topic Re: How to model integral beam and slab? in Robot Structural Analysis Forum

How to model integral beam and slab?

Anonymous — Tue, 24 Nov 2015 14:02:13 GMT

Hi,

I am busy trying to model a beam that is integral with a slab essentailly forming a T-Beam.

I have attached a hypothectical model below to illustrate.

Looking at the 3D representation of the model this looks incorrect. Please be so kind as to assist.

Regards

Dane

Re: How to model integral beam and slab?

Artur.Kosakowski — Tue, 24 Nov 2015 14:09:44 GMT

You may want to check this topic:

http://forums.autodesk.com/t5/robot-structural-analysis/api-macro-creation-of-quasi-t-amp-l-beams/m-p/5485265/

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

Re: How to model integral beam and slab?

Anonymous — Tue, 24 Nov 2015 15:04:38 GMT

Thank you for the response Artur.

I have gone through these posts and from what I understand is that a beam is defined with no offest from the slab. The IY prop of this beam, with its defined dimensions, is then modified such that an equivalent T-beam stiffness is obtained.

I have attached an illustration below.

Following this method is there not a double definition of the concrete area with in the hatched portion of the model?

I would assume that the bending moment distribution in the beam would be correct if the IY prop has been modified correctly?

As for modal analysis, I would assume that the modal frequencies might be affected due to a double definition of the concrete area with in the hatch.

Regards,

Dane

Re: How to model integral beam and slab?

Artur.Kosakowski — Wed, 25 Nov 2015 11:35:41 GMT

Following this method is there not a double definition of the concrete area with in the hatched portion of the model?

You may define a rectangular beam of the height equal to the part of the web under the slab only.

I would assume that the bending moment distribution in the beam would be correct if the IY prop has been modified correctly?

This is indeed the the idea behind IY change.

As for modal analysis, I would assume that the modal frequencies might be affected due to a double definition of the concrete area with in the hatch.

You may consider making a test and compare models with offsets, "whole web" rectangular beam and "web under slab" one and see the difference in obtained results

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

Re: How to model integral beam and slab?

Anonymous — Wed, 09 Dec 2015 13:16:51 GMT

Sorry for the late reply.

Cool, I will give this a go.

I see that the spreadsheet attached in the first reply that the macro creates a beam equal to the 'whole web' and not 'web under slab'. However, the process is the same for both so this is okay.

Regards,

Dane

Re: How to model integral beam and slab?

Anonymous — Wed, 10 Oct 2018 15:42:36 GMT

Hi @Artur.Kosakowski

I have since returned to this thread (apologies for reviving a dead thread) to refresh my memory on the matter.

Please would you help me understand the following:

When modifying Iy of the beam element we are attempting to obtained the correct bending moment distribution in the beam element itself so that we can import the rectangular section into RC module for design. When we look a the bending moment diagram of the bar element this should be the bending moment for design purposes say if hand calculations are anticipated.

However, my confusion comes in when looking at the bending moment in the slab along the axis of the beam (Same direction as the bending in the beam meaning that load is being distributed into both elements). How does one deal with the bending moment in the slab? and is this a realistic bending moment?

Regards

Dane

Re: How to model integral beam and slab?

Artur.Kosakowski — Thu, 11 Oct 2018 10:39:35 GMT

Hi @Anonymous

I assume you mean the bending in the slab in the direction perpendicular to the beam. It is realistic as by increasing IY of the beam you make it to work as being a support for the slab which otherwise would not happen (unless you use an offset instead).

If I managed to answer your question(s) press the Accept as Solution button please. This will help other users to find solution(s) much faster. Thank you.

Re: How to model integral beam and slab?

Anonymous — Thu, 11 Oct 2018 12:19:19 GMT

Hi @Artur.Kosakowski

Thank you for the response.

No, I am referring to the bending moment in the slab in the direction longitudinal to the beam. I have set up a small model to illustrate. You will see that the beam carries the majority of the load (My bending) however there is a small amount of bending in the slab (Mxx bending - with reference to the local coordinate system of the slab). I have attached snapshots of the model to illustrate.

