Plasma nitriding is almost never homogenous, each part has its own geometry, so is the plasma distribution dependant on the different edges. (edge effects on plasma sheaths). Now your nitriding machine should at least treat all similar parts in a narrow band of results.. if not, you have leakage,pumping or whatever gas distribution problems (or electric current flow problems?).    your problem is to homogenise the flow/pressure in your chamber..   (prof. K.T.- Rie , TU-Braunschweig used to supervise some good Thesis (E.Menthe, P.Kaestner, J.Woehle))  ,  lower the pressure is most of the time a good way, but let you reoptimise all parameters (Temp,Time,Voltage,  different sheath thickness, new edge effects).

without garantee..

Also you may use Active Screen Nitriding to avoid (or reduce) edge effects.

I have been involved 10 years ago with the numerical simulation of the nitriding of a gear tooth. The results have been published in the following reference that you can download on my RG page:

J M Bergheau, J Devaux, P Duranton, R Fortunier, M Larreur, J B Leblond, « Numerical simulation of superficial diffusion and precipitation of chemical elements, with application to some nitriding process », Surface Treatment VI, edited by C.A. Brebbia, J.T.M. de Hosson, S.I. Nishida, ISBN 1-85312-962-3, 2003, pp. 297-306.

May be this work can help you.

More recent results can be found in :

E. Feulvarch, J.M. Bergheau, J.B. Leblond, « An implicit finite element algorithm for the simulation of diffusion with phase changes in solids”, International Journal for Numerical Methods in Engineering, n° 78, 2009, pp. 1492-1512.

Thank you, I will go through the article you have mentioned.

If your university subscribes to them, there are technical papers, guidance, and reports from the technical committees that would give you guidance at American Gear Manufacturers Association www.agma.org of at the British Gear Association www.bga.org.uk and the Design Group at Newcastle University are always helpful www.ncl.ac.uk/gears/