UDC 681.172.088

ON THE SERVICE LIFE OF HIGH-FREQUENCY TORCH PLASMATRONS

I. A. TIKHOMIROV, Vl. L. TEPLOUKHOV, Vik. L. TEPLOUKHOV,
V. A. VERNYAEV, S. M. GOGLEV, V. M. STAROVEROV

(Presented by the scientific seminar of the Faculty of Physics and Engineering)

As was reported earlier [1], low-temperature plasma and plasma jets of a high-frequency (h.f.) torch discharge are used by us in carrying out a number of plasmochemical processes. For this purpose, designs of h.f. torch plasmatrons have been developed and certain experience in their operation has been accumulated.

In principle, a high-frequency torch plasmatron is a coaxial line transmitting electromagnetic energy from an h.f. generator, the inner conductor of which terminates in a water-cooled electrode. A single-electrode h.f. torch discharge is excited from the electrode [2].

The gas-discharge zone is usually made in the form of a nozzle, whose walls are provided with water cooling. The high-voltage electrode is vacuum-tightly separated from the outer conductor by an insulator made of ceramic or Teflon, which simultaneously separates the gas-discharge zone from the feeder line. The plasma-forming gas is supplied axially or tangentially, ensuring axial stability of the plasma flow and leading to the formation of a protective gas flow in the near-wall region.

Studies of high-frequency torch-type plasmatrons of relatively low (up to 60 kW) power, intended for carrying out both homogeneous and heterogeneous reactions, have shown that they ensure high purity of the gas-discharge space, possess a practically unlimited service life, mechanical strength, and tightness (which is necessary when working with aggressive media), have small overall dimensions compared with ordinary chemical reactors of the same productivity, and can be used for processing and synthesis of various chemical compounds, especially for carrying out thermal decomposition processes.

However, up to the present time very little attention has been paid to questions concerning the use of an h.f. torch discharge. Among some researchers there even exists the opinion that the input of large powers at a high efficiency of conversion of electrical energy into heat in an h.f. torch discharge is an impracticable task. Studies carried out by us have shown that the overall efficiency of this type of plasmatron can reach 0.65–0.75.

Another parameter, no less important than efficiency, characterizing the operation of plasmatrons, is the service life (resource) of the apparatus.

Table 1

| No. | Electrode type | Electrode material | \multicolumn{2}{c}{High-frequency generator operating mode} | \multicolumn{2}{c}{Discharge characteristics} | Electrode losses | Discharge power | Channel diameter | Evaporation rate | Note, electrode diameter |
|---:|---|---|---:|---:|---:|---:|---:|---:|---:|---:|---|
| | | | $I_c$, A | $I_a$, A | $U_{\sim}$, V | $I_{\sim}$, A | $W_e$, kW | $W_p$, kW | $d_k$, cm | $q \cdot 10^3$, g/s | |
| 1 | Uncooled electrodes | Aluminum | 0.45 | 2.2 | 950 | 1.2 | 0.350 | 1.10 | — | 0.42 | Ø 10 mm |
| 2 | Uncooled electrodes | Copper | 0.53 | 2.4 | 1200 | 1.1 | 0.33 | 1.60 | 0.2 | 1.78 | Ø 10 mm |
| 3 | Uncooled electrodes | Steel 3 | 0.6 | 2.5 | 1100 | 1.0 | 0.35 | 1.45 | 0.2 | 1.58 | Ø 10 mm |
| 4 | Uncooled electrodes | Steel | 0.5 | 2.3 | 1100 | 1.1 | 0.34 | 1.53 | 0.2 | 0.51 | Ø 10 mm |
| 5 | Uncooled electrodes | Tungsten | 0.6 | 2.15 | 950 | 1.4 | 0.3 | 1.60 | — | 1.37 | Ø 10 mm |
| 6 | Water-cooled electrodes | Copper | 0.55 | 2.8 | — | — | 0.37 | 1.65 | 0.15 | $5 \cdot 10^{-7}$ | wall thickness $\delta = 3$–4 mm |
| 7 | Water-cooled electrodes | Copper, steel Kh18N9T | 0.45 | 2.7 | — | — | 0.43 | 1.57 | 0.15 | $10^{-5}$ | wall thickness $\delta = 3$–4 mm |

To determine the service life of high-frequency torch plasmatrons, studies were carried out on the destruction of both water-cooled and non-water-cooled electrodes made of Steel 3, Steel Kh18N9T, copper, and tungsten. The studies were conducted for a free high-frequency torch discharge in air using an LD4-10 high-frequency generator with an oscillatory power of 10 kW and a frequency of 39–40 MHz. The generator was operated in a mode with 50% anode voltage. The experimental procedure consisted in determining the rate of electrode destruction under various operating modes of the high-frequency generator. The measurement results are presented in Table 1.

The data obtained indicate that a high-frequency torch discharge exists stably with electrodes made of various metals. The power losses at the electrodes in our experiments amounted to 300–350 W; the dimensions of the near-electrode spot, determined by photographing the discharge channel and from the characteristic imprint on the uncooled electrode, were 0.15–0.20 cm.

On this basis it is not difficult to estimate the average heat-flux density in the spot, which was \((1.0—2.0)\cdot 10^8\ \mathrm{W/m^2}\).

