Heat Transfer Project

Proposal for Wind Turbine with Fins

Spencer Fulmer, David Maze, Amanda Ostberg & Bonnie Vue
Heat Transfer 4600:315 5:10-6:25
Dr. Jerry Drummond
December 10, 2013

Table of Contents
Problem Statement .........................................................................Error! Bookmark not defined.
Summary of Results ........................................................................................................................ 2
Discussion. ...................................................................................................................................... 3
Appendices. ..................................................................................................................................... 4
Nodal Equation breifing. ......................................................................................................................................... 4a
Nodal Equations. .................................................................................................................................... 4-1 thru 4-13
Figure 1 .....................................................................................................................................................................5
Figure 2 .....................................................................................................................................................................5
Figure 3 .....................................................................................................................................................................6
Figure 4 .....................................................................................................................................................................7
Figure 5a ...................................................................................................................................................................8
Figure 5b ...................................................................................................................................................................9
Figure 6 ................................................................................................................................................................... 10
M-file 1 ................................................................................................................................................................... 11
MATLAB Nomenclature ........................................................................................................................................ 17
Virus free disc including M-file 1 (Attached) ......................................................................................................... 18

Heat Transfer Project

ii

Amanda Ostberg, Spencer Fulman, David Maze & Bonnie Vue

Heat Transfer 4600:315
Dr. Jerry Drummond
December 6, 2013

Heat Transfer Project

Proposal for Wind Turbine with Fins

Team 2 is delighted to have the opportunity to propose a solution to

your wind turbine. We believe that we could dramatically improve
the energy efficiency just by adding fins to the nacelle. Adding fins
to the nacelle is necessary to permit the use of higher energy
dissipation requirements for the component inside the nacelle.
Energy dissipation is a physical process of the body cooling in an
ambient temperature and the energy is now unavailible and
irreciverable.
With the model above of the nacelle as a finned tube we will be able analyze to how the
temperature varies within the nacelle and the fins. The cylindrical model is removing the heat
while the components inside the nacelle produces 5 kW of energy that must dissipate. For
simplification we will use a small segment and some assumptions. The material for the nacelle
and fins are AISI 316 stainless steel where

We are assuming that the material properties do not change with temperature. The overall length
of the nacelle is 1.6 m and the designers intend to use 80 fins on the outside cylinder.
Throughout the next couple of pages we will use one-dimensional analysis of the finned tube to
estimate the inside temperature of the nacelle wall, the temperature at the outside surface, and the
outside tip of the fin. Also we will analyze the two-dimenional analysis to predict a more
accurate temperature by writing the governing partial differential equations for the steel along
boundary conditions. There is a matlab code to follow; which will solve for a steady-state
temperature to go along with the above two dimensional analysis.

Heat Transfer Project

Summarization of Finite Element Analysis Results

After completing a one dimensional analysis of the nacelle wall (Fig. 1), as well as an indepth two dimensional element analysis of the nacelle wall; one can interpret this data to obtain
results and conclusions. As shown in Fig. 5a & 5b a one dimensional analysis of the finned tube
can be completed by determining wall and fin resistances. While a numerical analysis can be
computed using the given data (M-file 1) Of course in order to correctly calculate heat transfer
some assumptions must be made including: Steady-State, Constant Properties, no present
Internal Convection, Finite lengths, and no Temperature variation in the angular direction.
With the given data and assumptions in mind, the wall resistances along with the given
values, we obtain the temperatures of the inside wall, outside wall, and fin tip temperature as
shown in Fig. 5a & 5b. Throughout the thickness of the wall, including one fin, a temperature
drop of about 5K is present due to the 5kW of heat transfer inside the nacelle as well as the
convection currents of the exterior. Results are shown in Fig. 5a & 5b.
However, using the thirteen derived nodal equations shown on Page 4 inserted into an
iterating M-file, greater results are obtained. Titled M-file 1 in the appendix, it is shown how this
given data can be used numerically to obtain higher accuracy. The precision of both methods
very close with a total temperature drop of 9K in the two dimensional analysis (Fig. 3) compared
to 5K in our one dimensional calculations (Fig. 5a & 5b), but the accuracy of our findings can be
increased by using a nodal analysis of one fin shown in Fig. 2. This fin is analyzed but placing a
symmetry line in the center and analyzing one half at a time. The code iterates about six hundred
and fifty thousand times before converging due to a stopping criteria. The stopping criteria in Mfile 1 is a nominal residue value of 0.01K. This means that the code will stop running once the
total change in temperature over about seventeen hundred nodes is less than one-tenth of a
degree Kelvin per iteration. The stopping criteria is shown graphically in Fig. 4 where the total
residual decreases in way such that the residual values approach zero. In others words, when
these residual values come very close to zero; that means that the temperature is no longer
varying in the nacelle wall, and we can assume that the temperature values in the wall will stay
constant from this point on. See Fig. 3 for the temperature values of each node upon
convergence.
From the obtained results, one can begin to appreciate the finite element analysis. The
results are easy to obtain if the correct algebra is applied at the time of nodal derivation, and
instead of solving out a PDE with boundary conditions such as the ones shown in Fig. 6; a code
can be programed to work for you.

