Table of Contents ECSS Model Page
Background Information Atmosphere and ionosphere models
Atomic oxygen erosion


Before the early Space Shuttle flights, interactions between spacecraft surfaces and the low Earth orbital environment were considered to be benign especially for missions of short duration. The principal concern has been the induced molecular environment in which sensitive instruments, such as those flown aboard Spacelab, must operate to obtain useful scientific data.

Space Shuttle flights (flights 5 and 8) have demonstrated that interactions between spacecraft surfaces and atomic oxygen, the major component of the low Earth orbit environment, can produce significant changes in mass and surface properties for many materials through erosion and oxidation. The orbiting spacecraft motion through the atmospheric atomic oxygen can generate a flux to the spacecraft surfaces with significant energy of about 5 eV.

The changes in surface properties are directly related to the atomic oxygen fluence or total integrated flux incident on material surfaces. This fluence in turn is dependent on many parameters such as attitude of the surface relative to the orbital velocity vector, altitude, solar activity, and life span of the spacecraft.

Assessment of the problem involves three areas:

  1. flight data on material degradation under atomic oxygen exposure;
  2. ground testing of degradation under atomic oxygen exposure;
  3. modelling of the atomic oxygen environment and the resulting fluences and effects on spacecraft surfaces during a mission.

Atomic oxygen effects

Low Earth orbital systems must be constructed of materials that are compatible with their operational environment to ensure successful long term performance and durability. Early space shuttle flights have demonstrated that polyimide Kapton, carbon coatings, and some paints suffer changes in optical properties as well as loss in mass when exposed to the low Earth orbital environment. Atomic oxygen has a number density and ram impact energy between the altitudes of 180 km and 650 km that is sufficient to pose a threat to the long term durability of solar arrays and other material surfaces being considered for use on the Space Station [Banks and Rutledge, 1988].

Numerous events can occur when atomic oxygen impacts a surface [Banks and Rutledge, 1988]. The atomic oxygen may elastically scatter from the surface in a specular manner, it may energy and momentum accommodate to the surface and then be ejected in a diffuse manner, it may attach to the surface and react with other arriving species to form excited nitrogen oxide, which de-excites to cause the glow phenomenon, or the atomic oxygen may chemically react with the surface. Space tests indicate that the probability of chemical reaction of atomic oxygen with carbon is only 13%, and with silver, greater than 62% [Banks et al., 1988; Gregory, 1986]. In the case of carbon, the unreacted atomic oxygen is predominantly ejected in a diffuse or near cosine distribution with an ejection peak shifted slightly in the specular direction. The actual probability of reaction of atomic oxygen with silver may be quite high for clean silver surfaces. However, silver oxide formation may contribute to shielding of the underlying silver or may serve as a catalytic surface for reassociation of the atomic oxygen. Possible atomic oxygen reactions with polymers in low Earth orbit include hydrogen abstraction, oxygen addition to form excited radicals followed by hydrogen elimination, oxygen insertion into the C-H bonds, and replacement by formation of alkoxy radicals.

Quantification of the susceptibility of materials to oxidation from atomic oxygen interaction is made possible by the fact that most oxidation products are volatile, and measurable surface recession occurs. In cases where oxidation products remain adherent, inhibition of further oxidation usually occurs. From a spacecraft designer's perspective, quantification of atomic oxygen interaction in terms of erosion yield defined as recession per fluence (cm3 atom-1) is most useful for determining mission suitability. Banks and Rutledge [1988] present a compilation table of erosion yields of materials tested in low Earth orbit. On a microscopic level, spacial variations in erosion yield tend to give rise to a dense pillar of conelike surface morphology if the reaction products are volatile species, as in the case of organic polymers and carbon. This surface morphology is much more pronounced for ram exposed surfaces as opposed to sweeping atomic oxygen incidence. These microscopic surface features, typically ranging in size from 0.1 to several microns, give rise to changes in optical properties, the most significant of which is a dramatic increase in the diffuse reflectance with an accompanying loss in specular reflectance [Banks et al., 1988].

