최종보고서_수정본.hwp

Size: px

Start display at page:

Download "최종보고서_수정본.hwp"

민수 도
5 years ago
Views:

10 . Title Development and Validation of Simulation Technique of Advection-dispersion of Water pollutants Using Field Measurement. Objectives and Necessity The main objective of the research is to execute tracer experiments in designated rivers and verify the results of the model estimation. Currently our country has various river contaminant accidents occurring at least 50 times per year, and due to the 4 Major River constructions the drastic changes in the river require a reliable prediction model based on numerical models for the accidents. Therefore for safe water supplement by rivers, accurate prediction of the water quality is necessary, and research on water quality models must be continued for the development of accurate and reliable prediction model for pollutant accidents. Simulation by numerical models using advection-dispersion of water pollutants are needed to determine the time of arrival and pollutant dispersion concentration so the development and application of simulation verification is necessary. In shallow water flow such as rivers the stratification affects by density is minimal and the vertical mixing of the pollutants is rapid, so a two dimensional advection-dispersion analysis is sufficient for the simulation of pollutant transport. But the necessity of three dimensional simulation for representation of hydraulic structures and their affects to the flow field and the pollutant transport are now becoming important, so for future analysis the expansional to a dynamic three dimensional model will be required.. Research range and details 1. Development of EFDC model verification and calibration To find the suitability of the EFDC model governing equations, the characteristics and types of three dimensional and quasi-three dimensional model were investigated. Also the EFCD model structure, pre and post processing, dynamic model, toxic transport model was put into research. Through this information understanding required for the application and

11 verification of the model using tracer tests were possible. 2. Sensitivity analysis of the dynamic model For the sensitivity analysis of the dynamic model, various simulated factors were determined and applied to various channels. The CFL conditions for the stability of the numerical solution, dispersion coeffecients, Peclet numbers were looked into for the sensitivity analysis of the mesh size, time frame, viscosity coefficients. For the objective of stable results for the dynamic model, artificial channels such as straight channel and confluent channels were tested, and dam break simulations were tested to find the changes that the mesh size, time frame, viscosity coefficients give to the simulation without non-convergence. Also EFDC models simulate three dimensional modeling by controlling layers so the affect of number of layers in the model was analyzed. With the sensitivity analysis results of the dynamic model, the parameters which affect the EFDC advection-dispersion were found out by the simulation of curved channels. EFDC is a quasi three dimensional model and the dispersion coefficients are changed by the viscosity coefficents. So the number of layers and the viscosity coefficients were changed for the sensitivity analysis of the advection-dispersion simulation. 3. Tracer tests This introduces the tracer tests developed in this research. The designated river was selected considering the accessibility, shape of ground surface and risk of pollutant accident. The test river located in Gamcheon in Gyeongbuk Gimcheon and in Gyeongnam Gimhae was selected and Rhodamine WT was used in the field experiment. The was pre tested twice, and the main field experiment was exectued in 4 times. The velocity was measured by the three dimensional hydrometer ADV Flow-Tracker, and the lateral lines were installed perpendicular to the main flow direction. The Rhodamie WT concentration were measured by water sampling by pump and location calibration was done by GPS. Other tracer test results than Rhodamine WT were also displayed and was executed in by NaCl twice. The stabilty and application of Rhodamine WT was described and hydraulic characteristics and concentration measurement methods were also introduced. 4. Analysis of the tracer test results The determination of the dispersion coefficients were depicted and dispersion coefficients were calculated by estimation and computation. There are computation methods using the measured field data and estimation methods not using the measured data. The two dimensional streamtube routing method which is applicable in natural rivers and able to calculate the

