Semester

Spring

Date of Graduation

2006

Document Type

Thesis

Degree Type

MS

College

Statler College of Engineering and Mineral Resources

Department

Industrial and Managements Systems Engineering

Committee Chair

Steven Guffey

Abstract

Experiments were performed to determine a predictive model for sliding gate damper resistance (Xdamper). Independent variables included eight ratios of insertion depth (I) to duct diameter (D), two different values of D, four different levels of air velocity, two damper orientations, and two damper edges (Concave/Straight). Analysis of Log (Xdamper) values showed a highly linear relationship with I/D and high statistical significance (p<0.05) for all independent variables and for most two-way interactions. Nearly the entire regression model (R2 = 0.985) was explained by I/D, so all but I/D and Edge were dropped from the final model: Y Log(Xdamper) = -1.434 + 0.089*(Edge) + 4.30*( I/D) (R2 = 0.978). The resulting predictive model: Xdamper = 0.037*10(0.089*(Edge )+4.3*I/D). However error for untransformed Xdamper increased sharply with increasing (I/D) and percent errors were as high as 125% across all values of (I/D). Hence the predictive models are useful for only initial rough adjustment and should be followed by final trial and error adjustment to goal air flow.

Share

COinS