Structural Analysis of Box Culvert

[ngneer]

I am reviewing the structural design of precast box culvert manufacturer and need some clarification. The culvert will be produced as "clam shell" sections (upper and lower halves) with shear key to prevent lateral movement. When analyzing as a frame using moment distribution, I am unsure whether to model the joint at the shear key as a pin, fixed, or roller support. The precast engineer assumes the culvert wall acts as a cantilever beam for fixed end moments with the horizontal thrust acting at the shear key. Also, could you advise as to sources of literature available for design of box culverts.

 

[stanfea]

In this case I would say that the curved "beam" should be modelled as a pin on one side and a roller on the other. This is a classic design case for that type of structure. If however the lower half of the clam was fully restrained by the earth or foundations etc an arguement could possibly be made for a pinned-pinned analysis. This would be the 'tied arch' analogy. This also assumes that the 'shear key' has the required shear capacity to resist the forces that result from a tied arch scenario. The safe anaylsis is the pinned-roller assumption. The fixed condition in my opinion could not be achieved with two part construction.

Good luck
Stan

 

[Ron]

For analytical purposes, a pin and roller configuration would be appropriate; however, there will be some rotational restraint along the joint axis, and partial rotation restraint at the shear key in the vertical plane, as well as lateral restraint from the key in two directions.

 

[yack]

I'd analyze it a couple of the ways you mentioned--to see how it could handle picking up the load--what types of reduncancies are available, etc.

 

[stanfea]

I agree with Ron that there will be some partial rotational restraint at the joint. However I would NEVER count on it in the design. I would consider this "gravy" on my design. The horizontal thrust at the shear key can only develop if the lower section is restrained from spreading (or is designed to restrain the upper section with increased wall thickness or reinforcing). To have a shear at the lower end of the upper half, you must have a support condition established by the lack of movement of the lower half. In other words a reaction point. If both sections are similar in design then I would still go Pin-Roller.

Stan

 

[Ron]

Stan...we agree for analytical purposes, just wanted to point out to ngneer that actual construction will produce the different conditions.

Ron