New Opening in Existing Brick Wall - Lateral Concerns 1

[melann189]

Here's the situation: Existing (3) story 15' wide rowhouse / brownstone (middle of the block), double wythe brick masonry rear wall. New addition to be added to the rear which will require large openings be installed in the rear masonry.

My concern is lateral design requirements. Per the IRC (NJ edition) it appears I do not have to consider the rowhouse as a structurally independent building because of the fire rated party wall. With that in mind I could easily say that any lateral load is simply taken by the adjacent buildings and no real harm done. The argument is that in a worst case scenario both neighboring buildings could at some point no longer exist. But I would think that the existing masonry as is would be well overstressed at that point as well.

However, I am certainly reducing the lateral capacity of the building with these openings (approx. 10' wide floor to ceiling opening on two levels), so instinct tells me that lateral reinforcement should be installed. So, I guess the question is, to what requirements do I design the reinforcement? If I plan for a steel moment frame and assume it takes full wind / seismic loading, I imagine it will become a cost and aesthetic issue, but I don't see a way around it.

I seem to be coming across this more and more lately and have been told by my clients that their other engineers just put in a lintel, and no additional reinforcement. Thus, I am concerned that I will be said to be "over-engineering" the renovation. Anyone else come across this or have a good code reference for why it needs to be reinforced (or not reinforced)? Also has anyone had experience with the NJ Rehab code? I am finding it very vague which is not helping this issue.

 

[CELinOttawa]

Lots of engineers do crappy work, and some eventually lose their licences over it... Don't compromise. You're doing the right thing here!

As to your rehab code, try asking the publishing authority or your AHJ.

You can also look at trying to justify a three sided box system. Also a relative strength and stiffness approach may be sufficient.

 

[KootK]

A middle ground approach: estimate the lateral strength and stiffness that you're removing and see if you can replace it. It's never quite as simple as that as I'm sure you know. Still, it's rational and might allow you to remain competitive. That's important. If you disappear, it will just be the negligent engineers left doing the work. And then what would become of us?

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Archie264]

If Bookowski weighs in, pay attention. He's had a lot of experience with these structures...

 

[bookowski]

This is a topic that drives me nuts (from dealing with it a lot).

I would second kootk on two points:
1. For the rest of us please do something that is at least justifiable. These are not very black and white situations so there is a lot of leeway, especially with the codes in this area. The NYC code is useless on the subject. There are a lot of engineers that just put a lintel, I've seen 100% wall removal with lintel party wall to party wall. These morons are not only putting themselves and the client at risk but also screwing the rest of us that have to explain this to the client. There are definitely a lot of engineers taking a more rational approach so you won't be alone.
2. The replace or make better what you took out route is one that is justifiable in the absence of any more strict code requirement. This gets you with something less crazy looking but probably still a lot more than the client expects.

I can post some photos of lots of moment frames that we've done for these. The frame solution has it's own set of problems, just to get started: is there a legit diaphragm to deliver the lat'l to the frame? compatibility of a 'soft' frame with a very stiff/brittle/weak masonry building, the front wall which you leave in place is probably rubbish so in a real design event will your shiny frame be standing while the rest is in rubble? What does the frame sit on - the crap rubble wall below? (when I do frames I put new concrete wall below as well or serious upgrades at least).

If you go the replace what you remove route you have a chance at a masonry solution - however it needs to be reinforced masonry, you'll get big tension forces in p/a+m/s which can't be justified easily in an unreinforced scenario. This introduces a new set of problems. In general it's not great to mix cmu with historic brick masonry - so now what. I have done many of these with reinforced brick masonry, this takes some contractor hand holding as it's not typical in the area. Basically it's not so dissimlar to cmu except it's all brick with areas omitted for vertical reinforcing (which needs anchorage into a legit base, i.e. conc. wall/footing). Is it perfect? No. Is it better than the original - probably, and you can sleep at night.

The neighboring building argument is rubbish. Are you going to send a letter to all the neighbors informing them that they can never modify their back walls?

This is a big/recurring problem in this area and the bldg departments need to step up and issue some specific guidance.

99% of the time these projects are not a little old lady repairing her back wall, it's someone turning these things into $5M homes, so if they need to drop $40k on the structural work at the opening it's not really the end of the world. Don't be pressured by their 'other engineer'.

 

[Archie264]

I can post some photos of lots of moment frames that we've done for these.

Hey, if you're willing then yes, please. I'd like to see them. Thanks.