The electric-current density in the spot was 30–110 \(\mathrm{A/cm^2}\). Consequently, the heat fluxes to the electrode in a high-frequency torch plasmatron can be removed by ordinary water cooling [3]. In addition, the nature of the high-frequency torch discharge determines the constant displacement of the near-electrode spot over the electrode surface (even in the case of a discharge with a heated electrode). This leads to a redistribution of the heat flux over the entire end surface of the electrode and correspondingly facilitates heat removal from its surface.

With an increase in the discharge power, the heat flux to the electrode also increases. This, in turn, may lead to more intense erosion of the electrode. Therefore, we carried out studies to measure the heat-flux density in the spot and the electrode erosion rate at discharge powers of 1.5 kW and 10–12 kW. The results are presented in Table 2.

Table 2

The data obtained indicate an increase in the channel diameter with increasing power, while the heat-flux density in the spot remains practically unchanged. Consequently, with sufficient cooling of the electrode, the rate of destruction of electrodes in powerful high-frequency torch plasmatrons should remain constant.

Attention is drawn to the fact that the combustion mode of the discharge with “hot” electrodes differs only slightly from the discharge mode with water-cooled electrodes. This indicates that near-electrode processes are not determining factors in maintaining a high-frequency torch discharge, or, at least, that the phenomenon of autoelectronic emission predominates over thermionic emission even in the case of tungsten electrodes. In order to confirm the indicated propositions, studies were carried out—

we shall carry out estimates of the near-electrode voltage drop and of the active and capacitive resistance in the layer.

Assume that the heat flux to the electrode is determined by the magnitude of heat release in the near-electrode region. In addition, we shall assume that in the near-electrode region the expression for the conductivity is valid:

\[ \sigma=\sigma_a+j\sigma_r=\frac{ne^2}{m_e}\left(\frac{\nu}{\omega^2+\nu^2}-j\frac{\omega}{\omega^2+\nu^2}\right), \tag{1} \]

where, for our conditions, the effective frequency of collisions of electrons with plasma particles \(\nu\) is much greater than the frequency \(\omega\) of the feeding h.f. field \((\nu \gg \omega)\). Then the active conductivity of the layer \(\sigma_a\) must be much greater than the reactive conductivity \(\sigma_r\). On this basis one may assume

\[ \Delta\varphi \approx I\cdot r \approx \frac{W_{\mathrm{e}}}{I}, \tag{2} \]

where \(\Delta\varphi\) is the voltage drop in the near-electrode layer;
\(r\) is the resistance of the layer;
\(I\) is the total current through the electrode.

The field strength in the near-electrode region may be determined as

\[ E_{\mathrm{e}}\approx \frac{W_{\mathrm{e}}}{I\cdot l_d\cdot a}, \tag{3} \]

where \(l_d\) is the Debye radius for the plasma of the channel of the h.f. torch discharge, determined by the well-known relation

\[ l_d=\sqrt{\frac{\kappa\left(T_e\cdot T_i\right)}{4\pi ne^2\left(T_e+T_i\right)}}, \tag{4} \]

where \(\kappa\) is Boltzmann’s constant;
\(T_e, T_i\) are, respectively, the electron and ion temperatures in the discharge channel;
\(n\) is the electron concentration in the channel;
\(e\) is the electron charge;
\(a\) is a numerical coefficient.

Table 3

The numerical values of the quantities are summarized in Table 3. Such values of the field strength in the near-electrode layer may indicate the predominant role of processes determined by autoelectronic emission. Thermionic emission from a water-cooled electrode, as is known, is hindered.

Thus, the heat-flux density in the near-electrode spot of the torch discharge is \(10^8\ \mathrm{W/m^2}\) and can be removed by ordinary water cooling. This substantially reduces the role of thermionic emission and makes possible the existence of the discharge due to

field emission. The erosion rate of water-cooled copper electrodes is \(\sim 5 \cdot 10^{-7}\) g/sec.

The data obtained make it possible to estimate the service life \(\tau\), including that of flare plasmatrons, in the case of reliable cooling of the electrodes:

\[ \tau = \Delta m / q, \]

where \(\Delta m\) is the permissible loss in electrode mass;
\(q\) is the evaporation rate \((5 \cdot 10^{-7}\) g/sec).

For \(\Delta m = 0.7\text{–}0.8\) g, the service life \(\tau\) can reach values of 450–500 hours. Operation of plasmatrons with a power of 20–40 kW confirms the calculated data.

In a number of technological processes, uncooled electrodes may be used. The service life of this type of plasmatron can also reach hundreds of hours.

In conclusion, it should be added that the service life, including that of plasmatrons, is determined by the degree of contamination of the solid reaction products by the electrode material. Thus, at a discharge power of 20–30 kW, which ensures the processing of 5–10 kg of products per hour, the fraction of impurities of electrode materials will be \(10^{-6}\text{–}10^{-7}\%\).

REFERENCES

1. Generators of Low-Temperature Plasma. Collection “Energiya,” Minsk, 1969.
2. L. A. Stolov. Scientific Notes of Kazan State University. Issue 2, vol. 117, 1957.
3. A. S. Korotaev, A. M. Kostylev, V. V. Koba, M. A. Lomovtsev, V. A. Kutsavalov, B. V. Cheloznov. Generators of Low-Temperature Plasma. Moscow, “Nauka,” 1969.