PEMB frame deformation due to foundation settlement 3

[bones206]

I'm evaluating a 40 year old PEMB that has settled over 16" from front to back. The building is 200' x 220', with 4-bay frames spaced at 44'. I am involved in a relatively high-profile retrofit project at this building, with new mezzanines and equipment foundations, but the settlement findings have paused the project while we await results of a geotech investigation.

Meanwhile, I had the building laser-scanned and noticed significant frame deformation in the scan model. Looking in plan view, the frame beams at the roof are deflecting out-of-plane in a sine wave shape (i.e., buckling mode shape). The columns are twisted at the top following the deflected shape of the beams.

Obviously, this is not a good situation. I think the fly braces are doing a lot of work to prevent a full on buckling failure and frame collapse. The fly braces are L2X2X1/4 angles field-welded at the frame bottom flanges and the bar joist roof purlin bottom chords. These connections do not inspire confidence at all, but are critical to the stability of the building.

I'm being asked by the client to assess if there is a safety risk and to propose mitigation and repair options. Preliminary geotech results indicate that piles would be required to arrest the ongoing settlement. I have conceptual pile designs in mind, but I am looking for opinions on the building frame aspect. My inclination is to declare the building basically "totaled". I don't see how you can undo the buckling shape in the frames, and I believe that deflected shape indicates the frame elements are already stressed beyond their design limits. Given there is a lot at stake with this project, the client wants to see some more robust analytical "proof" that the building is not salvageable.

I am content to be the bad guy in this case, since my gut instinct is telling me the building might not be able to sustain a design-level snow load. On the other hand, the building has probably stood this way for decades, and maybe I am overlooking some structural heroics that could allow the building to be safely occupied into the future.

So my questions are: 1) how can I "prove" that a building is beyond saving, besides relying on my reputation and judgement? and 2) any longshot ideas on how to save the building?

 

[SWComposites]

re 2) could you bolt in a series of transverse full depth beams between the existing frames? (yeah, might be tricky to get them installed if the frames are already bent; maybe it requires shorting up the existing frames first?)
but re 1) can you calculate how much additional dead weight you could add to the structure and still have acceptable margins?

also re 1) can you model the existing structure, first in the unsettled position, and see how much margin exists, then apply an enforced displacement to mimic the settlement and rerun the analysis?

 

[GC_Hopi]

The owner is paying you for your expertise. I would hope that they would value your judgement. Now I will get off my soap box, I would pose the following question if the members do not meet the AISC tolerances for straightness, etc. does it make sense to evaluate them using AISC methods or are you outside the bounds of the design criteria. I think if the members are not in tolerance then you are justified in putting corrective action on the table to "fix" the members, price that fix and see if that drives you to "total" the building. If the owner does not want to accept your judgement then walking away may be the final action.

Would the owner be open to a full scale load test. lol

 

[Eng16080]

Based on my interpretation of what I just read (which might be out of line), I think it's worth repeating more or less what GC_Hopi wrote above:

You're the expert here. The owner doesn't get to decide what counts as proof that the building is or isn't safe. Maybe you really aren't sure, which is fine, and in that case perhaps it makes sense to provide a more in-depth analysis, but it doesn't makes sense to do that because the client needs better proof. Also, there are never any guarantees and you don't owe anybody "proof." You can't prove anything anyway. None of us can. We can make very reasonable recommendations/observations/suggestions/whatever, but there is no proof. At least that's my take. Perhaps the closest thing to "proof" would be running an analysis of the members based on the deflected shape, and checking if they work by code, idk.

I barely know anything about this project, but just based on what you wrote above, it sounds really bad. It's settled 16"? Wtf! There's no way that frame is acting in the way is was intended when it was designed. It's almost certainly stressed in ways it wasn't designed for, and being a PEMB, we all know there's essentially zero reserve capacity anywhere. The beams have buckled into a sine wave shape? This sounds like the kind of engineering disaster you might read about.

