Ben Bowman | GrafTech International Holdings Inc. (retired), Parma, Ohio Klaus Kruger | Helmut-Schmidt-University, Hamburg Arc Furnace Physics Sp} STAHLEISEN COMMUNICATIONS.© 2009 Verlag Stahleisen GmbH, Diisseldorf ‘The work and all parts contained in it are protected by copyright. Any use outside the narrow terms of reference of the copyright statute without the written consent of the Publishers is illegal and a punishable consent. This is particularly applicable to reproductions, translations, microfilming and storing and/or processing in electronic systems, especially databases and networks The present work was diligently prepared. Notwithstanding, the authors, editors and publishers do no accept any liability for the correctness of the data, information and recommendations as well as any printing errors. Any design names, trade names and brands used in this book shall not be considered as generally useable terms according to the purport of the trademark and proprietary right act Any content based on regulations, directives or ordinances shall only be used provided that the applicable version valid from time to time is referred to Any additions, important notices or corrections that might become available after the publication will be posted on the Internet at: www stahleisen de/errata Printed in Germany ISBN 978-3-514-00768-0Preface Dieter Ameling 2007 was a record year in terms of world steel output A total of 1.351 bn tonnes of crude steel were produced during that year, thereof 422 million tonnes via the electric steelmaking route This represents a share of 31.2%. However, the share of electric steelmaking varies markedly in the regions of the world. The share is biggest in the Middle East, accounting for 85.9 %. In total figures this, however, amounts to only 13.8 million tonnes. In the United States of America, 58.9 % (57.8 million tonnes) of the steel is produced via the electric steelmaking route. The highest output, namely 84 million tonnes, is achieved in the 27 member states of the European Union, accounting for 40.2 %. These figures illustrate the importance of electric steelmaking in the various regions of the world. A total of 530 million tonnes of scrap were used to produce the 2007 crude steel output, making a major contribution to environmental protection and efficiency in resource and energy use. I estimate the scrap input to produce the world electric steelmaking output at about 415 million tonnes. The recycling rate of the total steel output is 34.5 %. The electric arc furnace improves resource efficiency. By the way, the recycling of steel is higher than that of any other metal Scrap is a secondary raw material. Almost 90 % of scrap is metallic iton. Steel serap has already undergone the process of iron ore reduction which in blast furnace/BOF steelmaking is almost completely (more than 90 %) effected by carbon. Accordingly, scrap-based steelmaking produces much lower CO, emissions. Whereas high-efficiency BF/BOF steelmaking generates about 1.85 tonnes of CO,, including cokemaking and sinter plants, the equivalent figure for an EAF is only about 400 kg CO, per tonne of crude steel, assuming a CO, freight of 570 g per kWh of supplied electrical energy. This is a significant advantage with a view to the objective of reducing CO, emissions Scrap will certainly remain the most important input material in electric steelmaking. Nevertheless primary raw materials from direct reduction of iron ore have been playing a gradually increasing role recently. Direct reduced iron can be produced, for example, by the Midrex process, which was developed at the end of the 1960s and during the 1970s. The most commonly used reductant is natural gas converted to CO and H,, In 2007, the record year of steel production, approximately 65 million t of DRI were produced, with India accounting for the biggest share, namely 18 million tonnes. High or ultra-high power electric are furnaces with specific transformer ratings of 1,000 kVA per tapped tonne achieve tap-to-tap times of some 40 minutes today. A modern ultra-high chemical power electric arc furnace (UHCP), which can be operated in a burner mode using eight 8 MW natural gas-O, burners, achieves production rates of 320 tonnes per hour. The trend towards ever higher output rates started in the late 1960s when it became clear that it was not a technologically reasonable option to retrofit the existing open hearth furnaces, which were major consumers of scrap, with dust removal equipment. This trend gained additional momentum in the 1970s when electric furnaces with water-cooled walls were developed. However, the fact that the secondary voltage was initially limited to 1,000 V brought the continuing capacity growth to a temporary halt. When this limit was overcome, high-capacity furnaces for 100 to 150 tonnes charge weight could be built. Higher secondary voltage resulted in increased radiation from the electric arc on the furnace walls. When foaming slag practice was introduced, the electric arc could be covered during flat bath operation, shielding the furnace walls against excessive radiation. The dramatic improvement in efficiency, from tap-to-tap times of about three hours in the 1960s down to approximately40 minutes, was accompanied by a corresponding decrease in energy consumption. Blectric energy consumption was brought from 630 kWh per tonne down to 345 kWh pet tonne, or -45 %: electrode consumption was cut by 83 %, fiom 6.5 kg per tonne to 1.1 kg Although the dramatic reduction in tap-to-tap times was the main driver of this positive development in consumption figures, it must not be left unmentioned that the low rate of electrode consumption was also enabled by the simple, though ingenious practice of spray cooling the graphite electrodes We certainly also owe this impressive efficiency improvement to a great deal to the cooperation work between the Electtic Are Furnace Committee of the Steel Institute VDEh, the USA and Japan. The momentum triggered by exchange of ideas and experience is not to be underestimated. Since 1982 the Electric Arc Furnace Committee of VDEh has organized seminars with a special focus on electrical engineering aspects of the electric arc furnace. Under the expertise of Prof Dr. Klaus Timm, Professor Emeritus at the University of the Federal Armed Forces Hamburg (nowadays Helmut-Schmidt-University) and, since the mid- 1990s, co-organized by Prof. Dr. Klaus Kriiger, University of the Federal Armed Forces Hamburg, these seminars ate intended to provide metallurgists important knowledge about electrotechnical aspects of electtic are furnaces, The German-speaking seminar was last held in May 2009, for the 30th time. Since 2000 there is also an English equivalent. The most recent one in a line of eight was held in May 2008 The initiative to publish this book came from Dr. Ben Bowman, when he retired as consultant to GrafTech. Dr. Ben Bowman, born in 1939, studied physics at the University of Liverpool where he was awarded a B.Sc. in 1961, followed by a Ph.D, in electrical engineering in 1965 based on research into electric ares under the guidance of Professor Harry Edels. He commenced his involvement with arc furnaces in 1965 at the Arc Furnace Research Laboratory in Sheffield, England, of the United Steel Companies (later British Steel Corporation), at that time the operator of the largest electric steelshop in the world (6 furnaces, 7.3 m dia., 40 MVA). It is now a museum! He joined UCAR Carbon Company Inc.