Apart from that all my hand calculation maximum displacement values are in e-06 figures while Abaqus outputs are in e-05 figures. I used 2.077e+11 as elasticity because my models are in meters. I used 0.3 poison ratio and load of 5.5E+6 (5.5MN). 

what is more disturbing is that my circular hollow pipe model of outer diameter 120 mm and 10 mm thickness produces output data abut 3 times larger than my hand calculation.

rectangular hollow section of wall thickness 10 mm with outer height 120 mm and width 60 mm. curved fillet corners have external radius 15 mm and internal radius of 10 mm. this model produces output data about four times my hand calculation. 

I use the displacement formula U = (F * L) / (E * A).

Is this normal or am I missing something?