Functions
	config ()
	drawing ()
Override	droop (float value, uncertainty=None)
	test_an_override_lands_in_the_drawing_with_its_provenance (drawing)
	test_the_restatement_table_reads_and_takes_absolute_pressures_in_psia (drawing)
	test_a_restated_parameter_reaches_the_twin_as_measured (config, drawing)
	test_the_store_wants_a_source_and_one_value_per_parameter (tmp_path)
	test_a_measured_uncertainty_replaces_the_default_range (config, drawing)
	sweep (config, drawing)
	_case (sweep, key, side)
	test_the_sweep_runs_every_factor (sweep)
	test_impulse_is_blind_to_the_feed_and_thrust_is_not (sweep)
	test_the_band_is_told_each_way (sweep)
	test_nozzle_efficiency_moves_thrust_and_nothing_upstream (sweep)
	test_worse_mixing_costs_impulse_and_more_cd_flows_faster (sweep)
	test_a_measured_engine_input_reaches_the_card (config)
	test_restatements_follow_the_drawing_through_an_edit (tmp_path, drawing)

Variables
	reason
str	FIXTURE = Path(__file__).parent / "fixtures" / "ethalox_6500N_doublet_cad_2026-09-28.yaml"
str	GN2 = "copv_study_gn2"
list	pytestmark

Detailed Description

Layer X phase 5: measured inputs, and which unmeasured ones move the burn.

* An override writes a stated number into the drawing before the twin reads it, with its
  provenance, so the twin's own assembly report counts it as measured.
* The uncertainty sweep varies each unmeasured input on its own. Checked against what each
  input can and cannot do, not against a stored answer:
  * nozzle efficiency changes thrust and nothing upstream of the nozzle;
  * a lower mixing E_m costs impulse;
  * a larger LOX Cd flows faster and burns out no later;
  * a measured input's stated uncertainty replaces the default range.

Function Documentation

◆ _case()

test_layerx_phase5._case	(	sweep,
		key,
		side )

protected

Definition at line 154 of file test_layerx_phase5.py.

Referenced by test_impulse_is_blind_to_the_feed_and_thrust_is_not(), test_nozzle_efficiency_moves_thrust_and_nothing_upstream(), and test_worse_mixing_costs_impulse_and_more_cd_flows_faster().

◆ config()

test_layerx_phase5.config ( void )

Definition at line 46 of file test_layerx_phase5.py.

◆ drawing()

test_layerx_phase5.drawing ( )

Definition at line 51 of file test_layerx_phase5.py.

◆ droop()

Override test_layerx_phase5.droop	(	float	value,
			uncertainty = None )

Definition at line 55 of file test_layerx_phase5.py.

Referenced by test_a_measured_uncertainty_replaces_the_default_range(), test_a_restated_parameter_reaches_the_twin_as_measured(), test_an_override_lands_in_the_drawing_with_its_provenance(), test_restatements_follow_the_drawing_through_an_edit(), and test_the_store_wants_a_source_and_one_value_per_parameter().

◆ sweep()

test_layerx_phase5.sweep	(		config,
			drawing )

Definition at line 148 of file test_layerx_phase5.py.

◆ test_a_measured_engine_input_reaches_the_card()

test_layerx_phase5.test_a_measured_engine_input_reaches_the_card ( config )

The card samples the engine with the config's measurements applied, as the runner and the
replay do. A measured LOX Cd 5 % above the model's has to show up as more LOX flow at the
same line pressures; before this was fixed the card's fast path sampled the unmeasured
config and the measurement reached the replay but not the card.

Definition at line 203 of file test_layerx_phase5.py.

◆ test_a_measured_uncertainty_replaces_the_default_range()

test_layerx_phase5.test_a_measured_uncertainty_replaces_the_default_range	(		config,
			drawing )

Definition at line 128 of file test_layerx_phase5.py.

References droop().

◆ test_a_restated_parameter_reaches_the_twin_as_measured()

test_layerx_phase5.test_a_restated_parameter_reaches_the_twin_as_measured	(		config,
			drawing )

The override is not a label: the regulator component the twin builds carries the
restated value, with the measured provenance, and the burn plan is prepared from it.

Definition at line 98 of file test_layerx_phase5.py.

References droop().

◆ test_an_override_lands_in_the_drawing_with_its_provenance()

test_layerx_phase5.test_an_override_lands_in_the_drawing_with_its_provenance ( drawing )

Definition at line 60 of file test_layerx_phase5.py.

References droop().

◆ test_impulse_is_blind_to_the_feed_and_thrust_is_not()

test_layerx_phase5.test_impulse_is_blind_to_the_feed_and_thrust_is_not ( sweep )

A fixed load burnt dry gives the same impulse whatever the valves: the tornado ranked by
impulse said "the top bar is the next measurement" while the feed was invisible to it.

