No abstract

This study investigated the thermal conditions preceding ignition of three dense woody fuels often found on structures by firebrands, a major cause of home ignition during wildland-urban interface (WUI) fires. Piles of smoldering cylindrical firebrands, fabricated from wooden dowels, were deposited either on a flat inert surface instrumented with temperature and heat flux sensors or on a target fuel (marine-grade plywood, oriented-strand board, or cedar shingles) to investigate critical conditions at ignition. The former provided thermal data to characterize the time before and at ignition, while the latter provided smoldering and flaming ignition times. Tests were conducted in a small-scale wind tunnel. Larger firebrand piles produced higher temperatures at the center of the pile, thought to be due to re-radiation within the pile. Ignition was found to be dependent on target fuel density; flaming ignition was additionally found to be dependent on wind speed. Higher wind speeds increased the rate of oxidation and led to higher temperatures and heat fluxes measured on the test surface. The heat flux at ignition was determined by combining results of inert and ignition tests, showing that ignition occurred while transient heating from the firebrand pile was increasing. Ultimately, critical ignition conditions from firebrand pile exposure are needed to design appropriate fire safety standards and WUI fire modeling.

The two-dimensional thermal elastohydrodynamic equations were numerically solved for a Ree-Eyring type lubricant under pure rolling conditions. Profiles of lubricant pressure, film thickness, and temperature were obtained for medium to heavy loads and moderate to high rolling speeds. The pressure results generally show a small secondary peak near the outlet, but at the highest load considered no pressure spike is obtained and the pressure profile is almost Hertzian. The film thickness results show an increase in minimum film thickness with increasing rolling speeds, but at a lesser rate than those predicted for a Newtonian fluid under isothermal conditions. It is found that unless the lubricant becomes non-Newtonian in the inlet region, the reduction in minimum film thickness at high rolling speeds is completely due to thermal effect. The lubricant temperature profile and the amount of heat generated and dissipated in the contact region were also calculated. The lubricant temperature reaches a maximum just before the entrance to the Hertz contact region. Both shear and compression heating are found to be important in raising the lubricant temperature in the inlet. As the lubricant enters the Hertz contact zone, the temperature first drops rapidly, because of the rapid heat conduction to the rollers, and then remains almost constant for most of the Hertz contact. Near the exit where the pressure gradients are large, the lubricant temperature drops rapidly below the ambient because of lubricant expansion. The lubricant then heats up rapidly before leaving the contact area as a result of heat generated by shear stresses.

The stress intensity factors for short straight and branched subsurface cracks subjected to a Hertzian loading are calculated by the finite element method. The effect of crack face friction on stress intensity factors is considered for both straight and branched cracks. The calculations show that the straight crack is subjected to pure mode II loading, whereas the branched crack is subjected to both mode I and mode II, with ΔKI/ΔKII < 0.25. Although KI is small, it strongly influences KII by keeping the branched crack faces apart. Based on the ΔKII values and Paris’s crack growth model, the number of stress reversals required to grow a crack in a rolling component from an initial threshold length to the final spalling length was estimated. It was found that the crack propagation period is small compared with the expected bearing fatigue life. Therefore, crack propagation is not the rate controlling factor in the fatigue failure of bearings operating under normal loading levels.

No abstract