Air-fuelled Plasma Propulsion System

Canterbury University Department of Electrical and Electronic Engineering

Air-Fuelled Plasma Propulsion System

2008 Final Year Project Report by Alexander C. Kendon Supervisor: Dr Wade Enright

Abstract Ion thrusters are propulsion systems that utilise the forces that are exerted on charged particles by electric and magnetic fields. Current designs are not well suited to use in-atmosphere because the fuel cannot be readily ionized at atmospheric pressures. This project focuses on the development of an ion thruster that can be operated in air at normal atmospheric pressure. In order to achieve this, the lightning arc drawing concept was to be adapted so that a strongly ionized air path could be created and excited into a plasma channel by a high energy electrical discharge. Initial tests focused on the development of a system that could plasmerise a corona ring by induction, but this resulted in major damage to the arc drawing electrodes and little or no energy was transmitted to the corona. Attempts were then made to use a corona generated between two sheets of NMN insulation by electrodes on either side to form a plasma channel. Initial attempts failed, so a study of the effect of corona excitation voltage on breakdown strength was performed. These tests showed that this method could not be made to work with the voltages available in the lab. In the absence of any method for creating the breakdown path needed in air a thruster design was tested using a plasma path generated by exploding wire. This design was able to direct some of the plasma cloud and shows promise for further development. It is likely that the use of an arc drawing which creates streamers along the desired path will allow a plasma channel to be created and further development will likely produce a successful result.

2

1.

INTRODUCTION

Ion thrusters are an emerging technology that utilizes the reaction of charged particles to electric and magnetic fields to create thrust. Current designs use various different configurations that take advantage of Electrostatic, Lorentz and Hall Effect forces to repel ionized gasses from the thruster body [1]. These thrusters are designed exclusively for use outside the atmosphere, where fuel is easily ionized due to the fact they are operating in a near vacuum. They require special fuels, such as ammonia or lithium, which can be easily excited. The force generated by these units is also extremely small, with perhaps only a few micrograms of fuel ejected per-second. For space vessels a lack of power is often acceptable because of the excellent efficiency this technology provides [2]. The goal of this project is the development of an ion thruster that can produce meaningful thrust in atmosphere at normal pressures. For this to be achieved a shaped ionized path must be created in the air so that high current pulses may be delivered through the air. These pulses will plasmerise the air and allow it to be ejected from the thruster electrically. The lightning arc drawing concept developed at Canterbury University is ideally suited to this application and investigations will focus on its adaptation to propulsion systems. 2.

THEORY OF OPERATION

2.1 Concept of the In-Atmosphere Ion Thruster In order to overcome the limitations of current ion thruster technologies in-atmosphere a method of ionizing the air to a degree where significant force can be applied to it electrically must be found. This project combines the lightning arc drawing concept with plasma coils to achieve this. Ionized air created by a lightning arc drawing is subjected to a powerful magnetic pulse created by a plasma coil, which exerts both magnetic and, to a lesser extent, Lorentz forces on the air molecules. Figure 2.1 illustrates this concept.

Figure 2.1 – Illustration of a basic in-atmosphere ion thruster In the ultimate realization of the thruster, the plasma coil will be replaced with a conducting plane and the ionized air will be used as the plasma coil. When current begins to flow in the ionized air it plasmerises and becomes highly conductive. The resulting current pulse induces and opposing current in the conducting plane and the interaction of the magnetic fields created by these two currents creates a force repelling the plasmerised air from the conductor [3]. This is illustrated in fig. 2.2.

Figure 2.2 – Illustration of conducting ion path thruster 3

Under ideal conditions, the force on the plasma conductor is given by F = ( B × I ) Nπd (1) Where B is the magnetic field due to the current induced in the conductor, I is the current in the plasma coil, N is the number of turns of the plasma coil and d is the diameter of the plasma coil. In a real-world implementation the force would be affected by factors such as the degree of coupling between the coil and the conducting plane and electrostatic forces, but formula (1) remains useful for calculation of the direction of force. Because both B and I are vectors, the direction of the lines of magnetic flux must be known before the direction of thrust can be calculated. This can be difficult to do accurately without an analysis tool such as Mag-Net, but effective approximations can be performed using a generalized flux-pattern such as that shown in fig. 2.3.

