Reinforcing a PEMB 6

[dik]

Background
One of my projects is the addition to an existing six storey building. The original building was used for storage, and it’s being re-purposed to Residential. It is a flat slab construction with round columns and capitals and drops. The only sign of distress observed is on the 3rd floor with a narrow, but long, flexural crack at mid span. The crack is nearly half the length of the building. The floor with the crack had almost wall to wall 10’x10’ masonry partitions to create storage compartments. I suspect that this is the cause of the flexural crack observed. The original building is over 100 years old.

A PEMB was added to the top floor (to make it a 6 storey building) several years back. The exterior columns are supported on the original building brick parapet which is about 4’ high and 24” wide. The horizontal thrust is accommodated by diagonal angle struts fastened to the existing roof and the base of the exterior columns. I looked at replacing the struts by reinforcing the structure to behave like a continuous simple steel structure and found the framing members to be a little ‘light’. I calculated the capacity using a design dead load equal to 13psf; the original supplier may have used 10psf; I don’t know, but I expect that sort of design for PEMBs.

Now the Crux
The proposed renovation includes adding new load to the PEMB in the form of added insulation, ceiling and sprinklers. This will likely double the existing dead loading. Because the original design loading is light (or, very tight) I’m considering treating the existing PEMB regular steel building with continuous roof beams and adding BAR material to reinforce it for the moment caused by the added loading without moments at the exterior columns (like a PEMB), without going into a complicated rigid frame design. Can anyone suggest an alternative method or any pitfalls?

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[XR250]

I imagine the "design fees" are just a BS way of them making it look like the building itself is cheaper. Would be interesting to hear from some of the PEMB folks on here. I mean, what is involved? Inputting some design criteria and hitting return?
Hopefully the MPC truss folks won't catch on!

 

[straub46]

I mean, what is involved? Inputting some design criteria and hitting return?

Oooooof. You realize the big fabricators staff PEs and SEs right? The whole thing is signed off by someone who went through the same process every other PE or SE has to go through to get their license.

The simplest buildings are pretty much what you suggested. Most (75%+) go through a design process that in terms of QA/QC is better than most, if not all, engineering consulting firms.

Engineers determine the loads on the frame (yes, using specified design criteria, but through spreadsheets similar to other firms), and then tweak frame geometry to come up with an economical (read, as close to 1.000 as possible) design. They'll work on layout of bracing and columns to work with the customer to give them what will be easiest to erect and work from a layout perspective.

A lot of engineers doing the intial design work have 0-4 years of experience, and are hired straight out of college. All those designs are checked, but they are the ones tasked with answering most questions, so a lot of interaction with PEMB engineers from outside is probably with someone with less than 3 years of engineering experience. You'll understand then why some "don't know what they're talking about" and aren't as familiar with concrete breakout and the intricacies of foundation design.

Go Bucks!

 

[dik]

I like the engineering that goes into PEMB... just not the product.

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[straub46]

As someone who's worked on both sides, I understand that aspect

Go Bucks!

 

[dik]

I had a venerated and reviled prof once tell me that if a structural engineer were to ever design an aircraft, it would never get off hte ground.

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[hokie66]

Who do you think designs the structure of aircraft? It's not architects or electrical engineers.

 

[SandwichEngine]

I'm a PEMB guy. The first thing I'd do is really drill down on what the actual loads are. DL for most PEMB frames calculates out to 2 psf or less. Purlins, panel and insulation typically calc out to 2-3 psf. So if that's all that was supposed to be in the original building, that's what it would have been designed for.

It may make sense on a normal job to assume heavier loading to leave a building with extra capacity when you design it but if I'm the owner I would want to know how the building works for true realistic loading and then maybe get options on what it would take to bring it up to a higher allowable capacity in the event that it's actually lower than what you've calculated.

I've often been on calls with the EOR and GC about what loads go onto the PEMB. On the one hand, it would be nice to know that your building is rated to take additional loading down the road if something changes. On the other hand, how much are you willing to pay for that now considering that may never happen or you may not be the owner if that's desired one day. Most owners aren't interested in paying for future loading.

