PEMB - How to Control Displacement 3

[TAK19]

Good Morning,

I am designing the Pre Engineered Metal Building. I am using a Etabs 2016 to perform the design.
I have noticed that the displacements value I am getting are too high with respect to L/360,L/400 requirements.(Owner require the exterior walls to have deflection limits of L/360) I need help in controlling this high displacements of the joints.
I know the common methods used in the industry to control the lateral movement but my problem with them are as follow:
1. Cable or Rod Bracings - Cant use as I have lots of windows in most of the walls
2. Portal Frame - When I added a portal frame to the rigid frame the software gives me warning that the columns are overlapping and also its not reducing the displacement noticibly.
3. Wind Column - Don't know how to add them in Etabs.

Questions :
1. How can I fixed the problem in item number 2
2. I have tried all known methods to make the rod bracings tension only but I am unable to.(in Etabs)
3. The largest displacement value I am getting from the software is 7 inches .I am sure I am missing something important because this building is only 4000 sqft with mono slope roof eave ht of 18 ft.

I would Highly appreciate if you all experts can guide me on this.

 

[phamENG]

ETABS doesn't seem like a great choice to design a PEMB. If you're working for a PEMB manufacturer, you should be using specialty software - you're unlikely to realize the economy the big boys get if you don't use the the right tools.

First of all, meeting typical steel building deflection criteria with a PEMB is challenging. It can be done, but you end up losing most of the economy they're known for. PEMBs shine when you have an industrial/agricultural use and a drift limit of H/60. Much beyond that and a standard steel frame is going to get you there just as well.

I haven't used ETABS in a couple of years, but to fix #2 I'd think you'd need to create a node offset from your current columns. I'd put it roughly the distance away that the portal frame columns would actually be placed.

A wind column is just a column at the ends of the building with pin connections at top and bottom to take girt loading down to the foundation and up to the roof 'diaphragm'

 

[Aesur]

I have very limited experience with ETABS (used twice years ago), however I recently worked with a PEMB manufacturer with a similar design requirement due to finish requirements from the local jurisdiction as well as the geometry of the buildings requiring the PEMB's to be connected by a non-PEMB building required seismic separation considerations.

You mention the deflection limits, if we are talking about finishes - I assume Stucco for L/360, can you use 2L/360 in lieu of L/360? This will give you better results as the way I see it is the members spanning from the foundation to the eave shall be limited to that deflection criteria as they directly support the stucco and the overall "drift" can be seen as a cantilevered structure off the ground, hence the 2L/360. This would need to be verified as acceptable with the EOR on your project.

Normally when I see PEMB calcs, they run the analysis in 2D, not sure I have ever seen a 3D analysis. This consists of designing the individual moment frames, portal frames and braced frames for the loading. Maybe a 2D approach would work for your lateral?

 

[Brad805]

I am curious why you want to design all of the PEMB frames and secondary components? The big pre-eng companies have a very sophisticated system that works from pricing to production. It will be a very large undertaking to be competitive, and at the end of the day someone must build what your customer needs. If you customer goes price shopping I doubt many will look at your design beyond the specifications. The normal process for the EOR is to provide specs and review their calc package. If you are the PEMB supplier, I think you are bringing a knife to a gun fight with ETABS.

 

[canwesteng]

I've seen PEMB use off the shelf big name software before (not ETABS), I don't think the analysis is quite as sophisticated as you are all making it out to be. Really your big need is the standard calculation sheets, and the knowledge of how to really shave meat off the bone like they do. You'll never be competitive without their drafting/production system though, it's a well oiled machine.

 

[Brad805]

A lot of the suppliers I have worked with use a lot of FE design for the connection design. It might not be all that sophisticated, but it is a complete loop from the re-seller to the production floor. I too have bumped into some that are more old school, but it is all about pounds of steel. The days of charts and tables seem largely done for most now. One could come close to the connection design with Idea Statica, but egad, the hours for something that is unlikely to get used.

 

[JedClampett]

Yeah, the PEMB guys (and girls) know how to squeeze every last psi out of their designs. They're not over the code allowables, but there's no daylight between them either.

 

[TAK19]

First of all I want to thank all of you for your valuable inputs.

I want to clarify that I am working for the PEMB fabricator and material will be purchased by the Owner. Therefore this is not just for quoting purpose but instead its a done deal .

I have tried to explain the Owner that for meeting the unrealistic requirement of L/360 wall deflection will defeat the purpose of the economical design but he is not willing to make any exceptions.(I don't know why)
PhamENG .I will try to do the way you explained.

Aesur: You are correct the Owner wants to put stucco directly onto the Wall R-Panel .I have explained them that this is not a feasible option as expansion and contraction limits of metal is not that equal to stucco and this will eventually crack the stucco but they still not willing to make any exception on the L/360 requirement, There is no EOR as I am also designing the foundation of the same building. Can you please advice from which code book you've found this 2L/360 requirement (it seems sensible and logical but would require some documentation to prove to the Owner).

Brad:I am designing the rigid frame in Etabs and I have manually designing the secondary members such as grit and purlins.

Can you please suggest any software which would be more compatible for the PEMB design ?

 

[Aesur]

@TAK19, take a look at 2018 IBC, Chapter 16, Table 1604.3 footnote i. 2018 IBC

Also, take a look at specifically what stucco they are using, for instance the last time I looked at Western 1 kote, their ICC report allows for L/240 instead of L/360.

The few times I have designed tapered members, of varying thickness, widths, etc, similar to PEMB members, I have used RISA 3D software.

 

[Brad805]

"Can you please suggest any software which would be more compatible for the PEMB design?" That will depend on how far your company wants to go with automation. The beauty of pre-eng is being able to automate the process from start to finish. If you want to go back and forth between general frame analysis software, spreadsheets, and other software that will work. It all depends on your company goals and how much they want to invest into your end of the process. As for a specific piece of software, I would need to look back thru submissions to see if any list what they use behind the scenes. I have listened to a few different Butler guys on the Tekla forum, but I do not recall if they mentioned the analysis software.

 

[phamENG]

It's also important to make sure you're not confusing deflection with drift. The H/60 I quoted earlier is drift, not deflection - a lot of people mix the two and the primary issue with PEMBs is usually the drift, so I jumped to the conclusion that we're talking about drift.

For information on drift (and deflection, for that matter), take a look at AISC Design Guide 3: Serviceability Considerations for Steel Buildings. If you have an empty warehouse with some hanging lights, maybe a big HVAC duct, and a sprinkler pipe, H/60 is all you need. If you have interior finishes like gypsum wall board, H/400 to H/500 is a better number to shoot for. Deflection of girts should be controlled by the table Aesur linked to. L/360 isn't going to inhibit an economical design if it's an appropriate criteria, which it certainly can be.

Stucco directly on a metal wall panel? That doesn't seem like a good idea, but maybe there's a system to do that?

 

[TAK19]

PhamENG,

I spoke to one of structure Engineer and he told me same thing that you are pointing at.

The software has provided the max drifts and avg drift value for each loads case and I am comparing them with the allowable story drift given in Table 12.12-1 of ASCE 7- 16 .This allow me to have 0.02 H value which comes out to be 4.32 inch(0.02*18*12).

Also Structural Engineer told me that deflection criteria is used for the gravity loads and all the limits in IBC are for the gravity load not lateral.(please provide your thoughts on this)

This Structural Engineer also advice me that I should design this building as Ordinary Concentrically Braced Frame.Can you all please confirm if I should do that or advice otherwise.

Aesur : Thank you
Brad805 : I have used Tekla but not professionally but I have found Etabs much simpler and user friendly

 

[phamENG]

Well... are you a structural engineer? If the answer is no, then I'm not sure why you're driving etabs.

The drift limits you're referencing are seismic. Important, yes, but that's not typically a serviceability criteria. Serviceability criteria beyond those listed in the IBC are not codified. It's up to the engineer to design a functional structure that doesn't crack up the interior. Read the AISC design guide.

Those criteria are for gravity loads on beams, etc. as well as lateral loads on wall studs (or girts in your case). They are not directly applicable to drift.

No. I've never seen a PEMB that qualified as a concentric braced frame. You could have an odd one there, but I would be surprised. Since I know virtually nothing about your building, though, I can't say for sure.

If you are not a licensed professional engineer with expertise in this kind of structural design, I would advise you to stop and discuss this with the engineer who will be responsible for the design.

