Dear Dr Mostafaei

please check the attachments if it is helpful for you.

thank you

The first attachment provided by Dr. Pradhan helps to understand the meaning of the curves shown by Dr. Mostafei. Pressure filtration (PF) is an inclusion measurement instrument developed for ensuring clean molten aluminum in making high integrity, high performance automotive castings. In general molten aluminum alloy is degassed and refined/modified using a powder flux. As a result the weight molten metal in the crucible increases over a period of time. If the molten metal is now filtered the filtrate will be free of the products of fluxing. The unique feature of the PF test is that the scum left behind on the filter is subjected to accurate weight measurement instead of being thrown away. Also the weight of molten metal sample prior to fluxing is accurately determined. The bottom curve in the figure shown by Dr. Mostafei represents this. If the weight of the scum is added to the weight of the metal before fluxing, the top curve will be generated. The more the scum the  greater will be its weight and the steeper will be the slope of the top curve. In short, if there is a large deviation of the top curve from the bottom it implies that the molten metal would have given rise to castings with undesirable dross and inclusions if not treated with the flux prior to pouring into the mold. 

I am not an expert in the use of Prefil. However, it seems reasonable to me that if the melt is clean (i.e. it contains no inclusions in suspension) then the filter will not block. Since the filter will not change its characteristics during the so-called filtration, the rate of flow of metal through the filter, given a constant pressure differential across the filter, will not change. Thus the weight of metal transferred will increase perfectly linearly (i.e. there will be no change in rate of flow).

All this contrasts with the more normal circumstance in which the melt contains inclusions in suspension. Now the filter will progressively become more blocked, so that the rate will progressively fall, and the Prefil curve will exhibit the usual curvature, slowly flattening as the rate of transfer of the metal decreases.

I understand there are interesting details of the changes in slope with some types of inclusion (for instance whether films or particles or liquid droplets etc.) and the type of blocking mechanism, whether in-depth filtration or cake build-up on the filter surface.  These can sometimes help to identify the type of inclusion involved.

Best regards

John Campbell

Thank you for your excellent answer, Dr. Campbell. My answer was concerned with providing an explanation for the linear plots shown by Dr. Mostafei. I fully agree with you that when the melt contains inclusions in suspension the Prefil curve should gradually flatten out as time increases. Thank you for sharing the information that by studying the shape of the Prefil curve it may be possible to identify the type of inclusion involved. On a personal note, I feel proud that I worked in the Aitchison Laboratory (I hope it is still in existence) at the University (of Birmingham) under the mentorship of Late Dr. Voya Kondic; he would often mention that you were his star student.

I was flattered to learn of my stardom - I did not know about this. However, I admired VK and was enormously grateful to him since he was generous and confident, allowing me to go my own way, since we almost never agreed on any technical issue.  We always remained the best of friends. As a supervisor his commitment to his work, his interest in his students, and courtesy and kindness were examples I shall always try to emulate.

The Aitchison Laboratory is no more. I agreed to move the foundry to the main metallurgy and materials building on the far side of Pritchatts Road. I felt we needed to be together with the electron microscopes and the ambience of physical metallurgy to ensure we did some real science. I am happy that, in general, the move has been good, and the department continues to flourish.