12 longitudinal and transverse dispersion coefficients at each section was especially used. Also using the measured hydraulic data a dispersion coefficient was derived by theoretical equation and validity of the results were examined. Using tracer test results instantly applied dye mixing simulation was executed for the verification of dispersion and arrival time. The simulation area starts from the Gampo bridge to Sonju bridge, the upstream area has a gradual curvature connected to the straight area, and the downstream area has two steep curvatures. The total length of the simulation area is 23 km and flow simulation and dye advection-disperson simulation was conducted. The mesh used in the numerical simulation was 8736 cells. Using the measured results of the fourth field test, the C-t curve from each measured site was calculated, and pollutant cloud maximum concentration arrival time( ), 30% concentration arrival time( ), concentration departure time( ), duration of pollutant cloud( ) was compared. Using the C-y curve, the average, variance, skewness was calculated to analyze the EFDC simulation results. And for the pollutant accident simulation such as the 2008 March Nakdong phenol river accident a advection-dispersion simulation was conducted.. Conclusion In this research, a mid size natural river with developed meander and natural river bed was selected for analysis of hydraulic characteristics such as water depth, velocity, turbulence and tracer tests were performed to calculated the dispersion coefficients. Also using the field experiment results, the change of contaminant concentration changes due to coefficient changes were analyzed. With this, the model for pollutant accident prediction (EFDC) was used and field measurement results were applied to verify the dispersion coefficients and arrival time, and sensitivity analysis of the mesh changes were conducted. Also for EFDC sensitivity analysis, the model was run in various artificial channels such as straight, confluent channels and dam break situations. 1. Investigation of the accuracy of the simulation according the Cr number The Courant-Friedrichs-Lewy (CFL) condition was introduced for the stability condition for advection and the Cr number max limit is usually defined at 1. But the research on the EFDC simulation results show that the Cr number close to 1 results in low accuracy solutions. In simulating EFDC with 1 layer, the Cr number should be lower than 0.5 as shown in Eq(5.1.1) for the stability and accuracy of the solution. TC and TM Case result comparisons show that the two cases can represent the tendency of the velocity changes, but in the case of TM the velocity changes beginning at the separation zone were more accurate. This shows that if the Cr number is below 0.5, a finer mesh would increase the accuracy of

13 the simulation. 2. Simulation stability according to the horizontal dynamic viscosity coefficient The EFDC senstivity analysis in one layer simulation is the objective of the topic. Therefore experimental simulations were conducted for the dynamic simulation sensitivity analysis by changing the horizontal dynamic viscosity coefficient which affects the depth averaged velocity field in the model. For the stability conditions according to the dynamic viscosity coefficient, the diffusion number condition and Peclet numbers were used. The confluent channel and curved channel simulation results show that if the number exceeds a certain limit, than the simulation results in non-convergence. For detailed simulation, a dam break simulation which has high transverse velocity gradient was conducted for the sensitivity analysis of the horizontal dynamic viscosity coefficient. In dam break simulation, the number and the mesh size and time interval was chosen to fit the diffusion number condition. But the Peclet number is changed in certain sections according to the simulated velocity results, so the lower limit which can satisfy the Peclet number condition cannot be determined beforehand. So continous numerical simulation must be conducted in order to determine the that is acceptable in the simulation. 3. Comparison of one layer and multi layer simulation EFDC is a quasi three dimensional model and through the adjustment of the layer number it is able to simulate both two and three dimensional modeling. The layer was increased from one to 2, 4,8 in the confluent channel for the flow simulation. In the case of multiple layers, the flow simulation shows transverse and longitudinal velocity profiles that cannot be reproduced in the single layer simulation and the size of the perpendicular velocity could be measured. With this, a secondary flow that is developed at the downstream of the confluent section could be measured. With the simulation results, the increase of the vertical velocity due to the difference between the elevation between the main and tributary section could be predicted, and the secondary flow could be observed. Also, the dam break simulation showed the vertical distribution velocity magnitude of the upstream and downstream sections of the dam. The velocity at the vertical direction was observed and showed that if using two layers the average of layer 1,2 and 3,4 were used in comparison of the case using four layers. This shows that increased layers will result in more detailed vertical velocity observation. But multiple layers will increase the time of simulation, so if the depth is shallow or the vertical velocity profile has minimal affect on the solution than a single layer simulation would be time and cost efficient.

