Constraints on fourth order generalized f(R) gravity

A fourth order generalized f(R) gravity theory (FOG) is considered with the Einstein-Hilbert action $R+aR^{2}+bR_{\mu \nu}R^{\mu \nu},$ $R_{\mu \nu}$ being Ricci\'{}s tensor and R the curvature scalar. The field equations are applied to spherical bodies where Newtonian gravity is a good approximation. The result is that for $0\leq a\sim -b<<R^{2}$, $R$ being the body radius, the gravitational field outside the body contains two Yukawas, one attractive and the other one repulsive, in addition to the Newtonian term. For $a\sim -b>>R^{2}$ the gravitational field near the body is zero but at distances greater than $\sqrt{a}\sim \sqrt{-b}$ the field is practically Newtonian. From the comparison with laboratory experiments I conclude that $\sqrt{a}$ and $\sqrt{-b}$ should be smaller than a few millimeters, which excludes any relevant effect of FOG on stars, galaxies or cosmology.