250 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS particle separated from its neighbors by a thin film of air will flow freely, while the same particles without this air film tend to hang up and clog. Another way to improve handling characteristics is to granulate the blended material. Granules, as we use the term, are uniform, spherical clusters of very small particles. These clusters are held together only by such effects as a slight surface wetness and sometimes, we think, by electro- static attractions. They are soft agglomerates that can usually be broken up very easily. Granulation is most often a preliminary stage to making a paste or dry cake, such as a powder pack. It is most important to avoid letting any confusion creep into our talks with a prospect about the job to be done. It is possible, if this happens, to do a tremendous amount of work, yet come up with a com- pletely wrong answer. Several years ago, we had just learned how to disperse a very small amount of liquid throughout a mass of fine powder with extremely high uniformity. We showed samples of this dispersion to a prospect with considerable pride. The stuff was dry to the touch, but the liquid was there, and every particle had its share. The man shook his head and showed us what he really wanted. It was a granular mix suitable for compacting--not a dry powder. All I could say was, "But that's so easy! We could have done that for you three years ago !" As a matter of fact, we always seem to be able to do new things each year--things we could not do the year before. One reason for this is that we are learning more. We understand a little better how particles behave. There is still too much we do not know, and we are always dismayed at the number of facts we cannot find in published literature. But where we do not have facts, we make up tentative theories, and some of these blue- sky theories have served us very well on a practical level. Another reason we can do more difficult blending operations each year is that we have continued to make changes in equipment design. Let me trace our blender design evolution for you briefly. We have concentrated on tumble blenders, because they are gentle and because the entire charge moves twice during each revolution of the vessel. Gentleness of blending action is desirable because it avoids breakdown of delicate particles. Sometimes gentleness is essential, because too much scraping or agitation will give particles so much electrostatic charge that they become completely unblendable. Moving the entire charge gently by gravity can be done in a variety of rotating vessels. Our experience has demonstrated that the shapes which present a different cross section for each half turn of the tumbling vessel are most effective. The double-cone blender meets this requirement.

APPLICATION OF THE TWIN SHELL BLENDER 25l It is a precision dry solids blender. It is relatively easy to clean and to maintain. A double-cone blender can be used for adding liquids to solids. How- ever, we run into problems of accessibility when we put a really so- phisticated liquid dispersing device into such a vessel. The V-shaped tumble blender we call the twin-shell takes care of the accessibility problem very well. We can install and service a rotating shaft device inside the vessel. Such a device can do either or both of two things. It can break up agglom- erates and it can disperse liquids. We call it an intensifier bar when its only job is to break up agglomerates. Then it is merely a shaft with pins or small knife-like blades on it. As it rotates at high speed, high-velocity impact breaks agglomerates ef- fectively and rapidly. It also serves to aerate the particle mass. Because the intensifier bar can be started and stopped independently of the tumbling action, it does its job without unnecessary dissipation of work and energy in the particle mass. As has been observed in connection with gentle- ness, too much work or violent scrubbing can make some mixtures virtually unblendable. When we use this shaft inside the twin-shell blender to disperse liquids, it is made as a hollow tube. It still has intensifier blades that impinge on the particle mass. The liquid goes in through a rotary joint and is dis- persed as a mist or fog by centrifugal force through annular orifices. The easiest way to defeat good liquid-solid blending is to let any of the liquid spray or fog touch the vessel wall before it reaches a particle. Once the liquid wets that inner wall, your have lost part of it. So, we create a hole in the middle of the tumbling particle mass and we spray from the inside outward toward the surrounding particles. Those whirling blades, that we used primarily for breaking up agglomerates on the intensifier bar, create this empty space inside the particle mass. They make the particles keep their distance while being sprayed. This has a great practical advantage in preventing the annular nozzle from clogging with solids. Easy accessibility is gained by having big covers on the ends of those two intersecting cylinders of a twin-shell blender. It is not hard to reach in through one of these openings with the cover off and remove an intensifier bar or liquid feed unit for cleaning. Also, the job of inspecting and taking care of rotary seals is simplified tremendously. There are some other features that can be built into a twin-shell blender. It can have a filtered exhaust connection for blending or processing under a partial vacuum. The vessel itself can be jacketed. That permits heating with steam to work at higher temperatures, or circulation of cooling water to reduce temperatures. These modifications transform what was merely a blending device into something more•actually it becomes a processing vessel.