(Now GrafTech Intemational) in 1971 at the European Headquarters in Geneva, Switzerland, as a customer technical service manager for are furnace technology. Dr. Bowman has authored more than 40 publications and seminars in the field of furnace ares, AC and DC furnaces, electrode and refractory consumptions. In 198] he suggested the addition of a reactor on the primary of are furnaces in order to stabilize the long ates being sought by steelmakers after the successful introduction of water-cooled panels. This idea was taken up in South Africa in 1983 and has since developed into the now standard High Reactance Are Furnace, That same year Dr. Bowman created an entertaining and instructive arc furnace computer game for steelmakers (at the Ist European Arc Furnace Conference in Aachen) In 1993, Dr. Bowman became a Senior Corporate Fellow and accepted a position at GrafTech’s Parma Technical Center in Parma, Ohio, where he added electrode development to his continuing studies of arc furnaces. In 2003 he retired but has kept an active interest in arc furnaces. He has acted as consultant to GrafTech while writing the book on Arc Furnace Physics with Prof. Klaus Kriiger. Klaus Kriiger, born in 1967, took up a career as an officer of the Federal Armed Forces after his graduation from secondary school. Doing so, he studied mechanical engineeringat the University of the Federal Armed Forces in Hamburg. On completion of his studies at the end of 1992 he was honoured with the Bottcher award for his results After 1% years in the army, Klaus Kriiger accepted to retum to the university in order to conduct research for a doctoral thesis. Being a scientific assistant to Prof. Klaus Timm from 1994 to 1998 he acquired profound knowledge about the electric arc furnace. In November 1997 Klaus Kriiger finished his dissertation about the development of a closed-loop power control for ‘AC are furnaces. Leaving the army Klaus Kriiger joined to the Robert Bosch GmbH, where he was in charge of the production of microhybrid automotive Circuit Units from 1998 to 2003. In 2003 the chair of Process Data Processing and System Analysis at the University of the Federal Armed Forces Hamburg was offered to him. Assuming this professorship, he reconnected to his roots of electric arc steelmaking, which now presents an important field in his research activities Ben Bowman and I first met in 1971, the year when I started my industrial career as an electric steelmaker, and I have held him in high esteem ever since. | first met Klaus Kriiger when he was working on his doctoral thesis under Professor Timm. With great pleasure have I been following his great achievements in not only continuing but greatly advancing the research issues commenced by Professor Timm, Jam glad that 1 have been able to bring together two authors from two generations and living some 8,000 km apart — in Ohio/USA and in Hamburg/Germany — to join forces for this work. Since the middle of 2008, they have bad a one to two hour telephone conference every Wednesday afternoon to exchange and coordinate theit ideas. The result is a highly comprehensive, excellent encyclopedia of the physics and electrics of the electric are furnace, a technology which still holds much development potential to be tapped. Congratulations for both authors on this remarkable work Dieter Ameling Chairman of Electric Arc Furnace Committee of VDEh from 198] to 1986 Chairman Steel Institute VDEh from 1998 to 2008 President German Steel Federation from 2000 to 2008,Preface Ben Bowman/ Klaus Kriiger It may seem surprising to find a book devoted to the physics of an industrial process, and one not familiar to many people. Yet the scope for applying physics to arc furnaces is wide indeed. Various states of matter are involved ~ the plasma of the arc itself, steel and refractories in the solid state, steel as liquid, slag as an ionic liquid, and the gases and fumes evolved within the furnace The are consists of electrically conducting, high temperature plasma in which the electrical power is released. Here the important heat transfer mechanism of radiation touches on quantum physics. Since high currents are used high magnetic forces are exerted on all the conducting components, including the arc and metal bath. A range of arc instabilities is observed which have important consequences on the power system, Unsteady currents also lead to vibrations in the mechanical secondary system. Sometimes resonances and near-chaotic motions are seen The electrical operation is complex. As an element of an electric circuit the arc exhibits an irregular, non-sinusoidal voltage waveform which cannot be treated either as resistor, capacitor nor inductor; it has unique characteristics. Modelling is usually performed in an empirical fashion The major input material, steel serap, comes in a large variety of shapes and size, apart from its chemical variability. Probably no two charges are alike so statistical methods are required. The physics of this material calls on particle mechanics, such as packing, and electromagnetic properties. Control of the process, which is via a relatively crude vertical displacement of the clectrode/ arm/cable system weighing perhaps 30 tonne per phase, is much slower than the variations within the furnace. The attainment of an optimum control strategy is still an open question, so open that each furnace manager seems to have his personal answer One of the difficulties with understanding what goes on inside an arc furnace is the problem of measurement; it is extremely difficult to obtain access to interesting parameters within the furnace. Apart from the actual physical danger, there is no easy way to protect an instrument in such an environment. An are furnace running at full power during scrap meltdown, with its noise, flames and vibrations, without doubt is initially sather terrifying. Today, as furnaces have become ever more powerful, there is no obvious way to ‘see’ what occurs inside. In fact rarely is an arc seen from outside; it is either surrounded by scrap, covered with foaming slag or obscured by fume. In certain periods, under very stable conditions, even the existence of an arc can be questioned One of us was fortunate to have started work on arc furnaces in the 1960's. At that time furnaces were of lower power, refiactory lined, using less oxygen, and foaming slag techniques were not known. As a consequence arcs were more often visible, or could easily be made so. Heat times were also much longer so that a delay in arranging for special measurements (high speed photographs for example) could easily be accommodated by a cooperative steelmaker, Much of the information obtained in this period has facilitated a physical interpretation of later data collected ‘extemal’ to the furnace, for example, electrical measurements Most of the engineers and managers who have responsibility for arc furnaces do not have the opportunity or time to study all the widely dispersed, previously published information on arc furnaces. It is our intention that this book supplies at least some useful information which is helpful in their continual quest to extract improved performance from our mutual acquaintance ~ the electric arc furnaceUnits SI units are used throughout this book but because the anticipated readership is expected to be mainly involved in the metal producing industries some more familiar units are also used In particular electrical energy will be quoted in kWh rather than MJ and of course metal weights are metric, written as tonne, to avoid any misunderstanding, In some cases quantities involving volume are quoted in cm? even though this is not a recommended subunit. This is justified for example for arc radiation since, of the alternatives, mare too large and mm? are too small. In these preferred units radiation falls into the kW/em? range, much