Heat Transfer Project

Discussion
In order to provide a thorough understanding of numerical analysis, below is a brief
explanation of how this process takes place. First, refer to Fig. 1, the analysis is based on a heat
flux on one side, and a convection current on the opposing side of the nacelle wall. Take for
instance a point in the center; this point can be enclosed by a control volume that has four
bountries, two in which conduction heat enters, and two sides where the conduction heat leaves
and heads towards the fins. Refer to drawing on Page 4-9. If an energy balance equation is
formed at this point; that equation can apply to each and every other inside point that we pick.
Repeating this control volume process and assessing the boundary conditions at each of the other
twelve differing nodes; a set of nodal equations can be formed (see Page 4b to view the
differing nodal equations that are created due to different boundaries and conditions).
An understanding of the task at hand and proper assumptions such as: Steady-State,
Constant Properties, no present Internal Convection, Finite lengths, and no Temperature
variation in the angular direction, must be made before the analysis can take place. Once
understood, these can be taken into consideration and the nodal equations can be derived. Once
each boundary and node has been assessed, an M-file can be created which will iterate this
system of equations to a nominal amount of times before the final result is reached. Since
internal heat is being dispersed via fins and a convection current; a stopping criteria must be set.
The nacelle is assumed to be at ambient temperature at iteration zero, and will increase in
temperature until the heat going is equal to heat going out. At this point, data is retrieved,
graphed, and put into a matrix which displays the heat distribution (Fig. 3). Under steady state
conditions, we can be sure that this is the maximum temperature that the nacelle wall will reach.

Heat Transfer Project

Nodal Equations
Shown on the following pages are thirteen nodal equations which describe the
geometry and boundary conditions of each side of the fin as shown:
For a sample derivation, see the following page:
Interior node: Page 4-9
Heat flux node: Page 4-2
Convection boundary node: Page 4-7

Heat Transfer Project

4a

Heat Transfer Project

FIGURE 3 (ATTACHED)

Heat Transfer Project

Heat Transfer Project

FIGURE 5a (ATTACHED)

Heat Transfer Project

FIGURE 5b (ATTACHED)

Heat Transfer Project

FIGURE 6 (ATTACHED)

Heat Transfer Project

10

M-file 1:
1. clc, clear, clear all
2. %TSC PROJECT
3.
4. %VARIABLES
5. Residual=1;
%Set nominal value for check below
6. qdp = 5000;
%(W)
7. ro = 8238;
%(kg/m^3)
8. k = 14.3;
%(W/m*K)
9. cp = 0.486;
%(kJ/kg*K)
10.
ho = 40;
%(W/m^2*K)
11.
T_ambient= 300;
%Ambient Temp (degree K)
12.
R1= 0.25;
13.
R2= 0.255;
%Radius(meters)
14.
R3= .285;
15.
X=.035;
X1=.005;
%X-DIMENSIONS(meters)
16.
Z=.01;
Z1=.0025;
%Y-DIMENSIONS(meters)
17.
delr = .0005;
18.
delz = .00025;
19.
r = 0 : delr : X;
20.
z = 0 : delz : Z;
21.
M=length(r);
22.
N=length(z);
23.
for i =1:M;
24.
if r(i)== X1
%finds M1
25.
M1=i;
26.
end
27.
end
28.
for i = 1:N;
29.
if z(i)== Z1
%finds N1
30.
N1=i;
31.
end
32.
end
33.
34.
r = R1 + r;
35.
36.
%Setting all points to ambeint temperature
37.
for i = 1 : M;
38.
for j = 1 : N1;
39.
T(j,i) = T_ambient;
40.
end
41.
end
42.
for i = 1 : M1;
43.
for j = N1+1 : N;
44.
T(j,i) = T_ambient;

Heat Transfer Project

11

45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.

end
end

%NODAL EQUATIONS
for number_iterations = 1:1000000;
A=T;
if Residual>=.01;

81.
82.
83.
84.
85.
86.
87.
88.
89.

