AutoMovie API
    Preparing search index...

    Variable analyzeAutoMovieEnvelopeConst

    analyzeAutoMovieEnvelope: (
        props: IAutoMovieAnalyzeEnvelope.IProps,
    ) => IAutoMovieAnalyzeEnvelope.IResult = ...

    Solve one envelope for heat and for surface condensation, as two runs.

    Two runs rather than one because heat and moisture are two domains a report rolls up separately: a wall may be warm enough and still wet, and one merged verdict would let either fact hide the other. Both read the same inputs and carry the same settings digest, so they are provably about the same envelope.

    The governing model is the one-dimensional steady-state network, stated exactly:

    • R = Rsi + sum(t / lambda) + Rse, and U = 1 / R;
    • The interior surface sits at Tsi = Ti - U * Rsi * (Ti - Te), so the temperature factor fRsi = (Tsi - Te) / (Ti - Te) reduces exactly to 1 - U*Rsi and needs no temperatures at all;
    • Fabric loss is sum(U * A) * dT and linear bridges add sum(psi * L) * dT;
    • The dew point comes from autoMovieDewPoint, the Magnus form, and surface condensation is Tsi <= Td.

    What it does not do, it says. Condensation inside the build-up, which needs vapour diffusion, and solar gain through glazing each appear as an unsupported metric with the exact reason, because an envelope study that quietly omitted them would read as a complete answer. Transient storage is excluded by the steady-state model itself, which the solver's model text states.

    Without an exterior boundary condition there is no answer at all, and both runs come back not-run naming the missing input rather than assuming a temperature nobody declared.

    The Magnus coefficients are fitted over roughly -45 to 60 degC, and an indoor temperature outside that range is not refused. Throws on malformed requests or a nonfinite computed metric; finite positive fields alone do not prevent resistance/temperature overflow, underflow or dew-point singularities.