The present invention relates to measurement of plasma parameters such as electron plasma density and magnetic field in a time-varying plasma.
Plasma density is a key metric used to quantitatively describe and classify plasma. The
Earth is surrounded by a sparse plasma in the ionosphere, denser plasma is used in the manufacture of semiconductor devices, and extremely hot and dense plasma is central to fusion research. A standard device used to measure plasma density is the Langmuir probe, a biased conductor placed directly in the plasma. See F. F. Chen, J. D. Evans, and W. Zawalski, “Electric probes” in Plasma Diagnostic Techniques, edited by R. H. Huddlestone and S. L. Leonard, Citeseer, (1965). However, while Langmuir probes are simple to construct and are in widespread use today, they are limited to invasive single point measurements that can disrupt the plasma in some cases or be destroyed by the plasma in other cases.
Because of the drawbacks of Langmuir probes, a variety of techniques have been developed to make spatial, non-invasive measurements of plasma.
Laser induced fluorescence (LIF) excites ions of the plasma to higher energy states, and observes the photons released when the ion relaxes to its original energy. See R. A. Stern and J. A. Johnson, “Plasma Ion Diagnostics Using Resonant Fluorescence,” Phys. Rev. Lett., 34, 1548, (1975). LIF can measure particle velocity distributions and is useful in the study of flows of particles in plasma, but can only give a relative, not absolute, measurement of density.
Plasma tomography uses measurements of plasma radiation along lines of sight to reconstruct spatial plasma density profiles. J. F. Camacho and R. S. Granetz, “Soft x-ray tomography diagnostic for the Alcator C tokamak,” Rev. Sci. Instrum., 57, 417, (1986). However, such methods require the high-density and high-temperature environments of fusion plasmas to produce the required plasma radiation.
Electrical impedance tomography (EIT) uses low-frequency measurements of boundary voltage and current to reconstruct an image of internal conductivity, see L. Borcea, “Electric impedance tomography,” Inverse Problems, 18, R99, (2002), but no reconstruction algorithm have been applied to plasmas.
Laser plasma interferometry can provide line integrated density measurements and multi-chord systems can invert this data to provide density profiles with good temporal resolution. See E. C. Merritt, A. G. Lynn, M. A. Gilmore, C. Thoma, J. Loverich, and S. C. Hsu, “Multi-chord fiber-coupled interferometry of supersonic plasma jets,” Rev. Sci. Instrum., 83, 10D523, (2012). This method is non-invasive and has provided good results when large refraction does not occur and if the assumption of rotational symmetry holds.
Polarimetry measurements using Faraday rotation can provide line integrated time-resolved measurements of magnetic field. See S. E. Segre, “A review of plasma polarimetry—theory and methods,” Plasma Phys. Control. Fusion, 41, R57, (1999). These measurements are sufficiently difficult to make that tomographic systems have not been developed to provide spatial information.
This summary is intended to introduce, in simplified form, a selection of measurement concepts that are further described in the Detailed Description. This summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter. Instead, it is merely presented as a brief overview of the subject matter described and claimed herein.
The present invention provides a method of remotely imaging the parameters of both static and time-varying plasmas by measuring the complex self-impedance and mutual impedance between successive pairs of antennas in a plasma impedance sensor array over a wide range of frequencies.
In accordance with the present invention, the antennas can be placed at the edge of the plasma, negating the need to insert a probe into the plasma, which can disturb the plasma being measured or lead to destruction of the probe. Mutual impedance measurements provide line-averaged spatial information across the plasma, which can be inverted to make images of the plasma dielectric. Plasma parameter profiles such as density and magnetic field can be determined from the plasma dielectric profiles. Using a pulsed measurement technique allows for measurement rates at approximately 1% of the plasma frequency.
This plasma impedance tomography (PIT) approach of the present invention can provide absolute plasma density and magnetic field measurements for laboratory and fusion plasmas and imaging of dynamic plasma density structures.
The aspects and features of the present invention summarized above can be embodied in various forms. The following description shows, by way of illustration, combinations and configurations in which the aspects and features can be put into practice. It is understood that the described aspects, features, and/or embodiments are merely examples, and that one skilled in the art may utilize other aspects, features, and/or embodiments or make structural and functional modifications without departing from the scope of the present disclosure.
The present invention provides a method of remotely imaging parameters of both static and time-varying plasmas by measuring the complex self-impedance and mutual impedance between successive pairs of antennas in a plasma impedance sensor array over a wide range of frequencies.
