Large column-point load on top of a basement wall? 4

[IJR]

My dear friends,

This one always drives me crazy. Perhaps it is because I have no great feel for this one:

You have 4 R.C walls immersed in soil, walls 1ft thick. At the corners or middle of one of these walls you have a steel column pushing in some 40tons.

Which one is more correct?

1) Take an arbitrary strip, say 2ft wide and analyze it for active pressure as well as the 40ton, design it as a stocky column(no second order effects) and tie the strip just like you would a column. Honestly I have no number to attach to my ties. Some minimum will do kind of rationality.

2) Carry out finite element analysis with any standard software, get membrane forces(pressure) and simply provide vertical reinforcement as you would with a wall. You have your min. reinforcement as that of any other R.C wall. In this case I dont tie.

I do both and I am not very happy with so much a difference.

Any discussion will do.

Thanks.

IJR

 

[Qshake]

Before I would jump into a finite element model of the situation I would check into a classical elasticity solution using Prantl's method. See Timoshenko's text: Theory of Elasticity. While at first it may seem to be mathematically imposing it is really very simple and you will be able to determine the stresses (and influence of the load) in the immediate area in the wall. You will also have the additional freedom to modify the load to a uniform pressure load or even a knife edge if you like.

I'm sorry this doesn't address the reinforcing issue. Also, check out Bowles' Foundation Engineering. Seems to me I've seen this application somewhere before in a text. I just can't place it right now. I'll be on the lookout for it though.

 

[IJR]

Thanks Qshake.

If you come out with anything extra, please let us know.

 

[JAE]

Well, the cookbook approach is to follow ACI Chapter 14.

Specifically, they require a concentrated load to be applied over a width of wall no greater than 4 times the wall thickness (plus width of bearing)....14.2.4.

Section 14.3.6 states that, "Vertical reinf. need not be enclosed by lateral ties if vertical reinforcement area is not greater than 0.01 times gross area...."

Section 14.4 requires walls to also be designed per numerous other sections in chapter 10 (column design).

With this, unless your wall is somewhat lightly loaded, you must design it as you would a column and use your option 1.

Qshake is right, the FE analysis would be a lot of work.

 

[IJR]

Respects as usual, JAE

 

[Riz]

IJR,
Are your basement walls topped up with a slab? If not, than its a cantilevering wall!
Fine, both approaches are correct. In both instances the vertical load on the wall creates a clamping effect. If this force can resist the overturning effect due to active forces by more than 4 implies u have an ideal prop.
Note the column load is dispersed at 45 degrees from the point of application. U need to check the maximum compressive stress on wall against allowed.
Riz

 

[IJR]

Thanks Riz,

JAE above mentioned code-specific solutions. I am going into that now.

As for your response, seems logical too. But me, used to detailing steel, tend to somehow treat dispersion of point loads the way we treat them in steel beams, I mean somewhere there is a critical section with a code-specified area, and we check bearing or buckling there.

What about your dispersion concept, where should I check and how?. In case you have time.

 

[Ron]

IJR...well this one has covered just about all the options for discussion. I would take the simplified ACI approach unless some criticality led me to FEA. One positive point to bring up and it is likely the reason you have disparity between your two approaches, is that the point load dissipates greatly before the lateral soil loading maximizes so you get a droopy, egg-shaped stress distribution if you do FEA. The net result is bi-axial shear near the bottom of the wall (inward,downward)that is highly resisted at that point anyway.

That stress distribution is significantly different than what you will get using the simplified column approach.

 

[IJR]

Ron,

Million thanks. You got a way to look at and make things simpler and clearer.

IJR

 

[IJR]

Ron, JAE

I have checked ACI 318 clause 14.2.4 which mentions that the effective strip can not exceed bearing width + 4 TIMES THE WALL THICKNESS.

When applied to a point load at some corner where 2 walls intersect, do you use 2 TIMES WALL THICKNESS ON EACH DIRECTION. I mean at location X in plan.

-----------X
.
.
.
.
.

 

[Ron]

IJR...If truly right on top at the intersection, consider only the long wall condition for shear and don't consider the intersecting wall. The intersecting wall will resist bending for the long wall as a fully braced section, but won't contribute a great deal to the angular shear distribution in the long wall.

 

[IJR]

Another great input Ron. It really does matter how one approaches code with behavior in mind, doesnt it.?

Thanks.

 

[Riz]

IJR,
45 Degrees either side of the column centre line. Shear concept!
Authority to check: Foundation Design by A. Hodgkinson, 1986, Architectural Press, London.
Riz

 

[IJR]

Riz, great input. Cant get access to that reference quite easily, and I would like to compile as much info as I can while fresh at it.

Now, the point load is going to disperse over a distance say 2 m below covering a strip 4m, right?. I will have a beam-column section there of 400cm*30cm, if my wall is 1ft wide. A metre above I will have a lesser section, well with a lesser moment coming from lateral pressure.

Lets discuss your idea: How do I go finding section from which I can use Moment and axial force (M,N) to help me detail the wall as a column. Are you against treating the wall as a column?. Are you suggesting direct use of membrane forces (I mean force/m kind of approach)?

Thanks for consideration. Be back.

IJR

 

[IJR]

Riz, great input. Cant get access to that reference quite easily, and I would like to compile as much info as I can while fresh at it.

Now, the point load is going to disperse over a distance say 2 m below covering a strip 4m, right?. I will have a beam-column section there of 400cm*30cm, if my wall is 1ft wide. A metre above I will have a lesser section, well with a lesser moment coming from lateral pressure.

Lets discuss your idea: How do I go finding section from which I can use Moment and axial force (M,N) to help me detail the wall as a column. Are you against treating the wall as a column?. Are you suggesting direct use of membrane forces (I mean force/m kind of approach)?

Thanks for consideration. Be back.

IJR

 

[JAE]

IJR:
At your corner condition, I would approach the design of the wall "column" as 2T in one direction and 2T in the other...sort of an "L" column - for the column design. This would be an irregular column that could be designed by hand (difficult) or using software like PCACOL (IRRCOL module).

Now, for the footing design, I would agree with the other comments that you can distribute the load over a 45 degree (or 30 degree if your conservative) distance.

 

[IJR]

Thanks JAE

Your comment also makes Riz's more clear to me.

Also special thanks to Ron, Riz and Qshake. It has been a great discussion.

IJR