"A Field Generation Device" THIS INVENTION relates to a field generation device, and in particular seeks to provide an improved field generation device.

Accordingly, one aspect of the present invention provides a field generation device, comprising a field generation arrangement operable to: generate a first field having a first field angular momentum and a first field mass associated therewith, the first field mass having a centre of mass at a first point relative to the device, such that the device tends to rotate in a direction substantially opposite to that of the first field angular momentum; and generate a second field having a second field angular momentum and a second field mass associated therewith, the second field mass having a centre of mass at a second point relative to the device, such that the device tends to rotate in a direction substantially opposite to that of the second field angular momentum, wherein the first and second points are displaced from one another.

Advantageously, the field generation arrangement comprises a first field generator, operable to generate the first field, and a second field generator, operable to generate the second field.

Preferably, the first field generator comprises at least one first field generation solenoid, and the second field generator comprises at least one second field generation solenoid.

Conveniently, the field generation arrangement is operable to generate the first and second fields alternately.

Advantageously, the first and second fields are generated alternately at a rate of at least one cycle per second.

Preferably, the device comprises one or more reaction elements, to cause rotation of the craft upon generation of the first and second fields.

Conveniently, at least one of the reaction elements comprises an electret element.

Advantageously, the electret element comprises a simulated electret comprising an electric field maintained between a pair of electrodes.

Preferably, a plurality of substantially parallel plates are provided between the electrodes.

Conveniently, a neutral electrode is provided in proximity to one of the pair of electrodes.

Advantageously, the device comprises at least one first reaction element, to cause rotation of the craft upon generation of the first field, and at least one second reaction element, to cause rotation of the craft upon generation of the second field.

Preferably, the device further comprises a field cancellation arrangement, operable to cancel at least partially effects of switching off the first field and the second field.

Conveniently, the field cancellation arrangement comprises a first field cancellor, for cancelling at least partially effects of switching off the first field, and a second field cancellor, for cancelling at least partially effects of switching off the second field.

Advantageously, the first field cancellor comprises at least one first cancellation field generator, operable to generate a first cancellation field having a first cancellation field mass associated therewith, and at least one second cancellation field generator, operable to generate a second cancellation field having a second cancellation field mass associated therewith.

Preferably, at least one of the first and second cancellation field generators comprises at least one solenoid.

Conveniently, at least one of the first cancellation field mass and the second cancellation field mass has a centre of mass located within the device.

Advantageously, both the first cancellation field mass and the second cancellation field mass have a centre of mass located within the device.

Alternatively, at least one of the first cancellation field mass and the second cancellation field mass has a centre of mass which is external to the device.

Preferably, both the first cancellation field mass and the second cancellation field mass have a centre of mass which is external to the device.

Conveniently, the first cancellation field generator is operable to be switched on during the time when the first field is switched from being on to being off.

Advantageously, the first cancellation field generator is operable to be switched on before the first field is switched off, and to be switched off after the first field has been switched off.

Preferably, the second cancellation field generator is operable to be switched on during the time when the second field is switched from being on to being off.

Conveniently, the second cancellation field generator is operable to be switched on before the second field is switched off, and to be switched off after the second field has been switched off.

Advantageously, the first and second points are on substantially opposite sides of the centre of mass of the device.

Preferably, the first and second points are substantially the same distance from the centre of mass of the device.

Conveniently, the first and second points lie outside the device.

Advantageously, the first and second field angular momenta are in opposite directions.

Preferably, the device has a rest centre of mass, being the centre of mass of the device when neither the first nor the second fields is being generated ;

generation of the first field results in rotation of the device around a point which is substantially between the centre of mass of the first field mass and the rest centre of mass of the device ; and generation of the second field results in rotation of the device around a point which is substantially between the centre of mass of the second field mass and the rest centre of mass of the device.

Another aspect of the present invention provides a means of transportation having a device according to the above.

Conveniently, the device is provided within a cavity in the means of transportation.

Advantageously, at least part of a gap between the device and an internal surface of the cavity acts as a resonant cavity for at least one of the fields.

Preferably, the device is coupled to the means of transportation by at least one substantially frictionless coupling.

Conveniently, at least one of the substantially frictionless couplings comprises an electromagnetic coupling.

