API PUBL 937-A:2005 pdf download

API PUBL 937-A:2005 pdf download.Study to Establish Relations for the Relative Strength of API 650 Cone Roof Roof-to-Shell and Shell-to-Bottom Joints.
3.4 Summary of Responses
The results presented above highlight the primary features of tank response:
I. Balanced uplift pressures are a function of product level. Empty tanks uplift at a lower pressure than full tanks, for this 30×32 foot tank the uplift pressures arc 0.295 psi empty and 0.804 full.
2. The effect of buckling in the roof (and floor) can be significant. \Vilhout buckling, the roof-to-shell joint is predicted to fail at l.04 psi, with buckling the predicted failure pressure is 0.724 psi. It is expected that the actual fuilure pressure lies between these two bounds.
3. Significant uplift can occur at the pressure required to fail the roof-to-shell joint, for this tank the uplift was 4.64 inches (no buckling) and 3.65 inches (with buckling) when empty and negligible when full. This uplift could cause attachments to fail. It could also lead to loads on the bottom that could cause failure.
4. The difference between the pressure to cause failure of the roof-to-shell joint and the shell-to-bottom joint is relatively small for empty tanks. In this case, the pressure to cause failure of the roof-to-shell joint when empty was 1.04 psi and 1.27 psi for failure of the shell-to-bottom joint (0.72 psi and 1.06 psi respectively with buckling). When full. the pressures were 1.04 psi for the roof-to-shell joint and 3.26 psi for the shell-to-bottom joint (0.72 psi and 3.24 psi respectively with buckling). Thus, the joint failure ratio (ratio of bottom joint failure to top joint failure) for empty small tanks can be low. For full tanks, the joint failure ratio is larger.
This behavior must be considered when developing new design criteria. Since uplift may not be prevented, the new criteria must accommodate uplift. This introduces several new failure modes: shell-to-bottom joint yield, weld failure of the shell-to-bottom joint. and failure of the bottom lap joints. These are addressed in the new suggested design criteria.
Bottom plates are welded on the top side only, with a continuous full-fillet weld on all seams. Details of the bottom welds dcpcnd on whether annular plates are used. Without annular plates, the minimum distance between the shell and any three-plate laps is 12 inches. If annular plates are used, the minimum distance between the shell and any lap-welded joint on the bottom is 24 inches. During uplift. some of the bottom will be lifted off the foundation. As a result, these welds may he subjected to bending and in-plane loads that would not occur if the bottom remained flat,
In Section 5 of this report we propose criteria to prevent failure of these welds.
4.5 Failure of Attachments due to Uplift
In the event of uplift, attachments will be subjected to increased loads. This could lead to failure
of the attachments or of the shell at the attachment location. This is discussed further in Section
5, although development of an appropriate criterion is left as future work.
4.6 Fracture
Failure due to fracture is not addressed in this report. It is assumed that by the selection of materials specified in API-650, sufficient toughness is provided to avoid initiation and propagation of fractures. It should be noted that there are different consequences of failure due to fracture at the roof-to-shell joint and at the shell-to-bottom joint. At the roof-to-shell joint, failure due to fracture at a pressure lower than that required for yielding would have the effect of relieving internal pressure more rapidly. This is conservative, as long as the fraewre is confined to the region of the top angle. However. failure by fracture at the shell-to-bottom joint could result in bottom thilure bcfire over-pressure relief was provided by the frangible roofjoint.