My intuition tells me that the ratio between these bending moments is based on the relative stiffness between the beam and the slab. If we are only extracting the bending moments from the beam only how can we be sure that the the slab is not attracting to much load such that we have an unconservative amount of bending in the beam?

What happens to the bending moment in the slab if we are only designing for the bending moment in the beam?

Please also explain the sawtooth in the bending moment diagram as no offsets are used?

Regards

Dane

Re: How to model integral beam and slab?

Artur.Kosakowski — Thu, 11 Oct 2018 12:27:09 GMT

Hi @Anonymous

Do you mean the bending along the middle line of the slab parallel to the beams?

Re: How to model integral beam and slab?

Anonymous — Fri, 12 Oct 2018 06:47:25 GMT

Hi @Artur.Kosakowski

I mean parallel to the beam (Where the panel cut shown [SLAB_Mxx]).

From the illustration:

Bending moment in beam (midspan): approx. 2.12 kNm

Bending moment in slab (same position as the beam): approx 0.61 kNm/m

Any idea on the manifestation of the saw toothiness?

Regards

Dane

Re: How to model integral beam and slab?

Artur.Kosakowski — Mon, 15 Oct 2018 10:37:08 GMT

Hi @Anonymous

I mean parallel to the beam (Where the panel cut shown [SLAB_Mxx]).

From the illustration:

Bending moment in beam (midspan): approx. 2.12 kNm

Bending moment in slab (same position as the beam): approx 0.61 kNm/m

As there are two object which carry the load you can see the results for both of them. The point is how to 'increase' the part 'assigned" to the beam and reduce the one for the 'slab'. There is no 'integration' of them so that the bending moment is displayed as for an 'united' body.

Any idea on the manifestation of the saw toothiness?

Different function shapes for bar and surface elements. You should assume the average value of the two in the given location. For the better accuracy you need to generate denser meshing and do not use offsets.

If I managed to answer your question(s) press the Accept as Solution button please. This will help other users to find solution(s) much faster. Thank you.

The

Re: How to model integral beam and slab?

Anonymous — Mon, 15 Oct 2018 11:04:12 GMT

Hi @Artur.Kosakowski

So if I understand correctly, we are attempting to increase the stiffness of the "rectangular beam" by modifying its stiffness (Iy) so that it (the beam) attracts the majority of the load (away from the slab component) so that we can import the modified rectangular beam with correct bending moment into RC design? Hence, we are unconservatively ignoring the bending moment component in the slab?

Thank you, this is what I thought.

Regards

Dane

Re: How to model integral beam and slab?

Artur.Kosakowski — Mon, 15 Oct 2018 11:18:48 GMT

Hi @Anonymous

Correct.

Hence, we are unconservatively ignoring the bending moment component in the slab?

I'm not sure if I understand what you mean. Where wold you like to 'use' this slab bending component?

Re: How to model integral beam and slab?

Anonymous — Mon, 15 Oct 2018 11:56:56 GMT

I'm not sure if I understand what you mean. Where wold you like to 'use' this slab bending component?

What I mean by this is that we are applying a certain amount of external load which must be carried by the slab and in turn the beams which support it. So if, due the the relative stiffness between the slab and beam, 15% remains in the slab and 85% remains in the beam we are only designing the "t-beam" for 85% of the external load propotioned to it and not 100%.

If I were to do this by hand calculation I would proportion the load accordingly into the monolithic t-beam and design accordingly. Here, we are adjusting stiffness to best guess the load in the rectangular beam and discounting the percentage of bending (albeit small) remaining in the slab? Should the correct design bending moment for the "adjusted beam" therefore not be the summation of the bending moment components in the slab (0.61) and the beam (2.12) hence 100% of the external load proportioned to it?

Regards

Dane

Re: How to model integral beam and slab?

Artur.Kosakowski — Mon, 15 Oct 2018 12:10:04 GMT

Hi @Anonymous

See message 6 form https://forums.autodesk.com/t5/robot-structural-analysis-forum/beam-offsets/m-p/3332921

If I managed to answer your question(s) press the Accept as Solution button please. This will help other users to find solution(s) much faster. Thank you.