Advantageously, the coupling between the means of transportation and the device is activated periodically.

Preferably, the first and second fields are generated alternately at a rate which is substantially greater than the rate of activation of the coupling means.

Another aspect of the present invention provides a method of generating fields, comprising the steps of : providing a field generation device ; generating a

first field having a first field angular momentum and a first field mass associated therewith, the first field mass having a centre of mass at a first point relative to the device, such that the device tends to rotate in a direction substantially opposite to that of the first field angular momentum; and generating a second field having a second field angular momentum and a second field mass associated therewith, the second field mass having a centre of mass at a second point relative to the device, such that the device tends to rotate in a direction substantially opposite to that of the second field angular momentum, wherein the first and second points are displaced from one another.

Conveniently, the first and second fields are generated alternately.

Advantageously, the first and second fields are generated alternately at a rate of at least one cycle per second.

Preferably, the method further comprises the step of providing one or more reaction elements, to cause rotation of the craft upon generation of the first and second fields.

Conveniently, the method further comprises the step of providing a field cancellation arrangement, operable to cancel at least partially effects of switching off the first field and the second field.

Advantageously, the method further comprises the step of generating a first cancellation field having a first cancellation field mass associated therewith, and generating a second cancellation field having a second cancellation field mass associated therewith.

Preferably, at least one of the first cancellation field mass and the second cancellation field mass has a centre of mass located within the device.

Conveniently, both the first cancellation field mass and the second cancellation field mass have a centre of mass located within the device.

Alternatively, at least one of the first cancellation field mass and the second cancellation field mass has a centre of mass which is external to the device.

Advantageously, both the first cancellation field mass and the second cancellation field mass have a centre of mass which is external to the device.

Preferably, the first and second points are on substantially opposite sides of the centre of mass of the device.

Conveniently, the first and second points are substantially the same distance from the centre of mass of the device.

Advantageously, the first and second points lie outside the device.

Preferably, the first and second field angular momenta are in opposite directions.

In order that the present invention may be more readily understood, embodiments thereof will now be described, by way of example, with reference to the accompanying drawings, in which:

Figure 1 is a schematic representation of Feynman's disk; Figure 2 is a distorted version of the Feynman disk; Figure 3 is a schematic representation of a device embodying the present invention; Figure 4 represents a superposition of electric fields during operation of the device of Figure 3; Figures 5a, 5b and 5c show forces acting on the device of Figure 3 during operation thereof; Figures 6a and 6b are further representations of a device embodying the present invention; Figure 7 is a view of a part of the device of Figure 6a; Figure 8 is a graph of a probability of photon-photon scattering; Figure 9 depicts a simulated electret element; and Figure 10 depicts a means of transport incorporating a device embodying the present invention.

The construction and operational principles of devices embodying the present invention will be described below.

Conservation of momentum represents at a fundamental level the isotropy and homogeneity of space (L. D. Landau and E. M. Lifshitz, Course of Theoretical Physics : Mechanics, Butterworth Heinemann, Vol. 1, 3d ed. 1980).

This is expressed in the Lagrangian formulation succinctly as: The Lagrangian for the system being the sum of kinetic and potential energy terms expressed in terms of the generalised co-ordinates. Application of Lagrangian method and Relativity to the combined system of mechanics and electromagnetics (L. D. Landau and E. M. Lifshitz, Course of Theoretical Physics : The Classical Theory of Fields, Butterworth Heinemann, Vol. 2, 4th ed. 1982) renders the field a system with infinitely many degrees of freedom.

However between the mechanical and electromagnetic parts of the field conservation of momentum still applies if we admit the field has momentum.

The force density on a particle density distribution is given by: p", # - #E - µ0J # H - ##T + #/#t#0µ0E # H = 0 eqn. 2 The final"Poynting term", the cross product of electrical and magnetic fields represents the field momentum. The effect of the field term is well known in the dynamic or radiative regime and is the pressure of radiation that insures the stability of our Sun or the deflection of comet trails.

