basement vs retaining wall

[keyPitsimplE]

a continuation of thread 256-257568

I have debated over what is best to use for residential basements for years now. I ended up using a (cantilevered) retaining wall design because it was too much hassle to have the guys in the field build and make the connection to the floor diaphragm before backfill. Also, the connection to floor around stairway openings and parallel to joists is even harder. Not to mention the seemingly poor choice of having a wood floor and little metal clips support a concrete wall.

However, with the retaining wall you end up with a huge footing. Also, in reality, where there are corners, or the wall takes a jog, the wall does not act like a retaining wall at all. Basement wall vertical reinforcing goes toward the inside of the wall because that is where the tension is. For a retaining wall it goes on the outside (dirt side). In reality, the tension side changes throughout a typical basement, even more so when there are long runs of wall between corners.

I am half tempted to design a wall that has a 10"x30" footing with an 8" wall in the center, vert rebar in the center figured for worst case bending, for a max. of 9' basement height. However, I find no good way to justify a design like this with numbers.

Have any other engineers had this same debate and come up with a good solution that meets practical as well as theoretical considerations? Things also change when one side of the basement has parking above is like garage or driveway, thus vehicle loading. There is much more to this conversation, but if you have considered this, you know what I mean.

(Best I have done yet is use retaining wall design, but clipped the heels off at the outside of the footings to save concrete, since there is no overturning there. I also use 2 or 3 horiz bars near the top of the wall to resist "bowing" at the top, as a pseudo top restraint. With daylight basements I step the footing back based on actual backfill height. I'm just not crazy about the steel being on the backfill side of the wall because I know the tension is not there except far away from the corners.)

 

[dcarr82775]

Once or twice in the past for an investigative project I have looked at standard basement walls as fixed at the ends, pinned at the bottom, and free at the top with (1) layer of bars centered. If you do an RISA shall analysis where you adjust the thickness for effects of cracking you can actually show that it works.

 

[keyPitsimplE]

Thank you! I wondered if anyone had done that. It is definitely a finite element analysis to do it right, but that was beyond my capacity to do without spending days or weeks. What did you base your initial vert and horiz steel design on? What is "standard" for you? 8" thick wall, 8' tall, 10"x30" footing, #4 @ 12" o.c. each way?

 

[BigH]

Have you considered an "L-shaped" footing for the base? with the L going to the inside? - rather than a true cantilever. Drainage on the outside to take away hydrostatic pressures of course.

 

[keyPitsimplE]

Yes, that is how I have designed my basement retaining walls in the past, to save excavation. It is still a cantilevered retaining wall design though. The footings get even bigger when they are to the inside and don't have the weight of backfill to resist overturning. That is the crux - basement walls do not overturn like a text book one way retaining wall. They are restrained exactly as described by dcarr82775 above.

 

[dgillette]

Does anyone ever use external counterforts for new construction? I recently had some put on one wall of my basement where swelling clay had bowed it in. They were 6' long by 7'? high, and each had a helical anchor screwed into soft bedrock at its outboard end. (So far so good, but this has been a dry summer here.)

 

[tumbleleaves]

I haven't done a true basement, but I've done two houses that were embedded into a hillside with the garage covered on three-sides. Houses don't typically have basements in California.

I used the connection to the floor diaphragm. I've seen walls with a "theoretical" safety factor of less than 1 standing. It's noted on the plan for them to make the connection before back-filling. If this is a problem they can ask me to change it (they didn't). I suspect they probably back-filled before making the connection, but the compaction effort was likely mediocre. Haven't heard anything back on these and it's been at least ten years now.

I may have used empirical design methods ACI publications for the actual concrete and reinforcing design (can't remember for certain, sorry). I remember analyzing the wall like plates fixed at various edges (like counterfort retaining walls), but the elastic theory yielded unreasonable results. A great deal of thought and worry went into the specific analysis and conditions. One of the houses the front had to be on piles and the back embedded in the hillside bedrock to prevent differential settlement. These structures were multi-story and had a lot of issues. At the time I was still finding my way in the profession and took on the structural design of a few residential buildings. Since, I've settled in, far from building construction.

Does anyone else recollect having problems with the elastic plate theory giving overly conservative results? Seems like they were overly conservative (from a stress standpoint only) for bridge decks as well, but I went along with that sense the deck can use all the help it can get and is usually the first thing to go. I guess it's not important since nobody uses tabled values for elastic plates anymore.

I sure did work and worry a lot then, and didn't make good career choices or good money. I still don't make good money. But no worries--Hope you all are as fortunate.