Although the discussion below assumes that the antennas are dipole antennas, one skilled in the art would readily understand that other types of antennas such as loop antennas, microwave horns, and many others can also be used in the method for imaging plasma parameters described herein, and all such alternative configurations are deemed to be within the scope and spirit of the present disclosure.
In accordance with the present invention, the antennas can be placed at the edge of the plasma, negating the need to insert a probe into the plasma, which can disturb the plasma being measured or lead to destruction of the probe. Mutual impedance measurements provide line-averaged spatial information across the plasma, which can be inverted to make images of the plasma dielectric. Plasma parameter profiles such as density and magnetic field can be determined from the plasma dielectric profiles. Using a pulsed measurement technique allows for measurement rates at approximately 1% of the plasma frequency.
This plasma impedance tomography (PIT) approach of the present invention can provide absolute plasma density and magnetic field measurements for laboratory and fusion plasmas and imaging of dynamic plasma density structures.
If the stimulus applied to each antenna is above the ion plasma frequency ωpi2=e2ni/∈0mi, only the electrons can respond to the stimulus, and the relative frequency-dependent dielectric constant for an unmagnetized plasma can be written as:
where ωpe2=e2ne/∈0me, ni,e is the ion or electron density, ∈0 is the permittivity of free space, v is the sum of the frequency of collisions between electrons and electrons, neutral particles, and ions in the plasma, e is the charge on the electron, and mi,e is the ion or electron mass.
When a background magnetic field B is present, the simple dielectric constant given above is replaced with a dielectric tensor given by:
where Ωce=eB/me is the electron cyclotron frequency. When a current is applied to an antenna immersed in a plasma, the electric field of the antenna is modified by the presence of the plasma. Since the modifications are frequency dependent, this gives rise to the plasma impedance spectrum.
The basis of the PIT method of the present invention lies in the mutual impedance measurement between a pair of antennas in a plasma. For a pair of antennas 1, 2, the ratio of the induced voltage V2 on the second antenna due to the current I1 in the first antenna is the mutual impedance Z21=V2/I1. In an exemplary case, where the antennas are dipole antennas separated by a distance d and each antenna having a total length 2L, where L is the length of each “arm” of the dipole, the result of the self-impedance of a short dipole can be extended to determine the mutual impedance spectra between two short dipole antennas for parallel and perpendicular orientation with respect to the background magnetic field in a plasma:
(see K. G. Balmain, “Impedance of a Short Dipole in a Magnetoplasma,” IEEE Trans. Antennas Propagat., 12, 605 (1964)).
The plots in
The reconstruction procedure for imaging plasma parameters in accordance with the present invention can be illustrated by an exemplary small system of three dipole antennas, such as that illustrated in
The plasma region of interest is partitioned into pixels, with the approximate density and magnetic field being assumed to be constant within each pixel. In the example illustrated in
Z21(n, B)≈f21(n1, n2, n3, n4, B1, B2, B3, B4), (8)
where n1, n2, n3, and n4 are the plasma densities ni and B1, B2, B3, and B4 are the magnetic fields Bi corresponding to pixels A, B, C, and D, respectively.
Using a Taylor expansion of f21 around a suitable point ({circumflex over (n)}1, . . . , {circumflex over (B)}4) gives a linear approximation for Z21:
where Δni=ni-{circumflex over (n)}i and ΔBi=Bi-{circumflex over (B)}i, {circumflex over (n)}i and {circumflex over (B)}i being the points around which the function is expanded.
In vector notation this can be written as:
To reconstruct an image of magnetic field and plasma density from the measurements of impedance, we begin by repeating this procedure for each of the self- and mutual impedances Zij=fij(ni, . . . , B4) in the system, which yields the matrix equation
The functions fij have no closed form, so the values of impedances and the values of the derivatives at the point of Taylor expansion must be found numerically by simulating the forward problem. Using m for the vector of measured impedances, S for the Jacobian, or sensitivity matrix, and p for the vector of pixel parameters (density and magnetic field), this can be compactly written
m=Sp. (12)
In an exemplary system with A dipoles, where the region of interest has been divided into a grid of B×B pixels, m will be an (A2×1) column vector, S will be an (A2×2B2) matrix, and p will be a (B2×1) column vector.
In typical systems, there will be many more pixels than measurements of mutual impedance, and so S will generally not be full rank, and its inverse will not exist. However, approximate solutions can be found using a variety of methods to regularize ill-posed problems such as this one. One exemplary reconstruction method is to apply the Moore-Penrose pseudoinverse known in the art to matrix S, which gives
p=(STS)−1STm. (13)
Although the Moore-Penrose pseudoinverse reconstruction algorithm is used in the present disclosure to illustrate the PIT method of the present invention, one skilled in the art will readily understand that any suitable electrical impedance tomography (EIT) reconstruction algorithm, such as linear back-projection, Tikhonov regularization, Newton-Raphson and iterative Tikhonov methods, and projected Landweber, can also be used where appropriate. In some cases, the Gauss-Newton method can be used to find a solution without first linearizing the model of mutual impedance. One skilled in the art would readily understand that any suitable reconstruction methods can be used, and use of any and all such antenna inversion algorithms is deemed to be within the scope of the present disclosure.
The PIT method of the present invention can be accomplished by means of an apparatus comprising a suitable antenna array arranged around the plasma, an arbitrary function generator to provide a stimulus to one or more of the antennas, a switch to control which antenna is being driven and terminate the inputs of the receiving antennas, a measurement circuit for each antenna to provide the oscillating voltage and current measurements, a multichannel oscilloscope to record the measurements, and a computer programmed with appropriate software that processes the results and produces maps, plots, and other visual depictions of the plasma parameters.
In the exemplary embodiments illustrated in
These assumptions about the antennas in this exemplary system simplify the presentation of the method, and one skilled in the art would readily understand that more complex antennas can be modelled, either analytically or numerically, and all such alternative antenna models are deemed to be within the scope and spirit of the present disclosure.
As described in more detail below, each of the antennas 301a, 301b, . . . 301n is coupled to a signal generator 305 by means of a switch 304 which allows a predetermined oscillating signal to be controllably and selectively applied to the antennas in a predetermined manner, e.g., in a predetermined order. A remote processor controls which output of the switch is connected to the input while the other outputs are terminated in the characteristic impedance of the system. Each of the antennas 301a, 301b, . . . 301n is coupled to a corresponding pair of digitizers configured to record the voltage and current produced by application of the oscillating signal and to output the recorded data, either directly to remote processor 306 or to a data storage device for later processing by a remote processor.
The schematic in
Thus, as illustrated by the block schematic in
It will be noted, however, that one skilled in the art would understand how to construct other measurements of voltage and current for different ranges of frequency or different types of antenna, and any such suitable methods for measurement of voltage and/or current are within the scope of the present disclosure.
Taking the sensor array shown in
In a first step (1) in a method for obtaining an image reconstruction of plasma parameters in accordance with the present invention, an appropriate pulse shape and center frequency for the signal generator is chosen based on a priori knowledge about the plasma under investigation and the voltage amplitude VS of the signal is chosen to be small, i.e., eVS<kTe, where k is the Boltzmann constant and Te is the electron temperature.
In the next step, step (2), signal generator 303 is connected to the first antenna 301a by means of switch 304, and using signal generator 303, an oscillating signal as determined in step (1) is controllably applied to a first of the antennas 301a, where the parameters of the signal are determined as described in step (1).
In step (3), using the digitizers coupled to that antenna, a time series of values for each of 301a_V, 301a_I, 301b_V, . . . , 301h_I is recorded and data of each of those time series, referred to herein as an “acquisition,” is output to remote processor 306 or to a data storage device for later processing.
Each of steps (2)-(3) is then repeated for each of the antennas 301b-301h in the array, with the switch controllably applying the signal to each antenna and the time series for each of 301a_V, 301a_I, 301b_V, . . . , 301h_I resulting from the application of the signal to that antenna being output for processing.
In the next step, step (4), using a processor such as remote processor 306, the fast Fourier transform (FFT) of each of 301a_V, 301a_I, 301b_V, . . . , 301h_I is computed, and in step (5) the self- and the mutual impedances produced by the signal are computed using the appropriate ratios of the FFTs. For example, the mutual impedance Z12 between antenna 301a and 301b can be approximated as Z12≈301a_V/301b_I.
In the next step, a vector m as set forth in Equation (12) is assembled by the processor, and an inversion in accordance with Equation (13) is carried out to determine the vector p, which is then reshaped by the processor into a two-dimensional image for display, where the two-dimensional image illustrates a plasma density, a magnetic field distribution, or both.