Intriguingly, though, the Poynting term implies the existence of field momentum in static situations such as when a steady magnetic field impinges on an electret. The well known Feynman disk thought experiment (figure 1) illustrates this. The device 1 comprises a disk 2, which is formed from plastic or another insulating material, and is provided with a number of charged metal spheres 3 distributed near the rim thereof. The disk 2 is supported on an

elongated spindle 4 passing through the centre thereof, perpendicular to the plane of the disk 2 and the disk 2 is free to rotate about the axis of the spindle 4.

On one surface of the disk 2, a coil 5 of conducting wire is placed around the spindle 4, and a battery 6 or other current source drives a current around the coil 5.

If the current flowing through the coil is suddenly switched off, a tangential electric field acts around the coil 5 and this generates a torque around the spindle 4 when acting on the metal spheres 3. The disk undergoes angular translation even in vacuo. An experimental account of this is given by G.

Graham and D. G. Lahoz"Observation of static electromagnetic angular momentum in vacuo", Nature 285,154 (1980) amongst others.

Various schemes to utilise the momentum from static fields and generate simple linear momentum have been dispelled. A full treatment can be found in V. Hnizdo"Hidden momentum of a relativistic fluid carrying current in an external field", Am. J. Phys. 65,92 (1997). They are prohibited by a relativistic effect by the static electric field on the charge carriers of the solenoid that give them exactly equal and opposite momentum to the field momentum.

However angular rotation around the field always results and this exposition shows a scheme to convert that to linear translation. Consider figure 2 in which a Feynman disk has been distorted. It is a generalisation of figure 1 in which the solenoid 5 is not just off-centre but tilted too. This device 7 comprises a part disk 8, which is formed from light strong plastic or another insulating material, and is provided with a number of charged metal spheres 3 distributed near the rim thereof. The disk 8 is for the purposes of argument

freely floating in space. The solenoid 5 tilted over from the perpendicular to the disk projects a field outside of the disk 8 whose centre is at 9.

On field transitions the relevant electric field circles around point 9 and once again acts tangentially on the charged metal spheres 3.

The device will undergo a rotation about the centre of mass of the device 7 and the centre of mass of the external field whose centre is at 9. Also the centre of rotation can be made to occur around a point that is located outside the device. Note that the electric field generated by the"beam"of the solenoid as it sweeps into new space only has an effect on charges moving relative to it and so will have no effect on the metal spheres 3.

Figure 3 shows a simplified arrangement of a distorted Feynman disk for analysis, in which the device comprises two balls 10 connected by a stiff rod 11. The field is projected outside of the device. The centre of mass of the system is shown shifted from its quiescent position equidistant between the balls 10, and lies substantially between the quiescent centre of mass and the centre of mass of the mass associated with the field. When switching on the field the device is shifted to the right (in this depiction) and the field to the left.

In an extreme case of substantial electromagnetic mass being projected outside the device, this combined centre of mass can even be outside of the device.

Shown too is the Poynting flow integrated over space and centred as a mass MB,, E though we neglect to show it in subsequent analysis in comparison to other masses.

The action of the electrical fields from the changing magnetic field produces torque on an electret element 12 which is shown situated at the leftmost baton mass by the circle around it ; arrows indicate the direction of rotation. Note that the Poynting field rotates in countersense to the device and its field arrangements, thus the sum of the angular momentum of the whole is zero: eqn. 3 L(m1) + L(m2) + L(mf) - L(mBxE) = ##BxEdt<BR> <BR> <BR> <BR> <BR> T At the end of the cycle the device returns to the system centre of mass equidistant between the two balls, since the electromagnetic energy is called back to the device. The Poynting flow ceases too and acts on the electret element 12 and the craft leaving it with zero angular momentum but the device has undergone an angular translation.

The aim of the device is to render the integral on the right-hand-side of equation 3 non-zero over a cycle ("decoupling"the electrical aspects of the system from the mechanical) leaving both the device and field with angular momentum at the end of the cycle-in short,'dumping'excess momentum on the field and its zero-point quantum basis in a mechanism postulated later. First we shall delve into the necessary conditions to make the time average of the Poynting force non-zero.

