1) $\tau = \rho l^2 (\frac{du}{dy})^2$ from this equation, the velocity distribution can be obtained if the relation between L, the mixing length and y is known.
2) Pranditl assumed that mixing length, L is linear function of the distance y from the pipe wall. i.e., L = Ky where k = 0.4 known as karma n constant.
3) Now, $\tau = \tau = \rho l^2 (\frac{du}{dy})^2$
$\tau = \rho \times (ky)^2 (\frac{du}{dy})^2$
i.e. on rearranging above equation in $\frac{du}{dy}$
we get, $\frac{du}{dy} = \sqrt \frac{\tau}{\rho k^2 y^2} = \frac{1}{ky} \sqrt \frac{\tau}{\rho }$ ---- (1)
For small values of y that is very close to the boundary of the pipe, prandtl assumed shear stress $\tau$ to be constant and approximately equal ?o
$\therefore$ putting $\tau = \tau_o \ in \ (1)$
$\therefore \frac{du}{dy} = \frac{1}{ky} \sqrt \frac{\tau \ o}{\rho }$ ---- (2)
now, $\sqrt \frac{\tau_o}{\rho }$ has dimension $\sqrt \frac{ml^{-1} T^{-2}}{ML^{-3}} = \sqrt \frac{L^2}{T^2} = \frac{L}{T}$ which is velocity, known as shear velocity (u).
Thus $\sqrt \frac{?}{\rho } = u , \ then \ \frac{du}{dy} = \frac{1}{ky} \ u$
on integrating above equation, we get,
$u = \frac{u}{k} log_e \ y + c$ ------ (3) (where c = constant of integration)
Above equation (3) shows that in turbulent flow, velocity $\alpha$ $log_e \ Y $ i.e. velocity distribution in turbulent flow is logarithmic in nature.
To find value of C, putting y = R and u = $u_{max}$ (boundary condition) in (3)
Hence,
$u_{max} = \frac{u*}{k} log_e$ R + C ( R is radius of pipe)
$\therefore$ Putting (4) in (3) we get,
$u = \frac{u*}{k} \ log_e \ y + u_{max} - \frac{u*}{k} \ log_e$ R
$= u_{max} + \frac{u*}{k} \times log_e \ y - log_e$ R
$= u_{max} + \frac{u*}{0.4} log_e \ (Y/R)$ ( $because$ k = 0.4 )
$\therefore u = u max + 2.5 \ u * log_e$ (Y/R)
Above equation is called as 'prandtl's universal velocity distribution equation for turbulent flow in pipes. This equation applicable to smooth and rough pipe boundaries.
Velocity defect:
$u = u_{max} + 2.5 \ u* log_e$ (Y/R)
i.e. $u_{max} - u* + 2.5 \ u* log_e$ (R/Y)
Dividing by u*
$\therefore \frac{u_{max} = u}{u*} = 2.5 log_e$ (R/Y)
$\frac{u_{max} - u}{u*} = 2.5 \times 2.3 log_{10} (\frac{R}{Y})$
$( \because log_e \ (R/y) = 2.3 log_{10} (R/Y) )$
$\frac{u_{max} - u}{u*} = 5.75 log_{10} (R/Y)$
The difference between the maximum velocity (u max) and local velocity (u) at any point i.e. (u max - u ) is velocity defect.