easier to grasp than say 10° W/m’. Also since radiation, arc power and voltage gradient are closely linked the latter is conveniently quoted as V/cm. Some spectroscopic and physical energies, such as dissociation and ionisation energies, are quoted in eV rather than J (1 eV = 1,6022:10"” J), Temperatures are quoted in either °C for items familiar to the steelmaker (steel, slag, etc ) or the scientifically more convenient K (0 °C = 273 K) for arc physics and chemical reactions Magnitude of the power of a large are furnace An are furnace is an impressively powerful unit. Those working around one each day may tend not to appreciate how powerfill it really is. A comparison with machines more familiar to the wider public helps to put it in perspective as shown in figure 1. Both the TGV (the French Train & Grande Vitesse, 300 km/h) and cruise liner Queen Mary 2, (100,000 tonne) have electric motors. The Airbus A380 power has been calculated at cruising speed from the speed (1000 km/h), weight (560 tonne) and lift/drag ratio (19.5). Tev Queen Mary 2 Airbus A380 Are furnace: 12Mw 80 MW 80 MW up to 150 MW Fig. 1: Comparison of powerful machines (*with permission from Steve Brimley) And the chemical power has not been included for the arc furnace! Let us compare the are furnace with, say, a 300 t basic oxygen furnace (BOF), for which only chemical power is released. Assume 4.5% C (13.5 tonne) eliminated in 0.4 h. For 80% CO /20% CO, generation the energy release by decarburisation is 4.2 kWh/kgC. The average chemical power is therefore 142 MW. So the BOF chemical and arc furnace electrical powers are of similar magnitude Precision Asa consequence of the lack of precise data and the overall variation in many of the important parameters, many calculations result in estimates. ‘To avoid a misleading indication of the precision it is then necessary to qualify these calculated values with such descriptive terms as — ‘about’, ‘approximately’, ‘of the order of”, ‘~’, etc., which can become unfortunately a little repetitive. In like manner, often a graph of measured data usually shows an obvious scatter so that a separate statement of the errors is unnecessaryAcknowledgements Many people busy in all fields of arc furnace steelmaking — steelmakers themselves, suppliers, researchers and colleagues ~ have supporied the authors by making measurements and trials at production facilities possible, by discussing phenomena and by providing results. Thanks to all of them for their help and their interest in understanding the fundamentals of arc furnace steelmaking Access to technical material is essential for a book of this type, so we would like to specially thank the following various groups and individuals: University of Liverpool, Department of Electrical Engineering and Electronics. We thank the late Professor Harry Edels, followed by Professor Gordon Jones, for the steady supply of the ULAP reports (University of Liverpool Arc Project). Thank you to Professor Klaus Timm, University of Federal Armed Forces Hamburg, who has provided a string of very relevant theses from 1970 to 2003 written by himself and by his assistants concerning are furnaces. Many of these are referenced in this book. He also introduced one of us (Klaus Kriiger) to the arc furnace The Commission of the European Communities has sponsored several research projects related to arc furnaces that are available publicly as reports, namely written by the British Stee! Corporation, the VDEh-Betriebsforschunginstitut and the Institut de Recherches de ta Sidérurgic Frangaise. Throughout Ben Bowman's years working for GrafTech International the librarians of the Technical Information Service in the research and development department at Parma have provided a very effective supply of information. In particular he would like to thank Linda Riffle, who often unearthed relevant material in answer to some rather unusual requests, and latterly Theresa Nawalaniec and Christine Martuch Some very valuable information is available on the internet In particular the National Institute for Science and Technology (NIST), a branch of the US Government, (http://physics. gov) offers, free access, much data on elemental physical and chemical properties, such as Partition functions, energy levels, line frequencies, ete. Thank you to Professor Yasunobu Yokomizu of Nagoya University who kindly gave us permission to access their database with calculations of electrical conductivity of gases containing iron and calcium vapours Last but not least we like to thank Professor Dieter Ameling who encouraged us to finish this book and the Verlag Stahleisen for publishing it,Contents XI Contents 1 Are generalities 1 1.1 General are physics 1 12 Introduction to industrial ares 2 13 Sources of information on ares 2 2 DC graphite electrode laboratory arcs 5 21 Ate stability 5 22 The plasma jet 5 23 Measurements of arc thrust 6 24 Mass flow rate 8 25 Are voltage and voltage gradient 8 26 Are shape 10 27 Power balance at the cathode spot ul 28 Current density and time constant of the cathode spot 13 29 Convective power of the cathode jet 13 2.10 Radiation 14 211 Power balance distribution between convection and radiation 15 212 Temperature profiles 7 213 Summary: The structure and power balance of the DC graphite are 19 3 Furnace arcs 20 31 AC arcs 20 3.1.1 Generalities 20 3.1.2 High speed photos of AC ares 21 3.13 Are voltage waveforms 23 3.1.4 Analysis of correlated electrical and high speed photographic records 25 3.1.5 Effect of ambient temperature on the arc voltage waveform 26 3.1.6 Diffuse ares 28 3.1.7 Waveforms recorded on large are furnaces 28 3.1.8 Arc lengthening measurements 30 319 — Voltage gradients in extremely long ares in air 31 3.1.10 Radiation measurements from arc and flame = 3.1.11 Are flames 32 3.1.12 Summary of measurements on AC ares 33 = Measurements on DC arcs in furnaces 34 3.2.1 Voltage gradients 34 3.22 Areshape 35XI Contents 3.23 Are voltage/length relationship 35 3.2.4 Arc voltage and current 36 3.2.5 Average values 37 4 Electrical conductivity, radiation and temperature in furnace arcs 38 _ Measurements of arc temperature 38 Plasma composition in steelmaking furnaces 38 4.2.1 Plasma composition for an arc buried in foaming slag 40 43 The electron density and the electrical conductivity 4l 4.3.1 Electron density of plasma containing iron vapour 4l 43.2 Electron density for an arc covered by foaming slag 41 43.3 Electrical conductivity 41 44 Estimate of arc diameters from conductivity 44 45 Radiation in high current ares 44 45.1 Radiation from neutral atoms 44 45.2 Radiation from ions and electrons 45 4.5.3 Free-bound radiation 45 45.4 — Optically thin and optically thick plasmas 45 45.5 Radiation from plasmas containing iron vapour 46 46 Radiation based estimation of the plasma temperature 52 4.6.1 — Theoretical estimates of axial temperature 52 46.2 Temperature gradients due to radiation diffusion 52 4.7 Radiation loss from large diameter ares 53 4.8 An ‘explanation’ for the voltage gradient 55 49 Conclusions concerning radiation power of furnace arcs 55 5 Arc terminations in steelmaking furnaces 56 5.1 Measurements of current density for graphite and steel terminations 56 5.2 Power balance at the anode spot on graphite 58 5.3. Analysis of the spots on steel terminations 59 5.4 Electrode consumption data for short duration ares on graphite and steel 60 55 Power balance at the electrode spots on solid steel 61 5.6 The cathode spot on liquid steel 63 6 Magnetic fields and forces in are furnaces 64 6.1 Magnetic force on AC furnace ares and electrodes 64 6.1.1 Magnetic force on the are due to a finite sized liquid steel bath 66Contents