Consider this for a plane electromagnetic wave: B=B", sin (wt) and E=Emsin (ut). Thus <p> will be BmEm/2 integrated over the volume and so an electromagnetic wave transfers momentum and this figure is independent of frequency. In our propulsion case the electric field is fixed to that of the electret's field, thus the integral is:

If we can cancel the second half of the cycle we will achieve a force proportional to the cycling frequency, otherwise the net force is zero. Put another way, since the force is the time derivative of the momentum this follows: Thus for a simple cyclical process the average force is zero. A way around this is to cancel the E-field around the collapsing magnetic flux on the second half of the cycle-the force on the electret drive element 12 is due to the tangential electric field of the magnetic flux collapse. We cannot just cancel the E-field from the electret (discharging it, etc. ) because the force from the Poynting vector will be the differential of the product of the B and E fields- returning to the earlier point, any form of simple cycling of these fields will lead to net zero force.

Reference to figure 4 shows we have to cancel the electrical field imposed on the electret drive element 12 by switching on at least one cancellation solenoid to generate a field whose field (shown as a dotted line) centre is different to that of the field generated by the selenoid of the distorted Feynman disk at the same instant as the electric field generated by the solenoid of the distorted Feynman disk (solid line) goes negative. Figures Sa, 5b and 5c

resolve the torques into forces acting at mi m2 and mf since these masses are called back to the device at the end of the cycle. Shown are the forces on switching on the field and then switching off the field. The resultant shown is the summation of these forces. We can see the reason why the field centre for the cancellation solenoid needs to act on a different centre-the furthest mass, m2 experiences an asymmetry. Figures 5a, 5b and 5c depict the cancellation field's centre as so distant that the forces acting at ml and m2 are almost the same-in short the torque vector fields have different curvature for the thrust and cancellation solenoids.

Not depicted in the diagrams of Figures 5a, 5b and 5c is the momentum acting on mass mBx as this has been"decoupled"from the system, however its mass-energy (but not its momentum) is returned to the solenoids (and hence the device) by the Lenz/Faraday Law. The situation is similar to a projectile being launched from a large mass (the base): in the limit of the base having very large mass, most of the kinetic energy goes to the projectile but the large mass receives as much momentum as the projectile.

Thus let: Mb be the mass of the base and V its velocity mp be the mass of the projectile and v its velocity At the end of the process there is momentum balance: MbV = mpv At the end of the process the energy is partitioned: E = 2 mpv2 it is easy to show by substituting for V by the momentum equation that, the projectile takes the lion's share of the energy if Mb is large.

The diagrams of resultant forces in Figures 5a, 5b and 5c show that the craft is left with angular and linear momentum from the one-sided process.

Figures 6a and 6b depict a symmetrical device using two such processes: two sets of field generation solenoids 13, cancellation solenoids 14 and electret elements 12 either side of it, setting up rotation clockwise on one side and counter-clockwise on the other. The sense of the polarity of the electret or solenoid fields to achieve the inverse rotation is clear to one in the art-for example, the right-hand electret can be of opposite polarity to the left-hand whilst the solenoid fields are still cycled in the same way. The device then is left with only linear momentum.

Figure 7 shows a construction to point out that the hidden momentum argument used against electromagnetic propulsion is not valid here. We note that hidden momentum effects arise from electrical potentials being applied to a relativistic charge carrying fluid. The solenoids 13,14 (both field generation and cancellation) are shown tilted projecting a field. These solenoids are shielded from the electret element 12 by a conducting box 15 at the device's potential. Thus since no field impinges on the solenoids 13,14, the hidden momentum argument is irrelevant.

Also we see that electromagnetic energy is projected outwards from the device but this is not the net mechanism by which the device is propelled forwards. The energy is recalled to the device at the end of the cycle. The net momentum from this setting up and removal of the field is zero-the time average is zero (similar to equation 6). Sure enough the flow of energy can be represented by a Poynting vector of the propagation of the changing E and B fields but it is distinct from the propulsive effect of the changing E field on the electret element 12. One displays momentum transfer by acting on the solenoid 13 (the radiation field, see later), the other on the electret 12 (the induction field). Thus we have two momentum density terms on a half cycle: <BR> <BR> <BR> <BR> <BR> <BR> gSolenoid = #0#/#t<BR> au<BR> <BR> <BR> <BR> gElectret = Eo (B x E) For the first expression it is easier to divide the known final expression for the energy density by c2 than compute the time varying Poynting expression.

