Virtual Joist Girders (VJG) as Moment Resisting Joist Girder 1

[STR04]

Has anyone used VJG's as a moment resisting joist girder in RISA 3D? What do you do with the DL moment if this is eliminated because the bottom chord is connected to the column after the DL is installed...well according to SJI there's no DL moment. How does RISA even perform a code check on this member? I don't know the correct way to model these elements when dealing with the effective length, boundary conditions, etc? Does it make sense to also use these as regular bar joists with pinned ends where applicable?

 

[JoshPlumSE]

When you say "DL moment" are you referring the moment from the joists self weight? If so, then just set the density of the material assigned to your joist to something like 0 kips per cubic foot.

Alternatively, you may be talking about something called "staged construction" where the structure is partly built, then load is applied, then the structure is significantly changed, then more load is added.

RISA does not directly handle staged construction. However, you can still use it to get member forces. Just in a more complicated way. First you create two models and apply the different load to the models and solve them to get member forces and deflections for each case. Then you can use the principal of super-position to come up with your final design forces and deflections.

 

[STR04]

DL is referring to the selfweight of the joists, deck, roofing, MEP, etc. How does RISA perform a code check on the VJG's?

 

[JAE]

Josh - I believe that STR04 is talking about cases where the joist is installed, dead loads like roofing, mechanical, etc. are slowly applied during construction - then at the end the bottom chords are attached so that they only take Live, Snow, Wind or Seismic forces. Thus the joist girder is pinned for the dead loads and fixed for others.

 

[STR04]

Well said JAE. Thanks!

 

[JoshPlumSE]

That's the staged construction example. As I said before, this is not addressed directly by RISA-3D.

The member receives load as a simply supported beam (that gives you solution results #1). The load and stresses remain. But, then you change the structure and all the load applied after that point gets applied to a fixed - fixed beam (which gives you solution results #2). The total load in the beam ends up being the super-position of the two solutions.

 

[JAE]

I suppose you could run two models - one with pins and one without and grab the output and paste into spreadsheets and start combining shears, moments, deflections, etc. to get a superimposed solution.
But with second order effects and a redundant system the superposition may become troublesome or confusing.