Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations TugboatEng on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Buckling of Anchor bolts 1
- Thread starter magmath76
- Start date
OK, then it is structural, not mechanical, or in any event not just mechanical. The column should have been grouted before loading, but that didn't happen, so the bolts had to take the load, or some of it probably found an alternate load path. The question remains as to whether all those bolts are indeed necessary in the completed structure. I suspect not, but that will require reanalysis. If the bolts are not doing much, then jacking the column to the proper level and grouting may be all that is required, provided the design can live with the loss of bending capacity in the base plate due to the slotted holes.
Tanks like this are commonly built with a scalloped bottom, that is intentional, and allows relatively thin bottom plate spanning over a much wider area.
Magmath, once upon a time, I got involved on some projects similar to this. We "inherited" a design method, and that method consisted of designing the radial beams for all the vertical loads involved and basically neglecting all other loads. When I began looking at it in more detail, it presented some real problems. Specifically, the scallop plates develop membrane tension, and much of that tension results in a large vertical load, for which the beams are designed. But there is also a net inward load on each beam from that tension, and if you start trying to find the leg loads, that inward component also produces a large moment that affects distribution of the loads on the legs, and shifts loading more to the inner legs. That kind of thing may or may not be part of the problem here, just be aware that the load at the center may have been miscalculated due to unwarranted simplifications in the design. An additional issue is that some amount of load from the scallop plates must be transferred to the shell, but you don't know how much. An additional issue is that if the scallop plates are detailed as cylindrical sections, you get sort of an indeterminate shape out of them in the assembled structure.
Where I ran into this, was we had circular ringwalls under those legs, and were getting differential settlement of one ring relative to the others- with the effect that when the tank was unloaded, it would pop anchor bolts out of the concrete on the inner legs, and base plates would be suspended a couple of inches above the concrete.
Magmath, once upon a time, I got involved on some projects similar to this. We "inherited" a design method, and that method consisted of designing the radial beams for all the vertical loads involved and basically neglecting all other loads. When I began looking at it in more detail, it presented some real problems. Specifically, the scallop plates develop membrane tension, and much of that tension results in a large vertical load, for which the beams are designed. But there is also a net inward load on each beam from that tension, and if you start trying to find the leg loads, that inward component also produces a large moment that affects distribution of the loads on the legs, and shifts loading more to the inner legs. That kind of thing may or may not be part of the problem here, just be aware that the load at the center may have been miscalculated due to unwarranted simplifications in the design. An additional issue is that some amount of load from the scallop plates must be transferred to the shell, but you don't know how much. An additional issue is that if the scallop plates are detailed as cylindrical sections, you get sort of an indeterminate shape out of them in the assembled structure.
Where I ran into this, was we had circular ringwalls under those legs, and were getting differential settlement of one ring relative to the others- with the effect that when the tank was unloaded, it would pop anchor bolts out of the concrete on the inner legs, and base plates would be suspended a couple of inches above the concrete.
if alignment of the bolts & base pl is anticipated beforehand, then, that might explain the extra length of exposed bolts to enable field alignment of the bolts....as JS pointed out this is indeterminate structure and sensitive to the relative deflection of each radial steel frame and it's col/fdn which could put unintended loads on the center col..also, if it is subject to seismic loads, I can not see how all this load is reacted by the perimiter bracing and that the centre col could see a proportion of this seismic load based on it's relative stiffeness....I would rather see a pinned base at the center col...back to the problem at hand, I can not definitively say what caused the existing problem, but, would hazard a guess that the length of the exposed bolts was a major factor in the result...
vinegaroon
Industrial
- Mar 17, 2014
- 1
I see there is a visible rotation which indicates twisting or torsion. Is it possible the outside support envelope had swollen up with ground water causing the center to raise? Is it possible it could retract under dry loading conditions?
charliealphabravo
Structural
When I see this many anchor rods in a base plate it makes me wonder if the connection was designed by an ME who was familiar with bolting steel-to-steel flanges typical in process design but not so familiar with base plate-to-concrete foundation embeds. This part of the design where the steel process interfaces the concrete foundation should involve a structural engineer.
I am also wondering why the center column should be designed to attract so much lateral load. It is unlikely that the connection between the radial beams and the center column can generate the kind of moment for the lateral resistance that is anticipated by the shear key and anchor rods. In an indeterminant structure like this I guess it is not surprising that center column bolts appear to have failed due to some unexpected lateral translation and racking under compression.
Ideally the lateral load should be taken care of with diagonal braces in the column array. The center column would then be modeled for bearing only with perhaps a horizontal slider connection to avoid attracting lateral load. A pinned connection could work if the center column is tied into the diagonal brace system.
Just a knee-jerk response based on two photographs
Sorry to be a debbie downer. Maybe the design is fine and they just should have grouted the anchor rods before loading.
I am also wondering why the center column should be designed to attract so much lateral load. It is unlikely that the connection between the radial beams and the center column can generate the kind of moment for the lateral resistance that is anticipated by the shear key and anchor rods. In an indeterminant structure like this I guess it is not surprising that center column bolts appear to have failed due to some unexpected lateral translation and racking under compression.
Ideally the lateral load should be taken care of with diagonal braces in the column array. The center column would then be modeled for bearing only with perhaps a horizontal slider connection to avoid attracting lateral load. A pinned connection could work if the center column is tied into the diagonal brace system.
Just a knee-jerk response based on two photographs
Sorry to be a debbie downer. Maybe the design is fine and they just should have grouted the anchor rods before loading.- Status
Similar threads
- Locked
- Question
- Locked
- Question