The two expressions are not equal and so clearly the argument that the propulsion comes from ejection of field energy of the solenoid is not valid.

Inevitably with cycling of electrical or magnetic fields electromagnetic radiation is produced but this is not the main mechanism by which the device is propelled-a photon rocket is a puny thing; the majority of the energy developed goes into the rearward beam and not the kinetic energy of the device.

In no way is it implied by the classic Feynman disk example that radiation emanates and provides all the momentum balance. In L. D. Landau and E. M.

Lifshitz, Course of Theoretical Physics : The Classical Theory of Fields, Butterworth Heinemann, Vol. 2, 4th ed. 1982 we see that radiation effects are put off to the second order in the vector potential and the force is of the order: The E field from the above expression is then E=- (l/c) 3A/9t. The force on charged entities constituting the solenoid current from the radiation field is thus of the order of l/c3 down on the forces generated on the electret element 12 by the induction fields:

#0##/#t (B # E)dV Ref. eqns. 2 and 4 In fact the radiation is so minuscule that we didn't even calculate it and can say with very high accuracy that all the useful energy transfer goes into the kinetic energy of the device, much as though we were pushing against a very large mass.

To stress the point again that we are dealing with two different effects, the momentum exchange from the radiation field and the force on the electret from the induction field, note that for a photon rocket momentum flux is independent of frequency, however for the device it is linear in frequency (equation 5).

Clearly by the action of the cancellation solenoids 14 the momentum of the Poynting flow is decoupled from the system. It cannot be recalled to the device by the magnetic field energy being recalled-the sense of the vectors is wrong. Also in the proceeding section it was remarked how small the radiative transfer of momentum is. We believe that this momentum is"dumped"on the ground state of the field, the so called zero point energy of the field. Consider the oscillator expansion of the electromagnetic field from the solution of the wave equation in the vector potential (summarised from L. D. Landau and E. M.

Lifshitz, Course of Theoretical Physics : Quantum Electrodynamics, Butterworth heinemann, Vol. 4, 2nd ed. 1982): #2 #2A

In a finite volume we can represent the solution in terms of a series of travelling plane waves: If all the vectors ak are defined, the field in the volume is completely determined. We can make a transformation to canonical variables suitable for Hamiltonian treatment: Leading to: We can substitute the vector potential by E=-d A/t and H=curlA into the Hamiltonian giving the total energy of the field, The above expression is directly analogous to the Hamiltonian for a one dimensional harmonic oscillator. The quantisation of the free electromagnetic field is now an easy step as we now make the canonical variables (generalised co-ordinates Q and generalised momenta P) operators subject to the commutation rule: #k#k - #k#k = The Hamiltonian operator for the field then is: Giving the energy levels by analogy to the harmonic oscillator as: Use of the classical Poynting expression for field momentum (equation 2) and replacement of E and H by operators leads to the following expression for quantised field momentum: In both expressions the summation is performed up to a maximum frequency cut-off of the Planck Frequency of order 1043Hz. This frequency is the point when the electromagnetic interaction merges with the weak force. We can see that both expressions contain the'/2 constant of the zero-point. Physical reality informs us that the constant is highly homogeneous and isotropic but it gives us a zero point mass of the order of 1093g/cm3, R. P. Feynman and A. R.

Hibbs, Quantum Mechanics and Path Integrals, McGraw-Hill, NY 1965).

It is well known and indeed experimentally proven by the Casimir effect that the zero-point field is real. Advanced calculations in Quantum Electrodynamics calculate the effect of"vacuum radiative corrections" resulting from the pair creation of electron-positron pairs out of the vacuum to normal classical electrodynamics. One of these processes is photon-photon scattering whereby a real photon and a virtual electron-positron pair undergo Compton Scattering.

