For the thermal simulation of the base points in the EDM different approaches of heat entry are selected in the electrode. In 80 percent of the articles a Gaussian heat distribution is assumed, without notice. What is the physical explanation for this?
I think the only right way is to determine the extent of the plasma channel diameter, depending on the pulse parameters, and a temperature-dependent enthalpy.
A simple reason for the Gaussian heat distribution is the shape and size of the craters formed during EDM. It is very similar to (not exactly) Gaussian shape. However, we have evaluated the evolution of plasma channel at an interval of 1 microsecond and it depends, as mentioned by you also, on cutting conditions and thermal properties of tool and workpiece.
References:
1. P. Shankar, V. K. Jain and T. Sundararajan (1997), "Analysis of Spark Profiles during EDM Process", Machining science & Technology, vol. 1(2), pp. 195-217.
2. Jain V.K., Advanced Machining Processes, 2002, Allied Publishers, New Delhi (India).
With best wishes.
VKJain
Dear Prof. Jain,
First of all thank you very much, that someone has found the answer to this question.
Of course I can make such an approach in a rough approximation, but I mean that it can not really be justified. If one looks in the radial direction of the heat input in the z-direction (tool-workpiece) then exist with short, steep pulses distributions of the heat input, which does not correspond to the Gaussian distribution principle. Any designated crater walls can be described much exater by the spread (diameter growth) of the plasma channel. This is often done with experimental formula, but again include some constants, the ever new combinations of the material (dielectric tool-workpiece) must be to bestimmtw. I admit that it is possible to describe an exact thermal simulation not an "easy" task, but essentially always the enthalpy dependence on temperature is neglected.
Best regards
Hans-Peter Schulze
Dear Prof. H.-P. Schulze,
I agree that it need not be a Gaussian distribution to have a particular shape of the crater. We have done the plasma channel evolution and that can be used to find exact shape of the crater. With such approach, one should try with Gaussian, uniform, point and triangular heat source and then compare the predicted shape with the experimental shape. It is open area for research.
With best regards.
V.K.jain
I thought that the interest in this question is somewhat greater, especially if you want to describe the erosion-mechanism (melting and boiling volume) in more detail.
My opinion is that the Gaussian distribution is not correct, because the assumption, which is the largest energy density and thus the highest removal in the center of the discharge, is refuted in practice. The Gaussian distribution is only a token application (approximation) when no plasma channel propagation is taken into account.
Why not. One can try with other shapes of heat sources and compare the evolved crater shape with the experimental one. In the following paper, no particular shape of the Heat Source was assumed to understand the plasma channel evolution, and consequently crater shape development which depends on the thermal properties of the tool and workpiece both. It turned out to be an interesting problem. However, the weakness of this work is that it did not compare the results with the experimental findings.
P. Shankar, V. K. Jain and T. Sundararajan (1997), "Analysis of Spark Profiles during EDM Process", Machining science & Technology, vol. 1(2), pp. 195-217.
Dear Prof.
In my opinion the Gaussian distribution (approach) is the better way that research found to simulate the diferent electrical resistivity inside the plasma and that in same cases is a gaussian distribution......
With Best regards,
José Duarte Marafona
Thank you once for their expression.
The Gaussian distribution is usually applied without further radius changing, ie the heat source is considered to be fixed and unchangeable. This assumption, however, can be accepted only in a few parameters constellations.
That corresponds to the differential electrical resistances in the plasma, is rather unlikely. In plasma, we can expect an equal distribution or radiofrequency pulsings by a skin effect (Gauss then applies not at all) only once. With the Gaussian distribution I construct a "apparently ideal" transition zone, but which is covered by the constantly rising real plasma radius.
Best regards
Hans-Peter
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers