Shear at the base of concrete walls

[dutchie]

I am looking for some ideas how to cut down the thickness of concrete walls I am designing for a rectangular concrete tank. The walls are about 5m high and are to be poured integral with the base slab, free at the top. Because of the height, the thickness of the wall to resist shear is quite large, based on CAN A23.3 requirement that Vc/2>Vf to avoid placing shear reinforcement.

I am not averse to placing shear reinforcement, but have heard various thoughts on it's effectiveness. I have not been able to find any documents that relate directly to this situation, and was wondering if anyone else out there has any thoughts.

Thanks

 

[StephenEd]

------"the member is part of a redundant structural system that allows substantial redistribution of load and will therefor display adequate ductility"------

If the entire wall sees the maximum load at the same time (and in a constant fashion) is it really a redundant system? If the loads are from fluid contents rather than earth pressure then I don't think you have a redundant system. There are no lesser-loaded areas which can assist those areas seeing some overload. Especially with such a great aspect ratio.

 

[dutchie]

Subject to fluid load, and assuming a fixed base, the max moment occurs at the base (obviously). We have reinforced the walls for that max moment, therefor they have that capacity up the full height. In reality, the base is not fully restrained against rotation, and the actual condition will be somewhere between fixed and hinged. This moves the maximum moment up the wall closer to the middle. In either case, there is room for redistribution of the moment, either further up the wall or further down the wall. for that reason, I believe that the quoted statement does apply to this case. I would relate it to a slab subjected to snow load. this load can be present in it's full capacity for some time, yet the slab assumptions still apply.

that's my two cents, any other thoughts are welcome...

dutchie

 

[StephenEd]

I would agree regarding redistribution if there is a lid on the system, but you mentioned that the top is free.

How does the moment redistribute up the wall? If there is less than fixity at the base, the wall is not stable.

As for the analogy to the snow on the slab - I've never really understood that one either! With concrete there is a certain element of "it's always worked when we did it like this in the past".

 

[dutchie]

The top is free, but the slab is restrained on either side (considering a 1m strip), so the moment just becomes more My instead of Mx. As far as less than fixity at the base is concerned, what i'm referring to is fixity against rotation. The wall is still stable.

the reason the slab analogy is valid is that slabs have such low percentages of tension reinforcement, therefor behave is a ductile fashion, which is what leads to moment redistribution. In a beam, reinforcement ratios are higher, and ductility is not as great, therefor, redistribution does not occur before concrete crushing and eventually failure.

cheers,

dutchie

 

[Focht3]

On November 3, 2003 [blue]tincan[/blue] opined:

A 5M X 60M wall integral with the bottom slab would be more apt to be under cantilever action at a distance of 7.5M to 10M from the ends of the walls. At the ends one wall supports another and there is considerable horizontal moment and horizontal restraint at the wall juncture. As you move away from the end of the wall the horizontal moment lessens and the restraint lessens. The wall then tends to act more and more as a cantilever.

I agree completely. Early in my professional career (~1984), I assisted in the evaluation of a failed 15 ft wall of a sewage treatment plant in Houston. In that case, the designer used USD (14" wall) rather than ASD (20" wall). What was relevant to this discussion? I modeled the wall using DSLAB30 and got excellent match between the predicted maximum moments and the locations of the cracks in the wall. In fact, the cracks were orthogonal to the orientation of the maximum moment values. The transition from a restrained edge to full cantilever (for both the actual wall and the computer model) occurred within 20 to 25 feet of the edge restraints - roughly 1.5 times the wall height.

How did a geotechnical engineer get so involved in what was clearly a structural issue? The designer claimed that the edge of the foundation had settled, causing rotation of the wall. (It had not - we did a full scale load test to confirm.) After I ran DSLAB30 to evaluate the shallow foundation system's performance, evaluating the wall was a relatively straightforward matter. I also have a lot of structural engineering in my background -




Please see FAQ731-376 by [blue]VPL[/blue] for tips on how to make the best use of Eng-Tips Fora.

 

[dutchie]

Nice post Focht3,

I should clarify a few things here, as the discussion has wandered somewhat from the initial post. The initial post was an effort to validate a thought that in a tank, Vf<Vc/2 is not required and in fact, Vf<Vc is suitable. We have wandered into discussions on restraint and moment distribution which have been very helpful. These principles are in use only where applicable, i.e. where length to height ratios in the tank are within acceptable boundries for the model to be effective (ie. cantilever action within 1.5x wall height from restrained edges). Otherwise, in the middle of a long stretch of wall, cantilever action is assumed for sure.

Also, just a note, I have noted before that I am using Robot Structural software to model the tank using plate FEM analysis. I have done some more verification against PCA published coefficients and have found excellent correlation (which I would hope). Also ran some models against CSA Appendix B (coeffecients for flat plates supported on 4 sides) with similarily good results. These models are the basis for design, and do predict the cantilever action at the same points discussed above.

One more note, the latest version of ACI350 publishes load factors (sanitary co-efficients) that are to be used with USD, with the intention of getting the design stresses to better match the working stress values. My experience is that it does ok, but crack width still ends up determining wall width

thanks for the input everyone, keep it coming if there are more thoughts...

dutchie

 

[MBruggink]

I couldn't let this pass by without getting more info on this..

JedClampett said:
<<The PCA publication is &quot;Rectangular Concrete Tanks&quot; and it's HORRIBLE. There's the small matter of the &quot;ERRATA&quot; >>

There exists such a thing, an errata?!?... If I could get my hands on a copy. It would be great! I notice that in the PCA Book there are no shear values for Case 5 of the Rectangular Plate. Along with a whole host of other errors...

Mark

 

[JedClampett]

I'll fax you the Errata if you give me your fax number. I'm not aware of a problem with the shear values, but it wouldn't surprise me.

 

[MBruggink]

Thanks Jed.

Email me at xhare2000 at yahoo dot com and I'll email you the fax number here.

Thanks.

MBruggink