Steady-State Analysis 1

[natedawg88]

I am trying to perform a steady-state analysis on a dam model that I have created in SAP and unable to get results that make any sense at all. I am trying to mimic the excitation of the dam with a shaker but when I put a joint load on it of approximately the same input force in the default units and measure the acceleration at given points I get absurd values back like 7.5 g's. I have already checked the scale factors on the analysis options and it is 1. Also if I change the input force by a factor of 2 the resulting accelerations do not even come close to changing by a factor of two. In fact, in some cases it hardly changes at all. Additionally, I was wondering if the effects of gravity were taken into account in this analysis or if I had to add it in somehow.

Anyway I have been working on this for about 20 hours now and was wondering if anyone had any suggestions at all.

Thanks!!!!

 

[masomenos]

I confess I haven't used the steady-state functionality yet, but it seems to be a harmonic analysis. If you put a joint load on your dam under a particular load case, you then need to create a steady state load function, which applies a multiplier at different frequencies which you define. You didn't mention how you defined your function.

You then need to define/add a steady-state analysis case, in which you select the load (assuming the load case you assigned your point load) and matching it with the function which you defined. It looks like the 'advanced parameters' lets you put in loads at different phase angles.

Also, it looks like instead of applying a function to a point load or area load, you also have the option to apply a steady state function to Accelleration which you would need define your direction (U1, U2, etc) instead of a load, and then apply a scale factor to g based for whatever units you want to use. Digging a little into your inquiry on my own version, I notice that SAP has added automatic accelleration loads for rotational degrees of freedom with the latest version.. I hadn't seen that before in previous versions. I suppose these rotational accellerations would be quite useful in analyzing ship-like movements for pitch, roll, yaw. kinda cool to know.

This is new to me too, so let us know if this advice helps or if you're still getting unusual answers. Also, you can always email your SAP .sdb file to CSI tech support with your comments if you continue to have problems and let the people there help you figure it out.

 

[natedawg88]

Thanks a lot for your suggestions masomenos.

I defined the steady state loading function according to a function that describes force input by the actual shaker that was used on the damn. This is a discontinuous function but SAP2000 seems to have dealt with this just fine when it graphed the function in the steady-state functions formula dialog box.

Some of the methods for analysis do seem very useful as you mentioned but I simply cannot get results out the analysis that make any physical sense. For example applying a load that is never any greater than 20 Kips results in accelerations at joints that are over 3000. And assuming these results are in my base units (Kips, ft, F) this means that I am receiving over 90 g's right? Anyway thanks for trying but I am still receiving nonsensical results.

Also I cannot figure out a way to make sure that gravity is being taken into account. If I add dead to the loads to be analyzed does that mean that I am simply adding a constant gravity force downwards or does that mean that it will be harmonically applied as well?

Thanks again!!!

 

[masomenos]

natedawg, if you're trying to mimic a shaker, why not use SAP's time history function instead of steady-state? That approach makes more sense to me. You can create a time history function and apply it to an accelleration load. Using accelleration instead of point loads (or area load for that matter) automatically applies a time history accelleration at the base of your structure.. SAP knows the masses in your structure, and the accelleration type load automatically accounts for the mass. You need to define a time history function or import one from a file which is applied to your accelleration in one or more directions. If you decide to go that route, you might familiarize yourself with the difference between transient time history analysis and analysis and periodic time history analysis. Periodic time history is kind of like harmonic load analysis in that the specified loads are assumed to repeat indefinitely with a period given by the length of the analysis.

By Defining a load combination, you can add self-weight Dead loads to the reactions from other load cases