¡@¡@¡@¡@¡@¡@¡@­^¡@¤å¡@ºK¡@­n¡@¡@¡@¡@¡@¡@¡@ Heat transfer of a molten splat to a thin layer rapidly solidified on a cold substrate and the heat transfer coefficient at the bottom surface of a splat is extensively and self-consistently investigated. Rapid freezing in the splat is governed by a nonequilibrium kinetics at the solidification front in contrast to the melting in the substrate simulated by the traditional phase change problem. Solving one-dimensional unsteady heat conduction equations and accounting for distinct properties between phases and splat and substrate, the results show the effects of dimensionless parameters such as the dimensionless kinetic coefficient, stefan number, latent heat ratio, initial, equilibrium melting, and nucleation temperature, and conductivity, density, and specific heat ratios between solid and liquid and splat and substrate on unsteady temperature fields and freezing and melting rates in the splat and substrate and on unsteady variation of Biot number are presented. The unsteady variation of the heat coefficient or Biot number can be divided by five regimes: liquid splat-solid substrate, liquid splat-liquid substrate, solid splat-solid substrate, solid splat-liquid substrate, and the nucleation of the splat. Appropriate choices of dimensionless parameters to control the time for freezing and melting of the splat and substrate and an understanding and estimation of the heat coefficient at the bottom surface of the splat therefore are presented. The velocity distribution function and transport variables of the positive ions and electrons in the collisionless presheath and sheath of a plasma near a wall partially reflecting ions and electrons are determined from a kinetic analysis. Since velocities of the ions and electrons near the wall are highly non-Maxwell-Boltzmann distributions, accurate predictions of transport variables such as density, fluid velocity, mean pressure, fluidlike viscous stress and conduction require kinetic analysis. The result find that dimensionless transport variables of ions and electrons in the presheath and sheath can be exactly expressed in terms of transcendental functions determined by dimensionless independent parameters of ions and electrons reflectivities of the wall, ion-to-electron mass ratio, charge number and electron-to-ion temperature ratio at the presheath edge. The effects of the parameters on transport variables at the wall are also obtained. The computed transport variables in the presheath and sheath show agreement with available theoretical data for a completely absorbing wall.

The modelling of single-molecule experiments is of vital interest to gain new insights into processes which were hitherto not accessible by measurements performed on bulk systems. In the first part of this thesis, the kinetics of a triple-branch DNA molecule with four conformational states is investigated by employing pulling experiments with optical tweezers and theoretical modelling. Probability distributions of first rupture forces, which are calculated by applying transition rate theory to a free energy model, show good agreement with experimental findings. Permanently frayed molecules could be identified by analysing the number of opening base pairs in force jumps. In the second part of the thesis, DNA hairpin molecules with periodic base sequences are studied. Their unfolding kinetics allows an analytical treatment, because they exhibit a regular coarse-grained free energy landscape as a function of the number of opened base pairs. A procedure is developed for determining all relevant parameters of the landscape, which relies on probabilities that can be easily sampled from the unfolding trajectories. By means of Monte Carlo simulations it is shown that already 300 trajectories, as typically measured in single-molecule experiments, provide faithful results for the energetic parameters. The approach in particular opens a new access to improve loop contributions in the free energy landscape. In the third part of the thesis, a simulation method is developed for modelling the unfolding kinetics of DNA molecules with arbitrary base sequences. The method is validated against experimental data for five DNA hairpin molecules with different length of the end-loop. Applications of the method enable one, among others, to improve the parameter determination in functional forms suggested for the tail behaviour of work distributions. Such work distributions enter detailed and integral fluctuation theorems, which are useful for estimating free energy differences between folded and unfolded states from nonequilibrium measurements.