XU 6.12 Summary of fields 67 613 Deflection during scrap boredown 68 62 Magnetic force on the arc in a mono-electrode DC furnace 68 621 Arc deflection correction for mono-electrode DC 69 6.2.2 Boredown 70 63 Conclusion on are deflection 710 6.4 The Lorentz force in a bent arc column 70 65 Skin and proximity effects nN 6.6 Magnetic field shielding due to scrap n 6.7 Proximity effect in furnace electrodes 14 6.8 ‘Torque on electrodes in an AC furnace ge 6.9 Magnetic fields around AC and DC furnaces 75 7 Arc instabilities 77 TA Observed arc instabilities 77 ap) Reasons for arc movement ag 7.2.1 Instability due to cathode spot ‘crater’ growth B 7.2.2 Asymmetrical conductivity near the cathode spot 78 7.23. Arc jumping over pieces of scrap 78 7.24 The scrap termination effect 79 7.25 Effect of DC tip shape 80 7.2.6 — Are plasma instabilities 80 13 ‘The helical rotation of the arc column 80 73.1 Frequency of arc rotation 81 7.3.2 Instability of very long ares 81 74 Aerodynamic instabilities 81 7.4.1 — Reynold’s number for an arc 81 74.2 Turbulence in the arc column 82 74.3 The stabilising effect of the plasma jet 83 1S Rates of change of arc voltage in AC furnaces 84 76 Effect of slag 85 8 The electric circuit of arc furnaces 86 81 The AC circuit 86 8.1.1 Magnetic fields and reactance 88 8.1.2 Reactance in the three-phase circuit 89 81.3 Effective phase reactance of a three-phase circuit 89 8.1.4 — Reactance calculation 90 8.1.5 Equivalent reactance 91xv. Contents 8.1.6 Measurement of phase reactance 92 8.1.7 Actual furnace reactances = 82 The DC are furnace power circuit 93 82.1 Six-pulse rectifier system 93 8.2.2 Line current ina 12-pulse system 96 823 — 18-pulse and 24-pulse systems 97 8.2.4 — Electrical characteristics of the DC furnace 98 825 Active-reactive power characteristic 98 9 Electrical performance of the arc furnace circuit 100 9.1 Measurable parameters 100 9.1.1 Voltage and current 100 91.2 Power 101 9.1.3 Three-phase induction meters 101 9.1.4. Definition and measurement of reactive power 102 9.1.5 Measurement of reactive power with a kVarh meter when harmonics exist 102 9.1.6 — Alternative definition of reactive power 103 9.2 The measurement of arc voltage 103 921. Magnitude of the induced error voltage 104 922 Effect of mast height variations 105 9.2.3 Variables that affect performance 106 92.4 — Effect of unsteady arc voltage 107 9.2.5 Effect of power factor on current waveform 108 93 The three-phase arc furnace circuit 109 93.1 The third harmonic 109 9.3.2 Easier passage through current zero 110 94 Electrical characteristics during operation 110 9.4.1 Establishing the ares 110 9.4.2 Variation of parameters through a heat Wl 9.43 Electrical measurements on arc furnaces: dimensionless relations il 9.4.4 — Relationship between active and reactive power 14 95 Unbalanced conditions in the three-phase AC circuit 115 9.5.1 Range of arc voltage and current imbalance 116 9.5.2 The three-dimensional operating space of arc voltage 118 9.6 Modelling of the electrical operation us 9.6.1 Arc asa linear ohmic resistance 118 9.62 Rectangular arc voltage waveform 119 9.63 — Cassie (circuit breaker) models 120 9.6.4 — Are voltage as a system of harmonics 121 9.6.5 The “operating or atc reactance‘ model 123Contents x 9.6.6 The measurement and definition of operating reactance 124 9.6.7 The measurement of the ‘arc reactance’ 125 9.6.8 Relation of operating reactance to power factor 126 9.6.9 An empirical link between the arc reactance and the arc resistance 126 96.10 Asymmetrical are voltage swings with harmonics 126 9.6.11 Misfiring 128 97 Summary 128 10 Physical characteristics of steel scrap 129 10.1 A tube model for scrap 129 10.1.1 Packing characteristics 129 10.1.2 Tubes 130 10.1.3 Number of contacting neighbours 130 10.1.4 Current flow between pieces 130 10.1.5 Contact shape and pressure 131 10.1.6 Resistance 131 10.1.7. The effective resistivity of scrap consisting of tubes 134 10.2 Results using the tube model 134 10.2.1 Effect of scrap size and aspect ratio 134 10.2.2 The effect of temperature on effective resistivity 135 10.23. Magnetic properties of scrap 135 10.2.4 The skin depth in scrap 137 10.2.5 Application to other shapes of scrap 138 10.3 DC furnaces 138 10.4 Scrap impedance 138 11 ‘The meltdown process 140 11.1 Heat transfer are to charge 140 11.1.1 Pattition of melting time: Arcing directly to scrap or to liquid 140 11.1.2 Heat transfer to solid steel arc terminations 140 11.1.3. Heating by the impact of the plasma jet 141 11.1.4 Normal impact of the are against a steel plate 142 11.1.5 Heat transfer from an angled arc plasma jet 144 11.1.6 Heat transfer to a bar 145 11.1.7 Summary of heat transfer to a steel plate 146 11.1.8 Radiation exchange between the arc and the charge 146 11.2 Heat transfer from are to liquid 146 11.2.1 Jet impact onto a liquid 146XVI Contents 11.2.2 Jet impact onto a liquid covered with a lighter layer (‘slag covered steel”) 147 11.3 Splashing by the are as a heat transfer mechanism 147 11.3.1 Splash modes 148 11.3.2 Splashing rates 148 11.3.3 Splash distribution and momentum 149 11.3.4 Size distribution of splash drops 150 11.3.5 Splashing by an arc 150 11.3.6 Liquid depression transient 151 11.3.7. Splashing due to a mobile arc 151 11.3.8 Splash rates due to the are 153 11.3.9 Heat transfer due to splashing 154 11.4 The general scrap meltdown pattern 154 11.4.1 Process stages 154 11.4.2 Diameter of the boredown pit 155 12 Liquid bath phenomena 158 a Stirring in arc furnaces 158 12.1.1 Current distribution and self-magnetic stirring 158 12.1.2 Stirring by the current in DC furnaces 158 12.1.2 Velocity measurements in electromagnetically pumped liquid metals 159 12.2 Stirring forces 160 12.2.1 Lorentz force in AC furnaces 160 12.2.2 Lorentz force in DC furnaces 161 12.2.3 Comments on Lorentz force for AC and DC 162 12.3 Stirring due to arc jet 162 12.4 Heat tansfer due to electromagnetic pumping 163 12.5 Depressions in the liquid due to the are jet 163 12.6 Electrode lifting from short circuit 164 12.7 _ Lifting of the bath surface by CO bubbles 165 12.8 Effect of CO bubbles on arc length 166, 13 Slag effects in are furnaces 167 13.1 Effect on are stability 167 13.2 The effect of foaming slag on furnace operation 167 13.3 Detection of are burial by foaming slag 168 13.3.1 Reference signal 169 13.3.2 Thermal status of water-cooled panels 169 13.3.3 Bath level measurement 169 13.3.4 Light emission 169Contents XVI 13.3.5 Electrical characteristics 170 13.3.6 Sound emission 171 13.3.7 Vibrations of the furnace shell 171 13.3.8 Use of the slag signal 172 13.4 Slag coating on water-cooled panels in 13.5 Slag properties and arc existence 173 13.6 Current through foaming slag 115 14 Mechanics of the secondary system 176 141 Vibrations and damping 176 14.2 Forced vibrations of an electrode column 177 14.2.1 Natural (Eigen-) frequencies 178 14.2.2 Response of the column to a cyclic driving force 178 143 Electromagentically forced vibrations of the secondary system 179 14.3.1 Magnitude of the electromagnetic forces 180 143.2 Deflection of the electrode/arm/mast system 182 14.3.3 Response of the electrode/arm/mast system to cyclic forces 182 143.4 Calculated resonant frequencies 183 14.3.5 Resonant frequencies measured on arc furnaces 183 14.3.6 Rotation of the electrode tip 185 14.4 Forces occurring during furnace operation 185 14.4.1 Examples of forces generated during scrap melting 185 14.4.2 Resonant conditions observed on some arc furnaces 186 14.43 Avoidance of resonance 187 14.44 Comments on modes of vibration 187 14.5 Vibrations due to movement of the mast 187 14.5.1 Electrode arm vibrations 187 14.5.2 Hydraulic system vibrations 188 