 

[CURVEB]

^+1 - I agree with what's been posted so far. Very difficult to make these renovations work, especially when clients are used to paying low fees to bad engineers that didn't go through the proper checks to show that it works (or doesn't). It's especially hard in low-seismic areas where you can't easily point to actual cases of building failure due to lateral loading.

 

[KootK]

@Bookowski: is the objection to mixing mixing brick & CMU simply an aesthetic / heritage issue? Or are you worried about shrinkage swelling compatibility too?

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[jfmann]

New Jersey here (for longer than I recall sometimes).......Rehabilitation Subcode of Uniform Construction Code (UCC) surely can be daunting on first parsing .....well even on fourth or tenth reading also ;-).....yet there is some semblance of sense after you flip pages for a while. Lack of index has always been a pain.
As with any code, Rehab Code simply does cover all the bases all the time........however, structural effects of alterations are addressed;

5:23-6.6 Alterations
(c)The work shall not cause any diminution of existing structural strength, system capacity or mechanical ventilation below that which exists at the time of application for a permit or that which is required by the applicable subcode of the UCC, whichever is lower.


As for design of new header / lintel beam to support brick over new opening in back wall.....other than in areas where seismic loads govern (not all that much in NJ),design for vertical loads (only) is a rational approach when you consider the thousands upon thousands of such installation that have been performed since time began without any major (or even minor) problems reported with collapse due to lack of lateral stability. Perhaps main reason for this successful history is because...surprise.... adjacent houses remain in place.....which itself is a reasonable conclusion to make during design.

I doubt any vintage rowhouse with brick walls has adequate capacity (by itself) to resist seismic forces per code.......severe damage and collapse during earthquakes is clear evidence of that.

As for wind resistance......with floor joists spanning across width of house (parallel to back wall)......there is very likely (almost certainly) no connection between wood floor joists and back wall........and even if there were, such connection almost certainly does not have anywhere near adequate design capacity to provide for transfer of wind force (that of course would only occur if adjacent houses are torn down) into brick wall. So lack of design for shearwall (or moment-frame) is not a design deficiency since there is no reduction of existing shear capacity.

However......the much more practical issue of concern is design for wind perpendicular to this wall.....even though this issue is often neglected entirely (especially by architects). If wind force is distributed (by windows or doors) to lintel beam......there is very likely no resistance at ends of lintel beam to distribute force down to foundation through edge of remaining wall.

For overall lateral stability...although I do not think it necessary for interior rowhouse........you might consider adding statement to plans (put in big box to highlight)....that design has been performed based on key condition that adjacent house remain in place.



John F Mann, PE

http://www.structural101.com

 

[jfmann]

Somehow omitted "not" in previous post;
"As with any code, Rehab Code simply does NOT cover all the bases all the time........however, structural effects of alterations are addressed;"


Also....for vintage brick bearing wall construction......existing wood joists very likely bear directly on (actually embedded in) brick wall....so that there is no diaphragm "chord" element.

Bottom line is that individual row houses do not have floor diaphragms as for wood-framed construction today.



John F Mann, PE

http://www.structural101.com

 

[bookowski]

I agree that there is not a great/clear solution to this issue but offer the following responses to jfmann's comments as I've had these arguments may times before:

New Jersey Rehabilitation Subcode of Uniform Construction Code (UCC)
Not sure about NJ codes. In NYC renovations are allowed to be filed under 'old code' which is essentially useless and can be interpreted to get you wherever you want to end up.

As for design of new header / lintel beam to support brick over new opening in back wall.....other than in areas where seismic loads govern (not all that much in NJ),design for vertical loads (only)
I am never clear on what people mean by "except when seismic loads govern".... govern what? If your wind is 10k and your seismic is 2k does that mean that you throw out the seismic? Whatever the seismic is you still need to provide capacity to resist it - even if it's small. Seismic loads are low in this area but they are not omitted from the code.

is a rational approach when you consider the thousands upon thousands of such installation that have been performed since time began without any major (or even minor) problems reported with collapse due to lack of lateral stability. Perhaps main reason for this successful history is because...surprise.... adjacent houses remain in place.....which itself is a reasonable conclusion to make during design
Past performance should not be used as any argument except where conditions are unchanged - material, condition, geometry, loading, and the neighboring buildings that you are counting on all remaining unmodified. In my experience these renovations almost always come with a complete gut that removes most interior walls. Varying from 14ft to 20ft for most residential lots (in nyc) people want them fully open in the short direction. The original buildings were chopped up into loads of rooms with plaster and lath partitions, which I would guess contribute greatly to the stability of the building. Removing all the interior walls, removing the back wall, having masonry that is 100 years deteriorated, and having this process repeated over numerous properties in the 'row' and then saying that it worked so far is meaningless because it hasn't worked in the condition that you are providing.

