Underpinning Brick Party Wall - Stability 2

[bookowski]

See attached rough sketch. This is a four story brick party wall (3 wythe) shared by two rowhouses. Currently they have approximately matching cellar levels and a floor height of about 8ft. On our project we will be lowering the cellar slab by 4 feet and also bringing the floor above down by the same amount. This means 4 feet of underpinning at the party wall.

My concern/question is the stability of the wall with the underpinning. It essentially has a hinge at midheight. A check of P/A + Mc/I with a fairly conservative value of P (i.e. no live load, light estimate of dead loads) shows no tension on the wall face by a factor of about 1.5.

I've seen this done many times but I don't feel very comfortable with it. The alternative, building a 'liner wall' inboard of the underpinning is an expensive proposition and I don't want to force it unless it's definitely required.

Any thoughts on this? All other basement walls on this and surrounding properties are brick masonry, and probably rely on a similar principle of the P overcoming the flexural forces.

 

[dik]

What are you supporting the new concrete wall on? Is the construction OK with the lateral soil pressure. Are you removing the existing firecut floor joists and replacing them with eccentric loaded ff ledger? Have you increased the unsupported wall by 4' to make it 12' and introduced an eccentric load at the 4' line? Seem to be doing a lot to a wall system that has been likely constructed for several decades.

Working with old structures can often have unanticipated issues.

Dik

 

[bookowski]

The underpinning will bear directly on soil, essentially the same as the original wall footing - this part I'm ok with. As shown in the sketch the unbraced height of the wall will remain the same, the floor above moves down as well so the end result is the same overall height. There are obviously sequencing and bracing issues. I am more concerned with the permanent condition, and whether as I stated a P/A + Mc/I check for tension is sufficient. The M is coming from soil lateral load (including possible surcharges etc.). This condition is done quite often, I've seen it many times by others but I don't particularly like it. I am just wondering if I am being overly conservative by assuming that it doesn't work even though the numbers imply it is ok.

 

[PEinc]

Bookowski, you might want to look at the PowerPoint presentation on recent building collapses in Philadelphia. It can be found under presentations on my web site. Pay attention to the part about lateral movement of party walls.

http://www.PeirceEngineering.com

 

[tony1851]

If the wall is 4 stories high, surely dead load alone will prevent any tension developing?

 

[shobroco]

I agree, I've seen it done lots of times and I'm not necessarily comfortable with it either. What about a few augered-in soil anchors to tie the top of the underpinning back to the loaded side?

 

[bookowski]

tony1851 - Yes, by calculation it does not see tension. As I mentioned in the original post, even with very conservative estimates of load there is no tension. I think that is the argument of why it works and why it is used in practice. Even so, it looks a bit shady and doesn't feel right.

shobroco - I'm not sure we're allowed to anchor under the neighbor's property and count on it for a permanent condition.

PEinc - I took a look through your presentation, I see similar stuff happening here in NYC. I agree that these buildings are precarious. The question is what is an appropriate check and reasonable factor of safety. If we were lowering the slab by 1 foot would you still be concerned? I am in favor putting a new liner wall in front of the underpinning and original wall - the pushback, as usual, is that this is done all the time with no problems. I don't want to be overly conservative, if the calcs show that it works and it's done often where is a reasonable line to draw?

 

[tony1851]

Personally, I suspect you are worrying needlessly, though admittedly we don't know the ground conditions and the state of the brickwork.
As a compromise solution, could you utilise some of the ground-bearing capacity on your side by underpinning with rc, and casting a toe on your side, like a retaining wall. If your calcs show their would be no tension anyway (by a reasonable margin), this might give a little extra support. (see sketch)

 

[EngJP]

Is there flexibility architecturally in the basement? Perhaps adding buttresses?

 

[msquared48]

Two things here:

1. I do not like the four to fve feet of soil on one side of the wall and the inherent knuckle joint. Just does not feel right regardless of the no tension result. I would add embedded ties from the wall to the existing slab, epoxied into a cut slot at intrvals before the underpinning process began.

2. The diagram implies no spread footing for the brick wallk. Seems like there is an excellent chance that the footing is overstressing the soil here, even in the current situation, unless it is on rock, which I doubt.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[JAE]

bookowski,

Another idea for a "compromise":

1. Do your underpinning as you have shown...but,
2. At some interval, include a series of vertical embedded steel tubes or small WF shapes that would extend from your
lower concrete up inside a cut vertical slot in the brick to the new floor above.
This could be set inside the concrete and brick so there would be no pilaster sticking out...
or you could build out your new concrete 4 to 6 inches inside of the brick and extend the vertical members just inside the existing brick.
3. This provides vertical wall girts that would provide a limited amount of redundancy in the wall system....
4. Your original design has no redundancy...if the dead load is lower than you think and the soil load is higher,
you have only the limited P/A + My/I safety factor to fall back on.
With the vertical girts you have an additional load path.

