• Advances in aircraft and spacecraft science
• /
• v.6 no.6
• /
• pp.529-550
• /
• 2019

The effect of multiple process parameters on a set of continuous response variables is, especially in experimental designs, difficult and intricate to determine. Due to the complexity in aeroacoustic and vibroacoustic studies, the often-performed simple one-factor-at-a-time method turns out to be the least effective approach. In contrast, the statistical Design of Experiments is a technique used with the objective to maximize the obtained information while keeping the experimental effort at a minimum. The presented work aims at giving insights on Design of Experiments applied to aeroacoustic and vibroacoustic problems while comparing different experimental designs and approximation models. For this purpose, an experimental rig of a ducted low-pressure fan is developed that allows gathering data of both, aerodynamic and aeroacoustic nature while analysing three independent process parameters. The experimental designs used to sample the design space are a Central Composite design and a Box-Behnken design, both used to model a response surface regression, and Latin Hypercube sampling to model an Artificial Neural network. The results indicate that Latin Hypercube sampling extracts information that is more diverse and, in combination with an Artificial Neural network, outperforms the quadratic response surface regressions. It is shown that the Latin Hypercube sampling, initially developed for computer-aided experiments, can also be used as an experimental design. To further increase the benefit of the presented approach, spectral information of every experimental test point is extracted and Artificial Neural networks are chosen for modelling the spectral information since they show to be the most universal approximators.

• Journal of the Korean Society for Precision Engineering
• /
• v.30 no.10
• /
• pp.1079-1086
• /
• 2013

Optimal mount position of a front pillar trim considering heat resistant characteristics can be determined by two methods. One is conventional approximate optimization method which uses the statistical design of experiments (DOE) and response surface method (RSM). Generally, approximated optimum results are obtained through the iterative process by a trial and error. The quality of results depends seriously on the factors and levels assigned by a designer. The other is a methodology derived from previous work by the authors, which is called sequential design of experiments (SDOE), to reduce a trial and error procedure and to find an appropriate condition for using artificial neural network (ANN) systematically. An appropriate condition is determined from the iterative process based on the analysis of means. With this new technique and ANN, it is possible to find an optimum design accurately and efficiently.

• Journal of the Korean Statistical Society
• /
• v.29 no.2
• /
• pp.219-229
• /
• 2000

Oh et al.(1999) 3$\times$2 crossover design for assessing bioequivalence when two new generic drug formulations and innovator are simultaneously considered. This design is not only more efficient than 3$\times$3 one, proposed by Lee et al.(1998), in practical sense, but also more ethical in medical sense. However, the general statistical methods are not directly applicable to both designs when subjects are dropped out in the experiment, even though it is always possible in bioavailability and bioequivalence studies because of some administrative and economic reasons. In this research we propose an inference to drug effects when subjects are dropped out in the planed-for 3$\times$3 and 3$\times$2 crossover experiments. An example is given for illustration.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.7 no.2
• /
• pp.101-106
• /
• 2006

The optimum design of turbine blade at minimized fatigue life can be derived by the statistical fatigue analysis in this study, The optimum value of positions in the axes of X and Y at turbine blade can be found by design of experiments on the condition that the value of fillet radius is fixed to minimize the fatigue life. The degree of uncertainty about process at the factors in the axes of X and Y can be calculated by six sigma analysis. The optimum value of fillet radius is determined by utilizing the robust design at uncertain condition. It is concluded that maximum von Mises stress can decreased by 20% and the fatigue life can be double.

• Korean Chemical Engineering Research
• /
• v.61 no.2
• /
• pp.322-327
• /
• 2023

A design of experiments was evaluated in optimizing MOF-5 synthesis for acetylene adsorption. At first, mixture design was used to optimize precursor concentration, terephthalic acid, zinc acetate dihydrate and N,N-dimethylformamide. More specifically, 13 conditions with various molar ratios were designed by extreme vertices design method. After preparing the samples, XRD, N₂ physisorption and SEM analysis were performed for their characterization. Moreover, acetylene adsorption experiments were carried out over the samples under identical conditions. The optimal precursor composition for MOF-5 synthesis was predicted on a molar basis as follows: terephthalic acid : acetate dihydrate : dimethylformamide = 0.1 : 0.4 : 0.5. Thereafter, multi-level factorial design was designated to investigate the effect of synthesis reaction conditions such as temperature, time and stirring speed. By the statistical analysis of 18 samples designed, 4 reaction parameters were determined for additional adsorption experiments. Therefore, MOF-5 prepared under the synthesis time and temperature of 100 ℃ and 12 h, respectively, showed the maximum adsorption capacity of 15.1 mmol/g.

