection. Use the method guide to interpret a domain result and evidence limits before claiming validation or qualification.