Space tests have shed some light on the degree to which erosion yield is dependent upon various environmental factors. The effect of atomic oxygen impact angle for Kapton and Mylar in space indicates that the rate of material recession depends on the impact angle with respect to the surface normal raised to the 1.5 power as opposed to the 1.0 power as on might expect [Visentine et al., 1985]. The effect of material temperature has been shown to influence the erosion yield of graphite, and activation energies have been predicted [Gregory, 1986]. However, the range of temperatures used for the evaluation of Kaption, Mylar, and Tedlar has been too small to ascertain any significant dependence of erosion yield upon material temperature. Space tests have not identified any significant influence of solar radiation, charged species, or polymer thickness on erosion yield [Banks et al., 1988]. Erosion yield of Kapton H due to space exposure indicates no significant dependence upon flux or fluence based on the limited information available from STS flights 3, 4, 5, and 8 [Banks et al., 1988]. Variations in the erosion yield of Kapton as a function of oxygen ion or atom energy are not available from space testing to date because the low Earth orbital configurations produce energies only in the 4-5 eV range. However, the dependence of erosion yield on impact energy is of utmost importance for the designers of ground simulation systems. An erosion yield dependence proportional to the 0.68 power of the oxygen atom or ion energy has been identified for Kapton H by means of a compilation of both in-space and ground simulation tests over a wide range of energies [Ferguson, 1984].

To facilitate accurate prediction of materials performance in low Earth orbit, one must be able either to accurately simulate the low Earth orbital environmental conditions or at least be able to quantifiably extrapolate how the performance of materials under simulated conditions relates to that which would occur in space. Identification of significant factors and the dependency of erosion yield upon these factors will clarify the suitability of various atomic oxygen simulation techniques for prediction of the long term performance of spacecraft materials in low Earth orbit. Based on the evaluation of materials tested in space, there appear to be three classes materials:

  1. materials of high erosion yield which include most of the hydrocarbon organic materials;
  2. materials which either do not react with atomic oxygen or form self-protecting oxides which allow the underlying material to appear durable to atomic oxygen;
  3. materials with low but non-negligible erosion yields, such as fluoropolymers.
In the low Earth orbital environment, materials with high erosion yields should either not be used or be protected by materials with negligible erosion yields. Verification of atomic oxygen durability of protected materials or materials with negligible erosion yields can be accomplished by the simpler, less exacting atomic oxygen simulation systems when it is desired to know whether a material is oxidized by atomic oxygen rather than the exact rate at which it is oxidized. Evaluation of materials with low but non-negligible erosion yields requires simulation systems which produce quantifiable agreement with the results of space tests.

Implementation in SPENVIS

The ATOMOX tool developed for ESABASE accurately calculates the atomic oxygen fluence on arbitrarily oriented spacecraft surfaces and includes all relevant parameters of the incident particles. The tool accounts for the thermal motion of the particles as well as the spacecraft motion through the corotating atmosphere. In addition to atomic oxygen, ATOMOX is able to analyse flux and fluence of other species (atoms and ions) of the atmosphere.

The ESABASE implementation of ATOMOX considers shadowing and scattering by other spacecraft surfaces and is capable of analysing atomic oxygen effects on a spinning satellite. These capabilities use a full geometric analysis of the satellite configuration, which is outside the scope of SPENVIS. Instead, SPENVIS uses the non-geometrical version of ATOMOX for a preliminary assessment.

The following operations have to be executed by the analysis tool at each orbital point:

The atomic oxygen density is evaluated with standard models of the Earth's atmosphere. The orientation of the impacted surface is defined in terms of the attitude vector specified for the orbit generator.

Physical modelling

The final goal of an atomic oxygen analysis is to assess the fluence of atmospheric particles to outer surfaces of satellites. Resulting material erosion is taken as proportional to the fluence of particles on these surfaces.

The fluence is defined as the integration over time of the flux crossing a surface. The flux on a totally exposed surface is the product of the particle number density and the average velocity of the particle.