Beyond the above argument there is a problem with using past performance in that 90% of what we design for never occurs, i.e. design level loading - so almost nothing ever fails. Take away construction accidents and problems related to water damage and how often do you see structural collapses outside of seismic events? If you went to haiti one month before the earthquake would it have made sense to say that nothing has collapsed yet so that's evidence that all the unreinforced masonry is fine? This is more of a contractors viewpoint in my mind rather than an engineers.

there is very likely (almost certainly) no connection between wood floor joists and back wall........and even if there were, such connection almost certainly does not have anywhere near adequate design capacity to provide for transfer of wind force (that of course would only occur if adjacent houses are torn down) into brick wall.
Iron ties have been required and implemented from parallel walls since prior to 1900 here. How much capacity do they have to transmit diaphragm forces? Not sure, probably not much but something. Could they transmit 1 kip each? I would bet yes, so lets say 2k per floor min?

you might consider adding statement to plans (put in big box to highlight)....that design has been performed based on key condition that adjacent house remain in place
I would also suggest putting a note that the wind shall not blow and that the client shall cook you dinner every Sunday. How can you dictate that the adjacent building can not be modified? In NYC all of these back walls are getting blown out - you can stand in a rear yard and count them.

Also....for vintage brick bearing wall construction......existing wood joists very likely bear directly on (actually embedded in) brick wall....so that there is no diaphragm "chord" element.
Outside of the simplifying assumptions that we make in new design you don't need a distinct and designed chord to transmit forces. Could the diaphragm deliver current code loads? Probably not. However those joists have wood lath and heavy plaster ceilings on the bottom and ~1" plank subfloor above, both perpendicular to the joists. I would guess that those components can deliver some forces > 0, probably quite a bit when push comes to shove.

When party walled buildings are demo'd for a new building it is very common here to get a lean and movement from the adjacent building(s), with permanent bracing from the new building being required. On these types of jobs we always warn the client early that we may end up having to carry the neighbor. This is evidence that the stability is dependent on the row, and since you have no ability to limit modifications to other buildings you should not be diminishing your capacity under the argument that the other buildings take care of it.

I am ok with a wide spectrum of design for these - limiting lateral load to what you think the diaphragm can take, or the walls can take, or some other rational argument - but I don't see how anyone can say they've engineered one of these modifications with no consideration of lateral forces.

 

[bookowski]

Archie - Here you go. Sorry, not great - I just grabbed a few quick.
1st two photos: full party wall to wall removal. Upper floors were originally supposed to be shored and resupported but contractor found it faster to remove and rebuild. On this one both neighbors (right and left in photo) already had their lower two floors blown out in past renovations.
3rd photo: typical
4th photo: frame is at location of original rear wall, beyond is an extension.


Kootk - CMU/brick is a compatibility issue because of the age of the brick/mortar relative to the stiffness of cmu, concern is popping the firing face off the bricks, cracking etc. Usually we specify a mortar sample and test be taken (only costs around $300 to do this) and a compatible repair mortar used for the work, pretty much always comes out similar to a type O. I don't get too worked up about it but most architects here recognize the issue and will steer clear of cmu with old brick.

 

[jfmann]

brief response to comments by bookowski;
Appears fair to characterize your position as at least acknowledging, by various comments, that good old "engineering judgment" is required for design in these cases.

If you "hear" me chuckling between the lines though, it is because I find taking the "other" side of this issue ironic......and somewhat new actually........since I am almost always addressing such issues from the more conservative angle (usually with owner and contractors, but also with architects).

Conservative approach is very fine..........yet, the code (and logic, which are not always in sync, but do sometimes merge) does not require upgrades (especially costly upgrades) to structural capacity as long as we are not causing an obvious safety problem........or, as the UCC states, we are not reducing existing capacity.

Attempts to make case that old wood-framed floors somehow provide diaphragm capacity.......while acknowledging that there are no chords.......is not consistent with conservative approach.

As for iron ties......in all my experience with vintage buildings with brick walls, I can not recall even one instance of iron ties. But even if such ties were installed (which I have no doubt you have seen in your journeys).......how much capacity could they really provide?.....especially considering they would be embedded in mortar that often is more like sand than anything else...........surely we would not rely on such connection for seismic resistance.

