Using Composite Floor in Footfall Analysis

[Anonymous]

I am attempting to carry out a footfall analysis of a number of composite floors, but I am unsure how to best model the composite floor.

Modelling the composite slab is straightforward, but it is when trying to achieve composite action between the slab and steel members that I run into problems.

I have defined the panel thicknesses as concrete slabs composed with metal decking, but presumably I would need to offset members from the underside of the slab in order to achieve the correct composite section properties. I believe this offset should be equal to the depth from neutral axis of the composite slab to the underside of the composite slab + half the depth of the steel section supporting the slab.

As a test, I have modelled a simple composite slab arrangement in Robot and compared the calculated deflections again the output from the same arragnement modelled using 'BDES' composite beam software. The Robot model appears to overestimate the stiffness of the composite floor by approximately 25% compared to the BDES software.

Can anyone suggest a reason for this, or any factors/parameters that I may have overlooked?

Is Robot even capable of modelling composite floors correctly?

In this particular case, the spacing between adjacent beams is small, such that the effective width of slab that can be utilised in composite action is equal to the beam spacing. So the discrepancy is not due to Robot considering a greater width of slab than is permitted in the Eurocode.

This is my first time posting, so apologies if I've missed a relevant answer on an existing thread, but my searches did not throw up any satisfactory answers.

[Rafal.Gaweda]

check these:

http://forums.autodesk.com/t5/Autodesk-Robot-Structural/I-want-creative-Flool-COMPOSITE/m-p/3564166/...

http://forums.autodesk.com/t5/Autodesk-Robot-Structural/Composite-Floor/m-p/3448285/highlight/true#M...

http://forums.autodesk.com/t5/Autodesk-Robot-Structural/Stress-Skin-Design/td-p/4367294

[Anonymous]

Thanks for the prompt reply.

I guess the thread from that first link wasn't thrown up from my searches due to spelling!

So, this suggests Robot cannot perform a footfall analysis on a floor system which utilises composite floors. Is this correct?

[Rafal.Gaweda]

So, this suggests Robot cannot perform a footfall analysis on a floor system which utilises composite floors. Is this correct?

Why do you think that ?

[Anonymous]

The response to that first thread states:

"You can define an orthotropic panel on steel beams but mind that there is no composite floor design module in Robot."

Perhaps I have misunderstood, but if there is no way of accurately modelling a composite floor, then I don't see how I could carry out a footfall analysis and achieve valid results.

To clarify, I am talking about a composite floor - i.e. a slab acting compositely with steel sections - and not just a composite slab, which can be easily defined in the 'panel thickness' dialog box.

I have read through the other threads as well, but they do not answer my original quesiton. I have already modelled the slabs as FE panels, as is suggested in the second thread, and using rigid links - as suggested in the third thread - between the slab and beam would not be a correct way of modelling the composite section.

Applying an offset between slabs and beams in Robot increases the stiffness of the combined elements as you would expect, but the magnitude by which the stiffness increased is greater than expected, leading me to believe the results are invalid.

If it is in fact possible to model a composite floor accurately, it is still unclear to me whether applying this offset is the correct approach. If you can clarify this for me, I would greatly appreciate it.

[Rafal.Gaweda]

---

@Anonymous wrote:

The response to that first thread states:

"You can define an orthotropic panel on steel beams but mind that there is no composite floor design module in Robot."

It means there is no !design! , not FEM calculations.

To clarify, I am talking about a composite floor - i.e. a slab acting compositely with steel sections - and not just a composite slab, which can be easily defined in the 'panel thickness' dialog box.

http://forums.autodesk.com/t5/Autodesk-Robot-Structural/Beam-offsets/m-p/3332921/highlight/true#M2762

http://forums.autodesk.com/t5/Autodesk-Robot-Structural/Overlap-between-slabs-and-beams/m-p/3586652/highlight/true#M6832

http://forums.autodesk.com/t5/Autodesk-Robot-Structural/T-slab/m-p/3185418

[Anonymous]

Thanks again for your respose, but again, these don't really answer my original question.

These all refer to concrete beams acting 'compositely' with concrete slabs.

I have modelled my composite slab in the same way as the first of these three threads, by applying a member offset such that the top surface of the steel member is level with the bottom surface of the composite slab.