What we postulate to the best of our knowledge now, for the loss of momentum at the field cancellation step for the present device, is that circulating flux of the Poynting field can be represented as an expansion and summation of low frequency plane waves representing the real photon flux of our field: Here cycle is proportional to the cycle time of the device. This real photon flux can be considered one system in juxtaposition to the virtual photon flux occupying the same space. The virtual photon flux series is however summed to the Planck Frequency. These two systems can behave like two isolated oscillators but they are forced to interact (or couple) by the photon scattering effect: the zero-point field creates virtual pairs off which the real photons scatter (Figure 8, cross-section versus energy). Overall in the absence of sources to sustain the real field, or real charges in the immediate vicinity to exchange momentum to (we cut-off and isolate in the field cancellation step) or the small"drainage"of momentum by electromagnetic radiation, it is statistically much more likely for the real photon momentum flux to partition itself into the zero-point system with its massively greater states (up to FPlanck) The collapse of the field in the absence of charges to sustain it can be viewed as

analogous to the loss of induction in a paramagnetic gas by Brownian motion- homogeneity and isotropy are statistically most likely. Overall though, momentum has been"dumped"to the very large mass of the zero-point field.

We shall now with reference to figures Sa, Sb, Sc and 6 calculate the magnitude of the force of propulsion. In doing so we shall see the effect of the field strengths, the frequency of cycling, and the volume of the electret and shifting of the craft from the initial centre of mass by projecting the thrust and cancellation fields.

Figure 5a shows the thrust forces on their own as the act on the representation of the device as two masses MI, m2 separated by a rod of length L (figure 3). The field mass has been projected back to the centre of the device but its momentum contribution is negligible and shall not be mentioned from here on. The forces at the masses ml, m2 are symmetrical on the cycle. Figure 5b shows the cancellation forces at the masses ml, m2. They act at a different centre since the field has been projected to a different location and shown here for exaggeration ; the masses ml, m2 are so distant from the centre that approximately the same force is experienced by both. In short, the torque vector field of the field generation and cancellation solenoids have different curvatures.

Figure 5c shows the resultant forces when the cancellation scheme of figure 4 is applied. Note that it is not a case of merely superimposing figures 5a and 5b as the cancellation field removes the forces on the masses ml, m2 in the off-phase of the field generation solenoid by superimposing the on-phase of the cancellation solenoid-thus only the on-phase of the field generation solenoid and the off-phase of the cancellation solenoid act at the masses ml, m2.

Conveniently we have shown the forces at one of the masses mn ancelling (by

contrivance of having the same field strength present at the electret element) leaving only the force acting at the other of the masses m2 (mf has been neglected as mentioned earlier). This force on its own would leave the device to rotate but the dual sided symmetrical device (figure 6) has another set of field generation and cancellation solenoids on the other side of the device leaving a true linear force.

First we shall proceed on the basis that the fields generated by the field generation and cancellation solenoids are projected outside the device, this is not essential. Calculating the centre of mass from the matter distribution (the notation is by figure 3, ml=m2=m) : Note that according to figure 5c and the arrangement of forces mentioned earlier, we need only calculate the force on m2, 6 is the torque: # fmZTH-L _ r eqn. 11 cmTH If the cancellation field is projected externally too, the same procedure to the derivation of equation 10 gives the derivation of the centre of mass when the cancellation field is present (to a good approximation neglecting mfth) in the same form.

Once again we calculate the force at m2 due to this rotation and arrive at the same form as equation 11 with r. f r L - rcmC The force, then, for a dual-sided device is twice the sum of fm2. th and fm2. c and substitution of expressions for field mass and torque due to the Poynting force give, after routine algebraic manipulation and approximation the following expression for the force on the device: m f soBzV Let B(t) = BmaxFt E(t) = EmaxFt #TH = #0EBFvercmTH (#C similar) Thus # 4#0EBFve(rcmTH - rcmC) eqn. 12 L Other forms of field projection-internal or a combination of internal and external lead to essentially the same formula. We have represented the fields as a symmetrical sawtooth going up to a maximum value which we neglect as a subscript in following formulae. The volume of the electret element is ve. The force is proportional to the frequency of cycling and reliant on the field centres for the fields generated by the field generation and cancellation solenoids being different. Thus we benefit from the mass of the device being small; the shifting of the centre of mass of the device is set against the small mass of the fields. Later we will see that it is better to have the device

as a"propulsion sub-assembly"inside a greater craft towed intermittently by smaller sub-craft; decoupling the large mass of the greater craft allows a large force and momentum to be developed by the smaller craft-the centre of mass shift can then be large even though the greater craft is massive.