Can you fix it? Probably! You just need to add piles and reinforce every member.

Sorry for the rant.

 

[phamENG]

You're the expert here. The owner doesn't get to decide what counts as proof that the building is or isn't safe.

I agree with the first part, partially disagree with the second. If bones says it's totalled, and reports it to the building official and the building is ordered condemned and demolished, but meanwhile the owner gets somebody to prove it could have been salvaged, guess who the owner's lawyers will come after to pay for that building...

But to bones' question...I'd analyze it with a forced displacement representing that settlement. I'd guess that will show that the frames as they are exceed the allowable stresses. From there, you could get into the risks involved with attempting to straighten or reinforce them in that condition, and then encourage them to obtain a second opinion about your conclusions.

 

[Eng16080]

If bones says it's totalled, and reports it to the building official and the building is ordered condemned and demolished, but meanwhile the owner gets somebody to prove it could have been salvaged, guess who the owner's lawyers will come after to pay for that building...

My point is more that bones's engineering judgment shouldn't be dependent on the client's request for "better proof."

The scenario you mention in the quote would definitely be messy. Recommending that a building be demolished isn't something that should be taken lightly. Ideally, I think it should be presented more along the lines that there are several items which need to be fixed in order to salvage the building. Normally, I would give my findings directly to the client and not to a building official, under the assumption that they would act appropriately given that information. I would also assume that it's possible that the client may hire another competent engineer who might arrive at a different conclusion. I wouldn't necessarily assume the other engineer to be wrong in that case. I would try to give the benefit of the doubt.

If the building was clearly deficient and immediate repairs were needed and I became aware of those repairs not being made after some time had passed, only then would I contact the building official, after first warning the client.

For projects that seem really bad from the beginning, I often try to give the prospective client a good idea of how I will likely proceed. If I think there's a good chance I'll recommend demolition, I let them know that possibility up front. Depending on the particular client, this often gives them enough information to know if they want to use me for the project. It makes me a good fit for certain clients and not for others. Some clients don't care if a building is safe and will shop engineers until they find somebody willing to say what they want.

 

[bones206]

I appreciate everyone's input so far. Just to add some more context, this is a long term client with multiple ongoing projects. When I say the project is "high profile", I mean to say that their projects tend to have several stakeholders who provide up-front funding for the renovations and expect an immediate return in the form of manufactured goods. I can't get into too much detail due to a strict NDA, but suffice it to say that there are a lot of eyes on this and the ripple effects of this project not going forward would be significant.

That being said, I am able to deliver the bad news and take any blowback if that is what the situation warrants, but I want to make sure I have done my due diligence and turned over every stone on behalf of the client. I've definitely given them a heads up that the building may not be salvageable and I believe they respect my judgment, but my fee is like 0.1% of the value of the project, so they are understandably anxious and want all options on the table. I may very well recommend a second opinion or hire someone myself to do a 3rd party peer review.

I barely know anything about this project, but just based on what you wrote above, it sounds really bad. It's settled 16"? Wtf!

I know, it's a silly amount of settlement... and that's just relative to the highest point of the building. Absolute settlement relative to the original datum is probably even more.

I have thought about doing the enforced displacement analysis, but I see a couple issues: if I do this in software like RISA, I don't think it would capture the out-of-plane deformed shape. I think I would have to do a global buckling analysis on the frame, which I have not done before. Additionally, do I really NEED to do an analysis if I can already visually confirm that the frame has deformed to this degree? I'm finding it hard to assess what the acceptance criteria is in this case, since it is really a global stability mechanism at play rather than a straightforward allowable stress comparison.

Note that each frame line is stepping down in elevation with the settlement profile. So in the plane of each frame, the support elevations are relatively constant. I suspect what has happened is the horizontal roof X-braces (flat straps) have lost their tension and thus their effectiveness. So the frame beams are free to wobble as if there is no roof diaphragm. But there is so much complexity that it is hard to say for sure. Maybe the fix could be new horizontal bracing that can be tension enough the bring the frame back in line? Seems dicey at best.