Heat Transfer Project

%C1=
%C2=
%C3=
%C4=
%C5=
%C6=

%number of iterations
%Stopping criteria loop

i+
C1*T(i+1,j)
iC2*T(i-1,j)
j+
C3*T(i,j+1)
jC4*T(1,j-1)
T(i,j)
additional term such as q" or convection term

%Node 1
i = 1; j = 1;
C1 = (r(j)+delr/4)*(delr/(2*delz));
C2 = 0;
C3 = (r(j)+delz/2)*(delz/(2*delr));
C4 = 0;
C5 = C1 + C2 + C3 + C4;
C6 = (qdp/k)*r(j)*(delz/2);
T(i,j) = (1/C5)*(C1*T(i+1,j) +C3*T(i,j+1) + C6);
%Node 2
j = 1;
for i = 2: N-1;
C1 = (r(j)+delr/4)*(delr/(2*delz));
C2 = (r(j)+delr/4)*(delr/(2*delz));
C3 = (r(j)+delr/2)*(delz/delr);
C4 = 0;
C5 = C1 + C2 + C3 + C4;
C6 = qdp*r(j)*delz/k;
T(i,j) = (1/C5)*(C1*T(i+1,j) + C2*T(i-1,j) +
C3*T(i,j+1) + C6);
end
%Node 3
j = 1; i = N;
C1 = 0;
C2 = (r(j)+delr/4)*(delr/(2*delz));
C3 = (r(j)+delr/2)*(delz/(2*delr));
C4 = 0;
C5 = C1 + C2 + C3 + C4;

12

90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.

Heat Transfer Project

C6 = (qdp/k)*r(j)*(delz/4);
T(i,j) = (1/C5)*(C2*T(i-1,j) + C3*T(i,j+1) + C6);
%Node 4
i = N;
for j = 2 : (M1-1);
C1 = 0;
C2 = r(j)*delr/delz;
C3 = (r(j)+delr/2)*(delz/(2*delr));
C4 = (r(j)-delr/2)*(delz/(2*delr));
C5 = C1 + C2 + C3 + C4;
C6 = 0;
T(i,j) = (1/C5)*(C2*T(i-1,j) + C3*T(i,j+1) +
C4*T(1,j-1));
end
%Node 5
j = M1; i = N;
C1 = 0;
C2 = (r(j)-delr/4)*(delr/(2*delz));
C3 = 0;
C4 = (r(j)-delr/2)*(delz/(2*delr));
C6 = T_ambient*((ho/k)*r(i)*(delz/2));
C5 = C1 + C2 + C3 + C4 + C6/T_ambient;
T(i,j) = (1/C5)*(C2*T(i-1,j) + C4*T(1,j-1) + C6);
%Node 6
j = M1;
for i = N1+1: N-1;
C1 = (r(j)-delr/2)*(delr/(2*delz));
C2 = (r(j)-delr/4)*(delr/(2*delz));
C3 = 0;
C4 = (r(j)-delr/2)*(delz/delr);
C6 = T_ambient*((ho/k)*r(j)*delz);
C5 = C1 + C2 + C3 + C4 + C6/T_ambient;
T(i,j)= (1/C5)*(C1*T(i+1,j) + C2*T(i-1,j) +
C4*T(1,j-1) + C6);
end
%Node 7
j = M1; i = N1;
C1 = r(j)*delr/(2*delz);
C2 = r(j)*delr/delz;
C3 = (r(j)+delr/2)*(delz/(2*delr));
C4 = (r(j)-delr/2)*(delz/delr);

13

133.
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.
168.
169.
170.
171.
172.
173.
174.
Heat Transfer Project

C6 = T_ambient*((ho*r(j)*delz/(2*k))
+((r(j)+delr/2)*(ho*delr/(2*k))));
C5 = C1 + C2 + C3 + C4 + C6/T_ambient;
T(i,j) = (1/C5)*(C1*T(i+1,j) + C2*T(i-1,j) +
C3*T(i,j+1) + C4*T(1,j-1) +C6);
%Node 8
i=1;
for j = 2: M-1;
C1 = r(j)*delr/delz;
C2 = 0;
C3 = (r(j)+delr/2)*(delz/(2*delr));
C4 = (r(j)-delr/2)*(delz/(2*delr));
C5 = C1 + C2 + C3 + C4;
C6 = 0;
T(i,j) = (1/C5)*(C1*T(i+1,j) + C3*T(i,j+1) +
C4*T(1,j-1));
end
%Node 9
for j = 2 : M-1;
for i = 2 : N1-1;
C1 = r(j)*delr/delz;
C2 = r(j)*delr/delz;
C3 = (r(j)+delr/2)*(delz/delr);
C4 = (r(j)-delr/2)*(delz/delr);
C5 = C1 + C2 + C3 + C4;
C6 = 0;
T(i,j) = (1/C5)*(C1*T(i+1,j) + C2*T(i-1,j) +
C3*T(i,j+1) + C4*T(1,j-1));
end
end
for j = 2 : M1-1;
for i = N1 : N-1;
C1 = r(j)*delr/delz;
C2 = r(j)*delr/delz;
C3 = (r(j)+delr/2)*(delz/delr);
C4 = (r(j)-delr/2)*(delz/delr);
C5 = C1 + C2 + C3 + C4;
C6 = 0;
T(i,j) = (1/C5)*(C1*T(i+1,j) + C2*T(i-1,j) +
C3*T(i,j+1) + C4*T(1,j-1));
end
end
%Node 10
j = M; i = 1;

