How to do seismicX and seismicY load combinations in ETABS?

[milkshakelake]

In ETABS, I can make seismicX and seismicY load patterns (say for different R-values). How would I actually make load combinations for these? The way I see it, there will need to be two different ETABS models with different R-values using just one seismic load pattern instead of separating into seismicX and seismicY...which is also fundamentally a different analysis, and doesn't achieve the same thing.

The first issue is that I need to make different ASD and LRFD load combinations with seismic for X and Y direction. We're at about 20 load combinations now. Then we add overstrength combinations. We're at 40. Then add seismic drift. We're at 60. It's barely manageable at this point.

The next step is to do the 45 degrees combinations (0.707EQX + 0.707EQY kind of thing). Now we're at 120 load combinations for seismic only that need to all be checked and managed, and applied for analysis and member design. (I didn't consider the -eccentricity and +eccentricity cases because I assume it'll be handled in the load pattern itself.) It becomes unwieldy and easy to screw something up, like missing one combination or putting the wrong factor somewhere, or even forgetting to designate one of the combinations into an envelope.

Is there something simple that I'm missing? I'd much rather have the typical 20-30 load combinations when using R-values in different directions, but I don't see an easy way to do it.

 

[ggcdn]

Im confused about why different R values would need different models. An R value is just a scaling factor. For instance, you could run EQx and EQy with R=1 and then combine them in combos with appropriate factors, 1/R.

Also maybe it’s a different code but are you sure you need a 45deg case? Usually the directional combo is 100+\-30.

If you are doing member design outside of etabs, then sometimes the easiest thing to do is run all your cases, extract them, and then do your scaling/combos in a spreadsheet.


-JA (working on [link calcs.app]Calcs.app[/url])

 

[milkshakelake]

@ggcdn The problem with putting the R-value into the load combination is that it would be project-specific. You'd have to modify the factors every time, which increases the chance of a mistake.

You're right, it's 100 and 30. I just figured that the number of load combinations would increase by 2 times if I used that (for example, I'd need Ex+0.3Ey, 0.3Ex+Ey, Ex-0.3Ey, etc). So I'd be at around ~150 to 200 load combinations, which is too much. My way is about 7% more conservative, which is fine for making the checking easier.

I don't do member design outside of ETABS for steel design. For concrete, I do, but concrete is a lot easier for seismic because it'll always be same R-value in both directions.


For anyone running into this same issue, I just went ahead and made all the load combinations. It took a full day of work. Triple checking, assigning envelopes, checking the base shear results, doing hand calculations to verify it. The list of load combinations is massive, but it works, and I only have to do it once in my life. So this is solved!

 

[structSU10]

Just curious but ETABs doesnt have auto load combinations? RAM would create the 300 or so load combos based on the assigned code and I can't forsee R's for each direction making that not work correctly.

 

[milkshakelake]

@structSU10 ETABS does have auto load combinations, but they're just not that good. They're named something like Combination1, Combination2...which makes them very hard to check. They also don't make overstrength cases. I usually prefer ETABS over RAM, but they need to learn something from RAM in this regard.

 

[Luceid]

@milkshakelake unfortunately, additional requirements of orthogonal combinations, 100% + 30% combinations, 45 degree combinations, overstrength requirements - etc.etc.etc. tend to just exponentially increase the number of load combinations and make it very difficult to track what is really going on (especially when looking at enveloped results).

If you have to manually enter it - definitely use Excel and the interactive database. Defining load combinations in ETABS repeatedly is somehow more mind-numbingly slow.

 

[milkshakelake]

@Luceid Got it, thanks for the tip! I had to enter them manually because I use ETABS 2016, before they had the interactive database. I have ETABS 21 but there's a bug with steel design that they haven't fixed, and I had a support ticket for years. Maybe when they finally fix it, I'll do the 100/30 combinations with Excel.

 

[Luceid]

Out of curiosity, what is the steel design bug?

 

[milkshakelake]

Deflection not designed properly under AISC code. Analysis results are correct, but design only considers strength, not deflection. It shows the correct result when you right click on beams one by one, which leaves lots of room for mistakes. They verified it's a problem, but just didn't fix it for years. I guess I don't have enough licenses to justify them spending time on it.