PC-based visual simulation of high pressure arc plasma

Zhan, Yiyi PC-based visual simulation of high pressure arc plasma. Doctor of Philosophy thesis, University of Liverpool.

	PDF (Yiyi Zhan Phd Thesis) Yiyi thesis pdf 2011/CHAPTER5_TOBE_SUB.pdf - Submitted version Access to this file is embargoed until Unspecified. Available under License Creative Commons Attribution No Derivatives. Download (150kB)
	PDF ZhanYiy_Jan2011_3433.pdf - Author Accepted Manuscript Available under License Creative Commons Attribution No Derivatives. Download (4MB)

Abstract

The work in this thesis is concerned with the visual simulation of high pressure arc plasmas in two different applications. The first one is an arc maintained in water for the formation of nano-structures such as nanotubes and fullerenes. The second application is high power switching using puffer circuit breakers. The fundamental behaviour of both arc plasmas was computationally studied. The structure of the thesis starts with an introduction to the background and objectives of the present work, the arc models used for the two applications are described in Chapter 2 together with a consideration of the radiation and turbulence models, material properties initial and boundary conditions, and issues relating to solution procedures. Chapter 3 is devoted to the modelling of the carbon plasma in water taking into account size of gas bubble growth, water evaporation and chemical reactions in the gas. Chapter 4 presents the computational simulation of the whole arcing process of a high voltage puffer circuit breaker where predicted arc voltage and pressure evolution in the storage volume are compared with the test results. Modelling of the current zero period and the calculation of the critical RRRV is also included. The proposed method for object based visual simulation is described in Chapter 5. Finally conclusions are drawn in Chapter 6 together with a discussion of the future work. For the arc in water case, it is the first time that an arc model is applied to study the carbon arc confined in water considering chemical reactions of water vapour and carbon vapour. The arc is generated between two carbon electrodes which are immersed in water with a gap length of 1 mm. The anode has a conical shape with a rod diameter of 8.2 mm and a tip diameter of 2 mm with a full conical angle of 60o. The anode has a cylindrical shape with a diameter of 12 mm. A local thermal equilibrium arc model was used to tackle the problem based without considering the sheath layer in front of the electrodes. The loss of carbon vapour due to formation of solid carbon structures is represented by a negative volumetric source. The erosion rate of the anode is derived from experimental result. Results show that the strong influx of carbon vapour displaced the original water vapour in the gap in a few microseconds, typically 5 μs. By 4 ms the arc burns in a carbon dominated environment with about 25% of water vapour in the region near the cathode. Our results thus suggest that for most of the time the arc can be regarded as burning in a carbon vapour dominated environment. Reaction rate coefficient between carbon vapour and water vapour was derived from that at low temperature (300 K) using the well known Arrhenius expression. The value of this coefficient is 1.054×106. It has been found that the mass concentration of H2 and CO peaks in a layer surrounding the arc column. The maximum concentration is only 0.2% and 2.8% respectively, in a region near the cathode surface. Although typically 30% of the carbon vapour is consumed by chemical reaction, water vapour still dominates the arc surrounding gaseous environment since the erosion rate of carbon anode (6×10-7 kg/s) is still much lower than the water evaporation rate (6×10-5 kg/s). The predicted voltage of the arc column at 30 A is 7 V while the measured arc voltage including the sheaths is 17 – 20 V. The difference is attributed mainly to the exclusion of the electrode sheath model in the present work. Considering that the first ionisation energy of carbon vapour at ground state which is 11.26 eV, the predicted voltage of the arc column is well placed in the expected range. A numerical scheme has been established in the present work to simulate the growth of gas bubble confined in water. The position change of the interface of the bubble-water is controlled by the pressure difference between that inside the bubble and the atmospheric pressure. The inertia of the water surrounding the bubble is taken into account in the construction of the dynamic response. Our results show that the growth of the bubble is dominated in the first half millisecond by the retreat of the water surface as a result of evaporation. Further on in the arcing process, the pressure difference makes a dominant contribution in the growth of the bubble. For the switching arc case, the arcing process for four test duties of a 252 kV puffer circuit breaker has been simulated. A novel method was proposed to improve the arc voltage prediction by mitigating the effect of Lorentz force near the tips of hollow contacts and their filling transparent contacts. The proposal of this method is based on consideration of the most probable practical scenario in reality. Results present in the thesis show that this method is effective and gives satisfactory arc voltage prediction for all four test cases at different current levels and arc durations. As an example the arc voltage in the 47 kA case at 16.8 ms with an instantaneous current of 55 kA is 220 V, almost the same as that of the measurement while without using the method the prediction is 340 V. Turbulence is important in the interruption process of high voltage SF6 circuit breakers. It has been found that Prandtl mixing length model with a constant turbulence parameter is not able to predict the important extinction peak in the puffer circuit breaker. In the present work this turbulence parameter is made variable, as a function of instantaneous current that is an indirect measure of the presence of cold SF6 flow around the arc column. The modified turbulence model was then applied to simulated the current zero period (from 15 kA before final current zero onwards to the current zero point) of a 47 kA case and a 10 kA case. The predicted arc voltage using the modified turbulence model starts to increase rapidly shortly before the final current zero point, qualitatively agree with the measurement. A detailed comparison is not sensible since the accuracy of the measurement is not known. The present work covers the simulation of the cold flow period, the high current, current zero and the post arc periods. The predicted pressure variation in the puffer cylinder and another three points including the middle of the flat nozzle surface all agrees reasonably well with the corresponding measurement. Detailed results on the high current phase were represented and discussed. The thermal interruption capability of a circuit breaker is represented by the critical RRRV, a value of the rate of rise of recovery voltage that the circuit breaker can with stands. Procedures were set up and some example results reported in the thesis. Accurate prediction of the critical RRRV is not possible at present. Nevertheless, the established procedure for the calculation of the post arc current under the influence of system recovery voltage can be of help in studying the behaviour response of a circuit breaker when a design parameter is changed. The shift of the critical RRRV in response to a design parameter change gives an indication of the effectiveness of the change in design. Computer simulation of electric arcs for industrial devices proves a difficult job in the sense that much coding and setting is required to implement the arc in a chosen software package. An approach has proposed to use Object oriented programming to change the model setup from manually to automatic. The feasibility of the approach has been demonstrated in the present work by a component builder and a Visual Analyser and Monitor (VCM). Information exchange between ISEE and the CFD solver is through memory sharing. Typical results for visual simulation are given in Chapter 2, demonstrating a range of useful functionalities it provides and the benefits for arc modellers. A number of aspects of the present work that require further work were identified. The importance of obtaining experimental results was discussed. This include the mass spectroscopic measurement of gases generated in the arc in water case, the more accurate arc voltage measurement just before current zero, and the post arc current measurement.

Item Type:	Thesis (Doctor of Philosophy)
Additional Information:	Date: 2011-01 (completed)
Subjects:	?? QA76 ?? ?? TK ??
Divisions:	Faculty of Science and Engineering > School of Electrical Engineering, Electronics and Computer Science
Depositing User:	Symplectic Admin
Date Deposited:	23 May 2012 09:51
Last Modified:	16 Dec 2022 04:36
DOI:	10.17638/00003433
Supervisors:	Yan, JD Spencer, Joseph W
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3433