Thanks very much Constantin, so you mean, although the gas temperature of the plasma is very high, it probably would transfer more energy to the gas flow that flows out, rather than to the electrode? 

I think, I can help with the second question: an anode spot has a higher electron temperature (which was measured several times), but the can transfer thermal energy very efficient to the neutrals in the gas inducing macroscopic gas flows as was shown  by Stenzel (Neutral gas dynamics in fireballs, Journ. Appl. Phys. 2011) and further investigated by Makrinich and Fruchtman (The Force exerted by a fireball, Phys. Plasm. 2014).

Thanks again, Constantin. I know your points now.

Thanks Johannes. By the way, do you know the difference between the anode spots and cathode spots? and If I change the polarity of electrodes, i. e. exchange the anode and cathode, will the properties of the plasma or spots be greatly changed? and what's the difference? Thanks very much for your time.

Hao, in the complement of what others said, it might be interesting for you to know that:

The temperature inside fluorescent bulbs is about 2000 k. but the gas density is not so high that you feel the temperature. Higher density means more impacts by gas molecules to the electrode and the transferred heat is just proportional to the number of impacts.

The main difference of cathode and anode spots is their creation mechanism. While  anode spots mostly ignite via electrons that are excited in the sheath that surrounds the spot and make additional impact ionisations, cathode spots are more complicated. The opinion of the formation of the latter ones include sputtering on the cathode surface, evaporation processes or even micro explosions on the cathode (along with other possible reasons).

Mohammad, Thank you for your reply. Your complement makes this problem understandable.

Dear, Johannes. Thanks for your patient response. You really help me a lot.

Hi, Hao, you should consider the cooling process of the electrodes when a high temperature plasma forms. Usually the anode and cathode are cooled by cycling water. An often seeing material making of the electrodes is copper for its relative high thermal conductivity and melting temperature. However, even though these cooling methods were applied, the electrode erosion could appear when the discharge current or current density was high. An example is the electrode erosion in the plasma torch, which limits the lifetime of a plasma torch.

As to the second question, I think Johannes gave good answers. There is very much difference between anode spot and cathode spot. In plasma torch, the anode spot is larger than the cathode spot. And the anode spot could become a "diffusion type" depending on the experiment conditions, which means the the anode spot is no more a spot, but becomes darker and circles the anode. This phenomenon had been observed in experiments.

Jinwen, Thanks very much for your reply

Dear Hao Zhang, If you specifically mention which type of plasma then I would be able to comment

according to my experience.....the excessive or high amount of power given to any electrode can itself make the electrode melt if it is beyond the melting point due to the power and the temperature rise because of the interaction of plasma with the electrode inside the reactor or the case if electrode is outside the reactor...peers you can comment on my answer if you we can discuss more on it...

 another answer can be there is an arrangement of casing the electrodes by quartz or high melting point materials so that the temperature rise wont make the electrodes melt

Dear Ajinkya, so in your opinion, the most probable reason that the electrode was not melted is that the temperature is not high enough to melt the electrode?

Dear Hao Zhang, there is a possibility of the electrode to melt if the power supplied goes beyond an extent in which the temperature of plasma reaches or goes beyond the threshold melting point of electrode...hope am able to convey 

and also depends the material used for electrode....

Ajinkya, thank you. I knew your point, but about this topic, you can see the comments of Mohammad and Constantin. I think they gave us a more reasonable explanation.

Hello all. Very intersting to read these discussions here. As to why the plasma does not melt the electrodes, the answer lies in the heat content of the plasma. Plasma may be very hot, but its number density (n) is normally so small that heat density ( nkT, k is Boltzmann constant) is quite low, not enough to melt materials. For example, in space plasmas in which satellites ( or space station) operates have temperatures larger than T = 3000 degree Kelvin. In geosynchromous orbits where communication satellites reside, plasma temperature is 100,000 degree Kelvin, but melting of objects in space is not a problem because plasma number density is less than a million electron and ion pairs per cm^3. It geosynchronous altitudes it is 1 pair per cm^3. In laboratory plasmas desnities are typically < 10^13 pairs per cm^3. Thus the heat content nkT is low to cause melting. In Fusion plasma machines, very hot ( more tha million degrees) plasma is confined by magnetic fields and not by the walls of a vacuum chamber; the reason is not that the chamber wall will melt away, but the plasma will cool down and vanish.

Dear, all, thanks very much for you patient and valuable comments, you really teach me a lot.

Hot plasmas can have an extremely high temperature, but temperature is connected to the gaussian distribution (mean velocity of the distribution) and temperature can be extremely high without melting the surrounding since the important element is not the temperature of the distribution but the energy transfer to the surrounding. If the number density of the hot particles are low (low density) the transfer og energy to the surrounding is also very low and the surroundings stay much colder.

hi,

Yes its true that temperature of some plasmas are very high greater than the melting temperature of vessel which one is using to produce plasma, but when we talk about glow discharge plasma, first it has low density that mean low pressure, though the temperature is very high but the transfer of energy from a charge particle to wall is reduced so there is less heating which doesn't  allow vessel to melt.

I agree with all of the above, but didn't see a fully proper treatment of heat fluxes and heat capacity. Even a small heat source could melt a larger object if there were no cooling mechanisms, so the key here is probably that the metal electrodes conduct heat away much faster than it is deposited by the plasma. The heat flux (W/m^2) of the plasma is proportional to particle flux * Temperature, or roughly density * Temperature^3/2 / sqrt(ion mass). The temperature of the electrode (or wall, or probe) is then governed by a differential equation relating its heat capacity (specific heat of the material * mass), heat conductance (conductivity * area / length), blackbody radiation (proportional to T^4), and possibly heat convection from coolant fluids like air or water. In steady state, the heat capacity is not really important and the heat source balances the losses.