Deck modeling with SAP 2000 1

[koo]

Hi,sap people,

I just started modeling a steel girder composite deck bridge in sap 7.12v. Say simple span, 4 girders 60ft long with 6ft spacing, 8 inch concrete deck. I saw posts here discussing modeling the deck with shell element, but I read the sap reference manual and it says that the lane load is applied to frame elements directly, not to shell and cautions using of shell element. I would like to know how would you apply a lane load to shell element, and how would you attach the shell to the girder frame elements? Is this more accurate? what about the confliction with the ref. manual? Is there a good sap book with comprehensive samples other than those on CSI's website? Thanks for answering.

Koo

 

[Qshake]

What you really need to do is research the literature on this subject. ASCE offers many good journals that offer different persepectives on modelling decks for slab on girder bridges. From those general type of dicussions or specific if they are you can conclude what you will be able to do with SAP.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[VoyageofDiscovery]

Hi koo,

I haven't used SAP for bridges, however pt. load moving loads and lane loads can be applied to FE. Use dummy members of relatively small inertia to the nodes of shells along the top of girder webs. Distribute the truck and lane loads to the girders and that's it. In most programs the lane load can alternatively be applied to the deck shell as a pressure load.

Perhaps SAP technical service can help you out as well.

Regards

VOD

 

[zliazmei]

Sorry KOO,

But even using dummy elements SAP 7.12 will not give you results from moving loads in a shell elements.
It is posible in the newest versions 8.2.* and above

You will need to use grilage models.
Zmei

 

[koo]

Zmei/vod/qshake,

Thank you for your replies. I tried to assingn shell elements between frame members. They did increase the stiffness (less deflection) of the bridge, but sap placed these elements not on top but middle of frame elements (webs). I've never done a grilage, but assume it's about tranforming deck to equivalent transverse frame members between girders, right? Can you recommend some literature? How would you transform the section (to steel? 1/8?), and where to put the transformed deck frame (same plane of girders or up a layer than connect girder layer to deck layer using dummy member)?

Koo

 

[edward1]

koo:
I have modelled several composite steel girder bridges using Staadpro and have compared the results with MDX using the high resolution option and the results are very accurate with the actual field results. You can connect the midpoint of the plates to the centroid of the girder by either rigid links or a concrete rectangular beam with one dimension equal the grid spacing along the girder and the other dimension equal the top flange width. The second option is easier for me since I can generate these beams easier than specifying the rigid links. The rigid links option gives a very small difference from the beam option.
Ed McFadden, P.E.

 

[koo]

edward,

Good to know you get good comparison results. What do you mean a "rigid link"? Is it a both end fixed member or a constraint requirement? For the concrete beam method, is the conc. beam defined with 2 nodes or 4? I think it serves to connect the plate nodes to the frame nodes. If this is the case, then 2 nodes from plate, 2 from frame, they form a vertical "plate" along the girder direction. Then the conc. beam is actually a plate member. Is this correct interpretion of what you meant?

I am thinking if it could be done this way: set up the girder frames, then set up the plate elements in a layer up from the frame plane where the centroid of the deck is supposed to be. Set nodal constraints (deflection constraints) to bond frame elements nodes to plate nodes above, so that the plate and the girder deflect the same. Then run it. Do you have any experience or commments? BTW, how does sap do the load distribution if it applies load to girders directly?

Regards,
Koo

P.S. I posted the same question on the Sap forum as well.

 

[edward1]

koo: A rigid link is a specification between two joints such that the displacements at the slave joint will be the same displacements as the master plus rigid rotation. If one uses the beam it is a 2 joint beam plus a beta angle for orientation. The beam connects a joint from the centroid of the plate to the neutral axis of the girder. This is better than your proposal of a 4 noded plate connection between these. Ed

 

[koo]

Ed,
The concrete beams add extra rigidity/weight to the super., but it seems negligible from your experience, right?
What mesh size you think gives good results? 2x2 say?

Koo

 

[edward1]

koo, The concrete beams are given a density so that the volume of the calculated beam x this density = the actual weight of the bolster between the bottom of deck and top of top flange. Yes, a 2'x 2' is usually ok. Sometimes, I even go to 1'x 1' if I have odd feet spans. If the beams are dimensioned with the grid spacing along the girders and with the top flange width the added rigidity is negligble. ed