THE MIXING AND BLENDING OF POWDERS 219 stability in a loose powder structure by the application of pressure. This could be achieved, for example, by placing the mixture in a container which is made of a non- porous flexible material so that essentially it will hold the structure together. Would this not overcome the subsequent segregation of the material? MR. LLOYD: Whilst I agree that placing the powder in some container which holds it rigidly would prevent segregation there will still be difficulties occurring in putting the powder into such a container without segregation occurring. I do, however, consider that concepts such as these can help us avoid the problem. The main problem to date seems to me to be that segregation exists. We must accept this and design our process so that its effects are minimised, either by control of the particle size of our components or by more sophisticated design of the basic operations in our process. MR. F. F. ADEY: IS the efficiency of mixing, which is apparent at 35% of critical speed, a function of the chosen materials or the type and dimensions of the chosen mixer? MR. LLOYD: I do not think that the figure of 35% of the critical speed is any magic number it is in fact a function of the materials chosen. Wehave shown that in a drum mixer there are two forms of segregation: one in which the large particles congregate in the centre of the mixer--this is found at one speed, and one in which the fine particles congregate in the centre at a different speed. Between these two conditions the pseudo-randomisation we have discovered will occur. With the material we used this happened to occur at 35% of the critical speed. I think other materials would have different speeds and that one could obtain some indication of what speed this would be by measuring the porosity of the rotating mass as a function of the speed of rotation. The indications are that where this porosity is a maximum the pseudo-randomisation effect will occur. MR. G. DUNCALF: I am interested in mixing and am rather concerned that at this Symposium only drum mixing has been considered. Would you or others care to comment on work going on now to consider theoretical aspects of other types of batch or continuous mixing, for example fluid bed mixing? MR. LLOYD: The drum mixer represents a very important group of tumbling mixers. This group includes such well known mixers as the rotating cube, the Y blender, the double cone blender. In studying the drum mixer, we are in effect studying all these forms of mixer. This is especially true if one is concerned with what happens to the particles within the mixer. My paper is concerned more with the basic mechanism of mixing within a tumbling mixer than with the outward mani- festations of such a mixing operation. I believe that other mixers are more suited to mix particles of widely different particle size. Such mixers as the Nautarnixer is eminently more suitable. However, manypeople do use tumbling mixers in an attempt to mix components of widely different sizes. This paper therefore is an attempt to illustrate 1. what can happen within the mixer to their materials 2. to give an indication of how a good mix can be obtained within the nfixer and 3. to draw attention to the importance of the design of the process after the mixer when even if a good mix is obtained in the machine this may be subse- quently demixed due to bad process design.

220 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS The philosophy of our work at Loughborough is therefore that we must first of all understand the properties of the materials and that with this knowledge we shall be able to operate upon those materials to obtain our desired approach. The second approach is to worry more about the comparisons of machines and the assessment of achieved mixers than the fundamental questions of particular behaviour. MR. ]. C. WILLIA•IS: I would like to reassure Mr. Duncalf that there is a con- siderable amount of work research into the mixing process which is not concerned with the performance of rotating drums. In the paper I presented here (7) a drum is being used only for the development o[ experimental and mathematical techniques. Future work will be concerned with other types of equipment. Perhaps I may be permitted to mention other •nixing projects in progress in the School of Powder Technology in Bradford. These are {i) a theoretical study of the statistical properties of non-random mixtures, (ii) a study of the mechanisms of mixing in a ribbon blender, (iii) a project, supported by the Ministry of Technology, in which the aim is to establish a simple method of testing powders which are to be mixed so as to assess their tendency to segregate this will be related to the performance of mixers of different types so that the best mixer can be selected for a given duty. (iv) A series of comparative tests into the performance of different industrial mixers for highly segregating materials has recently been supported. Other investigations into mixing are in progress in the University of Exeter (Professor Lacey), London School of Pharmacy, King's College London {Professor Rose), the University of Cambridge (Mr. Bridgewater) University College, London (Professor Rowe), and at •,Varren Spring Laboratory. (7) Williams, J. c. and Rahman, M. A., The continuous mixing of particulate solids. J. Soc. Cosmetic Chemists 21 3 (1970).