14.6 — Motion of the secondary of DC furnaces 190 14.7 Secondary mechanics and regulation 190, 15 Power control of are furnaces 191 15.1 Electrode regulation ae 15.1.1 General requirements 191 15.1.2 Controlled System ee 15.1.3. The importance of secondary dynamics 195 15.1.4 Controlled variable 196 15.1.5 Control strategy 198 15.2 Optimising electrode movements 200XVUL Contents 15.2.1 Acriterion for judging regulation performance 200 15.2.2 A computer model 200 15.3 Superposed power control 202 15.3.1 Transformer tap control 203 15.3.2 Reactor tap control 205 15.4 Ladle Furnace requirements of the regulation system 206 16 Heat transfer properties of fume in the are furnace 208 16.1 Fume formation mechanisms 208 16.2 Fume generation rates in steelmaking furnaces 209 16.2.1 Identity of fume particles inside the are furnace 210 16.2.2 Size distribution of fume particles 2 16.2.3 Fume generated by hand held metal cutting tools 212 16.2.4 Sectional area of fume 213 16.3. Absorption and scattering of radiation by fume 214 163.1 Absorption efficiency and depth 214 16.3.2 Significance of absorption and extinction depths 216 16.33 Role of fume in heat transfer 216 164 Power transfer bath to furnace without fume 217 16.5 Heat transfer during foaming slag or DRI operation, with/without post combustion 217 17 Noise from arc furnaces 220 17.1 Loudness and power 220 17.2 Sound emission of AC furnaces 221 17.2.1 Measurements around are furnaces 221 17.2.2 Measurements of sound pressure level near furnaces 222 17.2.3 Total power emitted outside the furnace 222 17.2.4 Acoustic power within the furnace 223 173 Sound emission of DC furnaces 223 174 Mechanisms of noise generation 224 17.4.1 Acoustic model 224 17.4.2 Reduction of noise through the heat 225 17.4.3 Noise in the harmonics of the AC furnace (pulsating cylinder) 226 17.4.4 Sound emission due to sideways displacements of the are 227 17.44 Broadband noise a 17.4.5 Turbulent jet noise 228 17.4.6 Influence of the electrical operating point, 229 2 Acoustic power in the meltshop 230Contents XIX 18 References 231 19 Appendix 243 19.1 Properties of steel 243 19.2 Properties of graphite 244 193 References (Appendix) 245P 1 Are generalities 1 Arc generalities The arc is a member of a family of electric discharges through gases, which includes glow discharges, sparks, lightning, corona discharges, ete. Arcs are in a sub-class of high-pressure discharges. The identifying characteristics of an arc include the follow- ing: High temperature plasma (5000 K, and hence strong radiation), constituents in local thermodynamic equilibrium, pressure above about 0.1 bar and high temperature electrode terminations, often melting or evaporating It is worthwhile mentioning that the name ‘are’ refers to the shape of the discharge at low current (say less than 10 A) between horizontal electrodes at atmospheric pres- sure. Due to the vertical convection currents created by the hot plasma the discharge path is dragged upwards at the centre of the gap so that the path of current becomes arched between the electrodes. But at higher cur- rents, say over 100 A, plasma jets form near the electrodes, generally directed along the electrode axis; the momentum of such jets usually dominates any curvature due to fransverse convection currents so that an arched shape is no longer seen 1.1 General are phy Between the two electrode terminations of an are the current is carried by a high tempera- ture plasma, an electrically conducting me- dium. The temperature is so high, of the order of 10,000 K, that a fraction of the gase- ous constituents (mainly atoms) becomes ionised in sufficiently energetic collisions. At this temperature the average velocity of, for example, a nitrogen atom is 3500 mis Molecular gases such as nitrogen Nz or oxy- gen O2 are already dissociated into atoms at lower temperatures, around 5000 K. Figure 1 shows the distribution of constituent energies of motion at $800 and 11,600 K, where the kinetic energy is expressed in electron-Volts, eV. This is the energy acquired by an elec tron in traversing a potential difference of 1 Volt. This scale is used because the energy to ionise an atom ~ the ionisation potential — is traditionally expressed in such units. | eV corresponds to a temperature of 11,600 K. ics The ionisation potential is a function of the electronic structure of the atom. For gases such as Nz, Q; and carbon-monoxide CO, dissociated into atoms, the ionisation poten- tial is in the 11 to 14 eV range. For metallic vapours it is much lower, 7.9 eV for iron for example. Even so only a very small fraction of the particles have enough energy to ionise an atom by collision. In the plasmas that con- cern us in industrial furnaces only a small fiaction of the constituents is ionised, the majority remain as neutral atoms. T_T 1 —"11600« === 6800 K Probability o 1 2 3 4 565 6 7 8 Particle energy in eV The Maxwell-Boltzmann distribution of particle energies Fig. By ionisation both an electron and an ion are generated, In the bulk of the arc plasma the negative charges of the electrons are bal- anced by the positive charge of an equal number of ions. Thus the plasma is electri- cally neutral. Current is carried by both the electrons and the ions but, because of their much lower mass and thus higher mobility, the electrons carry the majority. The imped- ance to current flow is due to collisions, principally between the electrons and ions and the electrons and neutral atoms. Between collisions the electrons are accelerated by the electric field, gaining the energy lost in the previous collision The work done by the external voltage in driving the current through the plasma resultsin heating, maintaining the high temperature Power transfer is via radiation, convection, thermal conduction and kinetic energy to the surroundings. The mean distance between collisions amongst the particles in a plasma — the ‘mean free path’ — in an are is of the order of some 10*m Compared to the thermal energy of random- ised motion, the directional energy change acquired from the electric field between two collisions is very small, One can therefore assume that the velocities of the electrons are still accurately described by the Maxwell- Boltzmann distribution and hence can be identified with a temperature. The same applies to the ions and the neutral particles. These equalities in temperature are com- monly referred to as ‘local thermodynamic equilibrium’ (LTE). ‘Local’ means that within any small volume of the plasma, much less say than | mm’, a temperature can be used to describe the particle velocity distri- bution. Thus it is legitimate to talk of tem- perature gradients, With decreasing pressure below atmospheric the mean free path increases. Assuming a temperature of 10,000 K, below about 0.01 bar it is no longer valid to assume that the electron and ion temperatures are equal; the electron temperature diverges from, and rises above, the ion temperature. Some authors have chosen to define this pressure level as the lower bound to the electric arc, the lower pressure discharges falling into a general category of ‘low pressure discharges’ or ‘glow discharges’. The common fluorescent lamp is an example of a low pressure dis- charge. Some important properties of the plasma, such as conductivity and radiation, are dependent on the identity of the constituent gas or vapour. In industrial furnaces produc- ing metals or alloys metallic vapours exert a very strong influence on the plasma properties; conductivity and radiation are much higher than for common gases like nitrogen, oxygen or carbon-monoxide. This 1 Are generalities fact means that much of the early work on arcs, with