Another idea that just occurred to me would be to add vertical strap ties on the inside face of the wall. Any "failure" of your system would be tension on your side of the wall so the straps could be used as tension-side reinforcement.

 

[hokie66]

The principle with modifications to party walls or their attached structures is to do nothing which makes the final situation worse than before. In the case here, there are two issues which violate this principle: 1) the hinge between the existing wall and new underpinning (which has been discussed); and 2) the lowering of the floor above means the wall above probably has an increased unbraced length, as these walls typically depend on the floors each side for compression bracing rather than relying on tension connections of the joists.

Whatever method is chosen, the only dead load which I would use as contributing would be the wall itself. The assumption should be that one rowhouse or the other could be demolished or otherwise removed without compromising the wall's integrity.

 

[PEinc]

bookowski,

What you are showing in your sketches is done all the time. You are underpinning for a depth of only about 5 feet. This is minor when supporting a four story party wall. The dead load of the party wall and floors, combined with the weight of the underpinning piers, should be enough to prevent overturning and sliding of both the underpinning and the foundation wall above the underpinning. You are removing only about 6 inches to a foot of soil from in front of the botton of the existing foundation wall. Therefore, the foundation wall should be stable above the underpinning piers. Because you are not removing any floor slabs from either side of the party wall, the party wall should remain laterally stable. The attached photo shows just what you are trying to do, except that there are no floor slabs framing into the front side of the party wall.

www.PeirceEngineering.com

 

[dhengr]

I absolutely agree with Hokie, these types of walls should be handled/considered with a great deal of care and caution. You don’t know what kind of surcharge might be placed on the other higher bsmt. slab, and you don’t know if the other building will always be there, or be reconfigured itself. These brick masonry walls depend on the floor framing, just as he suggests for their stability, and you have potentially/likely increased the unsupported height by 50% on reworked side. Before doing any work on these old walls they should be surveyed in detail for any out-of-plane movement which could cause an eccentricity that is just barely being counteracted by the floor framing which is going to be removed and lowered. Study any wall cracking for what it might reveal, and the condition of all that old powdery mortar which might not support any eccentricity. Some of the cities which have a lot of these row houses have a fairly extensive knowledge and experience within their bldg. and fire depts. about the potential problems with these walls.

 

[bookowski]

Interesting mix of opinions on this one.

I still don't love it but I've come around to living with leaving it as just underpinning. The party walls themselves are in very good condition, this is a very well maintained row-house. A full wall survey will be done and any problems will be repointed or repaired as required. For dead load on the wall I have counted wall self weight and only a total dead load of 7.5psf for floors, no live load. I agree that I am increasing the unbraced height of the floor above, but I am ok with this based on the condition of the walls and calculation. The new framing will use bolted ledgers which will provide some tension capacity in addition to the compression capacity.

PEInc - Not sure what you mean about removing only 6 inches of soil. The existing walls go down about 6 inches, and then have an approximately 12 inch deep stone footing (based on test pits).

Like I said, I still don't like it, but I've grilled a couple of SoE engineers and it seems like they are very comfortable with it and the calcs do work out.

 

[tony1851]

bookowski; obviously you are just relying on dead load, but as a matter of interest, what reduction factor do you use? Where I am, any beneficial dead load in situations such as this has to be modified by a factor of 0.9; just wondering.
You added an amount for the floors; would it still work without them, ie just relying on the weight of the wall?

 

[hokie66]

bookowski,
The bolted ledgers go against the principle of firecut joists, the idea of which is to allow the floor to collapse in a fire without pulling the party wall with it. Whether or not you believe in that concept, the building you are working on was built with that in mind.

Are you or your client involving the owner of the other rowhouse? You only own the half of that party wall to the centre line, so whatever is done needs agreement by both sides.

 

[bookowski]

In my opinion the ledger detail leaves the wall itself in better shape than pockets at 16 inches, which effectively creates a zipper along the wall. New construction is also in fire rated enclosures, as compared to the original construction (and building will be sprinklered). Both neighbors have reviewing engineers. I work on rowhouses all the time, I don't want to get into other details - there are a lot of ways to skin this cat and at the end of the day I don't like to overcomplicate it. It's really just a single span bearing wall building.

Mostly I was just rethinking the underpinning approach that is usually done and whether or not it was reasonable, and what factor of safety people were comfortable with. Based on the variety of answers it seems like there is a lot of variation on this one.

Tony1851 - I didn't apply any reductions to dead load. I went with straight loads and demands to get a feel for the factor of safety.