14 4. Tracer test results Using the hydraulic characteristics and concentration measurements in the dispersion coefficient was calculated with the streamtube routing method which is applicable in natural rivers and able to calculate the transverse and longitudinal dispersion coefficients in each river sections. The streamtube routing method results show that there is difference between the field experiments and simulation results, but comparison with the theoretical equation shows its similarity. A dispersion coefficient was calculated by the four experiments conducted in and the hydraulic characteristics and concentration measurements were used with the streamtube routing method. Small differences were found through the sections but some areas showed that it matches the equation results. Since the test bed contained a lot of sand beds and meandering sections there was a probabilty that Elder and Fischer equations produced solutions with different results than the simulation. As the water depth compared to the width increased it resulted in increased dispersion coefficients and increase in shear velocity compared to the average velocity. 5. Verification of the dispersion and time arrival by numerical model Using the advection-dispersion simulation in the EFDC model the dye mixing simulation was conducted at the field test site. Using the fourth test area, a detailed mesh was formed with the boundary conditions specified by the results of the field test results. The numerical simulation result comparisons to the tracer test were done to verify the EFDC advection-dispersion simulation test results for the range of dispersion and arrival time of the pollutant at certain sections. It was difficult to reproduce the dry areas in the actual river so small errors occurred but finer mesh construction was applied to have the results close to the experimental results. Using the flow simulation results a decaying pollutant phenol input accident was simulated and research was done on the pollutant cloud arrival time and range of dispersion. Results showed that the phenol input into the center of the river had movement developed on the right side of the river due to the flow changes in the river. Therefore phenol concentration exceeding the limits of drinkable water was observed in the center and right side of the river, and the left side showed results which was below the criterion. Therefore flow characteristics as such shown in the fourth field test showed that the water intake facilities must be installed in the left side of the river to decrease the outputs of contaminant accidents. Also, since the mixing movement of the pollutant cloud is changed by the direction of the river width, a two dimensional mixing analysis is necessary. To apply the numerical model simulations in natural rivers to prediction systems for pollutant accidents, accurate analysis of the results must be produced in minimal time. But natural rivers have irregular bed formations and dry environmental areas which results in

15 difficult timewise conditions in numerical simulation because it requires detailed mesh construction. Therefore if the one dimensional model can provide accuracy in calculation results, the short calculation time and simple mesh construction will enable the one dimensional model most fit for these situations. But observing the results of the phenol input simulation, the pollutant cloud is different in the left and right side of the river in accordance with the river flow changes, so for more accurate prediction of contaminant accidents, a two dimensional mixing simulation should be conducted. In preparation of contaminant spill accidents, bed formation of principle areas near factories, tributary connections, sewer discharge points should be measured in advance for the mesh construction to have a model ready for mixing simulation in preparation and reaction to spill accidents. If a spill accident occurs in areas that are not measured and mesh constructed, a one dimensional model should be used for quick reaction and management of pollutant treatment. This will supplement the shortcomings of the two dimensional and quasi three dimensional models in developing a model for a pollutant accident prediction system.

22 sec m s m sec

23 m

24 m m m

31 ε ε ω ε ε ε

32 ε ε ω

33 ε

34 ε ω ε ε ε ε ε ω

36 U X V + Y W + Z = 0 æ H = ö ç è ø wo uoo L

37 ε ε

38 2 U U UV UW 1 p 1 t ZX 1 tyx 1 t XX fv + f yw = t X Y Z ro X ro Z ro Y ro X u u u u H u u H t t t t L L L H L H L L 2 2 o o o o o o o o o,,,, fuo, f y,,, o ro ro ro τ ρ τ τ ρ 2 U U UV UW p 1 1 t ZX fv = - + t X Y Z r X r Z o o 2 V UV V VW p 1 1 t ZY fu = - + t X Y Z r Y r Z o o 2 W UW VW W r 1 p 1 t XZ 1 tyz 1 t ZZ f yu + g = t X Y Z ro ro Z ro X ro Y ro Z u H u H u H 1 u H u H u H u t L L L L L H L L L L L L o o o o o o o,,,, fu,,,, 2 o g o p = - gr Z

39 U X V + Y W + Z = 0 ö = fv - g - - dz + ç A + F ø 2 U U UV UW h 1 pa g h r æ U Z v t X Y Z X ro X r ò o X Z è Z ö = - fu - g - - dz + ç A + Fv ø 2 V UV V VW h 1 pa g h r æ V Z v t X Y Z Y ro Y r ò o Y Z è Z u q q q q + U + V + W = t X Y Z æ q ö ç K v + Fq, q =, Z è Z ø { T S} r = r( T, S) η æ u ö æ æ u v öö Fu = ç 2Ah + ç Ah ç + x è x ø y è è y x øø æ æ u F v öö æ v Ah 2A v ö = ç ç + + ç h x è è y x øø y è y ø æ ( T, S) ö æ ( T, S) ö FT, Fs = ç Kh + ç Kh X è X ø Y è Y ø A h = CDxDy éæ U ö 1 æ V êç + ç êë è X ø 2 è X U + Y 2 ö ø æ V + ç è Y 2 ö ø ù ú úû