carbon or graphite electrodes in air, cannot simply be scaled up to industrial applications At the electrode terminations the ate is usu- ally attached at a well-defined spot. Within the electrodes, metals or graphite, current is cattied by electrons only, not ions. At the cathode electrons must be extracted from the electrode whereas at the anode electrons can fall’ into the electrode. Thus the physical mechanisms that occur at the cathode are different to those at the anode 1.2. Introduction to industrial ares Probably the type of are familiar to most people occurs in welding. When the welding rod, usually a metal electrode, makes contact with the piece to be welded a short circuit occurs. Upon separation, an arc is initiated between the rod and the workpiece, emitting the familiar very bright, bluish light, creating a little hissing noise and some fume. The electrode melts at its tip, spraying drops of liquid metal onto the workpiece, which is simultaneously heated by the locally intense heat, To protect the weld from oxidation the electrode is usually coated with a material which forms a protective slag on top of the liquid metal Atos in steelmaking furnaces are similar in many ways to welding arcs but scaled up enormously in power. In a furnace the graph- ite electrodes do not melt; only the “work- piece”, that is the scrap, melts, Welding typi- cally operates with an arc voltage about 20 to 30 V, currents in hundreds of ampere range and powers in the kilowatt range In steel- making furnaces of different sizes arc volt- ages range between 100 to 800 V while cur- rents run usually between 10 and 150 kA. In large furnaces a power of over 100 MW is quite common today. 1.3 Sources of information on arcs As is evident from the welding example an are emits copious amounts of radiation When the electrodes are made of carbon or graphite, the fact that these materials can support temperatures up to 3800 °C — the| 1 Arc generalities sublimation temperature — means that the radiation is especially intense, particularly from the anode. Thus an early application, in the late 19” century, of the carbon arc was for public illumin However, it was in the first half of the 20" century that much of the basic physies of arcs was established. The main motivation was again the study of illumination, initially for searchlights and later cinema projectors. The publications by Finkelnburg [Fin44], [Fin47], [Find8], Finkelnburg and Maccker [Fin56] and Maecker [MaeS5] could be considered as the foundation of are physics. Ares occur when it is necessary to interrupt or ‘break’ currents in electric systems either for emergency or for normal switching In a circuit breaker the arc which inevitably forms between the separating contacts is undesirable since it has the potential to melt the contacts or damage the insulation; it must be extinguished quickly and safely. Accord- ingly a large effort has gone into studies of circuit breaker arcs. Interruption of the cur- rent flow in an AC circuit involves prevent- ing the ase from re-igniting after current zero, whereas the energy stored in the inductance tends to maintain current flow. The factors that determine the are plasma conductance decay rate are the important parameters stud- ied Of course in arc furnaces the opposite of cur- rent interruption is normally desired; passage through current zero should ideally occur smoothly and without hesitation; even so, circuit breaker research provides some indi- cation of conditions to be avoided if possible and, inversely, can point to conditions fa- vourable to stable, continuous arcing In an AC steelmaking furnace both graphite electrode and steel termination altemate be- tween anode and cathode. We therefore have four electrode terminations to discuss — cath- odes and anodes of graphite or steel ~ with the added complication that the steel may be solid or liquid Examples of constricted ares In metallurgical furnaces, with which we aie most concerned in this book, the ares are described as ‘open’, in the sense the arc movements are not purposely testricted. Ares however can be operated inside water-cooled tubes for example. When the tube consists of a stack of disks insulated from one another the resulting stable plasma has proved very useful for the study of the properties of high temperature plasmas By passing a gas through the are and allow- ing the hot gas to escape via a circular open- ing we create a plasma torch, which can be used for cutting or high temperature chemis- tty High pressure are lamps operate within an insulating, transparent tube. In a gas-blast circuit breaker the arc appearing between the separating electrodes is forced to run within a constriction through which a cold gas is also forced. This forced cooling extinguishes the are when the alternating current passes through zero. Ares under vacuum Several applications are common in steel- works — vacuum circuit breakers, ingot refining and vacuum are degassing. Of course the word vacuum is relative; it really means very low pressure. In the vacuum circuit breaker the low pressure ensures that there are few collisions between particles. The interrupting characteristics are then dominated by processes occurring at the electrode surfaces In an ingot refining fur- nace an arc operates between the lower end of an ingot and its melted pool below. Low pressure is maintained in order to facilitate the removal of gases as the ingot is melted. Similarly, in a vacuum arc degassing furnace graphite electrodes are used while arcing under low pressure Alternating and direct current The processes occurring at the two electrode terminations of the arc ~ the anode and cath- ode -- are significantly different. Often, but not always, more power is liberated at the anode than the cathode. In welding thisdifference can be exploited to advantage in certain situations hence welding employs both AC and DC power systems The majority of steelmaking are furnaces are AC but a significant minority are DC. Al- though the reasons for the development of DC furnaces have similarities to the use of 1 Are generalities DC in welding (anode/cathode differences) the power supply arrangement is very differ- ent AC furnaces operate with three graphite electrodes and three arcs whereas the com- mon DC arrangement is with one graphite electrode, one bottom electrode and one arc The subject is treated in more detail in a later chapter2 DC graphite electrode laboratory ares 2 DC graphite electrode laboratory arcs Free-burning DC ares in air between graphite (or carbon) electrodes serve as a good intro- duction to ares in general, They are easy to establish, the electrodes consume slowly and little fume is generated. For these reasons they are ideally suited to laboratory condi- tions. It was this type that was studied exten- sively by Finkelnburg, Maecker and others in the mid 20" century. Fig. 1: Photo of typical DC arc between a graphite tubular cathode and a graph- ite plate anode (current about 600 A) The photo in figure 1 shows typical features: A bright plasma core increasing in diameter fiom the cathode to the anode; a weaker aureole outside the core; bright cathode and anodes spots on the graphite terminations. What cannot be appreciated fiom a single photo is that there is a very fast plasma flow from the cathode to the anode. Its reflection off the anode plate gives rise to broad plasma disk around the anode spot. We will learn below that this plasma jet is a consequence of the cathode spot having a smaller diameter than the are column. According to Maecker’s spectroscopic measurements, figure 2, the boundary of the core is at a temperature of 9750 K while the