Incidentally, radiation resistance of the Lorentz frictional force from the moving electric field around dielectric drive element is negligible being an order C3 down (L. D. Landau and E. M. Lifshitz, Course of Theoretical Physics : The Classical Theory of Fields, Butterworth Heinemann, Vol. 2, 4th ed. 1982): <BR> <BR> <BR> <BR> <BR> <BR> fres = 2e# eqn. 13<BR> <BR> 3C3 Finally with reference to figures 9 and 10 we shall discuss additional manifestations or devices to assist the propulsion system.

Figure 9 depicts an active electret element. A high voltage power source 16 is connected by shielded wires 17 to electrodes 18 between which a plurality of field shaping, conductive elements 19 embedded in a high permittivity dielectric 20 exists. Around one electrode is a neutral, conductive electrode 21 such that the field projected into space is of substantially one polarity; field lines are depicted.

Natural electrets exist, such as Barium Titinate but these can age and lose their moment from dust and water ingress. They also have the inconvenience of the field being present all the time making them difficult to handle. An active electret element won't have these disadvantages to the extent of a passive electret and can have a higher field too.

The propulsive effect from the Poynting vector is proportional to magnetic, electric field strength and the volume over which the fields exist. If an unmodified dipole is used for the electret element then the effects from the positive and negative ends will cancel. Thus the element should project substantially a field of one polarity into space. This can be achieved by having one pole of the dipole shielded by a conductive electrode at neutral polarity.

Suitable elements will always be dielectrics as anything charged will tend to attract dust and water vapour and the like to form a dielectric in consequence.

Consider a slab of positively charged material to which cancelling negative charges are attracted to its outer surface, then the electric fields will be that more intense over the short distance on the surface between the negative charges and positive charges of the greater dielectric slab than in the dielectric itself; so one would think that due to these intense fields at the surface that the propulsive effect would overall be zero. However, in a good approximation between the line of charges we can consider that the potential varies in a constant manner so that the field varies in a linear manner. It is then a case of integrating over space to calculate the total Poynting propulsive effect: assume that they see the same constant magnetic field and the electric field is constant too (but scaled): The contribution from the agglomeration on the surface is BxEDxl3 The contribution from the slab is BxExD3 We have normalised the distances to unity for the surface charges and D for the slab. Thus we see that if we don't project too much of the other charge into space and cancel it within a short distance then net propulsion is obtained.

Figure 10 shows the device of figure 6 as a propulsion sub-assembly (P. S. A. ) inside a greater craft 22 coupled to the P. S. A. by a low friction electromagnetic, intermittent spherical joint coupling 23.

Since no torque is developed by the joint 23, the greater craft 22 is reduced to a point mass acting at the device's (P. S. A. ) centre. This doesn't interfere with the device's operation and its need to have the electromagnetic fields shift its centre of mass for"leverage"to translate; thus the greater craft 22 is towed and pushed along by the device. In addition, the coupling can be made periodic and intermittent to allow the sub-assembly to develop momentum before"catching"and then releasing it.

The space 24 between the craft 22 and the P. S. A can be made into a resonant cavity allowing the electromagnetic fields to be set up in operation at microwave frequencies and above. This offers advantages for higher force production because the device is linear in frequency (eqn. 12) given the higher speed over a LCR arrangement. Also the field energy is recouped too on each cycle by the natural action of a resonant cavity. The cavities are tuned to the Fourier components to allow a waveform such as figure 4 to be constructed.

Radiation pressure will not interfere with torque-less coupling between the crafts and because as we have seen (eqn. 13) radiative field effects are minuscule compared to the inductive field effects responsible for the motive force of device.

A skilled person would readily understand that the present invention provides and improved field generation device, which will find application in many different circumstances.

In the present specification"comprises"means"includes or consists of' and"comprising"means"including or consisting of'.

The features disclosed in the foregoing description, or the following claims, or the accompanying drawings, expressed in their specific forms or in terms of a means for performing the disclosed function, or a method or process for attaining the disclosed result, as appropriate, may, separately, or in any combination of such feature, be utilised for realising the invention in diverse forms thereof.