DAM, software implementation and Deflection/Drift/Periods

[bookowski]

When using software (Risa, Etabs etc) to implement DAM the software is reducing member stiffness, two problems I have:
- this likely results in a longer period calculation and therefore not conservative for eq loads
- in my area/type of work it's almost always wind drift under service level wind that controls, the reduced stiffness is overestimating service drifts

Two models? Or one model first with no dam to get periods and establish a system that works for drift and then turn on dam for design.... and then remember every time you fiddle with the model to know if dam is off or on?

 

[structSU10]

In a recent seminar I had from Bentley, in RAM SS the would run the model with no stiffness reduction to determine the period and check drift. They would then take the building period from the full stiffness model, input if manually for the automatic force generation, and run the model with reduced stiffness. It still requires you to reanalyze with full stiffness to check drift as you iterate through the design and such, but you can at least force it to use the period with full stiffness to develop the forces.

 

[JoshPlumSE]

I can't speak to others, but when RISA does the Dynamic solution, we do NOT use a reduced stiffness. So, the mode shapes and natural periods will be the same as if you did not use the DA method.

For deflection and drift you would have a few choices:
1) Deal with the larger deflections. Or, reduce them by 20% to approximately account for the stiffness reduction.
2) Create a load combinations that does NOT include HR steel for design and use it for deflection. Because RISA-3D only performs the DA method stiffness reductions for load combinations that are used for HR steel design.
3) Re-run the load combinations for the same LC's that you used for design, but with the stiffness reduction setting turned off.

 

[mike20793]

Agreed with the earlier posts. I do it exactly as structSU10 explained. In RAM, it's as easy as unchecking the box to reduce member stiffness. I will note that I do not like that RAM assigns the same tau factor for every member. If you require a tau less than one, then it applies it to every member, so it's often better to increase the notional loads instead of reducing tau until you get something that works.

 

[bookowski]

1) Deal with the larger deflections. Or, reduce them by 20% to approximately account for the stiffness reduction.

Maybe for something small, but for a decent sized building it's already a fight to get enough lateral to control drifts - I don't want to give away 20%

2) Create a load combinations that does NOT include HR steel for design and use it for deflection. Because RISA-3D only performs the DA method stiffness reductions for load combinations that are used for HR steel design.

Interesting option for Risa. There are a lot of drift cases already though, this would double that.

As a side question: risa generates a stiffness matrix for each load combo? What if p-delta isn't chosen? Why not formulate matrix once and then play with that for the combos? I guess this is why you can't change load combos in Risa without re running the analysis?

3) Re-run the load combinations for the same LC's that you used for design, but with the stiffness reduction setting turned off.

Yeah, seems like that's the Ram users are saying they do. That was the gist of my question, if people are keeping two models or toggling on/off. Both options seem prone to errors/bookkeeping issues.


Maybe I'm just really lazy but I find DAM to be a real pain the *ss. There are already a million load combos with all the different wind directions, eq, and then combined in the various strength combos. Add in the notional loads and it gets crazy. In the type of structures that I do I can't see the notional load actually ever making a difference.

 

[JoshPlumSE]

As a side question: risa generates a stiffness matrix for each load combo? What if p-delta isn't chosen? Why not formulate matrix once and then play with that for the combos? I guess this is why you can't change load combos in Risa without re running the analysis?

RISA-3D will re-build the stiffness matrix for every LC if it needs to. Using one stiffness matrix is fine for purely linear elastic solutions. But, if you allow non-linear items (tension only members, compression only springs, DA Stiffness reduction, variable levels of concrete cracking) then you have to change the stiffness matrix.

If it doesn't need to (say Tau = 1.0 for all members and you don't have any non-linear elements) then we use the same stiffness matrix for multiple LC's. However, we still solve each LC separately because the P-Delta effect can be different for each load combination.

Programs (like RISAFloor) that don't have to worry about these various types of non-linearity will build the matrix once and solve each load category (like DL or LL). Then use superposition to build results for the various load combinations.

 

[KootK]

Might be something useful here if you don't already have this: Link. I agree, anything that makes for more load cases makes KootK sad.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[bookowski]

Thanks for the link.

My thoughts about this line from the eng-journal paper about using a tau of 1 and adding an additional .001 onto the notional: "this notional load will typically be so much smaller than seismic loads that it may reasonably be neglected when seismic loads are present". For most building isn't even the .002 insignificant? I guess if you're doing something really slender, or very heavy gravity loads with high drifts this may make a difference but I can't imagine anything that I work on where a .003 notional load even moves the needle. Seems to me that it would be pretty safe to basically ignore the notional loads. All of the example have a 2 bay frame - that's nice but what about when you have thousands of members and all the various asce wind directions etc.

I'm becoming more and more convinced that the codes in their current form should be tossed in the trash and a new system implemented. That's a rant for another day though.

 

[JoshPlumSE]

Bookowski -

You've got that basically right. The big concern and need for notional loads is for structures that are really slender, have very high gravity loads in conjunction with low lateral loads.

One example comes from industrial. Where you are in a low seismic zone, and you have a "hydro test" load that is many times greater than the gravity load you see during operation. This could approach the buckling load of the structure with zero lateral load (since you don't often assume wind or seismic load during your controlled tests). In those cases, notional loads are very, very important.

Also, internal mezzanines in low seismic regions. They have significant gravity loads and no real lateral loads.