Problem Irodov 1.333
Viscous fluid of viscosity fills the annular gap between two long, coaxial cylinders of radii and , with . The inner cylinder () is held fixed while the outer one () is driven at a constant angular velocity , and the resulting flow of the fluid between them is laminar. The friction force per unit area on any cylindrical fluid surface of radius is given as , where is the fluid’s own local angular velocity. Using this, find (a) how the fluid’s angular velocity depends on for , and (b) the moment, per unit length of the cylinders, of the friction forces acting on the outer cylinder.
Top view down the shared axis. The inner cylinder (radius , ticked rim) is held; the outer (radius ) is driven at . The dashed circle is the variable shell at radius , carrying tangential velocity .
- Given
- ,
- Find
- (a) for ; (b) , the moment of friction on the outer cylinder per unit length
Solution
Step 1The torque crossing every shell is the same constant
The torque per unit length crossing a cylindrical shell of radius is the shear stress times the shell’s area per unit length () times the moment arm ( again). Setting this equal to a constant and substituting the given ,
Step 2Integrate the constant-torque condition
Dividing through by and integrating once gives up to two constants.
Step 3Apply the two boundary conditions
The inner cylinder is held, ; the fluid at the outer wall shares that cylinder’s own speed by no-slip, .
Subtracting the first equation from the second eliminates and isolates .
The first boundary equation then gives the remaining constant,
Step 4Assemble
Step 5The moment on the outer cylinder
Because the torque transmitted across every shell is the same constant of (1), the moment the fluid exerts on the outer cylinder, at , is that very same : no further integration is needed.
Check
As (the inner cylinder shrinks to the axis), the prefactor , so at every fixed : with no inner boundary to hold it back, the fluid simply co-rotates rigidly with the outer wall, and the shear everywhere, consistent with . In the opposite, thin-gap limit with gap , , exactly the flat-plate estimate (the shear acting over the area per unit length, and that force then carried on the moment arm ) for a narrow Couette gap, confirming the curved result reduces correctly to the flat one.
Remark
The whole problem turns on one fact: with no torque source hiding anywhere in the fluid’s interior, the viscous torque crosses every concentric shell unchanged in steady state, even though the stress carrying it thins out as . That single conserved quantity is what lets (1) set directly from the given , without ever writing down the fluid’s equation of motion; the same trick, applied to , is exactly what makes 1.332’s straight coaxial sliding problem solvable by inspection too. In both problems the quantity that crosses every shell unchanged is whatever the steady balance conserves, the axial force in that straight film and the torque here, and in both the stress itself still falls off with radius.