 

[TAK19]

With Structural Engineer I mean a friend who is also a Structure Engineer with more experience .

I had told him the same thing that drift is related to Seismic whereas my structure is govern by Wind loads.

I have PEMB designed as Ordinary Moment Frame.

But still the question is does joint displacement values will be used for the serviceability criteria check or something else.(this is a question for someone who understands Etabs software)

I am licensed P.E

 

[phamENG]

Wind load drift is not codified. Again, read the design guide I linked to earlier and the serviceability appendix in ASCE 7. There's some good information there about scaling wind loads and criteria for occupied structures.

As for what part of the model to look at...that should be pretty self explanatory. If you want to limit the lateral deflection to X inches, look at the deflected shape and see how far it went. If that's a node, or a member, or whatever then that's the number to look at. We don't know what your structure looks like and can't tell you where to look for the information. Though in a typical PEMB it would be the node at the eaves.

Ordinary Moment Frame? But you said it's "govern by Wind loads." Why do even that small amount of seismic detailing? Ordinary moment frames require design and detailing per AISC 341, the Seismic Design Manual, even if you're in SDC B.

 

[Brad805]

"I have used Tekla but not professionally but I have found Etabs much simpler and user friendly" Tekla does do any of the analysis, but can be used very efficiently from start to finish. The techs can start the job per your typical standards, produce prelim costing, and when approved send it to you for design. From there you send it to one package for frame analysis, return the connection forces for design, and then move the project down the line to others to complete the loop. Many of those responding are not overly interested in many of the steps a fabricator may be. The large companies boil the process into a much neater work flow to minimize errors and hand calcs. Anyway, I am off topic. I am not in the US, so I will leave it to others to comment about the specific ASCE q's.

 

[dold]

re: braced frame classification. There are provisions within 341 that allow you to design a OCBF with tension only members. There are some decent penalties though. I believe pretty much....everything has to be designed with overstrength applied.

But if I were you, and in SDC B, I'd be using R=3 Steel not detailed. For the moment frames and braces.

And Pham is right. There are no code requirements for wind drift. The commentary of ASCE 7 says something to the effect of "it is up to the engineer and the owner of the building to determine acceptable drift limits for wind events...". You choose a reduced windspeed based on whatever MRI you think is appropriate and that is your wind serviceability case. Surprisingly enough, even in very high seismic regions, wind serviceability typically controls member (column/beam) sizes, especially if you are using moment frames (for multistory buildings anyway).

 

[TAK19]

The building is SDC A.

The issue is , I am uncertain about the values I am getting from Etabs .There are joint displacement values which are confusing if they can be used to check the serviceability or not.

Let me rephrase my initial question in detail.
The software shows joint displacements values (in x and y direction) under each load case and load combination at the top 4 corners of the buildings .Should these displacements values be checked for serviceability or not ?
I am checking the secondary frames(such as girt and purlins) deflections manually against the wind load(for girts) and gravity load(for purlins). In my opinion the deflection of the primary frame is already checked by the software therefore no need to recheck them but when I look into the software design preferences there are no limits saved for the lateral force deflection (such as wind or seismic) which bring me to this question that should the joints displacement values are matter of concerns or not.

 

[dold]

Well...yeah. Why not? Do you not trust the results? That's like 50% of the reason that you took the time to make a model in the first place, right? Typically there arent 'limits' that the program will check against (except RISA will do this if you have defined a diaphragm and drift definitions). But it takes 5 seconds to look at the nodal displacement and compare it to H/400 or 0.02H, right? Software does a lot of the work, but you can't expect it to just spit out a green checkmark and you're done.

 

[phamENG]

dold is right, but I'll take it a step further...what load combinations? Is it the same one you used to check the strength of the frame? If so, then not unless it meets the specification you're trying to meet. 0.9D+1.0W (with wind based on the 700 year MRI for a risk category 2 building) is not a reasonable check for drift. Designing for that load, when combined with considerations for material and workmanship defects, means your building has a statistical chance of collapse roughly once every 3,000 years. While that is an appropriate threshold of risk for collapse that could kill people, it is not an acceptable threshold for cracked drywall that you can repair with a $5 container of spackling. You should use serviceability combinations. Again, these are not codified. You have to find them in industry documents (such as those already linked), code commentaries, experience working with other engineers, etc.