aureole edges ate at 7350 and 6220 K [Maes3] sok 4000 K 6220 K 7260 K e760 k +1000 k Fig. 2: Temperature isotherms for the 200 A carbon atc due to Maecker [MaeS3] 2.1 Are stability Finkelnburg [Fin44] had reported that when the current exceeded 400 A the arc became spontaneously unstable, The plasma column rotates with a corkscrew shape about the cathode/anode axis. Motion has very little amplitude at the cathode but increases with distance from the cathode The column sweeps out a cone in space with its apex at the cathode Although this instability is interesting in itself (it will be examined in more detail in the chapter on instabilities) it is inconvenient for many measurements and also extremely noisy. However Montgomery and Sharp {Mon69a] were able to eliminate this instabil- ity by machining a crater in the cathode. The necessary crater diameter inereased with current, reaching 35 mm at 2 kA. With such a shaped cathode they were then able to make measurements on stable arcs up to the limit of their power supply of about 2.2 kA. 2.2 The plasma jet The mechanism and theory of the plasma flow was demonstrated by Maecker [Mae55] in 1955 in a brilliant set of experiments The driver is the self-magnetic compression exerted by the Lorenz force. For a cylindrical2 DC graphite electrode laboratory ares conductor the compression is purely radial But the conducting plasma column in this arc does not maintain a cylindrical shape; it expands away from the cathode and as a result there is an axial component of the Lorenz force The force acts on the conducting particles, electrons and ions, and these carry the non- ionised gas with them, Because the high temperature plasma has low density the velocities reached are high, of the order of knv/s. At such speeds, not surprisingly the flow field is turbulent Ramakrishnan and Stokes [Ram78] were the first to show that the plasma flow is turbulent which has the result that the effective viscos- ity is increased by more than an order of magnitude over the laminar case. Turbulence also results in a greater rate of entrainment of surrounding gas into the flow system. This momentum of the jet affords some stability to the are column. If the cathode spot is immobile the jet is propelled steadily in a direction normal to the cathode surface But if for some reason the cathode spot moves it inevitably entrains a curvature of the are column as the plasma flow moves downstream. Magnitude of the plasma jet thrust It has already been emphasised that electro- magnetic forces are extremely large at furnace currents. The consequent jet momentum (thrust) is also large and plays an important role in furnace operation. By converting the integration over volume of the Lorenz force F=Q (vxB), 2.1) where v is velocity of the charge density Q and B is the magnetic flux density, into an integral of the magnetic pressure over a cylindrical surface enclosing a diverging are column, Maecker [Mae55] showed that the axial force Fy. is given by F=2P wf (2.2) it Here jlo is the permeability of free space (1.256 10° Vs/Am), / is the current, r, and r, are the radii of the arc column and cathode spot respectively. The In term (log to base e) is typically between 11 and 1.2 2.3 Measurements of are thrust Since this mechanism of plasma jet formation is so important to are furnace operation it is of significance to note that there are several measurements which support the form of the thrust equation. Maecker himself calculated the back thrust on the cathode from a meas- urement of its depression in liquid mercury, up to currents of 300 A. Hilgel [Hiig70] reported measurements of thrust generated in a device designed to accelerate gases Al- though there was an axial gas flow in this device the thrust at the higher currents, up to 3 kA, showed the expected quadratic dependence. In welding arcs Lin and Eager {Lin86] made measurements of the thrust exerted on the anode up to 600A for two different tip cone angles. Bowman [Bow72a] calculated thrust in the 0.5 to 2 kA range fiom measurements of the plasma velocity field 1000 100 10 z s 1 2 © Blackbam 3 Bowman © Hegel © Mocdter a Use Eq 2) ee es 0.001 — 01 1 10 400 Current in kA Fig, 3: Measured thrusts For an arc at 50 kA Ushio et al. (Ush81] created a finite-element model incorporating the full, turbulent magneto-hydrodynamic equations. The thrust has been calculated by2. DC graphite electrode laboratory ares the authors from the published flow field All these measurements are summarised in figure 3. There are two lines shown in figure 3. The lower one is Maecker’s equation with the In- term set at 1,25. This equation actually describes the impulse generated by the diverging plasma column. If the column does not diverge this force is zero. However a cylindrical plasma column is pressurised by the magnetic force and this pressure hump exerts a force on each electrode In the welding measurements of Lin and Eager and the thruster measurements of Huge! this extra term is included rater an which is shown as the upper line in figure 3 uf )eos) (23) _ There is not a large difference between these two equations. As shown in the figure the measurements demonstrate a clear current squared dependence with an acceptable fit in magnitude to the theory. It is revealing for a steelmaker to compare the magnitude of this thrust at furnace currents with that generated by say a simple 50 mm diameter, consumable pipe lance injecting oxygen (not coherent). The com- parison is shown in table 1 Table 1: Arc thrust compared to a simple pipe lance Currentin | Thrustin [Arc depres-| 0; flowrate kA N___|sion inmm| in m'h 40 190 82 2630 60. 430, 185 3950 80 770 332 5280 The depression caused by a stable, immobile arc has been estimated for an arc length of 400 mm, assuming a paraboloidal liquid shape. This topic is treated in more detail in the chapter 12, “liquid bath effects” We will see later that the consequences of such a large thrust are fundamental to many aspects of furnace operation, especially heat transfer in terms of splashing and stirring Similarity of an are plasma jet to a nozzle jet Measured plasma velocity distributions for ares of 520, 1150 and 2160 A from a graphite cathode have been presented by Bowman {Bow72a]. The comparison to nozzle jets was first made in 1982 [Bow82] but the data has since been re-examined by considering the concept of a ‘virtual origin’ to the flow, a common procedure for nozzle jets. This vitual origin to the start of the jet expansion results from the fact that the establishment of a Gaussian shape to the radial velocity distri- bution occurs a few nozzle diameters (about four to seven) downstream of the nozzle orifice. We should not expect the virtual origin for an are jet to be in a similar position since the flow is created by a completely different. mechanism (magnetic compres sion). On the contrary, it has been found, purely by experimenting with a range of values, that an origin with a negative value along the axis yields a much higher degree of correlation, as demonstrated in figure 4, 00 In(vivg) 40 000 0.01 002 0.03 Pin (zzF Fig.4: Are plasma velocities fitled to a Gaussian distribution 0.04 005 The plot shows that > 4) where 0(7,2) is the velocity at the position (7,2), 20 is the velocity on the axis, 29 is the tual origin position, and & is a constant This is the form of the ‘self-similar’ property2 DC graphite electrode laboratory ares of a turbulent jet flow. To maximize the correlation coefficient the data for each of the three currents was manipulated with different virtual origins. Optimum values were achieved with a origin position zp of -10 mm for 520A, -13 mm for 1150 A and -20 mm for 