Design of Anchor Bolts for Tower

[TomBarsh]

Has anyone ever built a vessel with anchor bolts sized on the "shifted neutral axis" method described in the books by Megyesy, Jawad and Farr, Bednar, Brownell and Young, Troitsky, etc?

I'll expand: Because of the different modulus of elasticity of the concrete foundation and the vessel's metal base plate & skirt the neutral axis of the tension and compression forces acting on the anchor bolts and base ring will shift away from the geometic centerline. Using the method described in these references results with a lower calculated bolt tension than is found using the old standby method "P = -W/N+48*M/(N*BC)".

It would seem that the cost savings from bolt material, installation, and size/thickness of the concrete slab would be considerable if the anchor bolt diameter could be reduced by 1/4" or more.

This method is briefly discussed in "Wind Loads and Anchor Bolt Design for Petrochemical Facilities" by ASCE (3.5.1.1) where they say the procedure "may be worth undertaking" for some cases.

Has anyone actually done this? Is it successful?

 

[jte]

I used to use the shifted neutral axis approach occasionally, usually with some sucess in solving the problem. This approach was usually reserved for making an existing design work rather than new construction, though it was also used on occasion for new construction where the anchor bolts were unusually large. Pretty straightforward approach.

I don't often say that books are cheap, but Design of Welded Structures by Blodgett published by the JF Lincoln Arc Welding Foundation is a bargain at twice the price. In its section 3.3 it derives the shifted neutral axis approach and simplifies it for repetitive use. The rest of the book is also very well written and easy to understand. $22 at

https://ssl.lincolnelectric.com/foundation/item.asp?prodnum=DWS

jt

 

[PVGuy]

Does anyone know if any of the software providers cover this method?

 

[TomBarsh]

COMPRESS pressure vessel software by Codeware does. Meca Software offers an inexpensive base ring-only design program. I would be surprised if other vessel software vendors don't provide this analysis.

Tom Barsh
Codeware Technical Support

 

[PVGuy]

TomBarsh,

Thank you for the quick reply. The method sounds messy from this discussion and it is good to know that someone some programs have covered it. I will investigate the method further.

PVGuy

 

[JStephen]

Tom, it's my understanding that this method treats the anchor bolts and concrete as a composite section as would typically be done with reinforced concrete design. There are some problems that I see with this. One is that the distribution of forces in a beam is typically derived assuming the section is not near the end of a beam, but in this method, the section always is. The second problem is that regardless of the design procedure, the distribution of force right below the concrete surface and right above the concrete surface ought to be identical. And lastly, this distribution ought to be the same as the distribution of uplift forces in the tank shell. The standard procedure of using MC/I + P/a is obviously a gross approximation, but I'm not convinced that this alternate method is any better of an approximation.

 

[jte]

JStephen-

In essence, the shifted neutral axis method correlates the strain in the steel baseplate and the strain in the concrete through Young's Modulus ratios. With the standard neutral axis procedure, you ignore that when the overturning load is applied, the concrete on the compression side - well, compresses. With steel being stiffer, the baseplate starts to dig in on one side while the other begins to lift off. With the concrete area in compression being stiffer than the anchor bolt area in tension, it takes more anchor bolt area*arm to counter the moment caused by the area in compression which is bearing on the concrete. This shifts the NA towards the compression side thereby increasing the moment arm of the bolts and typically adding two or more bolts to the tensile side of the axis.

Easier to understand when you see the derivation by Blodgett. I still highly recommend spending the $22 to get Blodgett's book which I recommended above. Any other engineering text, large and several hundred pages, would easily be $150.

jt