Also to speak to some of the other conversations in this thread, PEMB companies design for the lowest possible loading and take members as close to a unity check of 1.0 because they have to to be in business. PEMB companies only win a job if they provide the lowest cost building (at least most of the time.) Providing the lowest cost building means using the least amount of steel. Consulting engineers would be doing the same thing if they had to determine the total cost of all the materials in their solution and provide that in their bid against other consulting engineers and the one with the lowest cost solution would be awarded the design work. I realize that's not plausible but the point is that the market forces it to be this way.

Also, our design fees aren't that high, at least not for engineering. On any job, 15% of the in-house "design" work is engineering. 85% is detailing and 3D modeling. We estimate how many hours we think it will take and that part of the cost is factored into the price with a flat hourly rate. I've never seen that cost broken out and I can't imagine it would really be as high as $1 per square foot but I'll offer my guess. On whatever job Aesur saw, my guess is that there were a lot of changes and the GC wanted the PEMB supplier to break out every item of cost so that if some item of steel was removed, the GC could get credit back for it. Then the PEMB company converted some of the steel cost to appear as design fees to give back less money when steel was removed. I'm not saying I've seen that done but there is a lot that goes into negotiating changes between a GC and a sub and to me, that would be a much more plausible answer than a PEMB company legitimately charging $1/sqft and actually winning the job.

 

[SandwichEngine]

Dik,

There's a company named StruktureOne in Austin, Texas that does PEMB consulting. I also saw someone on a previous post saying that they were analyzing a PEMB structure using RISA because RISA has incorporated Design Guide 25 for analyzing tapered members like PEMBs use. I don't have RISA anymore so I can't speak to how well that works but I know design guide 25 was written by the guy who created the proprietary software for one of the larger PEMB companies.

 

[dik]

maybe aeronautical engineers?

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[dik]

I'm a PEMB guy. The first thing I'd do is really drill down on what the actual loads are. DL for most PEMB frames calculates out to 2 psf or less. Purlins, panel and insulation typically calc out to 2-3 psf. So if that's all that was supposed to be in the original building, that's what it would have been designed for.

Using actual loadings, the estimated DL on the PEMB was just a little over 12psf. I'd normally use 20, but used 13 for checking it.

It may make sense on a normal job to assume heavier loading to leave a building with extra capacity when you design it but if I'm the owner I would want to know how the building works for true realistic loading and then maybe get options on what it would take to bring it up to a higher allowable capacity in the event that it's actually lower than what you've calculated.

Concur... and with industrial useage I stipulate added loads to be applied. This is something I get the client to agree with based on earlier problems I've encountered with PEMBs. There is a price to pay for this reserve capacity. There's n free lunch.

I've often been on calls with the EOR and GC about what loads go onto the PEMB. On the one hand, it would be nice to know that your building is rated to take additional loading down the road if something changes. On the other hand, how much are you willing to pay for that now considering that may never happen or you may not be the owner if that's desired one day. Most owners aren't interested in paying for future loading.

On one project a couple of decades back... there was a horizontal C section providing restraint for the top of a 4' high 8" CMU wall... deflection they used was L/90... This was not acceptable based on code requirements at the time, and was fixed, but the client paid addtional.

Also to speak to some of the other conversations in this thread, PEMB companies design for the lowest possible loading...

Concur... that's their job, and many of them are very good at that.

Also, our design fees aren't that high...

I have no idea what part of the price is design fee. Due to the high competition I suspect fees are pretty close, pretty much like the wood truss industry. Most PEMB manufacturers have highly specialised software that does the work. That's the part of PEMBs that I like.


So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[Aesur]

@SandwichEngine - great information. I wanted to give a bit more information in regards to the project I mentioned. There definitely was some custom aspects to it, ie 12 to 15' tall parapets that we designed but supported on the PEMB roof (not sure why architects love such tall parapets - they absolutely suck), lots of coordination between us and them, but for the most part their structure was very simple and repetitive. We had to design the walls because the jurisdictions here have started pushing for stucco finishes on PEMB's, not panels which changes building deflection criteria and the PEMB guys apparently won't design for anything but metal panels (so we are told). I believe the biggest thing may have been that they can no longer do preliminary reactions (many jurisdictions in the state are now requiring this) for foundation designs and must give final, but this is done before they even have a contract to provide the building, meaning they have to recoup their engineering cost up front. I wasn't privy to the number of hours, etc.. but was shocked at their fee. I also suspect the local fabricator/installer was adding a cut for themselves as the building itself was designed by one of the largest companies in the industry.

It was nice when I finally got to coordinate with an experienced PEMB engineer instead of the new grads who promised a lot and never delivered.

I understand how the building is made to be the cheapest possible, but in the long run, in my experience all that is happening is pushing the cost into the foundations which are very expensive, especially with shortages in concrete. There was another project where we were asked to provide a conventional steel building design because the PEMB plus crazy foundation was so expensive, in the end we provided a better building with cheaper foundation and overall less cost. I agree with dik that PEMB's are throw-away buildings, especially for anything that isn't a rectangle (which is more and more common).

 

[dik]

Do PEMB manufacturers have guidelines on the design of CFS roof purlins?

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[dik]

SandwichEngine...
Do you have a link for that... there are several, but none seem to be PEMBs. dikcoatesatgmaildotcom. Just want to add it to my address book right now in case...

Just went for a quick tour through DG25... is that ever a 'book'.

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[SandwichEngine]

Dik,

The website to the company I know about is teamskdotcom. I know that they design PEMB but do not fabricate themselves which is what makes me think they do PEMB consulting work. I've never actually spoken to anyone there so that could be untrue.

PEMB companies do have guidelines to the design of the CFS coldform purlins--the AISI code. Just like everyone else. That is, unless they have some proprietary purlin product where they've derived the allowable loading through testing. Normal Cees and Zees though? Same as everyone else.

Aesur,

PEMB guys apparently won't design for anything but metal panels (so we are told)

I've never heard of a company that wouldn't be able to design to different deflection limits but I don't know how everyone works. It's hard for me to understand how a business model that limited could be profitable.

 

[hokie66]

I wonder why the North American market has not made cold formed purlins and girts into a commodity similar to open web steel joists. In Australia, where we don't have commodity OWSJ, we just use load tables for purlins and girts. They are provided by the manufacturers, and are based on a combination of calculation and full scale load testing.

 

[Aesur]

@SandwichEngine - I suspected as much. They said they could design the building frame to limit deflection, but couldn't design the girts for anything other than metal panels, so we ended up with CFS stud wall system spanning vertically. This was a local installer that uses one of the largest companies for the PEMB engineering and fabrication. If this was one project that it happened on I would chalk it up to something unique for them, but this has happened on many projects (all the same local installer however). I learned after the first proposal to exclude that piece (and have add services) or have a separate line item for the wall designs to stay competitive with the other local engineers who also exclude that piece. I will try pushing back harder next time to test the waters.

 

[dik]

Sandwich... thanks; I'll look them up. I use CUFSM for looking into CFS shapes...

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[dik]

Hokie... we used to have tables for OWSJ where they had specific models for load combinations and spans, just like CFS and grating tables now...

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[KootK]

Some thoughts:

1) If you let the frame joints go plastic under gravity, that might alter their availability to resist lateral loads. In the extreme, you might might be down to only one of your two frame ends resisting lateral loads. Load casing stuff would probably prevent things from ever getting quite that bad. You likely don't have the kind of roof diaphragm that would allow you to switch your VLFRS to braced frames only.

2) If there is space and fiscal appetite to do so, you might externally post tension / truss your rafters to offset the added dead loads. This could get analytically complex in a hurry too so I suspect that a fair bit of judgement would be required to make a go of it.

 

[dik]

Thanks Koot... I'm thinking of relaxing the exterior support haunch moments and treating it as a 3 span continuous, with X-bracing, this is because the end supports are sitting on a parapet already and I don't want to attract any added moment, and resulting horizontal load. I briefly looked at seeing how the post-tensioning of the system, but it got way too hairy, too quickly.

I have a meeting with them next week, in the interim, I'll take a gander at the roof purlins and see the best way to address reinforcing of these elements. I was just starting to do an SMath program to deal with purlins with cantilevered ends, using altenating loads of 1/2 (DL and LL) on the cantilevers... to mimic cantilevered condition that's not connected, but with doubled members at the support. I'm almost coming to the conclusion it's time to 'throw away' this PEMB, and that may be my recommendation.

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik