• No se han encontrado resultados

TOTAL 2.802 2.109 3.319 954 632 10.708 Fuente: Elaboración propia a partir de los datos de matricula facilitados por la Universitat de les Illes Balears.

4.3. Diseño del cuestionario.

4.3.2. Base conceptual y definición de las variables: recogida y medición.

4.3.2.4. Estilos de vida: Tabaco.

certain circumstances. During laboratory electron beam materials characterisation, an isolated, conducting sample tended to behave in a conduction-limited fashion, rather than the emission-limited fashion predicted on the basis of published bulk and surface conductivities for the insulating polymer. For a Teflon layer, the increased conductivity is attributed to the presence of surface contaminants which appear to increase the surface conductivity by as much as two orders of magnitude above the published value. For a thin Kapton layer, surface conductivity is also important, but the electric field enhancement to bulk conductivity accounts for most of the increased conduction current. Where the total conductivity (bulk plus surface) may be specified, and remains constant with sample potential (for a thick insulator), both one and three dimensional simulation codes give good agreement for sample potentials versus beam energy, and for charge time history development. Much poorer agreement is obtained where field enhanced conductivity is

important.

5. The EQUIPOT code was used to assess the effect of making small changes in the conduction current term of the current balance equation. The most important result to emerge was that for thin dielectric films, if the conduction current model includes the effect of field enhanced conductivity, then the results are very different from those obtained assuming pure Ohmic conduction. This is consistent with the experimental results of Chapter 6, where field enhancement effects dominate.

6. A new model for secondary emission of electrons due to electron impact was introduced and incorporated into the EQUIPOT simulation code. The secondary emission yield curve is described in terms of three parameters; the maximum yield and energy at maximum yield for normal incidence, and the energy loss power, which is assumed to be constant for a given material. A sensitivity analysis using these parameters revealed that the equilibrium potential is highly sensitive to small changes in n, the energy loss power. Furthermore, if all other material definition parameters remain constant, it is the value of n which is most critical in determining the plasma temperature above which negative charging occurs. The equilibrium potential was shown to be most sensitive to changes in energy loss power when this quantity just exceeded the threshold charging value for a given plasma temperature.

7. The EQUIPOT code was used to verify that, for a Maxwellian plasma, there exists a threshold plasma temperature, above which negative charging occurs. EQUIPOT was used to determine the locus in (n,T) space, where n is the energy loss power for a given material, which (for all

other material parameters being constant) represents the negative charging threshold boundary. Furthermore, for measured non-MaxweIlian electron spectra (taken from the Meteosat F2 data set), EQUIPOT was used to demonstrate that the mean spectral energy (equal to twice the plasma temperature) could be used in place of a Maxwellian temperature to determine a charging threshold. However, for a given material, where the plasma mean energy only just exceeds twice the threshold temperature, a "triple-root" current-voltage relationship tended to develop, which is not observed with single Maxwellian environments.

8. A statistical survey of Meteosat F2 barrier events on a "per-event" basis reaffirms earlier results that events tend to occur in the midnight to dawn sector of local time (plasmasheet conditions), become increasingly likely when Kp is high and are observed only when the radiometer primary mirror is shadowed throughout the whole satellite rotation. Further analysis reveals a clear electron energy threshold, below which barrier events are not triggered which is consistent with EQUIPOT results of Chapter 7. In addition, there is evidence that this threshold electron temperature reduces as ap increases which can be attributed to the increasingly efficient removal of cold ions at higher activity levels.

9. A three-dimensional charging analysis of Meteosat F2 using the NASCAP code has successfully simulated a "barrier event". A severe plasma environment caused the shadowed primary mirror to charge to about -12 kV creating a potential barrier around the whole spacecraft which leads to absolute and differential potentials of the order of several hundred volts to develop on the solar array surfaces. The predicted potential contour configuration around the electron spectrometer location will cause trapped secondary and photo-electrons to be accelerated and detected by the spectrometer with energies in good agreement with those observed experimentally.

Inevitably, this work leads to further questions concerning future areas of research. There are two immediate areas of work which would add greatly to the value of the results presented here. Firstly, it would be interesting to consider ion data as well as electron data for the identification of plasma regimes, substorm studies and EQUIPOT charging threshold calculations. Suitable ion data sets for comparison include the ion spectrometer on ATS-5108 and the SC9 ion detector on P78-2 (SCATHA)109 in addition to measurements of cold ions by the GEOS spacecraft91. The CRRES110 satellite, launched in July 1990 spends a large fraction of its time at near-GEO altitudes and will generate a large database of plasma measurements in the tens of eV to tens of keV range. However, these measurements were not made simultaneously and the effects of solar cycle variation may well prevent serious comparison with the Meteosat F2 electron data set.

Secondly, there is much scope for improved measurements of secondary electron emission for materials using the electron beam facility described in Chapter 6. Recent studies using a different facility99 have

proved that good isolation of the target sample may be achieved such that results are emission limited. Further study into the effect of sample surface preparation is necessary along with the adoption of a new technique due to von Seggern111 for measuring the SEE yield of dielectric materials.

Acknowledgement s

I would like to thank my employers (Space Department, Royal Aerospace Establishment, Farnborough) for the opportunity to carry out this research as part of my normal duties. Special thanks to my supervisor at RAE, Dr. Gordon Wrenn for his guidance and interest over the last five years, and to my supervisor at Mullard Space Science Laboratory, UCL, Dr. Alan Johnstone for his suggestions on the direction of research, and for providing the Meteosat electron spectrometer data. The electron beam tests would not have possible without the enthusiastic support of Mike Duck of Chemistry Division, Harwell Laboratory. Data processing, analysis and simulation were carried out on a PRIME 2655 mini computer which has proved to be ever-reliable thanks to the Space Computing Service, RAE. Last but not least, thanks to Helen for her patience and encouragement throughout.

n o o n n o n n n n n n n n n n n n Appendix A