Apply aerodynamic force to the part

Parameters

part	the part
rbPossible	the rigidbody (either part or part.parent)
mode	the mode of force to apply

Apply aerodynamic force to the part

Parameters

part	the part
rbPossible	the rigidbody (either part or part.parent)
mode	the mode of force to apply

< rotate the lift force into world coords

< exclude anything in the drag direction, we're not worrying about drag here

< finally, add the force.

Calculates a convective coefficient, for convective transfer = coeff * area * (air temp - skin temp) Handles the splashed case, then if in air lerps between Newtonian and Mach (Hypersonic) convection based on the Newtonian-Mach lerp and the craft's Mach number.

Parameters

situation

Set to a Vessel.Situation to use for the calculation.

Returns

Calculates analytically a steady-state temperature for the entire vessel using the vessel's emissive and absorptive-weighted radiative area, the incoming solar flux, any body flux, any atmospheric convection, and any internal generation. Conduction is entirely ignored (it is assumed to be near-instantaneous on these timescalse) and convection is also guesstimated on a final-analysis, steady-state basis (basically lerping between external temperature and pure radiative-balance temperature based on atmospheric density). This supports callbacks to parts using various interfaces.

Returns

< show 0s when in analytic

< get incoming radiation

< fourth power

< get solar flux

< get flux from albedo and emissivity of current body

< Handle callbacks

< find resting temperature.

< lerp between it and ambient

< now apply it

< fourth power

< apply it.

This is the temperature of background radiation yields correct value for sea level in Earth atmosphere, more or less other values are my best guess. Uses the densityThermalLerp

Parameters

ambientTemp

Returns

Calculate the basic constants (pressure, density, solar flux, etc) when in atmosphere. Does solar air mass calcs, applies atmosphere temperature offset, calculates shock temp and convection stats, etc.

< Get/set aero stats

< compute lerp between the two modes of convection (also used to calculate shock temperature)

< compute external (shock) temperature, as distinct from static ambient temperature

< set convective heat transfer coefficient

< to be multiplied by a part's dimension later. Maybe. Probably not.

Calculate the basic constants (pressure, density, solar flux, etc) when in vacuum

< no atmosphere to change solar flux

Calculates a convective coefficient, for convective transfer = coeff * area * (air temp - skin temp) Handles the splashed case, then if in air lerps between Newtonian and Mach (Hypersonic) convection based on the Newtonian-Mach lerp and the craft's Mach number.

Returns

Calculates a convective coefficient, for convective transfer = coeff * area * (air temp - skin temp) Handles the splashed case, then if in air lerps between Newtonian and Mach (Hypersonic) convection based on the Newtonian-Mach lerp and the craft's Mach number.

Parameters

useSituation

Set true will use the vessel situation (splashed or not). Otherwise will calc as if not splashed.

Returns

< else calculate both and lerp

Calculate high-speed convective coefficient. Simple function of const * dnesity^exponent * spd^exponent (the exponents differ)

Returns

Calcuklate a coefficient for simple Newtonian convection Coeff ~= (density>1, density, else dnesity^exponent) * (base + spd^exponent)

Returns

Calculates a lerp value for use in determining background radiation temp and for steady-state temperatures w.r.t convection in analytic mode. NOTE: Uses the post-shockwave density, not the static density

Returns

Calculates the shock temperature given a velocity

Returns

Calculates the distance to the sun, the solar flux, and the atmospheric temperature characterists and body albedo/emissive flux

< this does mean on the first frame we'll be in sunlight, but that's not a terrible assumption for the use cases we'll have.

Check and recreate thermal graph

< should be:

< (tonne/m^3 to kg/m^3, then to kPa from Pa, cancel)

< should be part vel-based, not vessel-based

Handle most physics integration

< double check for any vessel part changes in case events don't catch

< orbitdriver runs earlier now so SOI transition occurs at the start of frame not the end. Consistency!

< calculate pressure

< set vessel values

< done in Vessel.Awake - density = currentMainBody.GetDensity(staticPressurekPa, currentMainBody.GetTemperature(altitude));

< do this instead

Get radiative area facing the body

Parameters

ptd

Returns

Reimplemented from VesselModule.

rewrite of the part method; this one accepts chains, and uses what we calculate in VesselPrecalculate

Parameters

part

Returns

Get radiative area facing the sun

Parameters

ptd

Returns

Apply all forces to the part, both via the part's force list and from gravity (if done here not in Precalc) and from drag/bodylift. Also calculate angular drag both in air and submerged

Parameters

part

Apply gravity and drag to physicalobjects

Parameters

pObjs
atmDensity

< Safety check

Reimplemented from VesselModule.

Reimplemented from VesselModule.

Reimplemented from VesselModule.

Sets the conduciton multiplier and skin-skin transfer multiplier for a PTD The latter starts from sqrt(radiative area) and then tries to take the relative areas of exposed and unexposed skin into account, i.e. it's maximal when about half is exposed and half unexposed

Parameters

ptd

< get circumference, basically

Determine the final convective coefficient based on occlusion and convective stats

Parameters

ptd

< calculate convection

< W to kW, scalars

Precalc radiative stats like final coefficients and the final background rad temp exposed and unexposed. This is complicated because we need to figure out what portion of sun and body flux each apply to exposed and to unexposed skin (this is done by taking the dot of the sun vector vs the velocity vector and the -upaxis and the velocity vector respectively). All these get stored as incoming fluxes, since they are constant given a position and orientation and so needn't be recalculated during all the RK2 steps.

Parameters

ptd

< shared scalar

< W to kW

< shielding section

< sun flux

< do we have to deal with separate skin temps?

< do we have to deal with separate skin temps?

Sets the basic skin properties for this thermal pass. Calculate the total area, the fraction of which is exposed or unexposed, and handle any transfers in thermal energy when those areas change.

Parameters

ptd

< set base area properties

< get exposed, if so area, set ptd data

< keep in sync

< will be negative

Determine skin thermal mass from radiative area and part dry mass

Parameters

part

< this thermal mass will be subtracted from part thermal mass next tick.

< save from prefab, load on part.

< would be done soon anyway, but let's do it now.

Reimplemented from VesselModule.

Smooth velocity so frame shifts are not passed on to aero/thermo integration

< First, detect spikes

< are we in a spike?

< >100G is probably a frame spike

< replace the old one too

< reset spike detection

< Calculate the smoothed velocity

The actual thermal integration pass. First update conduction (this merely establishes fluxes). Then loop through all PTDs. Store previous temps if we're doing RK2 passes, then calculte convection and radiation for the part. Finally apply the fluxes.

Parameters

averageWithPrevious

Is this the second step of an RK2 pass

< calculate convection

< calculate radiation

< Apply fluxes

< FIXME add skin conduction menu items

< DEBUG - log if we blow up.

Any not-every-frame thermo calcs

Unifies the exposed and unexposed skin temps into a unified temp using the whole skin thermal mass

Parameters

ptd

Recalculate those bits that don't need to happen every frame (for now, the sun and body fluxes / air temperature offsets

Updates the aerodyanmics stats of the part and also handles calculating submerged dynamic pressure and total angular drag.

Parameters

part

< detect rb

< set stuff to zero

< precalculate all the drag vector stuff

< sidestep a potential div by zero

Apply thermal links based on compound parts

< us to target

The big one. Processes conduction across the vessel. Occurs in multiple passes. The first pass involves zeroing fluxes and getting unified skin temperatures. Then internal conduction happens (i.e. part internal <-> part internal, no skin). The PTDs are sorted by int temp, and then each PTD's links are checked. Heat pushes out to the links based on the conductive coefficient and the attach area. Checks are done to ensure stability. Since we start from hottest first, the hottest parts get priority flowing heat outwards. The same occurs for part skin <-> part skin conduction. Finally skin-skin conduction is computed for individual parts, between exposed and unexposed skin. Note that throughout this process only fluxes are stored.

< reset temporary vars

< figure out total system energy and store initial conductive fluxes.

< Now clamp the conduction.

< localInt is negative

< so tempDelta is negative

< half of internal

< when skin is not unified

< figure out total system energy and store initial conductive fluxes.

< the inside of a cargobay's skin doesn't count.

< fudge factor, to make up for contactArea<radArea and that conduction is the product of both conduction factors.

< local is negative

< so tempDelta is negative

Process convection in the simple flux = coeff * t_delta manner

Parameters

ptd

Precalculate thermal mass and resource mass while finding CoM and Velocity.

< now Subtract skin thermal mass, then calculate recip

< update data according to projection vector

< Ensure sorted ascending

< clear cone list and add a cone for our leading part

< for every remaining part..

< initially not occluded

< check against all cones..

< just for sanity

< reset multipliers

< update data according to projection vector (velocity)

< Ensure sorted ascending

< this is the maximum wedge angle at a given mach for an attached shock. If angle > this, detached.

< a decent approximation

< clear cone list and add a cone for our leading part

< Apply angle to FX shocks based on leading part angle.

< for every remaining part..

< initially not occluded

< check against all cones..

< just for sanity

< update data according to projection vector

< Ensure sorted ascending

< clear cone list and add a cone for our leading part

< for every remaining part..

< initially not occluded

< check against all cones..

< just for sanity

Apply incoming fluxes to part radiative fluxes and also calculate outgoing rad flux based on black-body radiation.

Parameters

ptd

< will not be used unless wasExposed = true

< after precalcing, we get the actual BRT each pass.

< Radiative flux = S-Bconst*e*A * (T^4 - radInT^4)

Rebuilds the thermal graph. This involves recreating all PTDs and PTLs, cleaing and recreating occluders, reapplying old settings on the new PTDs, and finally registering the various callback interfaces.

< update the drag occlusion

< set skin temps if we had them before

< resize occluder arrays

Main thermo method. First decides whether to run an analytic pass based off the last time a thermo pass was run (so when an unloaded vessel is loaded, analytic is run with the whole unloaded time, not just the current fixed frame time. If analytic, then that temp is applied, following any callbacks. If not, then all fluxes are zeroed and temps sanified, and then each of the Precalculate methods are run (for conduction, convection, and radiation). Finally some number of thermal integration paasses are run, either one pass using RK1 or a set of RK2 passes.

< i.e. set to analytic temp, period.

< twice the analytic min for unloaded.

< FIXME this will slightly speed things up at high rates.

< see above.

< clamp first, so Lerp doesn't have to.

< catch things that spawn with no set temperature, like asteroids

< only need to do this once, since mach and velocity vector are constant over thermal loop.

< Apply added fluxes.

< run each pass

Vessel velocity members.

based on latitude, sun angle, etc.

in kW/m^2

in kW/m^2

We have to keep track of compound parts specially

Used to prevent calculating external temperature several times before fixedUpdate is called in the flightIntegrator (i.e. if more than one part tries to in the same frame)

Are we in Analytic mode?

Vessel stats.

Solar, Body.

list of IAnalyticOverheatModule for analytic callbacks

The set of PTDs, one per part with nonzero thermal mass. This is sorted by internal temp

Same set of PTDs, sorted by skin temp

List of preview modules for analytic callbakcs

solar air mass multiplier

reduction to solar flux due to atmospheric absorption/scattering/etc

dot between terrain surface normal and sun.

the vector to the sun