Another Structural Software Thread

[lexeng18]

Hi all, my team is looking for a software that has the following (apparently hard to find) qualities:
[ol 1]
[li]The ability to capture flexible diaphragm behavior correctly. I say "correctly" simply because the software I've demo'd so far that claims to have this feature does not dump the diaphragm shear into chords/collectors as axial loads. While the overall lateral force distributes correctly, the load just materializes into the bracing elements. I'd like to be able to properly design chords/collectors.[/li]
[li]The ability for wind loads to be defined in a way that can capture wind column behavior correctly. For example, RAM Structural System only lets you use the wind generation feature to apply wind as if the cladding spans vertically from diaphragm to diaphragm. Most of our work is industrial buildings where girts span horizontally to columns, so the columns should be loaded in bending from the horizontally spanning cladding, and the columns should span vertically diaphragm to diaphragm.[/li]
[li]Be able to have features #1 and #2 above without weird, hacky workarounds that create other problems.[/li]
[/ol]

It seems that in 2024 it should be possible for such a software to exist, but currently with RAM Elements, RAM SS, STAAD its not possible. I tried RISA3d and once I figured out that they lock down the flexible diaphragm option behind needing also RISA Floor I was turned off to the entire ecosystem.

I've yet to try VisualAnalysis, ETABS, or any of the very new to market ones (SkyCiv? ClearCalcs?). I was hoping someone here would have some experience with other software to help me in my search!

The reason for this request is simple. While we currently have RAM Elements and get by just fine by having separate gravity/lateral models plus some hand calcs, more and more we are being asked for calculation packages as part of a project. It would be really nice to have a combined, non-clunky, working analysis model that behaves true to its design.

 

[SE2607]

I'm very interested in this thread, intellectually, because I believe it is difficult to incorporate chord and collector behavior when using a rigid or semi-rigid diaphragm. It seems to me that chord and collector behavior is more clearly evident in flexible diaphragms only. Yes, requiring a RISA Floor license to do lateral analysis is frustrating, but it might be one of your few options.

 

[JoshPlumSE]

I don't know what software would be best for this desired outcome. But, I will share some experience I have on this general subject.

1) Some of the smaller (simpler?) structures that we often design by hand can be VERY challenging to design using analysis software. I'm very specifically talking about structures with wood shear walls or wood diaphragms.

2) With wood diaphragms, I've often seen them not modeled at all.... Which can lead to instabilities that don't really exist. And, that happens even if the software got the load distribution correct.

3) Even assuming you get the load distribution correct, your program will often neglect to check the diaphragm itself. You might get the loads into the right shear lines and collectors. But, the diaphragm then needs to be designed by hand.

4) You always have to be careful of the assumptions an analysis package makes about how the wind load gets into the main lateral force resisting system. Though I think most programs can do what you are asking related to girts spanning between columns. It's just not necessarily automated as much as we like to see these days.

 

[lexeng18]

I don't mind the software not being able to design the diaphragm itself, I just want the behavior accurately captured on the analysis side. With metal deck diaphragms like I'm mostly referring to, their design is frequently dictated by manufacturer's specifications anyways and so that part of the design is better suited to specialized/dedicated software like Hilti's Diaphragm software for example.

 

[phamENG]

VA and Risa3D will do number two, though maybe not automatically. You'll likely need to define the loads themselves manually, but they can be done as uniform or variable area loads that the software will distribute to modeled members based on your input. I've used it for wind girt/wind column style buildings successfully in the past.

I haven't encountered on that does number 1. I've debated trying RisaFloor, but haven't had a justifiable need for it yet.

 

[lexeng18]

Just played with Risa3D + RisaFloor for the afternoon and I hate it. The back and forth between them is unintuitive and just outright sucks. Better yet, after finally getting a model setup correct I found this in the help docs:

Axial Load Limitation

Members in the plane of the diaphragm will have no axial load attributed to them. This is because the internal rigid links that are created to achieve rigid behavior take the entire load. In these cases, the axial loads in these members will need to be considered outside of the program.

https://risa.com/risahelp/risa3d/Content/3D_2D_Only_Topics/Diaphragms.htm#Flexible_Diaphragms

So, it's out now too.

 

[Celt83]

That limitation should only apply to rigid diaphragms, the same issue is present in all software for rigid diaphragms as they enforce consistent deformation with the center or rigidty to every node on the level.
In RAM Structural System for example you always need to release one of the frame nodes from the diaphragm so that axial deformation develops in the beams.

 

[Celt83]

For 1 the closest I’ve seen function properly for flexible diaphragms is woodworks shear wall program but only for rectangular simple structures.

For 2 Robot may be a satisfactory option once you go through the effort of calibrating the wind tunnel simulation. Full disclosure the interface is god awful and Robot does nothing to assist the end user in modeling but if you can tolerate it the underlying analysis engine is pretty decent.

 

[JoshPlumSE]

Members in the plane of the diaphragm will have no axial load attributed to them.

That's pretty much true of any program that uses a "rigid" diaphragm. This type of feature was originally intended to account for concrete slabs where the in-plane diaphragm stiffness was so much greater than any frame stiffness that it was a reasonable assumption.

I'd suggest that maybe the best thing to do in cases where you want to get the best "analysis results" for a flexible (or nearly flexible diaphragm) is to use a semi-rigid (i.e. semi-flexible) diaphragm.

The key is that with that type of diaphragm is that YOU (the engineer) controls the thickness and stiffness of the diaphragm material. The drawback (over a rigid diaphragm) is that it is more computationally intensive.

The big benefit for flexible diaphragm models is that you shouldn't get a ton of lateral instabilities in your model... because all the lateral frames are tied together loosely with the mostly flexible diaphragm.

