Food pastes and suspensions are complex materials which are difficult to characterise in conventional rotational rheometry due to their complex microstructure and their usual tendency to slip at the wall. Squeeze flow, in which material flows between two circular plates as they are brought together, has traditionally been used as a way of overcoming these problems. Two squeeze flow rheometers were set up incorporating two novel pressuresensing devices, a strip sensor and a grid sensor. The strip sensor which measures the pressure distribution in the material along a radius, was found to be too sensitive to test fluids. The grid sensor which measures the whole two-dimensional pressure distribution in the material, was more robust and was validat~d using Newtonian silicone oils. Excellent agreement was found between the pressure sensor data and the well-known Stefan's theory, over the range of dimensionless times 0.4 to 0.8. On extension to non-Newtonian fluids, reasonable agreement was found between the well-known Scott's theory and the pressure data for a model pseudoplastic fluid, over an equivalent range of dimensionless times. The results for a paste material, however, did not agree well with Scott's theory. The flow regime in squeeze flow is different to that found in the simple shear flow of rotational rheometers. In squeeze flow the flow pattern changes from elongational to shear flow as the experiment progresses, which has two consequences: (i) it is unlikely that a single rheological equation could describe the flow over the whole experiment and (ii) the different flow regimes mean that comparison with rheological parameters obtained from conventional rheometry is unlikely to be successful.

Experiments have been performed to investigate the relationship between the geometry of an idealised bed roughness and the structure of the turbulent boundary layer above. This work has particular applications in relation to architecture and environmental health, designing comfort for pedestrians and managing environmental pollution. Experiments were carried out in a laboratory water flume to study the turbulence structure generated under both rough and smooth bed conditions. The relationship between small scale turbulent eddies formed between the elements of a two-dimensional bed roughness and the intermittent turbulent structure of the boundary layer was established in the form of roughness length scale and friction velocity using LDV measurements from the mean velocity profile. Turbulence production was also determined from measurements of the Reynolds stress and velocity gradient. This was compared with the diffusion characteristics of the flow. Spacing between the roughness slats was varied for different tests to produce isolated (k-type) bed roughness, skimming (d-type) flow, and wake interference flow for transitional roughness. Transition between these regimes was further investigated using flow visualisation, and a clear correlation established between the vortices at the bed and the turbulence characteristics of the boundary layer. The effect of sudden changes in bed geometry on the turbulence structure was investigated for flows aligned perpendicular to the roughness elements forming simulated urban environment. These results were compared for a similar model street canyon with and without a large scale upstream approach roughness. The turbulence generated as a result of the change in local bed geometry has a significant influence on the dispersion of dye tracer used to simulate pollutant. It was found that uniformity in height of buildings along parallel streets in an urban environment promotes shear at roof-level, thereby trapping fluid and pollution within the canyon. On the other hand, non-uniformity in building height and the presence of substantial upstream buildings promotes turbulence which helps in ventilation of street canyons.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.