Flight Loads Survey

Flight Loads Survey (flight-test-operations/envelope/flight-loads-survey)

Use when the task is a flight loads survey for a flight test program: strain gauge calibration, maneuver points, and the load factor versus speed survey.

Domain quick reference

• Survey objective: a loads survey measures flight loads (wing bending, shear, torsion) with calibrated strain gauge stations and compares them against the analytically predicted loads envelope (FAR 25.301 / CS-25.301 load conditions context). The survey data validates the structural load predictions that size the primary structure.
• Strain gauge relation: strain epsilon = delta_R / (R * GF), with delta_R the gauge resistance change in ohm, R the nominal gauge resistance in ohm, and GF the gauge factor (dimensionless, typical 2.0 to 2.2). Compression gives negative strain.
• Calibration: during the ground calibration the wing or empennage is loaded with known applied loads L_i while each gauge records strain epsilon_i; the through-origin least squares slope K = sum(L_i * epsilon_i) / sum(epsilon_i^2) converts strain to load. The measured in-flight load at a station is L = K * (epsilon - epsilon_0) with epsilon_0 the pre-flight zero offset.
• Maneuver points: steady symmetric maneuvers (elevator pull-ups and pushovers) and rolling maneuvers build load factor at discrete survey speeds; each point records the normal acceleration and the strain at every station. Measured load factor n = L_meas / W_ref with W_ref the reference weight.
• Load factor versus speed survey: at dynamic pressure q = 0.5 * rho * V_eas^2 the achieved load factor is n = q * CL / (W/S); the survey flies (V, n) points and reduces each to CL = n * (W/S) / q. A point is feasible only while CL stays at or below CL_max, otherwise the maneuver stalls before reaching the target load factor. Example: W/S = 6000 Pa, rho = 1.225 kg/m^3, V = 100 m/s EAS gives q = 6125 Pa, and CL = 1.8 gives n = 1.8375.

Workflow

1. Install and verify the strain gauge stations; record the nominal resistance R and the gauge factor GF of each gauge.
2. Run the ground calibration: apply the known loads, record the strains, and compute the calibration factor per station with calibration_factor(loads, strains).
3. Reduce each survey point with measured_load(k, strain, zero_strain) and load_factor_from_measured_load(measured, reference_weight); keep the point only while its strain lies inside the calibrated range (point_in_calibration_range).
4. Convert each speed point with dynamic_pressure(rho, v_eas) and lift_coefficient_at_maneuver(load_factor, wing_loading, q), or predict the achieved load factor with symmetric_maneuver_load_factor(v_eas, rho, wing_loading, cl).
5. Check feasibility against the stall boundary with maneuver_point_feasible, compare measured against predicted loads with load_error_percent, and gate the survey program on the per-point results.

Pitfalls

• Flying the survey before the ground calibration is complete; the strain to load conversion is meaningless without the calibration factor and the zero offset.
• Mixing up the two calibration forms: calibration_factor fits the through-origin slope K = sum(L * eps) / sum(eps^2); feeding it an intercept fit silently changes every measured load.
• Ignoring the zero offset: L = K * epsilon without subtracting the pre-flight zero strain shifts every measured load by K * epsilon_0.
• Reporting a strain outside the calibrated range as a load; the extrapolation beyond the applied calibration loads is not valid.
• Forgetting that the load factor depends on the reference weight; n = L / W_ref with the wrong weight matches neither the V-n diagram nor the predicted loads.
• Treating a (V, n) point whose required CL exceeds CL_max as valid; the maneuver stalls first and the point does not represent the target load factor.
• Taking the calibration factor from a single lightly loaded point; the least squares slope needs a spread of loads across the calibrated range.
• Passing empty, mismatched, or zero-energy calibration data; the module raises ValueError instead of returning a nonsense factor.

Behavior contract (gate 3)

The strain calibration, load reduction, and load factor versus speed logic is exercised by the gate 3 contract test: scripts/test_flight_loads_survey.py against scripts/flight_loads_survey_logic.py (stdlib unittest, offline). Run: python3 scripts/test_flight_loads_survey.py

Compliance

• Standards referenced, not reproduced: the FAR 25 / CS 25 load conditions (25.301 context) and the loads survey practice are common flight test methodology, summary-only per standards-map.yaml.
• compliance: STANDARDS-REF, gated: false.