Question 1 of 7: Question 1 (Part A, Question A1): Drag and Power for a Flat Plate Towed Edgewise
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Notes on this paper
Paper format. National Exams, May 2014 — 07-Mec-B6
Advanced Fluid Mechanics. Three hours, open book, any non-communicating
calculator permitted. Part A holds four questions and the candidate answers any three of them
(42 per cent of the paper, so 14 marks each); Part B holds three questions and the
candidate answers any two (58 per cent, so 29 marks each). All seven questions
are worked here.
The paper supplies an aid sheet of compressible-flow, boundary-layer, Navier–Stokes and
potential-flow relations; the skin-friction coefficients used below are taken from that sheet
(not from the Blasius/White constants a textbook would give) so that the arithmetic
matches what a candidate had in front of them.
Reference texts.
F. M. White, Fluid Mechanics, 8th ed. — Ch. 5 (dimensional analysis and
similitude), Ch. 7 (external flow and plate drag), Ch. 8 (potential flow), Ch. 9 (compressible
flow, Fanno line, normal shocks).
F. M. White, Viscous Fluid Flow, 3rd ed. — Ch. 3 (exact solutions of the
Navier–Stokes equations, annular Couette flow), Ch. 6 (turbulent wall flow).
J. D. Anderson, Modern Compressible Flow, 3rd ed. — Ch. 3 (normal shock
waves), Ch. 5 (quasi-one-dimensional nozzle and diffuser flow).
P. K. Kundu, I. M. Cohen and D. R. Dowling, Fluid Mechanics, 6th ed. —
Ch. 6 (irrotational flow, method of images), Ch. 9 (laminar internal flow).
H. Schlichting and K. Gersten, Boundary-Layer Theory, 8th ed. — Ch. 21
(plate drag with a mixed laminar/turbulent boundary layer).
R. W. Fox, A. T. McDonald and J. W. Mitchell, Introduction to Fluid Mechanics,
10th ed. — Ch. 7 (similitude and model testing), Ch. 13 (compressible flow with
friction).
Question 1 (Part A, Question A1): Drag and Power for a Flat Plate Towed Edgewise (14 marks)
In arrangement A side a = 5 m is normal to the flow (it is the leading
edge, as drawn in Fig. A1), so the stream runs only \(L = b = 1\ \text{m}\) along the plate over a
width \(w = a = 5\ \text{m}\). In arrangement B side b = 1 m is the leading
edge, so \(L = a = 5\ \text{m}\) and \(w = b = 1\ \text{m}\). The plate is thin, so both faces
carry a boundary layer and the wetted area is \(2ab = 10\ \text{m}^2\) in either case.
Find. The towing force and the towing power for each arrangement in each fluid,
which arrangement costs less power, and by what percentage.
Figure 1.1 — The two towing arrangements. Only the streamwise
length changes; the planform and wetted areas are identical.
Approach. Integrate the aid-sheet local skin-friction law along each face,
splitting the integral at the transition point \(x_{tr}\) where \(Re_x = 5\times10^5\), to obtain a
mean coefficient \(C_F\); then \(D = 2\left(\tfrac{1}{2}\rho U_o^2 wL\,C_F\right)\) and
\(P = D\,U_o\).
Locate the transition point in each fluid. Transition sits where
\(Re_x = U_o x/\nu = Re_{cr}\), so
\[x_{tr} = \frac{\nu\,Re_{cr}}{U_o}.\]
In water \(x_{tr} = (10^{-6})(5\times10^{5})/15 = 0.0333\ \text{m}\); in air
\(x_{tr} = (15\times10^{-6})(5\times10^{5})/15 = 0.500\ \text{m}\). This one number decides the
whole question: in water transition happens within the first 3 cm of either plate, while
in air it takes half a metre — which is half of the 1 m plate in arrangement B.
Integrate the wall shear over one face. The drag on one face of width \(w\) is
\(D_1=\int_0^L \tau_w(x)\,w\,dx\) with \(\tau_w = \tfrac{1}{2}\rho U_o^2 C_{fx}\). Defining the mean
coefficient \(C_F = \frac{1}{L}\int_0^L C_{fx}\,dx\) and substituting the two aid-sheet laws,
\[\begin{aligned}\int_0^{x_{tr}}\!\frac{0.67}{Re_x^{1/2}}\,dx&=\frac{1.34\,x_{tr}}{Re_{cr}^{1/2}},\\\int_{x_{tr}}^{L}\!\frac{0.0266}{Re_x^{1/7}}\,dx&=\frac{7}{6}(0.0266)\!\left[\frac{L}{Re_L^{1/7}}-\frac{x_{tr}}{Re_{cr}^{1/7}}\right].\end{aligned}\]
Both integrals are elementary because \(Re_x\) is linear in \(x\); the \(7/6\) factor is the
exponent bookkeeping of \(x^{-1/7}\).
