The convective stability of a horizontal layer of water with salt and heat addition from below was studied experimentally. The layer was bounded above and below by porous metal plates through which heat and salt could diffuse. A well-mixed region of warm, salty water below the lower plate and another of cooler, fresher water above the upper plate set the temperature and concentration difference for the intervening quiescent layer. With a fixed, constant concentration gradient established between the plates the temperature difference was slowly increased until convective motions were observed. The instability boundary for this system lies within the unstable region predicted by linear theory for a horizontal layer with constant gradients and stress-free boundaries and approaches the linear boundary at high Rayleigh numbers.

Flow blockage in steelworks bunkers is a major source of trouble. The correct design of bunkers to ensure flow would result in reduced capital and operating costs. The Jenike method of bunker design, which claims a solution to this problem, has been critically examined for a range of iron ores using a variable-geometry wedge-shaped bunker. The main conclusion from these experiments is that the Jenike method is valid for dynamic pressure conditions and even shows some over-design. Further experiments using just one iron ore have confirmed the validity for a conical-shaped bunker under similar pressure conditions. The main limitations of the Jenike design method are as follows: (a) It does not provide a design which can accommodate impact filling. In certain cases, especially with conical bunkers, flow blockage can easily occur and (b) It does not provide a design which can be certain of eliminating arching at the transition point of a bunker with surcharge.

Bunkers used in steelworks sinter plants and other applications need to be designed and operated so that the feed emerging is unsegregated. A series of model tests has been carried out to discover which bunker shape, flow regime, and filling method give the least segregation for raw materials of different particle size and density. Results are compared with results from larger bunkers. The main conclusions are that an evenly-filled wedge-shaped mass-flow bunker is best at preventing segregation and that material density has little effect.