14

175.
176.
177.
178.
179.
180.
181.
182.
183.
184.
185.
186.
187.
188.
189.
190.
191.
192.
193.
194.
195.
196.
197.
198.
199.
200.
201.
202.
203.
204.
205.
206.
207.
208.
209.
210.
211.
212.
213.
214.
215.
216.
217.

Heat Transfer Project

C1
C2
C3
C4
C6
C5

=
=
=
=
=
=

(r(j)-delr/4)*(delr/(2*delz));
0;
0;
(r(j)-delr/2)*(delz/(2*delr));
T_ambient*((ho/k)*r(j)*(delz/2));
C1 + C2 + C3 + C4 + C6/T_ambient;

T(i,j)= (1/C5)*(C1*T(i+1,j) + C4*T(1,j-1) + C6);

%Node 11
j = M;
for i = 2 : N1-1;
C1 = (r(j)-delr/4);
C2 = (r(j)-delr/4)*(delr/(2*delz));
C3 = 0;
C4 = (r(j)-delr/2)*(delz/delr);
C6 = T_ambient*((ho/k)*r(i)*delz);
C5 = C1 + C2 + C3 + C4 + C6/T_ambient;
T(i,j) = (1/C5)*(C1*T(i+1,j) + C2*T(i-1,j) +
C4*T(1,j-1) + C6);
end
%Node 12
j = M; i = N1;
C1 = 0;
C2 = (r(j)-delr/4)*(delr/(2*delz));
C3 = 0;
C4 = (r(j)-delr/2)*(delz/(2*delr));
C6 = T_ambient*(((ho/k)*r(j)*(delz/2)) +
((ho/k)*(r(j)-delr/4)*(delz/2)));
C5 = C1 + C2 + C3 + C4 + C6/T_ambient;
T(i,j)= (1/C5)*(C2*T(i-1,j) + C4*T(1,j-1) + C6);
%Node 13
i = N1;
for j = M1+1 : M-1;
C1 = 0;
C2 = (r(j)*delr/delz);
C3 = (r(j)+delr/2)*(delz/(2*delr));
C4 = (r(j)-delr/2)*(delz/(2*delr));
C6 = T_ambient*((ho/k)*r(j)*delr);
C5 = C1 + C2 + C3 + C4 + C6/T_ambient;
T(i,j)= (1/C5)*(C2*T(i-1,j) + C3*T(i,j+1) +
C4*T(1,j-1) + C6);
end

15

218.
219.
220.
221.
222.
223.
224.
225.
226.
227.
228.
229.
230.
231.
232.

%Residual
A = T - A;
Residual = sum(sum(A));
plot(number_iterations,Residual,'x')
hold on
end
end
xlabel('number of iteratoins'),
ylabel('Total Residual'),
TT=flipud(T);
TTT=[TT;T];
disp('Node by Node Temperature Distribution at Convergence:')
disp(num2str(TTT))

Heat Transfer Project

16

MATLAB Nomenclature:
Residual - Sum of temperature changes from (i) iteration to
(i+1) iteration. [K]
qdp interior energy to be dissipated [W]
ro tube density [kg/m^3]
k heat transfer coefficient of tube [w/m*K]
cp specific heat [kJ/kg*K]
ho Exterior convective coefficient [W/m^2*K]
T_ambient Ambient temperature [K]
R1, R2, R3 Radii values for the: inside, outside, and fin
tip locations respectively [m]
X, Z, X1, Z1 Vertical and horizontal coordinates for the
cross section shown in figure 1. They account for half of the
area of figure 2 due to a symmetry line. [m]
delr, delz Values of the distance of each iteration step.
r, z An array of distances which split up the horizontal and
vertical directions into nodes.
M, N length of each array : r, and z listed above.
A Each nodal residue value which are summed into Residual.
T The matrix produced which resembles the shape of half of
figure 2 (after convergence).
TT The T matrix is mirrored about the horizontal.
TTT This matrix is shown in figure 3, it is T and TT added
together to resemble the fin portion shown in figure 2.

Heat Transfer Project

17

Virus free M-file (ATTACHED)

Heat Transfer Project

18