• Journal of the Korean Statistical Society
• /
• v.7 no.2
• /
• pp.109-119
• /
• 1978

When a response surface by a seconde order polynomial regression model, the stationary point is obtained by solving simultaneous linear equations. But the point is a function of random variables. We can find a confidence region for this point as Box and Hunter provided. However, the confidence region is often too large to be useful for the experiments, and it is necessary to augment additional design points in order to obtain a satisfactory confidence region for the stationary point. In this note, the author suggests a method how to augment design points "eficiently", and shows the change of the confidence region of the estimated stationary point in a response surface.e surface.

• Journal of the Korean Society for Railway
• /
• v.16 no.2
• /
• pp.144-150
• /
• 2013

This paper presents an analysis of door design for the Ui LRT based on experiments to predict doorway flow-time and their analyses results. A similar railway vehicle (from Gimhae LRT) and operational conditions are utilized to assess the doorway flow-time through repetitive experiments. Design of the experiments consists of four scenarios, and the experiments are repeated 39 times in total. We use the results of the experiments to verify the design of doors of Ui LRT (e.g. the required number of doors and their width). Various statistical analyses are carried out for the flow-time with respect to the number of boarding/alighting passengers. We make three category levels of boarding/alighting passengers, and analyze the mean and variance for each category, and then carry out One-Way ANOVA to analyze how the number of boarding/alighting and onboard passengers impact flow-time. The results of this paper can be used for making decisions about doors of the LRT vehicle.

• Communications for Statistical Applications and Methods
• /
• v.3 no.2
• /
• pp.103-117
• /
• 1996

In the Taguchi Parameter design, the product-array approach using orthogonal arrays is mainly used. However, it often requires an excessive number of experiments. An alternative approach, which is called the combined- array approach, was suggested by welch et. al. (1990) and studied by Vining and Myers(1990), Box and Jones (1992) and others. In these studies, only single response variable was considered. We propose how to simultaneously optimize multiple responses when there are correlations among responses, and when we use the combined-array approach to assign control and noise factors.

• Journal of the Korean Vacuum Society
• /
• v.1 no.1
• /
• pp.198-205
• /
• 1992

Abstract-Planarieation technology, which is essential to VLSI, has been developed using non-etch back Spin- On-Glass (SOG). Process factors for 1.5 micron double metal technology are optimized by the statistical design of experiments. Optimum conditions are found to be a process with twice SOG coating, sufficiently long hot plate baking at 300t, and furnace curing for 40 minutes below 400$^{\circ}$C.

• IE interfaces
• /
• v.18 no.4
• /
• pp.402-411
• /
• 2005

Secondary batteries with high performance are essential in widespread use of modern portable devices such as cellular phones and laptop computers. High energy density, long cycle life, and safety are some of important requirements for secondary battery. To achieve such characteristics, a mixing proportion of electrolyte solution ingredients in the battery should be carefully chosen. In this paper, using statistical design of mixture experiments (DOME), we attempt to find an optimum condition of designing the secondary battery. DOME has a distinct feature in that the experimental region is represented by simplex, rather than hypercube, because the sum of blend proportions should be unity. Several designs based upon this point have been proposed for mixture experiments. Among them, an extreme vertices design is employed in this paper because there are a couple of blend constraints to be considered. In order to investigate how the mixing proportion interacts with other manufacturing factors, a fractional factorial design is also included across the extreme vertices design. As a result, we find that the blend proportion of solution ingredients has a significant effect on battery performances. By simultaneously optimizing two battery capacities, this paper proposes an optimum blend proportion according to process factor settings.