Risa-3D V15.0 - Portal Frame - Direct Analysis Method

[LSPSCAT]

I have read through many of the threads from 2007 to about 2013 regarding the implementation of the Dynamic Analysis Method and Risa 3D. I have done several example problems on this topic and would like to finalize my understanding of the current state of the software.

I have an application more related to mechanical/industrial design requirements than a typical building frame so I would like to verify that I am understanding the final moments from the frame analysis correctly as presented by the software. I also want to verify that it is not necessary to introduce any additional notional load if a lateral load is present. As this is not part of a building, there is no true gravity load on the frame, so the use of a notional load would not be required.

The structure has a portal frame that has both both distributed loading across the main horizontal girder and a lateral loading applied. In the software, using the settings for AISC 14th Edition, allowing it to adjust the stiffness, and including P-Delta effects is it correct that the moments I obtain are the correct moments to use in an interaction equation or are there additional manipulations to these values. My conclusion at this point is that the values obtained using the above settings are the correct moments used for the design of the member.

For my investigation I started by attempting to validate a simple hand calculation of the frame loading:

First Iteration:
- Moment Frame Force Calculation - Exclude Direct Analysis Method - No P-Delta / No Stiffness Adjustment
- Software Setting - P-Delta effects turned off
- Software Setting - Adjust stiffness - Set to "NO"
- Columns model with only two nodes at the ends
Result - Moments in frame are exactly as anticipated from a hand calculation

I then progressively induced the P-Delta Effects, P- Little Delta, and Adjusted the Stiffness; the result was the moments increased in the frame as I would have expected.

Second Iteration:
- Moment Frame Force Calculation - Direct Analysis Method
- Software Setting - P-Delta effects turned "ON"
- Software Setting - Adjust stiffness - Set to "Yes - Iterative"
Columns model with several nodes along the length to introduce P-Little Delta
Result - Moments increase in frame, as expected considering P-Delta Effects

Now that I have a frame in the program with moments computed based on the effects of P-Delta, P-Little Delta, and the adjusted stiffness is that suitable to use those in the interaction equations "Mu" ? (This is in comparison to possibly a hand calculation where I might use moment amplification with B1 and B2 to compute the factored moment Mu.)

(For comparisons sake I had also run through breaking the load into a non-sway and sway frame and using moment magnifiers to obtain Mu, the final factored moment using the magnifiers was 15% greater than that obtained utilizing the software.)

 

[JoshPlumSE]

LSPSCAT -

Just to be clear, the notional loads are not there to represent anything related to gravity loads. They are there to create an "initial deformation" of the structure so that a geometrically non-linear analysis (like a P-Delta analysis) can capture the elastic buckling of the structure.

With building type structures, this is really easy because the buckling will almost always be a simple lateral deformation that can easily be captured by notional loads.

For non-building structures, the concept could be trickier to capture and some engineering judgment would be needed. What you're really trying to do is deform the structure slightly in the generalized shape of the most probable buckling mode.

In your case (portal frames), I would guess that these would be similar to the notional loads that you have for building structures. But, the magnitudes may be up to your engineering judgment. Or, may not be needed because if you have significant lateral loads of your own.

The structure has a portal frame that has both both distributed loading across the main horizontal girder and a lateral loading applied. In the software, using the settings for AISC 14th Edition, allowing it to adjust the stiffness, and including P-Delta effects is it correct that the moments I obtain are the correct moments to use in an interaction equation or are there additional manipulations to these values. My conclusion at this point is that the values obtained using the above settings are the correct moments used for the design of the member.

Yes, I would say that these moments are the most correct moments to use for the design of the member. Especially since you go on to say that you have added in the P-Little Delta effect, presumably by adding joints along the length of your compression members.

 

[LSPSCAT]

Josh,

Thanks for the information. Correct, I probably was not clear, I was referencing a load case without any lateral force component. Perhaps I should have stated I do not have a load case that does not include some form of lateral load.

Since all my load cases include some form of lateral loading then I do not need to perform an additional case that would include a notional load. My lateral loads provide the "initial deformation" to get things going.

Correct, I have split the members along the length by adding nodes and the "split members" dialog in order to capture the P-Little Delta. I have retained the property to "Use Physical Member". As I stated, it was informative to walk through each model and individual adjust the settings to see the moments increase.

Thanks