Figure 2.3 – Approximation forces on plasma conductors when flux vector is considered 2.2 The Arc Drawing The lightning arc drawing concept takes advantage of the extremely intense electric fields that are generated around electrodes placed on opposite surfaces of an insulating sheet to ionize the air and create streamers that track across the surface of the sheet. Figure 2.4 shows a cross-section of a typical arc drawing with the applied voltage being sufficiently high to create streamers in the air.

Figure 2.4 – Basic arc drawing with HV supply sufficiently high to produce corona and streamers The strength of the electric field in the air around the electrodes is extremely high because the relative permittivity of the insulation reduces the field strength in the insulation relative to the air, as dictated by the formula Eair =

xair

Vapplied + ( ε air / ε insulation ) xinsulation 4

(2)

At low voltages the electric field around the electrodes has no effect, but as the voltage of the HV supply is increased the air begins to ionize. The rate of ionization is proportional to Townsend’s first ionization coefficient (α), which is equal to

α=

pσ i − ( σ i / kT )( E / p ) e kT

(3)

Where p is the gas pressure, σi is the cross section for ionization, k is Boltzmann’s constant, T is the temperature and E is the electric field strength. Air typically ionizes when E/p is in the range of 100-800 [4], which shows that at low pressure ionization occurs at much lower electric field strengths than at higher pressures. When the ionizing voltage is increased even further, streamers begin to track across the surface of the insulator. These streamers are channels of air that are ionized to a much greater degree than in a corona. They form when clusters of charge carriers (predominantly electrons) gather on the surface of the electrodes in such concentrations that they modify the external electric field, as shown in fig. 2.5 [4]. This causes the charge carriers to be forced away from the electrode, with the electrons at the head because they are more mobile, in what is known as an avalanche. When the charge at the head of the avalanche is causes the space charge field intensity, Er = 5.7 × 10 − 7

αeαx x/ p

(4)

where x is the distance travelled by the head of the streamer, to approach that of the external field a directed streamer forms and continues towards the other electrode until it is arrested by recombination. For this reason streamers form predominantly on surfaces of the arc drawing insulation that have an electrode on the other side. The length of the streamer, xc, can be approximated by

αxc = 17.7 + ln xc

(5)

Secondary avalanches can also be formed due to photo-ionization of t the air by the intense UV light that is generated as charge carriers in the streamer recombine [4]. This is one of the reasons that streamers appear “forked”.

Figure 2.5 – Shape of the space charge region around the head of a streamer, where E is the applied electric field, Er is the intensity of the space charge field and xc is the length of the streamer Once the air has become sufficiently ionized, it requires some leakage current to trigger the final avalanche that plasmerises the air and allows a power arc to form. In normal sparking this current is supplied by the electrodes that provided the external electric field, but no current can flow between the arc drawing electrodes because they are isolated from each other by a strong dielectric. Instead, the current is supplied by the plasma coil power supply, which operates at a higher voltage and has a much higher current capacity than the arc drawing power supply. The pre-ionized path created by the arc drawing should allow the plasma coil power supply to generate an arc that is much longer than normally possible and can be shaped by design, but because the streamers are not forming directly between the plasma coil electrodes careful design is necessary to shape them into the desired path. This is the major area of investigation for the project. 5

3.

EXPERIMENTS AND DISCUSSION

3.1 Experiment 1 3.1.1 Aim and Experiment Details The aim of this experiment was to apply the theory behind the PIT thruster [5] used by NASA to a lightning arc drawing. Figure 3.1 shows a schematic of the test apparatus, which consisted of a bucket with a ring shaped arc drawing inside it and an exploding wire plasma coil below it. The magnetic field created by the plasma coil should theoretically have induced a large current in the corona ring created by the arc drawing and repelled it upwards. The power supply for the arc drawing, shown in fig. 3.2, consisted of a neon sign transformer connected to an inverter and battery. It was connected in this fashion so that it would be isolated from ground and thus protect the neon sign transformer from the high voltage on the capacitor bank. A sphere gap set to trigger just above the neon sign transformer’s peak output voltage [6] was also placed across the arc drawing supply output for protection. More detailed information about the experiment is located in the appendix

