I'm currently doing a project where all columns are terminated at the ground floor (basement-ground) and from ground to second floor to roof all columns are supposed to be planted. I am using STAAD on my design of RC structures and I find that the structure tends to require too many reinforcement and requires a really large dimension for a 3 storey residential whenever there is too many planted columns involved. Sometimes I release the base of the planted column on all direction if I think it's not critical. Can anyone recommend a god read to better understand planted columns? thanks.
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Planted Columns 3
- Thread starter minuj
- Start date
msquared48
Structural
The only planted column I know, God refers to as a tree.
Can you please clarify the term "planted"?
Mike McCann
MMC Engineering
Can you please clarify the term "planted"?
Mike McCann
MMC Engineering
- Thread starter
- #3
- Thread starter
- #5
"Planted" is not a common structural term in most areas of the English speaking world. Now that we know you mean the columns start on a suspended floor, perhaps a more detailed description of your building in regards to spans, etc. would give us enough information to comment. What do you mean by "too many reinforcement" and "really large dimension"? Could it be that your expectation was unrealistic?
- Thread starter
- #7
-
1
- #8
This surely has to do with the analysis. You are getting Vierendeel behaviour and so the columns take significant moment, then require big dimension and high reinforcement. The supporting beam at bottom of the nascent columns uses to be called a transfer beam (viga de transferencia). There's also the issue of staged construction with these structures, easy to deal with ETABS and SAP2000 and other programs, or -to some extent- even making successive models following the construction growth in less sophisticated programs.
If you want to forfeit the Vierendeel behaviour to some extent, the usual approach is to make the (supporting) transfer beam quite rigid, a really deep structural member, or even some proportionally thick transfer slab, maybe post-tensioned if the dimension so recommends.
You may also articulate and even make some support able to displce horizontally to further restrain moment development in the columns.
It is an usual assumption to analyze and reinforce the transfer beam as simply supported (but certainly not in all cases); it all dependes of the ends being able to take the negative moments developing, and an overall sound and stable path to the foundation of the loads that you decide on such things.
If you want to forfeit the Vierendeel behaviour to some extent, the usual approach is to make the (supporting) transfer beam quite rigid, a really deep structural member, or even some proportionally thick transfer slab, maybe post-tensioned if the dimension so recommends.
You may also articulate and even make some support able to displce horizontally to further restrain moment development in the columns.
It is an usual assumption to analyze and reinforce the transfer beam as simply supported (but certainly not in all cases); it all dependes of the ends being able to take the negative moments developing, and an overall sound and stable path to the foundation of the loads that you decide on such things.
Not sure how ishvaaag infers all that from the information you have furnished. I would agree that 10000 mm^2 (5.5% Ag) is unrealistic reinforcement for a 200 x 900 column. Maybe the column is really a vertical beam which is trying to cantilever? As I am not familiar with STAAD, I don't know how to advise troubleshooting of the program results. How do your hand calculations compare?
-
1
- #11
Minuj:
Damn these fancy-a$$ed computer programs which seem to imply that you can do anything, if you only grind the numbers long enough; and then do this without any need for structural experience or intuition about how things actually work, in the real world.
This is a three-plus story residential building, with the first floor framing being beams which must span to foundations at some locations. These beams carry everything above, both gravity and lateral loads on the structure, to the foundation. As an educational exercise, do this by hand at first for that educational value: make all connections pinned, except for beams which might be continuous over columns below; and size beams and columns to take the gravity loads only. This might take several iterations, because large beam deflections at the ground level will mean beam support settlements at upper levels, necessitating a redesign of those upper beams.
Then, as a second step, apply some x-bracing or shear walls at appropriate locations to account for the lateral loadings on the building, still in a fairly simple load transfer fashion. Pay some attention to how these might change forces in the beams and columns already preliminarily sized, and consider size changes as required by this step. By superposition of these two design steps you should have a pretty good preliminary design, or good input for a high powered analysis and design program like STAAD. You should also have a better feel for what joints to release, or fix, or partially fix. Obviously, in an RC frame, many of these fixities exist; try a long hand moment distribution on some of these frames (using your prelim. sizes) to get a feel for how this fixity changes things; or maybe try a simple stiffness matrix solution on the frame problems to eliminate most of the extraneous junk imbedded in STAAD. Now, you can start to think about final sizing and reinforcement design; or you have really good sizes, etc. to use as input for a final design in STAAD. With much more experience you will start to become proficient at prelim. sizing for program input, so that a few columns don’t attract all the lateral loading, or some such. And, you’ll see that some bracing or shear walls do wonders to reduce column sizes and reinforcing.
My hat is off to you for being astute enough to question the column sizes and reinforcing you were getting, but these programs today or so complex that they will give you garbage, for results, unless you understand every detail of what they do with your input, and how you set constrainsts, etc.
A good program does not necessarily a good design make. A good and experienced structural engineer with good sound judgement ultimately makes a good design with the help of good software. You run the software and computer, they don’t run you or your design, or you will be lost.
Damn these fancy-a$$ed computer programs which seem to imply that you can do anything, if you only grind the numbers long enough; and then do this without any need for structural experience or intuition about how things actually work, in the real world.
This is a three-plus story residential building, with the first floor framing being beams which must span to foundations at some locations. These beams carry everything above, both gravity and lateral loads on the structure, to the foundation. As an educational exercise, do this by hand at first for that educational value: make all connections pinned, except for beams which might be continuous over columns below; and size beams and columns to take the gravity loads only. This might take several iterations, because large beam deflections at the ground level will mean beam support settlements at upper levels, necessitating a redesign of those upper beams.
Then, as a second step, apply some x-bracing or shear walls at appropriate locations to account for the lateral loadings on the building, still in a fairly simple load transfer fashion. Pay some attention to how these might change forces in the beams and columns already preliminarily sized, and consider size changes as required by this step. By superposition of these two design steps you should have a pretty good preliminary design, or good input for a high powered analysis and design program like STAAD. You should also have a better feel for what joints to release, or fix, or partially fix. Obviously, in an RC frame, many of these fixities exist; try a long hand moment distribution on some of these frames (using your prelim. sizes) to get a feel for how this fixity changes things; or maybe try a simple stiffness matrix solution on the frame problems to eliminate most of the extraneous junk imbedded in STAAD. Now, you can start to think about final sizing and reinforcement design; or you have really good sizes, etc. to use as input for a final design in STAAD. With much more experience you will start to become proficient at prelim. sizing for program input, so that a few columns don’t attract all the lateral loading, or some such. And, you’ll see that some bracing or shear walls do wonders to reduce column sizes and reinforcing.
My hat is off to you for being astute enough to question the column sizes and reinforcing you were getting, but these programs today or so complex that they will give you garbage, for results, unless you understand every detail of what they do with your input, and how you set constrainsts, etc.
A good program does not necessarily a good design make. A good and experienced structural engineer with good sound judgement ultimately makes a good design with the help of good software. You run the software and computer, they don’t run you or your design, or you will be lost.
The FE model that you have created is assuming that the structure has magically appeared overnight, attained strength and is behaving as one unit. This is not the case because of staged construction. As ishvaag suggested, the large moments in the columns that you have encountered are probably due to some virendeel action that is created by the stiff beams.
I would be looking at the top of ground to roof as one model and then transferring the reactions into the ground level slab design. Whenever there are transfers involved it will slow down the construction speed a slightly compared to the construction speed of identical floor plates going straight up.
This staged construction will also be critical in the design of your transfer beams.
I would be looking at the top of ground to roof as one model and then transferring the reactions into the ground level slab design. Whenever there are transfers involved it will slow down the construction speed a slightly compared to the construction speed of identical floor plates going straight up.
This staged construction will also be critical in the design of your transfer beams.
mathcadboy
Civil/Environmental
I have done that kind of structure before and I also encountered that kind of problem. I am also from the Philippines and yes we call them planted columns. Fix your framing, its all about proportioning your columns and beams correctly. Remember, the stiffer the member, the more load it will take.
-
1
- #14
CarloStruct
Structural
This may be a late reply but I have had a similar experience. I don't use STAAD but I guess the same principle applies. As mentioned in earlier posts, the column may be experiencing Vierendeel effect. It is likely that the column may be "hanging" from other frame members above the story it is supposed to support.
So if the column is tension, you would expect really high reinforcement ratio as concrete capacity is ignored. For the sake of designing the column, you may pin the beam members connected to it at above levels. Or, as suggested in earlier posts, use a stiff transfer element to support the column. I am not sure if STAAD allows you to modify the stiffness of frame members without changing the beam dimension. Other programs like SAP200 do.
So if the column is tension, you would expect really high reinforcement ratio as concrete capacity is ignored. For the sake of designing the column, you may pin the beam members connected to it at above levels. Or, as suggested in earlier posts, use a stiff transfer element to support the column. I am not sure if STAAD allows you to modify the stiffness of frame members without changing the beam dimension. Other programs like SAP200 do.
- Status
Similar threads
- Question
- Question