40 Δ Δ Δ Δ q Uq Vq Wq t X Y Z æ q ö ææ u ö æ u ö ö 2g r q = ç A + 2A ç + ç + K F Z è Z ø ç Z Z è ø Z B l q v v q è ø è ø ro q l Uq l Vq l Wq l t X Y Z æ q l ö ææ u ö æ u ö ö E1l g r = ç Aq + E1lA v ç + ç + Kv Z è Z ø ç Z Z èè ø è ø ø ro Z 2 3 q æ æ l ö ö - ç 1+ E2 ç + Fl B ç 1 è k ( h - Z ) ø è ø ε ε ε

41 2 U U UV UW t X Y Z h D = fv - g - - dz + A - c u u + v X r X X Z Z H 1 pa g h r æ U ö p Z v p o r ò ç o è ø 1 pa g h r æ V ö p Z v p o r ò ç o è ø V UV V VW t X Y Z h D = - fu - g - - dz + A - c v u + v Y r Y Y Z Z H 2 2 Z - z z = b H * 1 = Z H z * * æ ö æ 1 æ 1 b, z H ö æ, z H b ö ö ç = ç - z z X Y X H ç X X z Y H ç è ø è è ø è Y Y ø z ø h HU HV w = 0 t X Y z 2 HU HU HUV HUw h H p 1 a gh æ t X Y z X r z o X ro è ò X r ö = fvh - gh - - ç H dz ø gh æ z H ö æ A U ö * v + b z ( rs r ) r ç ç + o è X X ø z è H z ø HF u

42 2 HV HUV HV HVw h H p 1 a gh æ t X Y z Y r z o Y ro è ò Y r ö = - fuh - gh - - ç H dz ø gh æ z H ö æ A V ö * v + b z ( rs r ) r ç ç + o è Y Y ø z è H z ø HF v * * * 1 é b b b z z z H H H ù w = æ ö W U V z U V H ê ç + + t X Y ú ë è t X Y øû ( ds ) x = mxdx ( ds ) y = mydy mxmyh myhu m m xhv xmyw = 0 t x y z 2 mxmyhu myhu m m xhuv xmyhuw t x y z æ m m x y ö h myh pa = ç mxmy f - u + v Hv - mygh - è y x ø x ro x m gh æ r ö m gh æ z H ö æ A u ö * y 1 y b v ç H dz ç + z ( rs r ) ç mxmy mxmyhfu r z o è ò x ø ro è x x ø z è H z ø

43 ( ) 2 * 1 x y y x y x y x a x x x y o x x v b s x y x y v z o o m m Hv m Huv m m Hvw m Hv t x y z m m H p m Hu m gh m m f u v y y y x m gh m gh A v z H H dz m m m m HF z y z H z y y h r r r r r r æ ö = ç è ø æ ö æ ö æ ö ç ç ç è ø è ø è ø ò { },, x y y y x y y x v h h x y x y m m H m Hu m Hv m m w t x y z m m K HK HK m m T S x m x y m y z H z q q q q q q q q æ ö æ ö æ ö = + + = ç ç ç ç è ø è ø è ø q y x y x v y x o v y x q y x y x y y y x HF m m B l Hq m m z K m m g z u z u H A m m z q H A m m z z wq m m y Hvq m x Huq m t Hq m m ø ö ç ç è æ ø ö ç è æ + ø ö ç è æ + ø ö ç ç è æ = r r l y x y x v y x o v y x q y x y x y y y x HF m m Hz l E B Hq m m z K m m g E l z u z u H A E l m m z l q H A m m z z l wq m m y l Hvq m x l Huq m t l Hq m m + ø ö ç ç è æ ø ö ç è æ ø ö ç ç è æ ø ö ç è æ + ø ö ç è æ + ø ö ç ç è æ = k r r

44 Δ Δ Δ

48 ( mxmyhu) + ( myhuu) + ( mxhvu) + ( mxmywu) - femxmyhv t x y z æ * ö æ Av ö = - myh ( p + patm + f ) + my ç zb + z H p + ç mxmy u x è x x ø z z è H z ø æ m ö æ m ö 2 2 1/2 HA u HA u m m c D ( u v ) u è ø è ø y x + ç H + H - x y p p + x mx x y ç my y ( mxmyhv) + ( myhuv) + ( mxhvv) + ( mxmywv) + femxmyhu t x y z æ * ö æ Av ö = - mxh ( p + patm + f ) + mx ç zb + z H p + ç mxmy v y è y y ø z z è H z ø æ m ö æ m ö 2 2 1/2 HA v HA v m m c D ( u v ) v è ø è ø y x + ç H + H - x y p p + x mx x y ç my y m m f = m m f - u m + v m y x x y e x y x y z -1 ( t xz, t yz ) = Av H ( u, v) φ

