A slightly different explanation that may help...
When a beam bends, e.g. a wing spar, the top surface is in compression and the bottom surface is in tension. In between there is a theoretical "neutral surface", where there is no axial stress, only bending. Because the top surface and bottom surface are under axial stress, there is also strain, i.e. the top of the wing spar gets a bit shorter, the bottom gets a bit longer.
Now imagine this with respect to the "neutral surface", which is not under strain so it has not changed length. Any material between the neutral surface and the top or bottom surface will be under shear stress as well as axial stress, because of the variation of axial stress and strain throughout the thickness of the beam.
The shear force on an element of the beam above or below the neutral surface is parallel to the beam top and bottom surfaces (due to strain) and also normal to the surfaces. If the element is a square, the top and bottom shear forces are in oposite directions, as are the vertical forces either side of the element. Do the vector sum and the resultant shear force is at 45 degrees to the axial stress.
A good reference article is here:
Bending - Wikipedia, the free encyclopedia
Regards,
WF
