Part (a)
The viscosity µ = ρν = 3.8 × 10^{-5} slug/(ft · s). The spacing is 2h = 2.4 in = 0.2 ft, and D_h = 4h = 0.4 ft. The Reynolds number is
Re_{D_h} = \frac{VD_h}{\nu} = \frac{(6.0 ft/s)(0.4 ft)}{0.00002 ft^2/s} = 120,000
The flow is therefore turbulent. For reasonable accuracy, simply look on the Moody chart (Fig. 6.13) for smooth walls:
f ≈ 0.0173 h_f \approx f\frac{L}{D_h} \frac{V^2}{2g} = 0.0173\frac{100}{0.4}\frac{(6.0)^2}{2(32.2)} \approx 2.42 ft
Since there is no change in elevation,
\Delta p = \rho gh_f = 1.9(32.2)(2.42) = 148 lbf/ft^2
This is the head loss and pressure drop per 100 ft of channel. For more accuracy, take D_{eff} = \frac{64}{96} D_h from laminar theory; then
Re_{eff} = \frac{64}{96}(120,000) = 80,000
and from the Moody chart read f ≈ 0.0189 for smooth walls. Thus a better estimate is
h_f = 0.0189\frac{100}{0.4}\frac{(6.0)^2}{2(32.2)} = 2.64 ft
and \Delta p = 1.9(32.2)(2.64) = 161 lbf/ft^2 Better
The more accurate formula predicts friction about 9 percent higher.
Part (b)
Compute µ = ρν = 0.0038 slug/(ft · s). The Reynolds number is 6.0(0.4)/0.002 = 1200; therefore the flow is laminar, since Re is less than 2300.
You could use the laminar flow friction factor, Eq. (6.64)
f_{lam} = \frac{h_f}{(L/D_h)(V^2/2g)} = \frac{96 \mu}{\rho V(4h)} = \frac{96}{Re_{D_h}} (6.64)
f_{lam} = \frac{96}{Re_{D_h}} = \frac{96}{1200} = 0.08
from which h_f = 0.08 \frac{100}{0.4}\frac{(6.0)^2}{2(32.2)} = 11.2 ft
and \Delta p = 1.9(32.2)(11.2) = 684 lbf/ft^2
Alternately you can finesse the Reynolds number and go directly to the appropriate laminar flow formula, Eq. (6.63):
V = \frac{Q}{A} = \frac{h^2}{3\mu} \frac{\Delta p}{L} = \frac{2}{3} u_{max} (6.63)
V = \frac{h^2}{3\mu} \frac{\Delta p}{L}
or \Delta p = \frac{3(6.0 ft/s) [0.0038 slug/(ft \cdot s)] (100 ft)}{(0.1 ft)^2} = 684 slugs/(ft \cdot s^2) = 684 lbf/ft^2
and h_f = \frac{\Delta p}{\rho g} = \frac{684}{1.9(32.2)} = 11.2 ft
This is one of those (perhaps unexpected) problems where the laminar friction is greater than the turbulent friction.