Figure 3.1 – Schematic of experiment 1 test apparatus

Figure 3.2 – Schematic of isolated lightning arc drawing power supply 6

Figure 3.3 – Experiment 1 test apparatus in the Canterbury University HV lab 3.1.2 Results When the drop switch was triggered the plasma coil exploded violently and the bucket containing the arc drawing was thrown into the air. Inspection of the plasma coil Nomex insulation showed that the copper wire had undergone complete plasmerisation. Copper deposits on the NMN placed between the plasma coil and the bucket confirmed this. Both of the arc drawing electrodes, shown in fig. 3.4, were destroyed and had a ring of carbon deposit around them. A photo of the bucket, shown in fig. 3.5, taken during the test showed that there some small particles leaving the bucket before it was thrown in the air. These particles were most likely fragments of the aluminium that became charged and were ejected by electrostatic forces.

Figure 3.4 – Arc drawing bottom (left) and top (right) electrodes after experiment 1 7

Figure 3.5 – Photo of the bucket showing the arc drawing during the explosion (left) and inverted close up detail of particles being ejected (right) An oscilloscope was also connected to the input of the plasma coil. The trace, shown in figure 3.6, shows the voltage quickly rising as the switch was dropped, then holding steady as current flows in the plasma coil and finally dropping off as the capacitor bank discharges. If the trace is approximated to a square pulse, and with knowledge of the charge stored in the capacitors, it was possible to approximate the current in the coil to 13.7 kA.

Figure 3.6 – Voltage waveform captured at the input to the plasma coil The fact that the bucket was thrown in the air suggested that the foil electrodes of the arc drawing, which were physically attached to the bucket, may have experienced an upward force. This would imply that if any current had been induced in the corona ring it would have also have been thrown upwards. The considerable damage to the electrodes, however, made this configuration impractical for generating sustained thrust. The fact that large currents circulated in the electrodes also suggests that most of the energy from the plasma coil was coupled into them, rather than the corona ring. 8

3.2 Experiment 2 “Bucket of Doom II” 3.2.1 Aim and Experiment Details In order to determine what effects the corona ring in the previous experiment had on the results the test was replicated, but this time with the arc drawing electrodes disconnected from its supply. The only other detail of the experiment that was changed was the addition of a sheet of white paper below the plasma coil to see if copper plasma was also ejected downwards. Comparison of the condition of the arc drawing, movement of the bucket and plasma coil voltages to the previous experiment should help to identify how the ionized air in the bucket had behaved. It was also hoped that aluminium fragments would again be seen leaving the bucket. Also, if the arc drawing electrodes were again destroyed this would confirm that the plasma coil was strongly coupled to them, which could prove to be useful in other applications. 3.2.2 Results When the switch was dropped the plasma coil exploded and the bucket was thrown in the air in the same way as experiment 1. Inspection of the plasma coil insulation showed that it had again undergone complete plasmerisation. Very little copper was deposited on the paper under the plasma coil compared to the insulation on top of it (fig. 3.7), which shows that most of the plasma was firing upwards and breaking through the top of the Nomex insulation. The arc drawing electrodes, shown in figure 3.8, suffered similar damage to the pervious ones, but this time there were no carbon deposits around the electrodes. This suggests that the corona ring did have an effect and probably had a current induced in it by the plasma coil, although current could also have been transferred by conduction from the aluminium electrodes. The photo of the explosion showed an even greater number of aluminium particles being ejected from the bucket, which could also be seen in one frame of the video recording of the experiment (fig. 3.9). Some aluminium flakes can also be seen stuck to the adhesive strips on the bottom of the arc drawing, which confirms that they were what was causing the streaks in the photos. The larger number of particles seen could indicate that the ionized air in the bucket had prevented some from being ejected in the previous experiment.