49 Y ( ds ) x ( ds ) y x y X

50 p = - ghb = -gh - 1 ( r r ) r - o o z ρ ρ t x y z ( mxmyh ) + ( myhu) + ( mxhv) + ( mxmyw) = 0 t x y z z ( mxmyhc ) + ( myhuc) + ( mxhvc ) + ( mxmywc) - ( mxmywscc ) æ m ö æ m ö æ K ö = ç HK C + HK C + ç m m C + Q x è m x ø y è m y ø z H z y x v H ç H x y c x y è ø A = f A ql v A o φ

51 f = A -1 ( 1+ R1 Rq ) -1-1 ( 1+ R2 Rq )( 1+ R3 Rq ) æ 6A ö 1 Ao = A 1-3C - = è ø R R 1 1 ç 1 1/3 B1 B1 1 ( B - 3A ) C ( B + 6A ) B1 1 = 3A è ø 2 æ 6A ö 1 ç 1-3C1 - B1 = 9A A ( ) R = 3A 6A + B æ 6A ö ç è ø φ K v K o 1 fk = - K o = f K ql 1 ( 1+ R3 Rq ) æ 6A ö 1 = A2 ç 1- è B1 ø R q 2 ghb l = - q z H 2 2 A1, B1, C1, A2, B2 0.92,16.6,0.08,0.74,10.1 t x y z ( mxmyhq ) + ( myhuq ) + ( mxhvq ) + ( mxmywq ) æ Aq 2 ö Hq = ç mxmy q - 2mxmy zè H z ø B1l 3 æ æ ö ö m m u + v + c D ( u + v ) + gk b + Q ç H ç ç ç è z ø è z ø z è è ø ø 2 2 A 3/2 v æ ö æ ö x y ç ç h p p p v q

52 t x y z ( mxmyhq l) + ( myhuq l) + ( mxhvq l) + ( mxmywq l) æ Aq 2 ö Hq æ æ l ö æ l ö ö = ç mxmy ( q l) - mxm ç y 1+ E2 ç + E3 z H z B1 ç k Hz ç k H ( 1 z) è ø è ø è - ø è ø æ æ ö ö m m E l u + v + gk b + c D ( u + v ) + Q ç H ç ç ç è z ø è z ø z è è ø ø 2 2 A 3/2 v æ ö æ ö x y 1 ç ç v h p p p l E1, E2, E3 1.8,1.33,0.25

53 n+ 1 lq æ H ö n+ 1-l lq æ H ö ( Hu) + x z xz ( ) x z xz 2 ç d f -d t Hu m = - x 2 ç d f -d t m è ø è x ø lq N + 1 -l/2 - { d x ( myhuu) + d x ( mxhvu) + d z ( mxmywu) } m m x x y * { d x ( atm ) ( d x b d x ) } n+ 1 n+ 1-l lq + lq ( fehv) - lq B + Q m m N l/2 N l/2 N + 1 -l/2 H B = p + p + z + z H gb m x y δ θ λ λ * n+ 1-l lq æ H Q ö ( Hu) = ( Hu) - ç d xf -d zt xz - fehv + B - 2 mx mxm è y ø n+ 1-l ** * ( Hu) = ( Hu) - { ( myhuu x ) + x ( mxhvu) + z ( mxmywu) } x y n+ 1 -l/2 blq * * * - { d x ( myu% ( Hu) ) + d x ( m ( ) ) z ( mxmyw( ) xv% Hu + d % Hu )} m m x y alq d d d m m n+ 1 -l/2

54 α β α β *** ** ( Hu) = ( Hu) blq n+ -l * * * - { d x ( m ˆ ( ) ) x ( ˆ( ) ) z ( ˆ yu Hu + d m mxmyw( ) xv Hu + d Hu )} m m x y 1 /2 n+ 1 lq æ H ö q 2 ç m è x ø 2 n+ 1 *** ( Hu) + d f -d t = ( Hu) - ( 2 - l ) ( f Hv - B) ( l -1) 2q ( f Hv B) lq - ( l -1) q ( fehv - B) + Q 2m m = F n+ 1 x x z xz e + - e n n- 1 n+ 1 x y n+ 1 lq lq n n ( Hu ) ( Hd f ) ( t t ) n+ 1 n+ 1 = - x + sx - bx + Fx 2mx 2 ( Hv ) n+ 1 φ