 

[phamENG]

Ah. I was under the impression that it was settling across the frame. That does make it more complicated. My thought was that, in showing the allowable stresses were exceeded by a significant amount, you could 'prove' that the observed deformation was the result of a buckling instability (which is captured in the allowable stresses).

Could it be that, rather than losing tension, the roof X-bracing actually has more tension? As one side drops away from the other, you no longer have a straight line but the hypotenuse of a triangle. Additional tension from that settlement starts pulling the beams laterally while the wall X bracing tries to hold the columns in place in one direction but pulls the top of the columns down the slope in the other.

If the beams have already experienced plastic deformation, I would not try to correct them. Point to the recent collapse in Boise or any of a myriad others where PEMB frame instabilities resulted in collapse.

 

[bones206]

That's a really good point. The sidewall X-bracing has shown some signs of overstress. The sidewall braces are angles welded across the flanges of the columns and are 2 feet eccentric from the concentric workpoints (poor detailing). The sidewall columns are in fact kinked near the base at the X-brace connection point.

I suspect the frame beams are still in the elastic range, even if they have deflected 5"-6" laterally. But the columns could be plastically deformed in their twisted shape. Hard to say.

The Boise collapse is a good comparison and I feel like any attempt to fix and re-square the frames here would be similarly fraught with risk. My problem is that I can claim imminent collapse is a real possibility and qualitatively point to all these warts in the framing, but it's probably withstood collapse in this condition since I was in elementary school.

 

[phamENG]

Every building withstands collapse...until it doesn't.

The geotech has already indicated that further settlement is likely, meaning it has been slowly moving closer to collapse since you were in elementary school. To stop that, they're going to have to disturb the slabs and ground around the foundation. Helical piles may not cause a lot of vibration, but it's not zero. All that work around the foundations just to stop the settlement is going to create a condition that the building has never seen before - or certainly hasn't seen while in this state of distress. And even if you preload those piles, you'll still get some small movement before the settlement is completely arrested. How much more can it take before a collapse is triggered?

 

[bones206]

How much more can it take before a collapse is triggered?

These are the exact types of questions the client is posing, and so far I can only give vague but conservative responses. Struggling to point to anything definitive that shows the building is unsafe. In the end I will probably make an economic argument about frame replacement costs, rather than trying to engineer a fix to the existing. Putting my insurance adjuster hat on...

It seems there is a peat layer that will continue to decompose. Probably from an ancient coastal marsh. I have no doubt it would continue to settle until eventually something gave way. The foundation stabilization would likely be driven H-piles around the perimeter and cased micropiles on the interior. The building would be jiggling a bit for sure. I can already hear the client asking me "so how much jiggling can the building take?"

 

[XR250]

I appreciate everyone's input so far. Just to add some more context, this is a long term client with multiple ongoing projects. When I say the project is "high profile", I mean to say that their projects tend to have several stakeholders who provide up-front funding for the renovations and expect an immediate return in the form of manufactured goods. I can't get into too much detail due to a strict NDA, but suffice it to say that there are a lot of eyes on this and the ripple effects of this project not going forward would be significant.

Hopefully that is not affecting your judgement on how to proceed.

 

[phamENG]

For those sorts of answers, your best bet is probably to rely on those allowable stresses and point to the fact that the building, in its current state, is no longer safe enough according to the building code.

Look at the cables and their connections. Based on relative displacements, what would you estimate their tension to be? What does that do the columns locally? If you've already exceeded the allowable stresses anywhere in that load path, the answer to 'how much can it take' would be:

"It can not reliably take any additional load, shaking, or vibration. While it's not possible to say definitively that it will or will not cause a collapse, it will create a work environment that is unsafe for the workers and, even if that could be mitigated, you'll still be left with a building that cannot be relied upon to withstand the loading required of it to keep the occupants safe."

