K.1.3 Transport of molecular contaminants

K.1.3.1. Transport between surfaces

K.1.3.1.1 General

The following clauses only deal with the methods and models for transport of neutral molecules. There is no available model of ion transport devoted to contamination.

Three levels of complexity and accuracy in modelling the transport of neutral molecular contaminants can be distinguished.

K.1.3.1.2 Simplest view factors

This model simulates collisionless transport. In such a case the fraction of contaminants coming from surface j to surface i is given by the view factor V_ij of surface i seen from surface j (including the cosine factor coming from the Lambertian emission law). These view factors are similar to the ones of radiative thermal analysis. They can be computed geometrically or by Monte Carlo ray tracing. The incident mass rate on a surface i is then given by

(K-12)

where

S_j sticking coefficient on surface j

j runs over all surfaces and

denotes the outgassing mass rate of surface j.

K.1.3.1.3 Simplified Monte Carlo

Collisions of contaminants are simulated in a simplified way; the density and speed of possible partners for molecular collisions are given a priori:

• for ambient scatter, the ambient density and speed are easily known, but wakes (or “shades”) are usually not treated;

• for selfscatter, the contaminant density is very simplified and usually taken proportional to r^-2 and with spherical symmetry.

This method is usually limited to one collision per molecule because the uncertainties due to the densities given a priori increase with collision number. This effective view factors can conveniently be computed by Monte Carlo raytracing method.

Both methods (K.1.3.1.2 and K.1.3.1.3) can include other contaminant sources such as vents and plumes. The view factors are then replaced by interception factors.

K.1.3.1.4 True Monte Carlo (Direct Simulation Monte Carlo, DSMC)

This computes multiple collisions in a realistic way. The collision probabilities are computed autocoherently from the densities given by the simulation. This method is more time consuming and requires more work for programming (in particular, it requires a meshing of volume and not only of spacecraft surfaces).

Either method can be better suited, depending on the spacecraft configuration. A potential contamination of a sensitive protected surface through multiple collisions requires a precise DSMC simulation. In simpler cases, when contamination essentially happens in lineofsight, it is more appropriate to use the less timeconsuming and more widespread methods of K.1.2.1.2 and K.1.2.1.3.

K.1.3.2. Surface transport

Reflections on surfaces and reevaporation are easy to implement and are usually included in models, the latter (reevaporation) often as part of the outgassing process. Migrations on surfaces on the contrary are complex processes and there is no commercial available model.

K.1.3.3. Transport of particles

As mentioned in 11.1.5 particulate transport is governed by several phenomena:

a. atmospheric drag

b. solar radiation pressure

c. differential gravitational effects (with respect to spacecraft) which result in tide effects

d. particulate charging and subsequent electrostatic effects

Among which the first three can be computed by methods similar to spacecraft orbit computing, whereas point d. requires specific modelling to access particulate charging in a plasma and potential map around spacecraft. The dominant phenomena are most commonly modelled: point a. atmospheric drag, first, and also point d. that gets important in GEO. Points b. and c. can become dominant in cases when points a. and d. become small (high altitude and no charging).

A last aspect of particulate transport is their interaction with walls. Sticking and accommodation coefficients are, however, very difficult to assess.

Most particulate contamination models remain in the field of research. Very few of them seem to be transferable to other users.