Definition at line 172 of file test_layerx_phase5.py.

References _case(), and max.

◆ test_nozzle_efficiency_moves_thrust_and_nothing_upstream()

test_layerx_phase5.test_nozzle_efficiency_moves_thrust_and_nothing_upstream ( sweep )

Definition at line 188 of file test_layerx_phase5.py.

References _case().

◆ test_restatements_follow_the_drawing_through_an_edit()

test_layerx_phase5.test_restatements_follow_the_drawing_through_an_edit	(		tmp_path,
			drawing )

Keyed by what the drawing is (its shipped file, its pid-designer document), not by its
content: an edited drawing has a new id and keeps its restatements, and says which revision
they were entered against. An old id-keyed file still reads.

Definition at line 222 of file test_layerx_phase5.py.

References droop().

◆ test_the_band_is_told_each_way()

test_layerx_phase5.test_the_band_is_told_each_way ( sweep )

Mixing E_m costs far more impulse below than it adds above (0.70 vs 0.85 about 0.80):
one symmetric ± overstated the upside.

Definition at line 180 of file test_layerx_phase5.py.

◆ test_the_restatement_table_reads_and_takes_absolute_pressures_in_psia()

test_layerx_phase5.test_the_restatement_table_reads_and_takes_absolute_pressures_in_psia ( drawing )

Rail -> Measured values listed 'COPV pressure 3750 psi' beside a rail reading 3765 psia, and sent what
was typed with the bare 'psi', which the twin reads as gauge: typing the rail's 3765 gave the burn 3779.7
psia (ADR 0004/0007). The table now carries the psia each absolute pressure means, by the drawing rule.

Definition at line 74 of file test_layerx_phase5.py.

◆ test_the_store_wants_a_source_and_one_value_per_parameter()

test_layerx_phase5.test_the_store_wants_a_source_and_one_value_per_parameter ( tmp_path )

Definition at line 118 of file test_layerx_phase5.py.

References droop().

◆ test_the_sweep_runs_every_factor()

test_layerx_phase5.test_the_sweep_runs_every_factor ( sweep )

Definition at line 161 of file test_layerx_phase5.py.

◆ test_worse_mixing_costs_impulse_and_more_cd_flows_faster()

test_layerx_phase5.test_worse_mixing_costs_impulse_and_more_cd_flows_faster ( sweep )

Definition at line 197 of file test_layerx_phase5.py.

References _case().

Variable Documentation

◆ FIXTURE

str test_layerx_phase5.FIXTURE = Path(__file__).parent / "fixtures" / "ethalox_6500N_doublet_cad_2026-09-28.yaml"

Definition at line 29 of file test_layerx_phase5.py.

◆ GN2

str test_layerx_phase5.GN2 = "copv_study_gn2"

Definition at line 30 of file test_layerx_phase5.py.

◆ pytestmark

list test_layerx_phase5.pytestmark

Initial value:

=  [
    pytest.mark.skipif(not (shipped_drawings_dir() / f"{GN2}.json").is_file(),
                       reason="feed-twin's shipped drawings are not next to this checkout"),
    # The sweep fixture runs every factor of the uncertainty study on four spawned
    # workers, each importing and JIT-compiling the engine cold: 32 s on a laptop with
    # warm caches, past CI's 120 s per-test guard on its runners -- and the guard's
    # thread method ends the whole run, not just this test. pytest-timeout honours
    # this; without the plugin it is an unknown mark and only warns.
    pytest.mark.timeout(600),
]

Definition at line 33 of file test_layerx_phase5.py.

◆ reason

test_layerx_phase5.reason

Definition at line 22 of file test_layerx_phase5.py.

Functions

Variables

Detailed Description

Function Documentation

◆ _case()

◆ config()

◆ drawing()

◆ droop()

◆ sweep()

◆ test_a_measured_engine_input_reaches_the_card()

◆ test_a_measured_uncertainty_replaces_the_default_range()

◆ test_a_restated_parameter_reaches_the_twin_as_measured()

◆ test_an_override_lands_in_the_drawing_with_its_provenance()

◆ test_impulse_is_blind_to_the_feed_and_thrust_is_not()

◆ test_nozzle_efficiency_moves_thrust_and_nothing_upstream()

◆ test_restatements_follow_the_drawing_through_an_edit()

◆ test_the_band_is_told_each_way()

◆ test_the_restatement_table_reads_and_takes_absolute_pressures_in_psia()

◆ test_the_store_wants_a_source_and_one_value_per_parameter()

◆ test_the_sweep_runs_every_factor()

◆ test_worse_mixing_costs_impulse_and_more_cd_flows_faster()

Variable Documentation

◆ FIXTURE

◆ GN2

◆ pytestmark

◆ reason