Cudaxtreme,
well, this is one of the most interesting topics and queries I have read since I am aboard on the forum. Thus, may I be allowed to try to response? Thank you!
Before I start I beg yours and all the others of physics loving colleagues forgiveness that everything you may read subsequently is tremendously simplified due to it is finally not been clarified what really occurs in the arc-plasma! The amount of what would be necessary - particularly in regard to physics - to explain what is going on while welding with different polarities on the electrode would blast - I am sure - the forums text length limitation. Although it is itching in my fingers I tie up myself here...
The longer the speech the less thought. Therefore...
Of course the behaviour of the electric current in a welding arc is always the same whether the welding-process being used may ever be. Why? Well, since the electric current underlies the same physical laws as every other matter also does. Therefore it is equal if you are using GTAW (non consumable electrode), GMAW or SMAW or SAW (consumable electrode). All these processes generate an arc. And all these arcs underlie OHM's Law, since they are electrical conductors. What the main difference is, are the electrical power supplies or better, the characteristics of the welding power sources being used for the processes mentioned above. Here you can use Constant Current Characteristic (CC-Mode) or Constant Voltage Characteristic (CV-Mode). Both are - of course - different in terms of process behaviour and showing different specifics in reaction to outer disturbances applied to the arc. But I assume this may be well-known and thus I avoid to deal with these elements, only thing I would like to mention here further is, SAW, SMAW and GTAW are processes using mostly CC-Mode and GMAW uses mostly CV-Mode (Pulsed Arc or modern Special GMAW-Processes explicitly excepted). Therefore is the reaction time for regulating the arc-length in particular in comparison between SAW and GMAW quite different. In using CC-Mode the time between an outer disturbance to the arc is being regulated for stabilizing the process again, lies at approx. ¼ of a second. One can say the reaction is relatively slow. Whereas in using CV-Mode for welding conventional GMAW, the reaction times between a disturbance applied to the arc (e.g. varying contact tip to workpiece distances) and its regulation to stabilize it again needs only a few thousandth parts of a second (>50 ms).
Well so far so good. Coming now, after we have spoken about the arc as an electrical conductor and thus underlying OHM's Law ( Resistance (R) = Voltage (U) / Current (I) ), to the details occurring between the region where the welding-current is being induced into the wire and the region where the arc roots on the workpiece. By doing so we can find five different places affecting the arc and its behaviour.
1. The place where the current is being induced into the wire electrode and the wire is thus being warmed up by the resistance heating.
2. The free wires end between the place mentioned above and the arc root on the wire. This piece of wires volume is being warmed up due to its ohmic resistance heating. On this distance the voltage is reduced firstly by the electrical resistance of the warmed up wire.
3. The region directly adjacent in front of the wire electrode where the arc root is being located. Well, and this region is, in regard to the topic you have posted, the certainly most important one. This region is called in DCEP the "anode drop", where I would like to come to later on. Firstly let us assume we are using GMAW and the wire electrode is DCEP as in most cases used.
4. The plasma- or arc column. Within this region approx. 50% of the voltage is being reduced.
5. Assuming we use Gas Metal Arc Welding with DCEP the final rest is the so called cathode drop. This is the region where you have mentioned that the heat is "directed" into the workpiece. Important is, that within the regions directly adjacent to the wire tip and the workpiece, i.e. the anode- and the cathode drop the voltage is being reduced by approx. 50%(!).
What a value, if you may consider that these regions are very narrow sheaths with a thickness of only a few thousandth parts of a millimeter thick(!). And you know what happens when this is the case? Right! There is a very strong electric field being generated and this field affects again the carriers of the electrical charge which are negative (electrons) and positive (cations). Whereas the electrons are influenced much more within the electrical field of the anode-drop due to their much lower mass of approx. 1/2000 that of an positive charged ion (cation).
Coming now to ions and electrons. Under normal surrounding conditions gaseous matter is electrical neutral. But, for igniting a welding arc, this gas (shielding-gas in GMAW or GTAW) or air in SAW or SMAW must be made electrical conducting. For achieving this condition of electrical conductivity the electrical neutral atoms or molecules, respectively, have to be charged with energy. Therefore a strong and energy rich radiation (e.g. cosmic radiation) or thermal energy can be used for. Three different sequences can be stated with regard to achieving the condition of conductivity of the surrounding gases which should not be treated further here. But, what "Greg" has predicated, the thermionic emission in GTAW welding is a mix of thermal and field emission of charge carriers, i.e. electrons. But when you are interested further in this matter and the intricacy of what is probably going on when electrons are emitted from the cathode, here's a recommendation from my side. There have investigations been performed by TANAKA, TASHIRO and LOWKE recently dealing with the "Predictions of current attachment at thermionic cathode for gas tungsten arc at atmospheric pressure". Believe me your heart would cheer when reading this paper and its ingenious content!