Previous impedance probe methods have been operated by sweeping the applied frequency and sequentially measuring the impedance at each frequency. Although using this sequential method would provide the impedance information needed to reconstruct the images of plasma parameters, the time resolution of the measurement would be poor compared to the present invention, and the time needed to make such measurements sequentially would be too long to be feasible, especially in a time-varying plasma. Consequently, instead of using sequentially applied signals, each at a different frequency, in the method of the present invention, a single short bipolar pulse is applied to each antenna as the driver stimulus, where each pulse is a Gaussian monopulse having a shape that is the derivative of a Gaussian:
This signal has a pulse shape in frequency space, and the center frequency ωc=2πfc can be positioned over the range where we expect to find the resonant frequencies of interest in the plasma. The Gaussian monopulse stimulus injects all the frequencies of interest at once. By digitizing the RFI and RFV signals for each probe simultaneously and Fourier transforming the signals, the mutual impedance spectra for all of the receiving antennas can be determined in one acquisition. In the exemplary system illustrated in
The limit to the acquisition time for the measurement is set by the frequency resolution that is necessary for the impedance measurement and the number of antennas. In many cases, a reasonable frequency resolution would be about 10% of the expected resonant frequency in the impedance spectrum. An antenna array size on the order of 10 would result in a measurement rate of about 1% of the expected resonant frequency. For example, at a plasma density of 108 cm−3 and a plasma frequency of 100 MHz, we could expect a measurement rate of 1 MHz or a time resolution of 1 μs. The damping rate of the stimulus in the plasma also affects the measurement rate, since the plasma response to the previous pulse must be allowed to decay away before starting the next acquisition.
In order to demonstrate the PIT technique of the present invention, numerical simulations were conducted to both determine the Jacobian and to simulate measurements of a known density distribution for which a reconstruction will be produced. For these simulations, it was assumed that the plasma was unmagnetized, and eight single dipole antennas were placed in a plane at the edge of a 50-cm diameter plasma in an arrangement similar to that illustrated in
which is equivalent to assuming a triangular current density.
To generate the Jacobian, a series of simulations were conducted using a uniform background plasma and a 1-cm diameter perturbation where the plasma frequency was increased by 10%. The location of the perturbation was scanned throughout the space and a numerical simulation was completed at each position, while recording the resulting mutual impedance measured by each antenna.
The PIT method in accordance with the present invention uses mutual impedance information for a wide range of frequencies. In order for these measurements to provide more than noise reduction, the paths of the oscillating current throughout the plasma must be different when the driver frequency is changed. The different current paths sample different regions of the plasma providing new spatial information for the tomography reconstruction.
Numerical simulations were conducted using three different density structures that each had a 10% higher density above the background for a single frequency far below the plasma resonances. The first structure was a 10-cm diameter region centered on the origin, the second was a 10-cm diameter region offset by 10 cm in the x and y directions, and the third adds a second 10-cm diameter region to the second case, this second region being offset from the center by −10 cm in the y direction. These simulations provided the mutual impedance measurements from the same eight antennas used to determine the Jacobian. A measurement vector m was construct as described in Equation (12) and multiplied by the Moore-Penrose pseudoinverse as described in Equation (13).
The resulting vector of pixel parameters p was reshaped, and the resulting images are shown in
A variety of inversion algorithms exist for tomographic reconstructions of EIT systems, which could be applied to PIT to improve upon the pseudoinverse. A tailored algorithm for PIT, which incorporates the multi-frequency spectra and our knowledge of the plasma dispersion is being developed. As of now, existing inversion and regularization algorithms, which rely on assuming that the system can be linearized and a suitable Jacobian function can be estimated, are being used. The Jacobian can either be determined from numerical simulations as demonstrated here or determined experimentally by physically applying the necessary perturbations. The resulting Jacobian gives the measured response to each applied impulse and serves as a basis for expanding subsequent measurements of an unknown plasma.
Advantages and New Features
The plasma impedance sensor array used in PIT are a non-invasive measurement as they can sit at the edge of the plasma and do not need to be inserted. For applications with insulating boundaries, one could place the impedance probe array outside the vacuum vessel. This method has clear advantages over methods that need to insert a probe into the plasma, which inherently alters the plasma that is being measured. The applied oscillating amplitudes are kept purposefully small (energy in oscillating voltage is smaller than the thermal energy in the electrons) in order to maintain a linear response from the plasma.
The multi-frequency impedance measurements provide a multiplicative factor to the amount of available information for the inversion. This is the equivalent of increasing the number of antennas making measurements at a single frequency. Higher spatial resolution images can be obtained with many fewer antennas by exploiting multi-frequency impedance spectra.