55 lq n+ 1 + { ( myghu x ) + y ( mxghv )} = 2m m n+ 1 f d d x y lq 2m m { ( myghu x ) y ( mxghv )} n+ 1-l f - d + d x y n+ 1-l lq lq n n ( Hu ) ( Hd f ) ( t t ) n+ 1 n+ 1 = - x + sx - bx + Fx 2mx 2

56 t x y z ( mxmyhcw ) + ( myhucw ) + ( myhvcw ) + ( mxmywcw ) i i i j j j ( å ds cs å dd cd ) i j æ Kv Cw ö = m K S K D xmy mxmyh + ç + x è H z ø ì i i æ Cw ö i i j j æ Cw ö j j ü - m w ˆ w ˆ xmyh í KaSS ç y cs - cs + KaDD y cd - cd + g Cw i f ç ý j f î è ø è ø þ å å χ χ φ ψ γ t x y z j j j j j j j j ( mxmyhd cd ) + ( myhud cd ) + ( myhvd cd ) + ( mxmywd cd ) j j æ Ab D c ö D j j æ Cw ö j j = ç m w ˆ xmy + mxmyhkadd y cd - cd x H z ç f è ø è ø - m m HK + g D c j j j x y dd D t x y z i i i i i i j j ( mxmyhs cs ) + ( myhus cs ) + ( myhvs cs ) + ( mxmyws cs ) i i æ Ab S c ö S i i æ Cw ö i i = ç m w ˆ xmy + mxmyhkass y cs - cs x H z ç f è ø è ø - m m HK + g S c i i i x y ds S C = D c j j j D D

57 C = D c i i i S S t x y z ( mxmyhcw ) + ( myhucw ) + ( myhvcw ) + ( mxmywcw ) i i j j ( å ds S å dd D ) i j æ Kv Cw ö = m K C K C xmy mxmyh + ç + x è H z ø ì i i æ Cw ö i i j j æ Cw ö j j ü - m w ˆ w ˆ xmyh í KaSS ç y cs - cs + KaDD y cd - cd + g Cw i f ç ý j f î è ø è ø þ å å t x y z j j j j ( mxmyhcd ) + ( myhucd ) + ( myhvcd ) + ( mxmywcd ) j æ Ab C ö D j j æ Cw ö j j = ç m w ˆ xmy + mxmyhkadd y cd - cd x H z ç f è ø è ø - m m HK + g C j j x y dd D t x y z i i i i ( mxmyhcs ) + ( myhucs ) + ( myhvcs ) + ( mxmywcs ) i æ Ab C ö S i i æ Cw ö i i = ç m w ˆ xmy + mxmyhkass y cs - cs x H z ç f è ø è ø - m m HK + g C i i x y ds S

62 sin sin

64 m s m s

68 m m s m s

69 m m s m m m

73 Ÿ Ÿ Ÿ Ÿ Ÿ

80 sin sin

82 m s m s m s

85 m m

89 m m

91 m s m s m s m s

94 m m

98 m m

99 m

100

101 m s m s

102

103

104

105 m

106

107

108

109

110 m

111

112

113

114

115 m

116

117

118

119

120 m

121

122

123

124

125

126

127

128

129

130

131

132 초

133 m s m m s m s m s

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167 μ μ 연결케이블 측정위치 측선 LGH RGH 0.6H Detector

168

169

170

171 Φ

172

173 m s

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198 Φ

199

200

201

202

203

204 Φ

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219 ln

220 ln

221 exp exp

222

223 ln ln ln ln ln

224 exp exp exp exp

225 exp exp exp

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248 km km km km km

249

250 km km km

251

252 ε ε ε

253 sec m s m m

254 m sec

255

256

257

258

259

260

261

262

263

264

PJTROHMPCJPS.hwp

PJTROHMPCJPS.hwp 제 출 문 농림수산식품부장관 귀하 본 보고서를 트위스트 휠 방식 폐비닐 수거기 개발 과제의 최종보고서로 제출 합니다. 2008년 4월 24일 주관연구기관명: 경 북 대 학 교 총괄연구책임자: 김 태 욱 연 구 원: 조 창 래 연 구 원: 배 석 경 연 구 원: 김 승 현 연 구 원: 신 동 호 연 구 원: 유 기 형 위탁연구기관명: 삼 생 공 업 위탁연구책임자: