Dear Seismostruct,
I've built a three-dimensional RC frame building using infrmDB elements. Running some different types of analysis, I've noticed strange output results. For example:
- in Static Pushover analysis: the moment registered in beam elements overlap the ultimate moment capacity of the element itself.
- in Static Analysis: the presence of conspicuous axial forces are pointed out in beams. But,in these elements, a negligible magnitude (near or equal to zero) sholud be expected since they are only subjected to their weight and to an additional mass/lenght load.
If you think usefull, I can send my file to allow your check.
Best regards.
Clara Caponi
Unexpected beam behaviour
-
seismosoft
- Posts: 1271
- Joined: 06 Jul 2007, 04:55
Re: Unexpected beam behaviour
Hi Clara,
Have you introduced rigid diaphragms to model the slab? If you did, then you will inevitably increase the capacity of your beams, since you are restraining their axial deformation - see the notes in the Help System on this issue.
We note that such increase in beam capacity, if caused by the modelling of the slabs, may not necessarily be wrong, especially in the case of cast-in-place RC structures, but one should certainly be aware of it, in order to then make informed modelling choices.
Seismosoft Support
Have you introduced rigid diaphragms to model the slab? If you did, then you will inevitably increase the capacity of your beams, since you are restraining their axial deformation - see the notes in the Help System on this issue.
We note that such increase in beam capacity, if caused by the modelling of the slabs, may not necessarily be wrong, especially in the case of cast-in-place RC structures, but one should certainly be aware of it, in order to then make informed modelling choices.
Seismosoft Support
-
claracaponi
- Posts: 2
- Joined: 19 Jan 2010, 04:55
Re: Unexpected beam behaviour
Dear Seismosoft Support,
Unfortunately the problems does not regard the presence of rigid diaphragms constraints.
In fact, in static pushover analysis can be registered a huge difference between the moment in beam elements and the ultimate moment of the element itself even if these constraints are removed.
Moreover, the values given just for the initial step of Static pushover Analysis (Load Factor equal to zero) overlap the ultimate moment of the element itself. The only confortable note is that these values correspond to those provided in the Static Analysis: are exactly the same!
But they are really overstimated with respect to the acting loads and the element's capacities.
In static analysis, remain also the the problem relative to the precence of cospicuous axial forces in beam elements; while, as stated before, a negligible magnitude (near or equal to zero) sholud be expected since they are only subjected to their weight and to an additional mass/lenght load.
Since I principally use SeismoStruc for vulnerablity assessment analysis, I'm quite worried about these problems: they can completely invalidate the checks relative to brittle/ductile collapse mechanisms.
Thanking you in anticipation, I send my best regards.
Clara Caponi
Unfortunately the problems does not regard the presence of rigid diaphragms constraints.
In fact, in static pushover analysis can be registered a huge difference between the moment in beam elements and the ultimate moment of the element itself even if these constraints are removed.
Moreover, the values given just for the initial step of Static pushover Analysis (Load Factor equal to zero) overlap the ultimate moment of the element itself. The only confortable note is that these values correspond to those provided in the Static Analysis: are exactly the same!
But they are really overstimated with respect to the acting loads and the element's capacities.
In static analysis, remain also the the problem relative to the precence of cospicuous axial forces in beam elements; while, as stated before, a negligible magnitude (near or equal to zero) sholud be expected since they are only subjected to their weight and to an additional mass/lenght load.
Since I principally use SeismoStruc for vulnerablity assessment analysis, I'm quite worried about these problems: they can completely invalidate the checks relative to brittle/ductile collapse mechanisms.
Thanking you in anticipation, I send my best regards.
Clara Caponi
-
seismosoft
- Posts: 1271
- Joined: 06 Jul 2007, 04:55
Re: Unexpected beam behaviour
Dear Chiara,
If you have no nodal constraints and if you are using a zero-hardening steel material model, then the scenario you describe can only be the result of a modelling error, and we would thus strongly suggest you to check your model carefully.
You may start by modelling a single beam element, and comparing its results to hand-calculations; we are convinced that such exercise will provide you with the insight you need to identify the source of the anomalies you are observing.
Seismosoft Support
If you have no nodal constraints and if you are using a zero-hardening steel material model, then the scenario you describe can only be the result of a modelling error, and we would thus strongly suggest you to check your model carefully.
You may start by modelling a single beam element, and comparing its results to hand-calculations; we are convinced that such exercise will provide you with the insight you need to identify the source of the anomalies you are observing.
Seismosoft Support
-
seismosoft
- Posts: 1271
- Joined: 06 Jul 2007, 04:55
Re: Unexpected beam behaviour
Dear Clara,
The model you sent us does use nodal constraints to model floor diaphragms, hence the comments in our first response still hold.
A very quick look at your model also allowed us to identify a number of inconsistencies, such as excessive discretisation of cantilevered columns at the roof (which actually don't really need to be modelled in any case), too coarse a discretisation of some beams (1 DB element in some steel beams), unadvisable choice of master node for the floor diaphragms (a beam-column joint node would be preferred, for increased numerical stability), separate diaphragms for stairs-case, etc etc.
We reiterate our suggesting for a carefully re-check of your model.
Seismosoft Support
The model you sent us does use nodal constraints to model floor diaphragms, hence the comments in our first response still hold.
A very quick look at your model also allowed us to identify a number of inconsistencies, such as excessive discretisation of cantilevered columns at the roof (which actually don't really need to be modelled in any case), too coarse a discretisation of some beams (1 DB element in some steel beams), unadvisable choice of master node for the floor diaphragms (a beam-column joint node would be preferred, for increased numerical stability), separate diaphragms for stairs-case, etc etc.
We reiterate our suggesting for a carefully re-check of your model.
Seismosoft Support