A unique aspect of this measurement method is that it can measure both plasma density and magnetic field. The other diagnostics described in the background section can only measure one or the other. In fact, often knowledge of one is necessary to interpret the results and make a measurement of the other.
Alternatives
In this patent disclosure, dipole antennas were chosen to measure the mutual impedance of the antenna plasma system. However, these measurements could be made using a variety of different antennas. The predicted mutual impedance would be modified by the choice of antenna, but the overall method would be the same. For example, the dominant mutual impedance between two dipoles is capacitive, while it is inductive for loop antennas, however, the plasma dielectric enters into the two mutual impedances in the same way.
There are alternate standard ways of measuring the complex impedance. The method discussed here uses the RF-IV method; however, the network analyzer method that uses reflection coefficients to back out the measured impedance is also a viable measurement technique.
The Gaussian monopulse has useful properties as the driver stimulus for the pulsed impedance measurement. Alternate pulse shapes, such as a square pulse or a raised cosine, could just as well be used in this measurement method. The pulse shape merely determines the frequency distribution that is output.
We chose to focus on a probe array that surrounds the plasma that we are diagnosing, however, the shape of the probe array is not an integral part of the measurement method. The shape should be chosen based on the application. A linear or planar array would be just as effective. In addition, these arrays could be stacked to give three-dimensional images of the plasma.
Plasma Impedance Tomography provides a non-invasive method for making real-time, spatial measurements of plasma density and magnetic field, taking advantage natural resonances in the plasma to provide higher spatial resolution images using fewer antennas.
Although particular embodiments, aspects, and features have been described and illustrated, one skilled in the art would readily appreciate that the invention described herein is not limited to only those embodiments, aspects, and features but also contemplates any and all modifications and alternative embodiments that are within the spirit and scope of the underlying invention described and claimed herein. The present application contemplates any and all modifications within the spirit and scope of the underlying invention described and claimed herein, and all such modifications and alternative embodiments are deemed to be within the scope and spirit of the present disclosure.
This Application is a Nonprovisional of and claims the benefit of priority under 35 U.S.C. § 119 based on U.S. Provisional Patent Application No. 62/916,813 filed on Oct. 18, 2019. The Provisional Application and all references cited herein are hereby incorporated by reference into the present disclosure in their entirety.
The United States Government has ownership rights in this invention. Licensing inquiries may be directed to Office of Technology Transfer, US Naval Research Laboratory, Code 1004, Washington, D.C. 20375, USA; +1.202.767.7230; techtran@nrl.navy.mil, referencing Navy Case # 111979.
Number | Name | Date | Kind |
---|---|---|---|
20100300620 | Kanda | Dec 2010 | A1 |
20160281670 | Ikeda | Sep 2016 | A1 |
20180053634 | Kraus | Feb 2018 | A1 |
Entry |
---|
R. A. Stem and J. A. Johnson, “Plasma Ion Diagnostics Using Resonant Fluorescence,” Phys. Rev. Lett., 34, 1548, (1975). |
J. F. Camacho and R. S. Granetz, “Soft x-ray tomography diagnostic for the Alcator C tokamak,” Rev. Sci. Instrum., 57, 417, (1986). |
L. Borcea, “Electric impedance tomography,” Inverse Problems, 18, R99, (2002). |
E. C. Merritt, A. G. Lynn, M. A. Gilmore, C. Thoma, J. Loverich, and S. C. Hsu, “Multi-chord fiber-coupled interferometry of supersonic plasma jets,” Rev. Sci. Instrum., 83, 10D523, (2012). |
S. E. Segre, “A review of plasma polarimetry—theory and methods,” Plasma Phys. Control. Fusion, 41, R57, (1999). |
K. G. Balmain, “Impedance of a Short Dipole in a Magnetoplasma,” IEEE Trans. Antennas Propagat., 12, 605 (1964). |
Agilent Impedance Measurement Handbook, 4th Ed, (2009), pp. 216-219. |
Edmund Spencer, David Clark, and Sai Krishna Vadepu, “A Time-Domain Impedance Probe for Fast Measurements of Electron Plasma Parameters in the Ionosphere,” IEEE Transactions on Plasma Science, vol. 47, No. 2, Feb. 2019, pp. 1322-1329. |
Number | Date | Country | |
---|---|---|---|
20210118644 A1 | Apr 2021 | US |
Number | Date | Country | |
---|---|---|---|
62916813 | Oct 2019 | US |