Mold filling was studied for vertical, thin, plate-shaped cavities, with the liquid entering via some vertical ingate system connected to the bottom. In this study, sand molds were filled up with molten aluminum and molten cast iron, while water was used to fill up perspex models.The front walls of the sand molds were replaced by glass plates to allow observations of temperature distributions and free-surface behavior of the melt during filling. Computational fluid dynamics (CFD) simulations of the filling process were carried out to study free-surface behavior, velocity patterns, and temperature distributions. Digital particle-image velocimetry (DPIV) was used to validate the computer simulations for water.Generally, visual observations of the molten liquids, CFD simulations, and DPIV results are in good agreement. Combining the three techniques has resulted in a better understanding as to how a plate-shaped cavity is filled up.The behavior of the free surface is different for water and the molten metals. An analysis of surface waves is presented that explains these differences. Current ideas as to the role of the Weber number must be rejected. Rather, instabilities are associated with low values of the Ohnesorge number, with surface tension providing the driving force for surface instabilities and with viscosity as the damping force.

Revascularization of the femoropopliteal sector is often performed by the placement of a bypass. In this paper, we have studied the effects of hemodynamics on patency of the bypass for different positions of the distal anastomosis close to a collateral artery. Computational fluid dynamics (CFD) are used for this study. The cardiac cycle-averaged wall shear stress (WSS) and oscillation index (OSI) have been analyzed. Low WSS and high OSI may increase the risk of intimal hyperplasia (IH), which may reduce bypass patency. From the CFD simulations, spots of low WSS and high OSI are found within and near the entrance of the collateral artery, near the suture line, at the floor, toe, and heel. We regarded flow ratios of 20:80 and of 35:65. It is found that for the high flow ratio anastomosis located proximal to the collateral artery is clearly more advantageous. However for the low flow ratio anastomosis located distal to the collateral artery seems to be slightly more advantageous, the results are less conclusive. One of the studied flow geometries has been validated by in vitro experiments using a time resolved particle image velocimetry technique. Velocity fields from these experiments are in good agreement with the CFD results.