I know ETABS has a semi-rigid diaphragm capability. It's been a few years, but I could swear RISAFloor has that capability. Though they may call it something different.

Caveat to prevent one particular company from suing me: I currently work for CSI (the company that makes ETABS) and I used to work for RISA. In fact, I was there for 16 years and was a Vice President before they were taken over by the German investment company. So, I am not an unbiased observer and you should take any comments as potentially biased.

 

[Aesur]

Good luck, the software you speak of does not exist.

 

[SE2607]

I know ETABS has a semi-rigid diaphragm capability.

JoshPlumSE: Does ETABS have the capability of modeling a wood diaphragm (semi-rigid) attached to wood shear walls?

 

[JedClampett]

I'm currently having this argument with one of my co-workers. We can kinda sorta model metal diaphragm stiffnesses as lateral load carrying members. It involves begging a manufacturer for some test or analytical data, tricking the analysis program into treating it like a series of springs and placing loads on it. Then checking the defections, back checking, etc.
There's a lot of approximations. And what happens if they buy another vendor's deck? It's probably the same, but will the supplier agree with that, especially if there's a problem? Or they use a different size weld? Or they RFI to shot in pins? Or? Or?
To me it's just not worth refining these things to an unrealistic level of accuracy.

 

[271828]

SAP2000 can do #1 well. Define the shell stiffness to have the right in-plane shear stiffness from the G' parameter, along with the other types of stiffness.

 

[STpipe]

271828,

Wouldn't that make it a semi-rigid diaphragm in that case as opposed to a flexible diaphragm?

 

[Flotsam7018]

Yeah, I agree with the others here that I'm not aware of a software package that can do both well.

My current approach is that for a relatively simple structure I use Woodworks Shearwalls, similar to Celt83, for a flexible diaphragm analysis. In terms of output and efficiency I'm not aware of anything else which can compete. It's likely mostly useful for residential and small commercial jobs only. You also have to know it's limitations and when you need to extract the output to address certain details like transfer diaphragms, etc.

For something more complicated, I use RISA and a semi-rigid approach. You do not necessarily need RISAFloor for this though, you can model the diaphragm with the appropriate properties. This approach is way more time consuming, but for the right project is warranted.

 

[271828]

Wouldn't that make it a semi-rigid diaphragm in that case as opposed to a flexible diaphragm?

Yes! Good point.

 

[JoshPlumSE]

JoshPlumSE: Does ETABS have the capability of modeling a wood diaphragm (semi-rigid) attached to wood shear walls?

Modeling? Sure. You won't get code / design checks though. ETABS will certainly be more geared towards Steel and Concrete structures than wood.

It involves begging a manufacturer for some test or analytical data, tricking the analysis program into treating it like a series of springs and placing loads on it. Then checking the defections, back checking, etc.

It's been awhile since I really tried to model a metal deck. But, I didn't think it was really this difficult.... I remember there being design guide for modeling metal decking. Maybe it was this one from SDI (Steel Deck Institute).

https://sdi.org/publications/public...aphragm-design-manual-fourth-edition-no-ddm4/

If it is, then it has rules related to the G value you use based on the thickness of the deck.... Also, I think that G value gets adjusted based on the type of connections and distance between them. If I remember correctly it felt very similar to the G' adjustment factor that the NDS uses to account for fastener spacing.

Wouldn't that make it a semi-rigid diaphragm in that case as opposed to a flexible diaphragm?

The terms can be a bit mis-leading. A semi-rigid diaphragm can achieve essentially the same behavior as a rigid diaphragm (if it is much more rigid than the lateral frames it connects to), or it can mimic essentially flexible diaphragm behavior (if it is much more flexible than the frames / walls it connects to). Finally, the semi-rigid behavior can give you an in-between behavior as well based on the relative stiffness of the various elements.

The key here is that you (i.e. the user) have control. If you enter in reasonable stiffness properties for a steel (or wood) deck. Then you will see the behavior based on relative stiffness. But, if you desire to see more flexible (or more rigid) behavior of the diaphragm, then all you need to do is adjust the stiffness of the semi-rigid diaphragm.

If you've got some free time, it can actually be a fun exercise... to test your traditional assumptions to see whether what you think is a rigid or flexible diaphragm actually behaves that way.

 

[STpipe]

JoshPlumSE,

My assumption reading the OP's post was they were looking for something that does the automatic calculation for the distribution of lateral loads based on the flexible diaphragm assumption. I think typical materials that are assumed to be flexible (wood or steel deck) if modeled with the appropriate stiffness values would typically behave somewhere between the flexible and rigid assumptions which at least how I interpreted, would defeat the purpose for what the OP is looking for.

As you mention, you could easily tweak the stiffness properties to get closer to the flexible behaviour, but it adds increased computational time to run the model since the shells need to be meshed. The OP might be looking for something simpler, similar to how rigid diaphragms can be modeled with constraints to save on run time.

 

[SandwichEngine]

Tekla Structural Designer does #2 easily. Haven't needed number 1.

 

[JoshPlumSE]

STpipe -

I understand. My point was merely that the Semi-Rigid diaphragm approach is a good solution for diaphragm behavior. You either:
a) Get essentially the flexible diaphragm behavior like you were looking for.
b) Get something that could be a bit more realistic than the purely flexible diaphragm assumption you were assuming.

If I remember correctly, that SDI design guide wasn't about assuming purely flexible diaphragm behavior. Rather it was more about calculating a reasonable stiffness / deflection for the diaphragm to assess whether it should be considered flexible (or rigid) based on ASCE or IBC assumptions.