Figure 3.7 – The carbon and copper deposited on paper placed below the plasma coil (left) and insulation placed on top of it (right) show that the copper plasma was ejected upwards

9

Figure 3.8 – Arc drawing bottom (left) and top (right) after experiment 2

Figure 3.9 – Negative image of the bucket (left) and a frame of video (right) showing a large number of aluminium fragments being ejected during the explosion 3.3 Experiment 3 “Bucket of Doom Control” In order to determine what forces were causing the bucket to fly into the air the previous experiment was repeated without anything inside the bucket. This resulted in the bucket being thrown into the air in the same way as it was previously, but this time the bottom was completely blasted out. This may have been because the arc drawing had reinforced the bucket in pervious experiments, or because it was already weakened, is not likely to have been due to a change in the forces acting on it. This result shows that the bucket was being thrown in the air by the force of the copper wire in the plasma coil exploding, which invalidates any speculation about thrust being created within the bucket. Clearly, another approach to the investigation was necessary. 3.4 Experiment 4 “Effect of a Plasma Coil on Aluminium Sheets” This experiment was intended to give an indication of the shape of the magnetic field around the plasma coil. It was theorised that the area of aluminium foil that suffered the most damage would be the best place for a corona ring to be generated, and that the area of least damage would be the best place for arc drawing electrodes to be placed. The apparatus consisted of four layers of aluminium foil attached to a paper backing for strength. These sheets were suspended between four Acetyl plastic rods with 15mm spacing between the sheets. A plasma coil identical to the ones used in pervious experiments was placed 15mm below the bottom sheet as shown in fig. 3.9. 10

Figure 3.9 – Schematic of the test apparatus used to determine the effect of a plasma coil on aluminium sheets When the plasma coil was fired the capacitor bank, which was charged to 40kV, only discharged to 25kV. It also had almost no effect on the aluminium sheets. A ring approximately the same size as the plasma coil was indented into the bottom sheet and the other sheets were unaffected. The reason for this result is that most of the energy from the plasma coil was coupled into the aluminium foil, which did not leave enough energy to sustain the conductive plasma path. The indentation in the foil was created by the magnetic field created by the massive currents that would have flowed in the foil sheet. While the results did not yield the desired information, they did show that large conductive planes can “snuff out” a plasma path created by exploding wire, which could prove to be important in the design of arc signs that will be used around plasma paths. 3.5 Experiment 5 “Corona Path Excitation” 3.5.1 Aim and Experiment Details Because of the damage to the arc drawing electrodes in previous experiments a method for creating an ionized channel electrically isolated from the electrodes was required. This would allow the capacitor bank to be discharged through ionized channel without damaging the ionizing electrodes or placing the capacitor potential across the ionizing supply. Several different electrode/insulation configurations were trialled and the one that produced the energetic corona was selected for the main experiment. A 300mm corona path was created between two sheets of NMN insulation with ionizing electrodes on opposite sides. Electrodes connected to the capacitor bank, which was charged to 20 kV, were then inserted into each end of the corona path. Protection resistors were placed in series with the ionizing power supply, as shown in figure 3.10. Their addition did not affect the corona formation as the full voltage is still presented across the open-circuited output.

Figure 3.10 – Schematic of the test apparatus used to test the suitability of a corona path as a plasma conductor

11

Figure 3.11 – Completed corona path excitation test setup (left) and the corona path itself (right) 3.5.2 Results When the drop switch was triggered the test apparatus exploded. Upon inspection it was found that an arc had tracked around from one of the capacitor bank electrodes to the ionizing electrodes and caused it to explode. The experiment was repeated with larger insulation clearances around the ionizing electrodes. This test did not produce a result, so the distance between the capacitor bank electrodes was reduced. Even with the corona path length reduced to 100mm no plasma path could be created. 3.6 Experiment 6 “Corona Path Characterisation” 3.6.1 Aim and Experiment Details After the failure of the previous experiment a series of tests was devised to investigate the effect of electrode gap and corona excitation voltage on the breakdown strength of the corona path. The tests may result in a set or workable conditions or reveal a trend that could allow one to be derived. A corona path with open sides was constructed in a similar way to the last experiment. Rod-type electrodes were fitted to a slider and inserted into the corona path. Each set of electrodes was then connected to a neon sign transformer which was, in turn, connected to a variac.