Evaluate the mean coefficient for each case. Adding the two contributions and
dividing by \(L\) gives
\[\boxed{C_F = \frac{1}{L}\left[\frac{1.34\,x_{tr}}{\sqrt{Re_{cr}}}+\frac{7}{6}(0.0266)\left(\frac{L}{Re_L^{1/7}}-\frac{x_{tr}}{Re_{cr}^{1/7}}\right)\right]}\]
which yields the four values tabulated below. Note that arrangement B in water has
\(Re_L = 7.5\times10^{7}\) — essentially an all-turbulent plate — whereas arrangement A
in air is 50 per cent laminar.
Convert to force and power. With both faces wetted,
\[D = 2\cdot\tfrac{1}{2}\rho U_o^2\,(wL)\,C_F=\rho U_o^{2}\,(wL)\,C_F,\qquad P = D\,U_o .\]
Because \(wL = 5\ \text{m}^2\) and \(U_o = 15\ \text{m/s}\) are the same in every case, the force
ratio between arrangements is the \(C_F\) ratio.
Fluid
Arrangement
\(L\) (m)
\(Re_L\)
\(x_{tr}/L\)
\(C_F\)
\(D\) (N)
\(P\)
(a) Water
B (b leading)
5
7.50×107
0.0067
0.002308
2596.5
38.95 kW
A (a leading)
1
1.50×107
0.0333
0.002833
3187.3
47.81 kW
(b) Air
B (b leading)
5
5.00×106
0.100
0.003140
4.239
63.58 W
A (a leading)
1
1.00×106
0.500
0.002879
3.887
58.30 W
The two fluids give opposite answers, and that is the point of the question. In water
the laminar run is negligible in both arrangements, so the only thing that matters is that a longer
plate has a thicker, slower-growing turbulent layer and therefore a smaller mean coefficient:
arrangement B wins. In air the laminar run is a full half of the short plate, and the laminar
contribution is so much cheaper than the turbulent one that arrangement A wins despite its larger
turbulent-layer growth rate. The plot below shows \(C_F(L)\) for both fluids with the two operating
points marked.
Figure 1.2 — Mean skin-friction coefficient versus streamwise
plate length at \(U_o=15\) m/s. The water curve falls monotonically through both operating points;
the air curve bottoms out at the all-laminar \(L=0.5\) m (where transition first appears), climbs to a
peak near \(L\approx2.7\) m and falls only slowly beyond it, so the 1 m plate sits below the 5 m plate.
Percentage power reduction, water. Arrangement B is the efficient one:
\[\begin{aligned}\text{saving}&=\frac{P_A-P_B}{P_A}=\frac{47.81-38.95}{47.81}=0.1854\\&\Rightarrow\ \boxed{\text{arrangement B uses }18.5\text{ per cent less power}}\end{aligned}\]
Percentage power reduction, air. Here arrangement A is the efficient one:
\[\begin{aligned}\text{saving}&=\frac{P_B-P_A}{P_B}=\frac{63.58-58.30}{63.58}=0.0830\\&\Rightarrow\ \boxed{\text{arrangement A uses }8.3\text{ per cent less power}}\end{aligned}\]
Result
Water
Air
Transition point \(x_{tr}\)
0.0333 m
0.500 m
Drag, arrangement A (L = 1 m)
3187.3 N
3.887 N
Power, arrangement A (L = 1 m)
47.81 kW
58.30 W
Drag, arrangement B (L = 5 m)
2596.5 N
4.239 N
Power, arrangement B (L = 5 m)
38.95 kW
63.58 W
More efficient arrangement
B (long side streamwise)
A (short side streamwise)
Power reduction achieved
18.5 per cent
8.3 per cent
Check: modelling assumptions. The plate is treated as a
zero-thickness flat plate at zero incidence, so only skin friction acts — no form drag, no
edge (finite-width) effects, and no wave making at a free surface. Transition is taken as abrupt at
\(Re_x = 5\times10^{5}\) rather than as a transition region, and the turbulent correlation
is applied from \(x_{tr}\) as if the turbulent layer had started at the leading edge. The
coefficients 0.67 and 0.0266 are the paper's own aid-sheet values; the more familiar textbook
constants (0.664 laminar, 0.027 turbulent) would change the forces by well under one per
cent.