Strukturerade spårväxelexperiment med hjälp av försöksplanering

Sammanfattning: The Swedish railway is an important part of the country's infrastructure, but it is also a part that requires large amounts of maintenance to maintain safety and functionality. A central part of the railway is switches that enable the interconnection of different tracks. Switches include many moving parts located in tiring positions that are affected by weather and temperature. As a result, these systems often suffer from problems that can lead to downtime. To reduce these problems, The Swedish Transport Administration is constantly working to develop an understanding of how switches are affected in different situations. That work is ongoing, but there are unfortunately few guidelines and tested methodologies for experimental studies. A usually powerful method for experimentally investigating and developing knowledge of processes or systems is design of experiments. Design of experiments has historically been used few times to investigate effects within railways and their systems but has recently been shown to be very useful in that area as well. A large portion of the problems that occur on railway systems is in switches. Often the maintenance of switches is more expensive than the actual procurement. To reduce the problems in exposed switches, sometimes called problem switches, various systems are used to extend the service life and reduce the need for maintenance. One such system is Ekoslides, which is intended to facilitate shifts by reducing the amount of energy required during a shift. The system lifts the switchblade in a switch to reduce friction during movement. Exactly how much effect Ekoslides have on the switches in Sweden is unknown and optimization of Ekoslides could lead to longer service life and more opportunity to plan maintenance, which results in fewer errors. The thesis, therefore, aims to use experimental planning in a practical experiment on switches to evaluate on the methodology and to increase understanding of how Ekoslides affects a switch. The master’s project has been conducted as a six sigma project according to the problem-solving method DMAIC (Define, Measure, Analysis, Improve, and Control) in combination with design of experiments. Two practical experiments were performed on switch 661 located in the vicinity of the Boden central station. During the experiments, Ekoslide settings were used as experimental factors, while energy use and conversion time in the switch was measured as responses. On Ekoslides there are wheels that lift the switchblade. The inner wheels were varied between 0 and 1 millimeters and the outer wheels between 0 and 3 millimeters. This way, the effect of Ekoslides in switch 661 was examined at varying settings. The results from the experiment showed that Ekoslides significantly affected energy use when a switch performs a shift. Ekoslides resulted in small but significant energy savings. It was found that the best results were achieved when Ekoslides were mounted in new positions and had the settings at the highest level, which means 3 millimeters for the outer wheel and 1 millimeter for the inner wheel. Furthermore, experimental planning was also a time-efficient and good method for performing experiments on switches. Despite challenges, two short experiments resulted in increased knowledge and reduced energy usage in switch 661. Based on the evaluation of the conducted switch experiments, it was found that pilot trials, separate preparations, and detailed feasibility studies were important if design of experiments should be used to investigate cause and effect in a switch.