My problem, is that in doing this, Robot seems to overestimate the stiffness.

I think the problem I am experiencing is due to the use of an orthotropic plate/composite slab, as opposed to the rectangular section plates in these threads. Robot uses the 'effective' thickness for a slab composed with a trapezoidal plate, which is obviously different in the two principle directions. I would guess that Robot assumes the centreline of the modelled panel is where the neutral axis of the actual slab would be. However, the position of the neutral axis is not the same for bending about both axes, so the offset of the steel members would need to be different for bending about perpendicular directions.

It is easy to determine the offset that should be applied when using a symetrical plate section (half the slab depth + half the beam depth), but I am questioning the magnitude of the offset that should be applied when using an asymetrical plate section, due to the uncertainty of the location of the 'centre' of the plate.

I hope this makes sense.

It seems that the only way of modelling this correctly, is to determine the section properties of the composite section externally and manually enter the 'composite' member properties as a new section size. I just wanted to avoid doing this, as I have quite a large number of different slab types and member sizes that I would have to calculate the composite section properties for, and it would be quite time consuming to manually enter all of these as new sections.

[Anonymous]

Guy

Did you manage to get to the bottom of this? I'm currently trying to figure this out also. Is it true that each memebr's properties will need to be entered manually? I have lots of beams and doing this would take a lot of time!

Thanks

Oscar

[Calum_Scott]

I'm also keen to find out about the outcome of this. Being a very common form of floor construction in various building types, I'm surprised that a composite floor analysis and design function is not available in a popular software package such as Robot.

[Rafal.Gaweda]

in 2016:

[Anonymous]

Thanks for the continued support here Rafal.Gaweda, it's great to see a composite desite module coming to Robot in future. It's also reassuring to see others commenting on this thread, indicating that I'm not the only person interested in finding a solution to my problem. However, can I please focus this discussion back on the original topic?

I was not questioning how the design of composite sections may be verified, as this can be acheived by a number of external methods, either by hand or other proprietary software package (or indeed within the upcoming 2016 release of Robot, it seems).

I was, however, hoping to gain clarification on how to best model a composite floor system consisting of steel beams and a slab composed with a trapezoidal plate, such that the correct stiffness is determined, which can then be used to produce accurate results for a footfall analysis.

As previously stated, Robot appeared to be generating stiffnesses that were significantly higher than that which I had determined by other means, thus throwing doubt over the validity of my footfall analysis model.

I previosuly stated that:

"It seems that the only way of modelling this correctly, is to determine the section properties of the composite section externally and manually enter the 'composite' member properties as a new section size."

Currently, I stand by this statement, as I have yet to see anything to suggest otherwise.

[Rafal.Gaweda]

See this thread:

http://forums.autodesk.com/t5/robot-structural-analysis/t-slab/td-p/3183798/page/2

to change panel stiffeness (but of course you have to calculate factor by your own):

[Anonymous]

Thanks for your response. That thread refers to the design of a concrete beam composed with a concrete slab, and does not raise the same issues that I am referring to.

Also, your comment regarding panel stiffnesses doesn't seem to be related to the thread. Why are you suggesting I modify the panel stiffness?

In the thread, another Autodesk Support made a very similar suggestion to my own conclusion:

---

@Artur.Kosakowski wrote:

I would propose yet another approach that you may find useful especially in cases when you actually want to design a T shape beam in the RC Beam Design module after importing it from the model. In such case use of offsets that correctly 'model' stiffness of a slab cause existence of axial force that we usually don't want to 'see' while calculating reinforcement inside the beam (offset cause the bending to be replaces by set of the compression force in a 'flange' /panel/ and the tension force in 'web' /bar element/ or vice versa /span or over supports/). My approach is to model a rectangular beam ('web') with no offset but with artificially increased moment of inertia (see the attached picture - IY of rectangular beam multiplied by ratio of IY of  T shape beam / IY of rectangular beam). This approach causes 'increase' of bending in the beam (T shape beam effect) while no axial forcse in the beam being generated.

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.

---

[m_sheinin]

I know it's old tread however what will be the best way to model composite section. In my situation I have steel joist and above I have composite section ( steel deck with the concrete topping ) . The way I did my model is to create steel joist and then apply shell elements using Orthotropic tab.