A distributed beam is considered with clamped ends. One segment of the beam is made of viscoelastic (Kelvin-Voigt) material, the rest part is elastic. This is called locally distributed Kelvin-Voigt (or interior) damping. It is shown that by Kirchhoff hypothesis, neglecting the rotary inertia, the beam is exponentially stable with respect to transversal perturbations and is not stable exponentially with respect to longitudinal perturbations. It is shown that the \(C_{0}\)-semigroup associated with the initial-boundary problem of the transversal vibrations is not analytic. The last result does not depend on the boundary conditions or the type of damping.