Skip to content

Estimate each injection's duration ahead of time, and add scheduling … - #876

Draft
mck1117 wants to merge 1 commit into
inj-sched-2-trackerfrom
inj-sched-3-estimate
Draft

mck1117 wants to merge 1 commit into
inj-sched-2-trackerfrom
inj-sched-3-estimate

Conversation

@mck1117

@mck1117 mck1117 commented Nov 19, 2026

Copy link
Copy Markdown
Collaborator

…tests

WallFuel::preview() returns what adjust() would, without changing the film. At the end of the fast callback, once the fuel mass, injector model and wall wetting are all up to date, each injection event now works out how long its next pulse would be: that cylinder's mass, the event's own latched injection mode, the wall-wetting preview, staging. The estimate matches what is then injected unless something changes in between. Nothing uses it yet; it is for scheduling injections by when they need to start.

New harness tests for injection scheduling, as acceptance tests for the upcoming change in how injections are scheduled:

  • the estimate matches the injected pulse through wall wetting
  • the start of injection moving back through the cylinder's cycle start
  • the phase table crossing 0/720 while running
  • rev limiter fuel cut and resume
  • batch fuel through the cam correcting the engine phase, in every injection timing mode
  • duty cycle at and beyond 100% and back
  • a large phase table step later in start-of-injection timing mode: disabled, as today's scheduling injects twice in that cycle

…tests

WallFuel::preview() returns what adjust() would, without changing the
film. At the end of the fast callback, once the fuel mass, injector model
and wall wetting are all up to date, each injection event now works out
how long its next pulse would be: that cylinder's mass, the event's own
latched injection mode, the wall-wetting preview, staging. The estimate
matches what is then injected unless something changes in between. Nothing
uses it yet; it is for scheduling injections by when they need to start.

New harness tests for injection scheduling, as acceptance tests for the
upcoming change in how injections are scheduled:

- the estimate matches the injected pulse through wall wetting
- the start of injection moving back through the cylinder's cycle start
- the phase table crossing 0/720 while running
- rev limiter fuel cut and resume
- batch fuel through the cam correcting the engine phase, in every
  injection timing mode
- duty cycle at and beyond 100% and back
- a large phase table step later in start-of-injection timing mode:
  disabled, as today's scheduling injects twice in that cycle

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@mck1117
mck1117 added this pull request to stack #875 November 19, 2026 21:51
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant