Computational Methods in Engineering

---

In this paper, a numerical scheme is proposed for the multi-fluid compressible flows. This method is applied to the problem of underwater explosion. The proposed scheme is basically the extension of Godunov method in gas dynamic problems to the multifluid environments and is second-order accurate in space. In this method, also, the problem of artificial mixing of two different phases on Eulerian grids is prevented by a front tracking technique. The numerical results of this study are in very good agreement with previous numerical and exprimental results

---

When rails are laid in a railway track, small gaps are left between the ends of adjoining rails to allow for the expansion of the rails when the temperature rises. Rail joints are provided to form a continuous rail path which are normally laid in standard lengths bolted together by fishplates. When two rail joints are exactly opposing each other, they are called square (side by side) joints. Manufacturers produce rails with different standard lengths. The inner line of rails in a curve is shorter than the outer line therefore, it would be complicated to have square joints by using only rails of standard length for both inner and outer rails in the curve. If some short rails were used in the inner line, it would be possible for all joints of rails to lie opposite each other in a precise manner. It is clear that equal rail lengths can easily create square joints along the straight portion of the track. In curves with larger radii, there is a substantial difference between inner and outer arcs of the cruve. Manufacturers producing different standard lengths also produce some short rails in two or three different lengths. This paper presents a simple method of finding the location of joints in the railway curve. Previous methods (used in European and American railways) use some short rails (3 or 4 short rails) with different lengths for the inner arc. In this method, short rails of one single length are used.

---

Investigation of vertical vibrations of a railway turnout is important in designing track components under moving loads of trains. In this paper, the turnout is simulated by a linear finite element model with modal damping. A section of the turnout has a length of 36 sleeper spans surrounding the crossing. Rails and sleepers are modeled with uniform Rayleigh- Timoshenko beam elements. The rails are connected via railpads (linear springs) to the sleepers, which rest on an elastic foundation. The rolling stocks are discrete systems of masses, springs, and dampers. By passing the trains at a constant speed, only vertical dynamics (including roll and pitch motions) is studied. The wheel-rail contact is modeled using a non-linear Hertzian spring. The train-track interaction problem is solved numerically by using an extended state space vector approach in conjunction with modal superposition for the turnout. The results show that the rail discontinuity at the frog leads to an increase in the wheel-rail contact force. Both smooth and irregular transitions of the wheels from the wing rail to the crossing nose have been examined for varying speeds of the vehicle. Under perfect conditions, the wheels will change quite smoothly from rolling on the wing rail to rolling on the nose. The impact at the crossing will then be small, giving a maximum wheel-rail contact force which is only 30--50 per cent larger than the static contact force. For uneven transitions, the severity of the impact loading at the crossing depends strongly on the train speed. The increase in the contact force, as compared with the static force, is in the order of 100 per cent at 70 km/h and 200 per cent at 150 km/h.

---

In present study, dynamic modeling and control of a tethered space robot system in trajectory tracking of its end effector is investigated. Considering variation of the tether length in the model, dynamics of the system is modeled using Lagrange’s method. Librational motion of the tether is controlled by adjusting the tether length similar to conventional manipulators,control of the robot is performed by its motors. It is clear that, in the trajectory tracking of the end effector, the tether length should be kept more or less constant, keeping them in a stable position. Limiting the tether length variation while using it as a tool for controlling the tether librational motion, is the main challenging part of the control system. To deal with this problem, a hybrid control  system is proposed to control the system. A nonlinear model predictive control approach (NMPC) is utilized to control the tether librational motion and a modified computed torque method (CTM) is used to control the manipulator motion. Initially the NMPC controller is developed for a simple tethered satellite system. Then it is combined with the CTM controller. The proposed controller is employed to control motion of a space robot’s end effector on a predefined trajectory. Performance of the controller is then evaluated by numerical simulations.