Distribute Overturning Moment to bolts?

[StructureMan44]

Are there any rules of thumb for distributing overturning moment to a silo base-ring? When the moment is applied obviously half of the ring will be in compression and the anchor bolts in the other half of the ring will experience varying amounts of tension. The bolts furthest from the center of the tank will experience the greatest tension which will be the uplift case to design for. The question is, does any resource provide a formula or rule of thumb for distributing the tension force among the anchor bolts and determining how much tension the furthest bolt(s) experience (for a given moment and ring diameter)?

 

[JoshPlumSE]

Below is a link to a good resource on the subject....

http://www.towernx.com/downloads/Technical_Manual_MP_BasePL.pdf

It's a bit long, but it goes over a number of different scenarios.

 

[BadgerPE]

I use this paper, but also use API 650 because their seismic information is quite extensive. API and Monopole method produce similar results for most applications.

 

[JStephen]

The assumption normally used in the tank codes is to treat the anchor bolts as an equivalent ring of metal, and just treat it all like it's beam bending, Mc/I +/- P/A. There are some alternative approaches, but their assumptions seem to be no more valid than the assumptions made in this approach, only more complicated.

 

[StructureMan44]

JoshPlum and BadgerPE: thank you all very much. This resource is immensely helpful.

JStephen: You're right this is like a beam.

Upon re-inspection my situation is not a continuously ring but instead channels (with their webs flat against the foundation) at each anchor point. I'm assuming only the anchor bolts resist tension, not the channels. There for I'm using the (M*y)/I +/- P/N equation instead of Mc/I +/- P/A.

If given "envelop" or "not-to-exceed" loads from the tank manufacturer, what is an appropriate factor of safety for tension in these bolts?

 

[rb1957]

The simple way is to put the couple onto the bolts.

This is simple, but "wrong" as the compression is better modelled as a triangular distribution (from the tension bolts to a peak at the free edge).

And then of course there's preload to consider, clamping the joint.

another day in paradise, or is paradise one day closer ?

 

[JStephen]

The design standards for liquid tanks define how the bolt load is calculated. For silos in the US, there's not an equivalent standard, so it's worth asking how the loads were calculated.
If you need to calculate factored loads, ask for the wind/seismic/dead/pressure loads per bolt to be broken out. Or if they're using the beam-distribution assumption, just get the total moments and total dead/pressure loads and work off that.
The design standards for liquid tanks require that the embedment strength be adequate for the full yield strength of the bolt, and that can control some of the design if applicable.
For the liquid tank codes, seismic and wind are normally service-level loads, not strength-level loads, so check on that as well.

 

[StructureMan44]

Thank you rb1957: Does the triangular distribution invalidate the equation (M*y)/I +/- P/N used for the maximum force in the furthest bolt? Are cast-in anchors bolts typically preloaded?

Thank you JStephen: The maximum moment for wind and seismic is given, for conservativeness I will assume they are service if I don't get feedback from the manufacturer.

On a per bolt basis; which is a less costly, moving from F1554 36ksi to 55ksi? or moving from 0.75"Ø to a 1"Ø 36ksi bolt?

 

[rb1957]

depends on what "I" is based on ... I of the fastener pattern ? or of some equivalent "tube" section ?

and you civil guys use that concrete stuff, good in compression ... so you tend to make a section of the compression side and the tension fasteners.

The triangular distribution is an attempt to represent the "plane faces remain plane" assumption, that on the compression side the reaction is proportional to the distance from the pivot line.

another day in paradise, or is paradise one day closer ?