Figure 3.12 – Schematic diagram of the corona characterisation test setup 12

Figure 3.13 – Completed test setup for corona path characterisation 3.6.2 Results The rod gap was adjusted in 2.5mm increments. For each gap the corona excitation was increased from zero to 15kV in 10% steps. The NMN insulation between the rod gap and the ionizing electrodes broke down during the 10mm gap test. The results are shown in fig. 3.14. The test setup was rebuilt using two layers of NMN insulation on each side and plate electrodes replacing the rods. The results from this test are shown in fig. 3.15. High voltage diodes were then placed in series with both supplies to achieve half wave rectification. Tests were performed with two (fig. 3.16) and then one (fig. 3.17). In all cases the breakdown voltage eventually falls at the highest excitation voltages, with the fall-off becoming more pronounced as the gap is increased. Increasing the gap increases the breakdown voltage but the roll-off is also steeper, which indicates that at very high excitations the breakdown voltage may converge for all caps. This is the point where the path is on the edge of breakdown and would be ideal for the creation of a plasma channel. Such a state was not realisable with the equipment available during testing. Interestingly, for larger gaps the breakdown voltage actually rose for lower corona excitations. The effect was likely due to the fact that the electric field was actually holding charge carriers in place, rather than freeing them.

13

Corona Breakdown Voltages 10mm corona gap, 1 layer NMN insulation 9 2.5mm 5mm

8

7.5mm 10mm

Breakdown Voltage (kV)

7

6

5

4

3

2

1

0 0

2

4

6

8

10

12

14

16

Corona Excitation (kV)

Figure 3.14 – Test results for the corona characterisation test using a rod gap Corona Breakdown Voltages 5mm corona gap, 2 layer NMN insulation 16 5mm 10mm 14

15mm 20mm 25mm

Breakdown Voltage (kV)

12

10

8

6

4

2

0 0

2

4

6

8

10

12

14

16

Corona Excitation (kV)

Figure 3.15 – Test results for the corona characterisation test using plate electrodes

14

Corona Breakdown Voltages 1/2 wave rectification, 5mm corona gap, 2 layer NMN insulation 16

5mm

14

10mm 15mm

Breakdown Voltage (kV)

12

20mm

10

8

6

4

2

0 0

2

4

6

8

10

12

14

16

Corona Excitation (kV)

Figure 3.16 – Test results for the corona characterisation test with half wave rectification Corona Breakdown Voltages 1/2 wave rectification, 5mm corona gap, 1 layer NMN insulation 16 5mm 10mm

14

15mm 20mm

Breakdown Voltage (kV)

12

10

8

6

4

2

0 0

2

4

6

8

10

12

14

16

Corona Excitation (kV)

Figure 3.17 – Corona breakdown test results for half wave rectification and reduced insulation

15

3.7 Experiment 7 “Exploding Wire Thruster with Aluminium Core” In the absence of a method for creation of an air plasma channel one possible configuration for the ion thruster was trialled using an exploding wire to create the plasma path. The apparatus consisted of two turns of copper wire wound on an aluminium tube with two layers of NMN insulation, as shown in fig. 3.18. The residue left on the end-plates and the core shown in fig. 3.49. During the test the plasma is mostly forced away from the core radially, but at the left end the plasma cloud is forced onto the end plate and at the right end it is pulled away. The greater amount of residue on the left end plate compared to the right indicates the same thing. This is because at the ends of the core the lines of magnetic flux are more vertical than along the length. As with experiment 4 the conducting plane coupled a large amount of energy out of the plasma path and the capacitor bank failed to discharge fully. Use of a higher resistance core material may prevent this.

Figure 3.18 – Completed axial ion thruster test apparatus (left) and photo of the test (right)

Figure 3.19 – Residue patterns left on the axial thruster core and end-plates after the experiment (dark residue on the left of the right hand end-plate is not from wire wound on the core)

16

4.