Velocity computation

The velocity v of a particle is the sum of its aerodynamical velocity vaero and the thermal velocity u. The thermal velocity is computed by means of the Maxwellian theory of gas dynamics. This theory assumes an isotropic distribution. The components of the thermal motion follow a normal distribution with probability density f given by:

f(ui) = 1/(pi1/2u) exp(-ui2/u2) ,

with ui the three velocity components.

The thermal velocity is given by:

u = (2RT/M)1/2 ,

with T the ambient temperature (K), R is the universal gas constant (8314 J kmol-1K-1), and M is the molecular weight (kg kmol-1).

The aerodynamic velocity is the difference between the wind velocity vector vw and the spacecraft velocity vector vs in an inertial frame of reference:

vaero = vw - vs .

Flux computation

Flux integration

The flux phi of atoms (cm-2 s-1) impacting a given surface assumed to be totally exposed is equal to the product of the local particle number density N0 (cm-3) and the average velocity (cm s-1) of the particles impacting the surface.

Reference frame for particle
Figure 1. Reference frame for the velocity of impacting particles

If the inward surface normal defines the direction of the x axis (see Fig. 1), one has to calculate the average velocity component vx of particles having vx>0:

Expression for vx

The flux phi then is given by phi = N0vx. Solving the above integral yields:

phi = N0 (RT/2piM)1/2 {exp(-s2)+pi1/2s[1+erf(s)]} ,

where s is the molecular speed ratio given by:

s = (vaero.x) / u ,

and erf is the standard error function:

Expression for erf(s)

Flux distribution

When thermal motion is taken into account, it is possible to compute the distribution of the impact flux relative to the surface normal elevation angle. As thermal motion coordinates follow a normal distribution, a large number of samples following a centred normalised normal law are computed.

Fluence computation

Fluence is defined as the integration over time of the flux crossing a surface. It can also be defined as the product of the average flux over a certain time, and this time. As flux is computed for a discretised number of times (corresponding to the orbital positions), this definition is used in ATOMOX. Fluences for each orbit as well as for the whole mission are calculated.

Orbital fluence

A recurrent trapezoid method is used. The fluence phii at orbital position is given by:

phii = phii-1 + 0.5 (fi+fi-1) (ti-ti-1) ,

where fi is the flux at orbital position i and ti the corresponding Julian date, and phi0=0.

Mission fluence

A similar method is used for the cumulated fluence from the beginning of the mission. The fluence Fj after orbit j is given by:

F1 = Phi1 (T1-t1)

Fj = Fj-1 + 0.5 (Phij+Phij-1) (tj-Tj-1) + Phij (Tj-tj) ,

where Phij is the average flux over orbit j, Tj is the Julian date at the end of orbit j, and tj is the Julian date at the start of orbit j. The mission fluence is the fluence after the last orbital arc.


Banks, B. A., and S. K. Rutledge, Low Earth Orbital Atomic Oxygen Simulation for Materials Durability Evaluation, Proc. Fourth European Symposium on Spacecraft Materials in Space Environment, CERT, Toulouse, 6-9 September, 1988.

Banks, B. A., S. K. Rutledge, J. E. Merrow, and J. A. Brady, Atomic Oxygen Effects on Materials, Proc. NASA/SDIO Joint Workshop on Space Environmental Effects, Hampton, VA, 28-30 June, 1988.

Ferguson, D. C., The Energy Dependence of Surface Morphology of Kapton Degradation Under Atomic Oxygen Bombardment, Proc. 13th Space Simulation Conference, Orlando, FL, 8-11 October, 1984.

Gregory, J. C., Interaction of Hyperthermal Atoms on Surfaces in Orbit: University of Alabama Experiment, Proc. NASA Workshop on Atomic Oxygen Effects (ed. D. Brinza), Pasadena, CA, 10-11 November, 1986. (JPL Pub. No. 87-14)

Visentine, J. T., L. J. Leger, J. F. Kuminecz, and I. K. Spiker, STS-8 Atomic Oxygen Effects Experiment, Presented at the AIAA 23rd Aerospace Sciences Meeting, AIAA-85-0415, Reno, NV, 14-17 January, 1985.

Last update: Fri, 13 Apr 2018