Girder Truss Boundry Condition

[pioneer09]

Have a question with regards to an existing steel girder truss that will be modified; portion removed and column added at approximate mid-span. Both the top and bottom chords are attached with double angles to steel columns; see attached. It is my assumption that the connections should be modeled as pinned connections at the bottom chord and roller connections at the top chord; this would ultimately allow only the connections to transfer vertical load as the horizontal axial force would be in equilibrium (Fx=0 due to equal and opposite forces) at the bottom chord support. Does this seem reasonable?

Secondly, with this assumptions, the chord forces change between compression and tension at the very ends of the chords when I model only the "Truss Section" and "Truss Section with Columns" as seen in the attached printout. Any thoughts on this as to why this is occurring?

 

[mike20793]

I would model the columns and pin everything. A vertical roller at the top chord isn't quite accurate because the column prevents it from moving horizontally. The column will rotate some in bending from the couple from the chords, so you will want to capture that in your model.

 

[Once20036]

By assigning a pin at the bottom chord and a roller at the top chord, you're creating a fixed condition that cannot rotate - hence compression in the bottom and tension at the top, or negative moment.
I suspect that you will actually get some rotation, based on the flexibility of the column. This may or may not eliminate the negative moment at the support depending on the relative stiffness between the truss and the column. I suspect that your truss is much more rigid and the negative moment will decrease, however, you may introduce flexural issues with your column.
A more appropriate model should include the column.

 

[KootK]

1) I think that the main thing requiring modification here is the use of pin-pin fixity at the bottom chord supports in your truss only models. Those would be much more accurately modeled as pin-roller. Those 40K/80K arching action thrusts will never see the light of day.

2) Looking at your column base reactions, I suspect that you have no more that 5k worth of compression being introduced into the bottom chord by the column connections in your models that include the column. Since the horizontal force in your end compression diagonals is going to be upwards of 30k, I don't think that you'll see any compression in your bottom chord at all.

3) If you'll see no bottom chord compression then I think that a case could be made for not bothering to model the columns. And, if this was 1950, that's exactly what I'd do. Nowadays, with software, I'd just model the columns too. The extra effort is of no practical consequence and it's fun to know stuff about your structure.

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.