You can try to act as follows. First, let us leave your parameters of the equation to stay symbols and solve the first, third and fourth equations:
eq1 = Reyn == (\[Rho] v Diam)/\[Mu]
eq2 = 1/Sqrt[f] == -2.0 Log10[(\[Epsilon]/Diam)/3.7 + 2.51/(Reyn Sqrt[f])]
eq3 = 17 == f Le/Diam v^2/(2 g)
eq4 = v == vdot/((Diam/2)^2 \[Pi])
sl = Solve[{eq1, eq3, eq4}, {Reyn, v, Diam}]
I guess you only need the solution in real values. The third of the solutions seems to be like this:
sl[[3]]
(* {Reyn -> (2 2^(2/5) 17^(1/5) g^(1/5) vdot^(3/5) ρ)/(
f^(1/5) Le^(1/5) π^(3/5) μ),
v -> (2^(4/5) 17^(2/5) g^(2/5) vdot^(1/5))/(
f^(2/5) Le^(2/5) π^(1/5)),
Diam -> (2^(3/5) f^(1/5) Le^(1/5) vdot^(2/5))/(
17^(1/5) g^(1/5) π^(2/5))} *)
Let us substitute it into eq2, and after that let us substitute your numbers instead of the symbols. For the latter aim let us first prepare a rule:
rule = {vdot -> 20, Temp -> 10, ρ ->QuantityMagnitude[
ThermodynamicData["Water","Density", {"Temperature" ->Quantity[Temp, "DegreesCelsius"]}]], μ ->QuantityMagnitude[ThermodynamicData["Water",
"Viscosity", {"Temperature" ->Quantity[Temp, "DegreesCelsius"]}]],ϵ -> 2*10^-6,g -> 9.81, Le -> 200}
Now let us substitute:
eq2New = eq2 /. sl[[3]] /. rule
(* 1/Sqrt[f] == -0.868589 Log[
4.24351*10^-7/f^(3/10) + 1.64012*10^-7/f^(1/5)] *)
Now it becomes clear, why Mma does not return you any solution. Such transcendental equations are typically not solved with the function Solve
. It can, however, be solved numerically.
Let us first draw the parts of the equation:
[![Plot\[{-0.869 Log\[4.2435*^-7/f^(3/10) + 1.64`*^-7/f^(1/5)\], 1/Sqrt\[
f\]}, {f, 0, 0.02}\]][1]][1]
we see that there is the cross-section somewhere below 0.01. Now we can safely solve it:
FindRoot[eq2New, {f, 0.05}]
(* {f -> 0.00783754} *)
Done. Have fun!