CONCLUSIONS

The tests performed throughout the course of the project have proven that a corona path cannot be excited either by induction or conduction. The final test, however, did indicate that a force can be exerted on a plasma conductor when it is formed around an aluminium tube. For this reason further work towards the realisation of ionized air plasma paths could be performed. The results of the corona characterisation tests suggest that a high degree of ionization is required for air to be used as a plasma conductor. When the theory of arc formation discussed in section 2 is considered, it seems likely that a plasma path can be formed, but that this path must be very close to breakdown. This can most likely be achieved by using an arc drawing to create streamers along the desired channel. The electrodes would need to be constructed from a high resistance material, or have a high resistance in series with them. The use of protection resistors in some experiments has proven that this does not affect streamer formation. The earth plane for the arc drawing could also be used as the conduction plane for the ion thruster. It is recommended that different configurations of conducting planes and exploding wires be trialled until an air plasma path can be created. Further experimentation is highly likely to yield a successful result.

5.

REFERENCES

[1] “Ion Thruster”, http://en.wikipedia.org/wiki/Ion_thruster, Accessed 11 September 2008 [2] C. Lee Dailey and Ralph H. Lovberg, “The PIT MkV Pulsed Inductive Thruster,” NASA Contractor Report 191155, July 1993. [3] F. T. Ulaby “Fundamentals of Applied Electromagnetics”, Prentice Hall, 2007 [4] E. Kuffel and W.S. Zaengl, “High Voltage Engineering, Fundamentals”, Newnes, 1984 [5] C. Lee Dailey and Ralph H. Lovberg, “The PIT MkV Pulsed Inductive Thruster,” NASA Contractor Report 191155, July 1993. [6] “Voltage Measurement: Sphere-Gap Method,” AS2886, 1996

17

APPENDIX: EXPERIMENT DETAILS A1 Experiment 1 “Bucket of Doom” Table A1.1 – Details of Experiment 1 Setup Date Ambient Temp

5/05/2008 18 °C

Plasma Coil

Arc Drawing

Core Diameter 32 mm Insulator Type NMN Turns 25 Electrode material Aluminium Foil Wire Diameter 0.355 mm Electrode Ring Diameter 75 mm Insulation Type Nomex Paper Electrode Ring Width 10 mm Insulation Clearance 30 mm Input Voltage 15 kV rms Coil Top Insulation 2 Layers NMN Safety Sphere Gap 8 mm

HV DC Supply Bank Capacitance Capacitor Voltage

17.1 uF 40 kV dc

Figure A1.1 – Dimensions of the arc drawing used in experiment 1

Figure A1.2 – Dimensions of the exploding wire plasma coil used in experiment 1

18

A2 Experiment 2 “Bucket of Doom II” Table A2.1 – Details of Experiment 2 Setup Date Ambient Temp

9/05/2008 18.5 °C

Plasma Coil

Arc Drawing

HV DC Supply

Core Diameter 32 mm Insulator Type NMN Turns 25 Electrode material Aluminium Foil Wire Diameter 0.355 mm Electrode Ring Diameter 75 mm Insulation Type Nomex Paper Electrode Ring Width 10 mm Insulation Clearance 30 mm Input Voltage None Coil Top Insulation 2 Layers NMN Safety Sphere Gap N/A

Bank Capacitance Capacitor Voltage

17.1 uF 40 kV dc

A3 Experiment 3 “Bucket of Doom Control” Table A3.1 – Details of Experiment 3 Setup Date Ambient Temp

14/05/2008 18 °C

Plasma Coil

HV DC Supply

Core Diameter 32 mm Turns 25 Wire Diameter 0.355 mm Insulation Type Nomex Paper Insulation Clearance 30 mm Coil Top Insulation 2 Layers NMN

Bank Capacitance Capacitor Voltage

17.1 uF 40 kV dc

A4 Experiment 4 “Effect of a Plasma Coil on Aluminium Sheets” Table A4.1 – Details of Experiment 4 Setup Date Ambient Temp

23/05/2008 18 °C

Plasma Coil Core Diameter 32 mm Turns 25 Wire Diameter 0.355 mm Insulation Type Nomex Paper Insulation Clearance 30 mm Coil Top Insulation 2 Layers NMN

