All 1 entries tagged Experiment
View all 10 entries tagged Experiment on Warwick Blogs | View entries tagged Experiment at Technorati | There are no images tagged Experiment on this blog
December 03, 2015
Today we learned about the Taguchi Method by experiment (building 8 different types of paper helicopters). In order to improve the process, we must experiment first by testing our theories. We must always be on target with minimum variation. If we start to move away from our target, that could cost a lot of money for the company. The one-sided tolerance level's formula are: L = ky2 and L = k(1/y2). For Taguchi, the off-line quality control is more important which is about the design of products and processes. It is essential to create a feasible design space, optimise it and innovate. There are two main factors that affect the design or the process: the control factors that can be set and controlled, and the noise factors that cannot be controlled. In our helicopter experiment, the control factors were the followings: the length of the wings, the width of the wings, we had an interaction, the length of the body, the width of the body, the width of the head and the paper material. Robustness is crucial in case of the uncontrolled noise factors because we have to make sure that our operation works consistently well with minimum variation. Taguchi method is great in terms of time and cost saving. By getting the result of the noise factors, we can suggest which experiment will have the best result by looking the signal-to-noise ratio of "larger the better". η= -10logMSD (mean square deviation)
By getting the best combination of the factors, we can calculate the prediction of our process.
This robust design focuses on the improvement of products and processes by reducing cost, improving quality and reducing intervals.