Soft distribution requires hard experiment

A reliable method has been found to deliver pairs of earplugs from a bottle in an industrial environment. The plugs are soft, tend to stick together, and exceptionally efficient at jamming machinery. The breakthrough comes from performing practical experiments in order to understand all the ways the plugs can interact with different items of mechanism and each other. The same approach and lessons learned can be applied to many other difficult to handle items. Richard Thom, Hothouse's chief designer, explained to Eureka that the development stemmed from the needs of a client, who wanted to be able to dispense four different kinds of earplug from the same device. Some were made of squashy PVC and were sticky, while others were made of polyurethane and not sticky, but even more squashy. Part of the specification was that the bottle from which the earplugs were to be dispensed was to be refillable and not disposable. A typical installation environment would be a car factory. The device should give out two plugs without too much trouble, but not release so many at once that plugs would be dropped on the floor. Thom said that in order to develop a handling device that could succeed, it was first necessary to understand the mechanical behaviour of all the plugs. "If dropped into normal types of dispensing machinery, they can be guaranteed to gum it up solid. Earplugs are a bit like people. They don't like to be forced, but they can be gently coaxed. They also quite like forming queues and archways." The first conclusion was that the design would need to keep the plugs gently agitated and stirred up as they were dispensed. Then followed a substantial period of experimentation, leading to ten different roughly prototyped designs. Whereas CAD may be used to achieve right first time designs to handle materials which are fully understood and can be readily modelled, materials which show more complex behaviours still have to be studied experimentally and designs perfected by the traditional, 'trial and error' process. Early on it was discovered that if the feeding container has the wrong aspect ratio, agitation forms an internal igloo space which nothing can fall into. It was also found that the agitator device needed to include small protruding features to catch the ends of plugs and cause arches to collapse. Trials included studies of various different bottle shapes and agitator configurations with differently shaped protuberances. An additional restriction was imposed by the existence of patents on mechanisms that might solve the problem, so that Hothouse's client wanted new solutions which it could patent itself. One of the pre-existing patents led to a requirement to only have a single hole in the rotating agitator, but with a multiplicity of stationary holes in the plate beneath it. Design solutions that worked satisfactorily with a few earplugs, worked less well when more were added, packing down the plugs just above the distribution arrangement. It was essential to come up with a design that would not catch plugs in the mechanism. In a previous solution, there was a wide gap between the agitator and the distributor plate so that caught plugs would be pulled through. In the new design, it was considered preferable to have a very narrow gap which plugs could not get into in the first place. The final design of the agitator unit was in the form of an inverted cusp shape funnel. Ten lobes with small fins on top had the right kind of interaction with the plugs. In the course of development, it was found that material and surface texture were both important. The final design was made in filled polypropylene. Distributor plate hole apertures were made sufficiently wide to allow one plug to go in easily, but small enough that two could not get in side by side. Six holes were found to be the optimum number, yielding on average, the passage of two plugs for every three turns of the agitator. The final material choice was ABS. The holes were given vertical sides, but kept short. Beneath is a shutter with a single, large hole. If the agitator is turned quickly, distribution is discouraged, so that excess plugs are not given out and wasted. The distributed plugs fall into a funnel, and into the hand of the worker rotating the mechanism. The device has now been installed in a number of factories, including those of Ferrari, Audi and Nissan. The overall product time from concept to launch was ten months. Dispenser has a spring in its step Not surprisingly, the company is also asked to undertake the design and redesign of other kinds of dispensing systems. A recent task was the redesign of paper towel dispensers for Kimberly-Clark. These are units into which rolls are inserted from below, passing spring loaded fingers which discourage stealing, while allowing the top units to remain sealed. This is particularly important in installations in food and pharmaceutical processing operations subject to washdowns. The redesign was required by use of a higher grade tissue with a textured surface. This led to a redesign of the feed plate and edges for tearing off the individual sections. However, at the same time, it was decided to improve the door latch, which was tending to not latch when slammed shut, there not being enough time for a gravity latch to work when the door was slammed shut quickly. It was therefore decided to add a spring element to the plastic. Quite a number of ideas were looked at, including grooved strips that would bend at the grooves. The final idea was to mould an inverted 'L' shaped protuberance on the acetal. This engages on a ramp. However, in order to deliver the right combination of return force and movement, without causing fatigue failure, it does not deflect in the plane of the 'L' but sideways. An additional advantage of the concept is that it could easily be added to the existing mould tool by machining an extra slot. Whatever the dispensed or delivered product: earplugs, tissue paper, foodstuffs or mixtures of stone and asphalt, there is no substitute for practical experiment in the product design and development process. Pointers Practical experiment was an essential part of the design and development process. Ten different sets of configurations were studied before the final design was arrived at CAD tools are presently unable to accurately model the complex behaviour of many items that have to be mechanically handled Hothouse Design Richard Thom