Re: How to model integral beam and slab?

Anonymous — Mon, 22 Oct 2018 12:15:47 GMT

Hi @Artur.Kosakowski

So its a matter of best approximation towards the end goal you want to achieve. By following this method, I assume the deformation response of the system will not be accurate as the stiffness of the slab is still present along with the scale stiffness of the beam? Hence, if the deflection response is of interest it would be better if offsets are used (but design forces will be incorrect).

Seems like a bit of a catch 22...

Thank you for the recap and assistance, Artur.

Regards

Dane

Re: How to model integral beam and slab?

Artur.Kosakowski — Mon, 22 Oct 2018 12:23:33 GMT

Hi @Anonymous

Mind that for static analysis you assume elastic material behavior therefore for a RC slab on RC beams it is inaccurate anyway isn't it?

If I managed to answer your question(s) press the Accept as Solution button please. This will help other users to find solution(s) much faster. Thank you.

Re: How to model integral beam and slab?

Anonymous — Mon, 22 Oct 2018 14:32:01 GMT

Hi @Artur.Kosakowski

I agree 100% - as mentioned previously we are also using beam and shell elements in an attempt to model solid geometry. This in itself as influences structural response.

I am just clarifying the draw backs of the modelling approaches. I am pondering over the influence on modal analysis (seismic response) as the stiffness will influence the design forces in the element. So, we need to model the development of the design forces correctly (no offsets) whilst still getting the stiffness correct (offsets).

I have been reading the following document on the analysis of "beam" bridges which have composite action between beams and decking:

https://www.steelconstruction.info/index.php?title=Special:ImagePage&t=SCI+P356.pdf

This document suggests that the beam webs are modeled as shells elements, beam flanges as bar elements with the upper flange offset from the the deck and the deck as shell elements. This is an interesting approach which is relevant to what we have discussed previously. Would you care to offer you opinion on this?

My only concern is how to do code verification of the beam elements which are both bar and shell elements. The document suggests that the "software" is able to output effective design forces from the shell. Is Robot capable of this?

Regards

Dane

Re: How to model integral beam and slab?

Artur.Kosakowski — Tue, 23 Oct 2018 09:40:37 GMT

Hi @Anonymous

Yes, in some programs especially for bridges (e.g. in Autodesk Structural Bridge Design) you create this sort of models using rigid links connections to 'put together' elements defined at different levels. The key option is the ability to 'combine' results from bar and selected (strip of) surface elements together and display them as if it was a single beam which cannot be done in Robot. In other words the approach you use for creating a model is often driven by the kind of the output you need and can have.

If I managed to answer your question(s) press the Accept as Solution button please. This will help other users to find solution(s) much faster. Thank you.

Re: How to model integral beam and slab?

Anonymous — Wed, 07 Nov 2018 09:08:30 GMT

Hi @Artur.Kosakowski

I have had a quick look into ASBD. It looks like an interesting software.

I am a part of a marine structures team where the majority of our work involves quay wall design (embedded walls, gravity structures) however portions of our work involve the design of marginal quays (deck on pile structures), loading platforms and access trestles where precast construction is the preferred method of construction. Here construction sequencing is critical - precast beams used to span between newly driven pile bents, CIP concrete is used to stitch together these beams for load continuity over the bents, precast panels are then used to span transversely between the beams followed by a CIP decking to complete the structure (this is fairly similar to beam bridge construction and hence my interest in the topic in an effort to improve my workflows and efficiency in RSA).

Hence my following question -

Is it possible to use staged construction in RSA to follow the construction sequencing above? IE track the development of design forces in the precast beams as they are placed on the pile bents (behaving as simply supported beams), continuity is then provided across the bents by the CIP concrete making the beams continuous under the dead load of the precast planks (bending moment in the beam will then be a combination of simply supported and continuous fixity), followed by the last stage where the beams work integrally with the CIP topping (TBeam). See the image below for reference.

Regards

Dane