Bottom line is.......if the entire building does not have lateral resistance that we can determine with any reliability.......there is no capacity to be "reduced".

As for the Haiti analogy......that is just silly........there are many, many (way too many) demonstrations throughout time of plain masonry buildings failing miserably when subjected to seismic forces........so that using such an example makes no sense in this context.







John F Mann, PE

http://www.structural101.com

 

[melann189]

Just wanted to say THANK YOU to all who responded. It is nice to know that I am not the only one who has spent far too much time thinking about this. My original plan was a steel frame (with new concrete foundation) and to deal with the other issues this creates as best as possible. Will design rationally with the ultimate goal being that the new construction will not be worse than the existing wall. We'll see what the clients have to say but at least I'll be able to get some sleep. Thanks again!

 

[KootK]

Gotta say, these things look great. Even as a structural engineer, if I owned one of these units, I'd be pushing for a Bookowski Frame too. The bump in resale value and space enjoyment would justify the extra risk to my life and limbs.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[jfmann]

Just one other thought about this topic / issue, which is quite interesting.

What about tiedowns for shearwall?..........for analysis of existing house as if it were stand-alone (not part of assembly of row houses)........I submit that design capacity for lateral stability is (would be) inadequate due to lack of tiedowns for what would then be short shearwalls (15-feet maximum, even without openings) for 3-story building, which is probably about 40 feet long or more.

For assumed 40-foot length, 10-foot story heights......and 20 psf wind pressure.....tiedown force (for wall without openings) would be on the order of 6,000 lbs before taking into account weight of wall, which of course is substantial.

However, for wall with openings........such that only relatively narrow "piers" between openings provide shearwall resistance.....tiedown forces are likely much greater. Point of course is that lateral design capacity of stand-alone row house is not going to satisfy current code requirements......for many reasons (lack of tiedowns being just one).

Installing rigid steel frame is fine......surely does no harm!........yet I doubt that it really provides much more capacity for the overall building.

John F Mann, PE

http://www.structural101.com

 

[KootK]

How does one tie the beam/column joints into the floor diaphragms for out of plane frame stability? Or do you engage the demising walls?

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[bookowski]

KootK - I've done it different ways but they generally all involve some kind of metal ties back to the floor diaphragm, these almost always have all new joists/subloor at least in the floors/areas of the rear wall work. The wf beams are in the same plane as the joists. I typically have holes at midheight of the beam web intermittently and at the joint, through this goes a threaded rod that extends back at least 4 joists (new LVL, hole for rod at midheight as well) with staggered blocking on each side and nailed from new subfloor above. I've done variations on this as well such as having the frame come with angles welded to the top flange which then gets straps back. All of the details are a little weird and not perfect. I keep trying variations and then circling back to the original. If you look through seismic retrofit docs from the west coast you can steal a lot of details/ideas. (You also need to provide connections for the shear transfer - again I don't have a great detail, lots of attempts)

John - I don't want to argue the technical details too much, there are many opinions on the subject. BUT - yes, of course it needs to be tied down.... this is how you design a lateral system. If it's a frame then the anchors need to anchored. If it's reinforced masonry then the reinforcing needs to be anchored... no different than any lateral system. If you need a #7 bar 1 inch above the lowest level then you also need that anchored, nothing revolutionary there. If you refer back to earlier post we always provide a new concrete base from which we anchor. This is able to transfer the T/C into the overall wall width. If you mean that the entire structure is subject to overturning - well, yes I don't go that far to tie down the building. Normal order of operations: shoring posts + needle beams --> party wall bracing ---> remove orig lower walls ---> remove rubble wall or grout and encase in concrete --> install frame or reinforced cmu ---> make it pretty ---> removing shoring/patch. As stated earlier there is a level of judgement in how much you are resisting. It seems like your argument boils down to throwing your hands up and implying that it's not possible to engineer it. I agree that it's not perfect but lots of people all over the world are modifying and engineered historic unreinforced masonry buildings. It's not perfect but better than nothing.

 

[bookowski]

OK, I'll stop after this. See attached first example I could grab. Elevation shows existing wall configuration with substantial masonry walls, keep in mind that the interior is also full of old plaster and lath cross walls. Now look at proposed plan - brick pencils on each side and entirely open interior. Discount the original diaphragm action all you like, count the original mortar as lime/weak, only a few anchors from floors to walls - even with all of that how could you look at that and put only a gravity lintel and put a stamp on it.

 

[CURVEB]

And stairs next to the new opening? Even less chance there is a way to tie the diaphragm into the exterior wall.