Let's try to explain this from another viewpoint.

For above-ground piping systems, the supports are placed 20-50 pipe diameters apart. This gives a formula for the natural frequency of a straight pipe between two supports of the form f = K/(2 * pi) * sqrt (E*I*g/(w*L^4)). As the span between supports gets longer, the natural frequency decreases. As the natural frequency decreases, the amplitude of the displacement envelope increases. As the amplitude of the displacement envelope increases, the stresses induced by the vibration increase.

Agreed?

OK, now for buried pipe the distance between supports is infinitely small. What, exactly, do you think will happen to the displacements and stresses?

As Richard notes above, there are obviously direct interactions between the soil and buried pipe. However, these are on a scale where local FEA is necessary - CAESAR II is not really valid for local effects.

So I would not recommend trying to solve this problem using CAESAR II. For the effects for which CAESAR II is valid, the stresses are by definition very small. For effects which may cause significant stresses, CAESAR II is not valid.