 

[bones206]

Thanks for the reassurance. This is the path I was already heading down, but I always value feedback from you guys and wanted to see if there were any contrary opinions I should take into account. I was also fishing a little bit for clever ideas and magic fixes, but alas with PEMB's there are never any good options.

 

[Brad805]

I think you will need to do a deep dive into the local distortions. The scans will collect some good data, but if they are really serious about saving this bldg this will take a lot of time and cost to look at areas in detail from a JLG or similar. My initial impression is the frames are toast. Now that I see the deformation is more complex, I am not sure. You may want to look for a pre-eng expert, but I doubt you will find many that have dealt with the type of distortions you may be looking at. We looked at some rusted out pre-eng frames not that long ago, and we ended up replacing the frames from below. We did not remove any of the existing cladding or secondary framing.

 

[LittleInch]

Looks fairly straight forward from a distance.

A 40yr old PEMB has served it time, especially when the uneven settlement is concerned / visible deformation of certain steel members which were probably always at ~100% of their allowable loading.

Structures like this and many others exhibit little pre warning of collapse and only small increase in loads or perhaps someone running into one of the columns with a vehicle or forklift would be enough to collapse the whole building, or even a snow dump or very heavy rain. So whilst the "It's been there 40 yrs and is still standing" argument needs to be addressed, that's one way to look at it. Given the importance, a second opinion / peer review may be seen as a useful addition so long as it doesn't say something different....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bones206]

My initial impression is the frames are toast. Now that I see the deformation is more complex, I am not sure. You may want to look for a pre-eng expert, but I doubt you will find many that have dealt with the type of distortions you may be looking at.

It seems like a mix of global buckling-type frame deformation, and localized distortion at the sidewall X-brace connections. As a mental exercise, picture a PEMB being erected, then someone goes and loosens up most of the brace cables in the walls and roof until they are slack. The frames would become unstable and want to buckle under their own weight a la the Boise collapse. Then to pham's point, the remaining taught cables would be overstressed trying to keep the frame standing up, resulting in localized distortions at the connections (even more pronounced due to the poor workpoint detailing).

We looked at some rusted out pre-eng frames not that long ago, and we ended up replacing the frames from below. We did not remove any of the existing cladding or secondary framing.

That sounds like a cool project! I'm fairly certain it would be cheaper to demolish and build new in my case.

@LittleInch - thanks for the feedback. Pretty much exactly where my head is at with this. Having a third party investigate and come to the same conclusion would probably reassure everyone involved.

 

[driftLimiter]

I can't really see how trying to model the deformed structure is worth your time. The frame was designed for AISC tolerances, and it was brought right up to the limit of capacity vs design. So unless the distorted shape would somehow improve the capacity, or the loading has been reduced it seems pretty obvious that you are not going to get a good result out of that.

I haven't done much ASCE 41 work, but I think there may be a more practical method to assess the structure in there, and you would have something more definitive to stand on as the ASCE 41 document prescribes a rigorous assessment of existing buildings. Not simply model the structure according to current code.

 

[Aesur]

This sounds like it may be close to qualifying for being brought up to code based on added loading? If this is the case, I would bet there is very little chance you will get this to calc out. That being said, I believe leveling the building and or modeling with enforced displacement are your only options to potentially salvage the structure. That being said, options to "fix" include adding interior columns so you can change the way the frames work. Potentially adding a new lateral system to avoid the need for the moment frames.

 

[Brad805]

" it seems pretty obvious that you are not going to get a good result out of that" A commercial real estate person will not see that being obvious. A $75k investigation proposal will bring them into reality.

Can you post any sections of the frame distortion from the scan? I am curious.

In our carwash repair there were new code implications if they were to replace. The cost of new frames was a fraction of the re-build cost when one considered architectural, electrical and mechanical code upgrades.