Composite Beam Section Properties

[maj.rentung]

Hi...

• What is the correct way to create section properties for a composite concrete prestressed beam and decksab.  Kindly advise either to  "joint" or "merge" elements?

• Is it correct that the modular ratio is not considered in SBD if we choose to "joint" structure elements to form a composite member.

• Im creating a simple grillage model and the longitudinal beam members were assigned by importing the composite beam section properties which were created using "Design Section" mode. Under properties, what should i use "gross section" or "transformed section? Any advise for my consideration if im choosing this method to assign composite structure to the member created in SBD grillage.

Please refer to the attachment for ease of understanding.

Thanks

[dave_geeves · Accepted solution]

HI,

I have taken a look at your attached pdf and this looks a very interesting structure and is far from what I call a "simple" grillage.  It has one curved edge and the other defined by two straights, and as the edges aren't parallel the slab parts of the composite girder are tapered.

From the pdf, I think the mesh looks well conditioned and the general layout looks good.  The central section looks as though it is a solid slab over a single pier/column but I'm not sure and whether you are modelling this as a plate finite element slab or that this is also modelled as a grillage (although this is not really relevent to your questions).

From your post it seems that you have chosen to define the section properties using "design sections" from the section modules - and there is nothing wrong with this (although the prestress beam design checks will need to be done externally).  You are quite right that the slab parts of the section and the precast girder parts should be kept separate, as they generally have different material properties.  In this case the merge operation to combine the two elements should be used as "join" will make them one section with one material property and when assigning the properties set them as "Transformed" so the modular ration will be considered.

However, I would have preferred to use the beam module "design beams" to define each of the structural composite beams.  In the beam module you cannot define tapered slabs but the beam can be made up rom a number of different sections, each with a different slab width, so you can define the same number of sections in the design beam as you have longitudinal segments in the mesh grillage and this will give a stepped representation of the tapered slab and curved edges.  Using this method the correct section properties are automatically assigned to each longitudinal beam with the appropriate transformed properties.  The results from the analysis can be transferred back to the beam module load tables so that the beam design and code checks can be carried out in the beam module.

I also notice that you are using the 2014 version of the software.  Although this is still valid there are a number of enhancements in later versions that should be beneficial to you. For example, from 2016 versions onwards, a number of design sections and/or a number of design beams together with the structure definition can be defined in one "project" data file.  The latest version is 2019, so I recommend that you upgrade to this. I also recommend that you look at example 10.1 in the examples manual which gives some general guidance. Although this is a two span parallel deck the process is the same but the beams will have multiple sections.

I hope this has been helpful to you.  If this has answered your query then please mark my reply as a solution so that others can benefit from this post.  Thanks

Kind regards

Dave Geeves

[maj.rentung]

Hi Dave,

You explainations are clear and realy helpful.

Thank you.

Im in the midst of modeling 2 spans semi integral bridge for the curve merging ramps. The central part of the bridge consist with mesh section represent the table top crosshead (wide crosshead) which monolithically connected with the composite beams and pier. The beams are also seat on bearings at both EJ locations.

Its quite challenging to define the best solutions and combinatios for the live load effects for a curve and tappered bridge. I also notice, we cannot model tapper lanes with SBD even if we want to self defined the live loads instead of using the influance surface module.

[dave_geeves]

Hi,

Thanks for marking my reply to you as a solution.

Yes, I agree with you that the analysis and design of curved/tapered ramps is not simple, especially when it comes to the application of traffic loads.  Although you do not specify which design code you are designing to, all of the codes supported by ASBD are not clear in how to define notional loaded lanes when the raised kerbs are not parallel with each other.  All of the rules and formula are related to constant width notional lanes so it is difficult to make an interpretation.  I have tried to discuss this with different road authorities in the past but have never received a sensible reply.  If you have any better definitive information I'm sure the developers would welcome a copy of this information for possible future implementation.

The curve of the carriageway is not an issue, as traffic loads can be applied to curved lanes, it is purely if they are tapered that the problems occur

With this lack of clear rules in the code it is very difficult to develop software with code based rules to cover this situation so it is necessary to use the tools available, with constant width lanes and divide  the carriageway longitudinally into a number of separate carriageways.  The transition from one carriageway to the next will be at a location when the number of notional lanes changes based upon the rules in the code.  Traffic loads can then be placed in each carriageway based upon the constant lane width rules. (If the load optimisation is used then manual adjustments will be required as the loaded lengths will only be related to the current carriageway rather than  the sum of the carriageways.  Any special vehicles generated may need to be adjusted/removed as necessary).  With this lane arrangement there will be triangular strips along the edges that will be unloaded, which is not a conservative solution.  In this case additional triangular UDL can be placed in these areas (if it will have an adverse effect).  In fact some engineers form two sets of carriageways, one set with lanes parallel to the southern edge and the other set with lanes parallel to the northern edge, and then envelope the results.

I am sorry if this is such a long explanation but it is quite a complicated procedure.  I should also add that these suggestions are my own suggestions and are not necessarily recognised as a definitive solution to the problem but I hope they are useful to you.

I would really welcome anyone else's thoughts on this.

Kind regards

Dave Geeves