1 reply to this topic

[carib1]

I am following the procedure in Coulson and Richardson volume 6. My design operates at 19 bar and is 3.6m in diameter and 18 m in height.

In the example one longitudinal stress is negative and the other is positive but both my longitudinal stresses downwind and upwind are negative. I have checked and rechecked my calculations.

In designing the skirt the maximum tensile stress is negative. I don't even know if this is possible or what the implications of this are. I've checked the mathematics and it is sound so I really don't know what to do. Could someone please advise me?

The basic criterion is satisfied in that the maximum compressive stress is less than the critical buckling stress. I tried decreasing thickness and I still get 2 negative longitudinal stresses, when I increase the thickness to 340mm then I get one longitudinal stress (upwind) positive and the other stress (downwind) negative.

Is this a feasible solution?

Does it matter if both resultant longitudinal stresses are the same sign?

[kkala]

1. Assuming you follow Coulson Richardson (vol 6) / Example 13.3 (preliminary thickness for a distillation column), design conditions for your case could be about 100 oC and 21 barg = 2.1 N/mm2 ; carbon steel design stress 125 N/mm2 (Table 13.2) and thickness to withstand pressure e=2.1*3600/(2*125-2.1)=30.5 mm (+corrosion allowance). A thicker wall will be needed to withstand wind and weight loads too, indicated procedure in example 13.3.
2. At vessel bottom tangent line, following stresses are developed.
α. Weight stress (L, negative); L=longitudinal, compressive stress is conventionally considered negative.
β. Bending stress due to wind (L), positive (i.e. elongation) upwind - negative downwind, concerning the tower section.
γ. Pressure stresses (L and circumferential, the latter being 2 fold L), both positive.
δ. Resultant longitudinal stress is the sum of above (L), different upwind and downwind. Stress (β) has opposite sign upwind and downwind, but this is not necessarily so for the sum.
3. Probably both upwind and downwind stresses are negative in your case because of the heavy weight of the vessel. Anyway you can recheck signs of partial stresses before summing.
4. Concerning the skirt, there is no pressure stress (2γ), thus
α. Stress (L) of max absolute value is negative and downwind (both partial stresses are negative, that is compressive). This is normal.
β. Increasing skirt thickness would decrease absolute value of bending stress (13.9.2, skirt supports), seeing that bending moment is same; so upwind bending stress (positive) would decrease and sum of stresses would remain negative. Probably recheck is needed for this case.
Apart from this, 340 mm thickness looks unrealistically high, steel plate so thick cannot be bevelled; but this was just for testing stress signs.
5. Hope it is helpful, comments are welcomed.

Edited by kkala, 18 January 2013 - 06:18 AM.