2160. The increase of virtual origin position with current accords with the comment by Fukushima et al, [Fuk02] that zy depends on the Reynold’s number, which, as shown below, increases with current The correlation coefficient for the plot of figure 4 (R’ = 0.86) is considerably improved over the original presentation [Bow82], where the concept of virtual origin was not used. The slope of the regression line has a value of -84.5, in excellent agreement with the value of -84.9 determined by Fukushima et al. with an isothermal water jet at a Reynold’s number of 3000, 2.4 Mass flow rate 12 ° Mass flow rate in gikAs 00 qc 0 2 6 Relative distance from cathode (2-25)F, Fig. 5: Correlation of mass flow rate with distance from the cathode The mass flow rate was also given in an earlier publication [Bow72b]. It increases with distance from the cathode and with current. Figure 5 shows a plot of the mass flow rate per current against the relative distance from the virtual origin (2-2))/r, where r, is the radius of the cathode spot calculated from a current density of 4kA/cm? Although the correlation coefficient is low, R’ = 0.49, the general trend is linear as it would be for a nozzle jet. The relationship for the actual mass flow rate is tn=0.0341 gkAs =; (2.5) K This analysis shows that the plasma flow in an are has much in common with nozzle jets The comparison of the different origins of the two flows is sketched in figure 6 The similarity to a nozzle jet, so familiar to steelmakers, allows us to estimate such parameters as thrust, mass flow rate, velocity field and entrainment, which will prove useful in later chapters, for example in heat transfer, liquid steel depression and splash- ing Fig. 6: Virtual origins of arc (left) and nozzle jet (right) 2.5. Are voltage and voltage gradient In general, arc voltage increases with the Jength of the arc. It has been found experi- mentally that the voltage increases at a rate of, between 5 V and 10 V per cm arc length. A figure in the neighbourhood of 10 V/em was reported by Finkelnburg for currents below 1 KA and similar figures have been reported since by several authors for arcs of much higher currents [Sch88}, [Tre08]. Even in arc furnaces, where some measure of voltage gradient can be obtained from elec- trical and electrode position measurements, voltage gradients of this magnitude have been deduced. However it has to be consi ered that the voltage gradient also depends on2 DC graphite electrode laboratory ares the arc current Otherwise it would not be possible to adjust the are current by changing the are voltage as it is done at DC-EAFs. Description the arc as an element of an electrical circuit will be discussed more in detail in a later chapter Are voltage gradient measurements Two methods have been used to measure arc voltage gradient; one by measuring the increase in voltage with increase in are length, the other by sweeping a small metal probe through the arc at different distances fiom the cathode These two methods do not yield the same results, the reason being that in the former case the anode characteristic does not remain fixed as arc length decreases. The effect of arc length on the plasma jet is one of the reasons why electrode termina- tions change. Measurements of arc voltage by Sharp [Sha67a] and Montgomery [Mon67a] are given in figure 7 120 100 Am Are voltage in V a Ny oot wo eabode L 0 2 4 6 8 ‘Arc length in cm Om arc letting Fig. 7: Arc voltage measurements for cur- rents between 1.6 and 2.0 kA It is to be noted that the two techniques tend to agree for the longer ares. This means that the electrode terminations are not being affected at the longer lengths Voltage gradient, the variation of voltage along the arc column, was determined by a combination of probe and photographic ‘measurements. The photographs were used to measure the variations in core diameter along the length, from which, assuming constant electrical conductivity, the relative gradients can be calculated For the current range 0.4 to 2.0 KA the results are given in figure 8 35 T T K sear * in AV ak Voltage gradient in Viem 0 + G 2 3 4 6 Distance from cathode in cm Fig. 8 Voltage gradient along the arc column The curves exhibit a strong decrease away fiom the cathode, falling towards asymptotes below 10 Vem, A small current dependence is apparent, in the sense higher gradients at higher currents Thus, at a given arc length, arc voltage increases with increasing current. This behaviour is typical for high current arcs © 50 A), it is caused by the almost constant plasma conductivity. At such high powers the conductivity just slightly increases with increasing are current Measurements by Strachan on pulsed arcs Strachan [Su76] worked with pulsed arcs, powered by a capacitor bank, lasting about 10 ms. The current waveform was almost triangular; rising to a peak in about 1 ms and then decaying linearly towards zero in the following 9 ms. Ares at peak curtents up to 21 KA were studied After peak current was passed the arcs proved to be conveniently stable The overall structure resembled a scaled up version of the arc studied by Maecker, that is, a bell-shaped plasma domi- nated by a strong cathode jet10 With the possible exception of the cathode spot, the arc demonstrated very little delay behaviour, ie it rearranged its structure so fast that the instantaneous current was the controlling factor rather than how fast the current was changing 60 Voltage gradient in Viem 0 25 5 78 Distance from cathode in om 10 Fig. 9: Voltage gradient along a 20 kA arc Strachan reported electrical, photographic and radiation measurements and generated empirical relations between them. Figure 9 shows an example of the voltage gradient correlation for a 100 mm long are at 20 kA Such curves can be integrated along the length to yield a total voltage. Strachan found that by adding 15 V to this integral agree- ment could be achieved with the measured total arc voltage. The implication here is that the 15 V addition is occurring at the electrode junctions, principally at the cathode 2 DC graphite electrode laboratory ares Again the dependence of the local voltage gradient £ on current was rather weak E~i®, (2.6) implying that the total arc voltage also has positive dependence on current; for example, a doubling of current to 40 kA is predicted to cause an increase from 160 V to 186 V for the 10 cm long are 2.6 Are shape The shape of the are as determined from photographs, including AC arcs to be dis- cussed later, was shown to be independent of current, over the range I to 10 KA, when expressed in dimensionless form [Bow94] Diameters were rendered dimensionless by dividing them by the cathode spot radius. The reason for choosing the cathode spot 1adius is that experimental data suggests that the cathode operates at generally constant current density. Thus its radius can be estimated fiom the current. Figure 10 shows the shape of the core and the conducting boundary as a function of distance from the cathode (also in dimensionless form) The separate visibility of core and conducting boundaries, as in figure 1, is a function of the radiating species. Radiation fiom the core is principally due to the electrons, the aureole due to ions from the electrodes With ares in a steelmaking environment most of the radiation comes from metallic elements, principally iron A separate core is no longer visible, only the conducting boundary. 4 T T | 3. | conducting | a boundary g2 € T 1 » Lezeeaney | 7 8 9 10 1 12 «13 «14 15 Distance from cathode in F, Fig. 10: Arc core and conducting radii as a fumction of distance from the cathode, all dimen- sions are divided by the radius of the cathode spot