Foil Sheets

HV DC Supply

Substrate Paper Bank Capacitance Electrode material Aluminium Foil Capacitor Voltage Electrode Spacing 15 mm Voltage After Firing Electrode size 265x265 mm

17.1 uF 40 kV dc 25 kV dc

Figure A4.1 – Test setup for investigation of the effects of a plasma coil on aluminium sheets A5 Experiment 5 “Corona Path Excitation” 19

Table A5.1 – Details of Experiment 5 Initial Setup Date Ambient Temp

19/07/2008 18 °C

Corona Path

HV DC Supply

Insulator Type NMN Electrode material Aluminium Foil Electrode Length 300 mm Electrode Width 25 mm Corona Path Height 3 mm Input Voltage 15 kV Safety Sphere Gap 8 mm Insulation Clearance 100 mm

Bank Capacitance Capacitor Voltage

17.1 uF 20 kV dc

Table A5.1 – Details of Experiment 5 Modified Setup Date Ambient Temp

31/07/2008 17.5 °C

Corona Path

HV DC Supply

Insulator Type NMN Electrode material Aluminium Foil Electrode Length 100 mm Electrode Width 25 mm Corona Path Height 3 mm Input Voltage 15 kV Safety Sphere Gap 8 mm Insulation Clearance 200 mm

Bank Capacitance Capacitor Voltage

17.1 uF 20 kV dc

A6 Experiment 6 “Corona Path Characterisation” Table 6.1 – Details of Experiment 6 Test 1 Date Ambient Temp

04/008/2008 15 °C

Corona Path

Spark Gap

Electrode Width 10 mm Electrode Type Electrode Length 300 mm Electrode Material Electrode Material Aluminium Foil Electrode Diameter Insulation Type NMN Max Excitation Voltage Insulation Clearance 100 mm Corona Path Height 10 mm Max Excitation Voltage 15 kV

Rod Copper 4 mm 15 kV

Table 6.1 – Details of Experiment 6 Test 2 Date Ambient Temp

05/008/2008 18.5 °C

Corona Path

Spark Gap

Electrode Width 30 mm Electrode Type Electrode Length 100 mm Electrode Material Electrode Material Aluminium Foil Electrode Dimensions Insulation Type 2 layers of NMN Insulation Clearance 100 mm Corona Path Height 5 mm Max Excitation Voltage Max Excitation Voltage 15 kV

Table 6.1 – Details of Experiment 6 Test 3 20

Plate Copper 30 mm W 80 mm L 1 mm H 15 kV

Date Ambient Temp

08/008/2008 18 °C

Corona Path

Spark Gap

Electrode Width 30 mm Electrode Type Electrode Length 100 mm Electrode Material Electrode Material Aluminium Foil Electrode Dimensions Insulation Type 2 layers of NMN Insulation Clearance 100 mm Corona Path Height 5 mm Max Excitation Voltage Max Excitation Voltage 20 kV dc pk

Plate Copper 30 mm W 80 mm L 1 mm H 20 kV dc pk

Table 6.1 – Details of Experiment 6 Test 4 Date Ambient Temp

08/008/2008 18 °C

Corona Path

Spark Gap

Electrode Width 30 mm Electrode Type Electrode Length 100 mm Electrode Material Electrode Material Aluminium Foil Electrode Dimensions Insulation Type 1 layer of NMN Insulation Clearance 100 mm Corona Path Height 5 mm Max Excitation Voltage Max Excitation Voltage 20 kV dc pk

Plate Copper 30 mm W 80 mm L 1 mm H 20 kV dc pk

A7 Experiment 7 “Exploding Wire Thruster with Aluminium Core” Table A7.1 – Details of Experiment 7 Setup Date Ambient Temp

5/09/2008 18.5 °C

Thruster Coil

HV DC Supply

Wire Diameter 0.355 mm Bank Capacitance Number of Turns 2 Capacitor Voltage Pitch 45 mm Voltage After Firing Insulation 2 layers of NMN Core Material Aluminium Core Thickness 2 mm Core Diameter 75 mm Core Length 200 mm

21

17.1 uF 35 kV dc 27 kV dc