Once again. The neutral gas is being changed for being electrical conducting by igniting an arc. I suppose the sequence of ionisation is well-known and thus I won't treat it further on here. When the arc burns one can say there is a plasma being generated consisting of non neutral ions having "lost" one or sometimes more than one of their negative charge carriers (electrons). Due to the wire electrode is DCEP, the electrons - having a much lower mass compared to the ions (~ 1/2000 of an ion) but due to that having a much higher velocity being induced by the electric field - are accelerated to the anode which is the filler-wire electrode. Here is in a narrow sheath the anode drop recognizable which is nothing more than a uncompensated room charge consisting of electrons. Within this anode drop the electrons being accelerated in direction to the wire tip (anode) and impinging onto the molten surface of the wire electrode, which is the arc root on the anode. Here they are generating ions again which are accelerated within the anode drop but now in the direction to the negative charged cathode (workpiece). Due to their higher mass (electron x ~ 2000) their kinetic resistance against the acceleration by the electric field is higher that one of the electrons. Owing to the lower kinetic energy of the positive charged ions they are slowed down in a narrow sheath before the workpiece arc root (cathode) from where electrons are being emitted, i.e. here is generated another uncompensated room charge which is the cathode drop, having only small dimensions but also high field strengths. That means, the strongly by the electrical field accelerated negative charge carriers (electrons) having indeed a much lower mass compared with the positive ions (cations) but having a much higher kinetic energy being altered on the wire tip into thermal energy - here is the hotter pole of both. The cations having indeed a much higher mass but being slowed down on their way through the arc plasma and being uncompensated in the cathode drop region can not induce the same amount of kinetic energy altered into thermal energy "directed" into the workpiece, which is the cathode and thus the colder pole of both.
When now the polarity is being altered between DCEP and DCEN, the ratio of voltage of what is being reduced within the regions I have tried to explain, which are the cathode- and the anode drop, is being altered also. Thus the direction of charge carriers (large mass cations) and electrons (low mass but high kinetic energy) is likewise altered. Now the workpiece is positive (anode) and the welding wire electrode is negative (DCEN) and thus cathode. The wire has to emit now the electrons which are being slowed down by the cathode drop (uncompensated cation sheath) directly adjacent and existing in front of the arc root on the wire tip. The voltage increases. Mentioned by the way, cathode drop's field strength is much higher that the anode drop field strength.
Assuming what you said and everything (voltage, current, stick out,...) but the polarity is constant, one can say that the ratio of voltage drops is being altered when the polarity is being altered. While the larger voltage drop could be observed on the workpiece when using DCEP and thus the specific power, calculated by Voltage (U) multiplied by Current (I), was higher on the workpiece, the cathode was hotter in comparison to the anode.
When switching the polarity to DCEN the larger voltage drop can be observed on the wire electrode, thus the specific power in the directly adjacent sheath before the wire is drastically increased and the share of arc-energy or power, respectively is being reduced. That means, only a reduced share of energy reaches the workpiece and combined with the lower voltage drop adjacent to the anode (workpiece) leading to a lower amount of specific anode power (lower U[Anode] multiplied by lower I[Arc] = lower Power) the total injected thermal energy into the workpiece is decreased. This again leads to a shallower depth of penetration on the workpiece to be welded, but on the other hand to a higher heat input into the wire electrode, caused by the higher specific power coupled into the wire, being DCEN.
I hope that I could find the right words for the attempt to describe in a simple way what in fact is really intricate. However, for a better understanding I would like to attach also a Portable Document File (I hope that I will succeed since John Wright was so kind at that time to explain to me how to realize this action :-)). This file shows a section from an investigation been carried out from one of the leading German Researchers in this field Dr. Gerd Huismann from the University of the Federal Armed Forces in Hamburg Germany. The picture I have attached shows in a visual way what I have tried to explain. You can see on the left side the wire electrode in DCEN. There you can see that the voltage directly in front of the wire has been measured with 9.5 Volts (which is the cathode drop). The measured arc-current was about 141 Amps. Whereas on the right hand side you can see the wire electrode positive, i.e. DCEP. The voltage directly in front of the wire (anode drop) when DCEP, is only at ~ 5.5 Volts, which is approx. only ½ of the DCEN case, and the measured arc-current was however ~ 206 Amps. Therefore the specific power being generated in the wire electrode is lower when DCEP than in DCEN.
My best regards,
Stephan