punshing shear and unbalanced moment 1

[CHAKKOUR]

hi guys.
An alternative for not taking into account the punshing shear is to consider the structure as a pinned one. If potentially, there will some moment, a crack will occur and then a plastic hing which will stop the transmission of the moment. What do u think ??

 

[JedClampett]

That's scary. In one sentence you're talking about punching shear (you have that no matter what) and in the next you're talking about moment. Please clarify.

 

[JoshPlumSE]

JedC -

Look up the punching shear calculations in ACI. There is a shear stress (due to unbalanced moment) that comes into play. I believe that's what rony111 is talking about.

I don't think I can authoritatively answer his question.... But, I'm interested in what other people say. Cause it's a good question.

 

[JedClampett]

OK, I'll bite:
"An alternative for not taking into accunt punshing (sic) shear..." To me that reads like he/she wants to not take into account punching shear. That's not unbalanced moment shear.
I wouldn't remove the unbalanced moment shear in any case. There's no easy way to build a column slab joint without some level of fixity. The price you pay for that fixity is a potential unbalanced moment due to a pattern loading and some additional shear on one side.
You don't want a crack forming at the critical shear plane in a slab/column system.

 

[JoshPlumSE]

I interpreted the original question as asking "Rather than accounting for the unbalanced moment portion of the punching shear calculation, can I just make the column-slab connection a pinned connection".

Re-reading the question, I guess it wasn't all that clear.

I almost replied back that if you detailed the connection out so that moment transfer were eliminated, then this would be reasonable..... Until I started thinking about whether it was possible to detail this connection that way. It is possible, but it sure wouldn't be very easy. Probably easier to just include those moments in the punching shear caculations.

 

[CHAKKOUR]

First of all, i am sorry because i wasnt so cleared, i must have formulated in an otherway. Thanks JoshPlum for reformulatng.
If u include those unbalanced moments in punshing shear calculations , u must also take into account the moments transfered from wind and earthquake (if there is) to the slab. So u must build an etabs model in the first step and extract the loads. I dont know if that is the easiest way.
U can have a pinned structure when the connection between slab and column doesnt allow for the continuity of the top reinforcement through the column. Consequently, there will be no unbalanced moment at the column and the checking of the punshing will just be via ultimat axial load.
All the moments will be transfered to the cores.
I am very interested in ur opinions. Thanks and sorry if the english isnt so good.

 

[hokie66]

You cannot just assume a pinned joint to eliminate the need to design for unbalanced punching shear. Don't even think about it. Joints in concrete structures are not pinned, whether you account for the moments in designing the slab reinforcement or not.

 

[rapt]

I agree with the scary thougths from above!!! Really scary that a practicing engineer could consider this.

Punching Shear is a Brittle Failure condition. The moment cannot redistribute away from it without a collapse occuring. Codes do NOT allow redistribution from a brittle failue condition for very good reason, they cause collapses. There is no elastic/plastic action, it is elastic/collapse for punching shear!

You must design the connection for the moments that will be attracted to the connection. There is no other choice.

So the only possibility is to build the column/slab connection as a pinned connection. This is not practical, so design for the moment transfer.

 

[CHAKKOUR]

Thank u guys, in practise, i do always take into account the unbalanced moment so do not worry
But this thought came to me because in some softwares (safe for example), u have the choice to consider the structure as pinned one by removing the include bending stiffness for the column. If the code deoesnt allow for such a practise, why to have the choice ??
A pinned struture will affect the deformed shape and the bottom reinforcement will be higher in the middle of the spans but as u said and i do totally agree, the unbalanced will not be taken and that is not safe.

 

[JedClampett]

rony111, you've been a good sport as we've slapped you around, so here's some advice: Software is developed by software developers. They made a choice not to do design. Don't trust their judgement. If it doesn't smell right, don't do it!
In addition to poorly designed software, some of the biggest mistakes I've seen and heard of had computer calculations. No matter what the software, it's no better than the input.

 

[hokie66]

The problem of unbalanced moment is obviously accentuated at edge columns. Removing the fixity at an external column makes the flexural design of the slab more conservative. There is nothing wrong with doing that. But you can't ignore the moment for punching shear purposes. Using a flat plate without edge beams or cantilevers is not good practice, in my opinion.

 

[rscassar]

All concrete software packages I have used have allowed the user to input the column stiffness (generally I go between 25 and 50%). Reducing the column stiffness will reduce the amount of moment which is transferred into the column.

Generally the unbalanced moment will be greatest for edge columns. The greatest amount of unbalanced moment which is transferred into the column for internal columns will occur for patterned load cases, in which case the axial force will be less.

I like to use the analysis package to determine column moments using full column stiffness and then perform the punching shear calculations by hand.

Have you considered using stud rails if punching shear is creating issues?

 

[rapt]

Hokie,

My opinion also (but I am old too, like you!!), but you try to convince a frame contractor in Europe or USA to do it. They want the simplicity of pure flat plate formwork and get it because they control the purse strings in the end. You want another project, then no edge beams!

Kikflip
You should only reduce column stiffness if the column is going to crack relative to the spanning members. If both are cracked, you need to estimate the relative cracking or use full stiffness for both. If the beam/slab is cracked and the columns uncracked (common in highrise for vertical load), theoretically you should allow for higher column stiffness relative to floor stiffness!
But you should never allow for reduced column stiffness if the column is not cracking of if it is cracking less than the floor.

JedClampett
Being a dreaded Software Developer I will take offense (not too badly because I am pretty easy!!) with your post. It does apply in some cases but not in others.
I originally wrote RAPT to make my design life easier. Unfortunately too many people like it and want it's developemnt to be continued, so after 25 years on a 2-3 year development project I am still developing. That does not mean I cannot still design. I am giving both RC and PT design advice continually and am on concrete code and technical report committes in Australia and UK to help with concrete design development.

On the software issue, a program must offer the user any option that they may require in different circumstances to carry out the design they want to do. The USER must know what options he can use in any particular instance and use the software appropriately. It is not the softwares fault if the designer wishes to make design decisions which are not logical. If the user does not understand design and the effects of using such design options, that person should not be using that software and probably should not be designing, even by hand, because they will make the same or worse mistakes. Rubbish input will lead to rubbish design in both hand calculations and software calculations. But the computer program in not an engineer. It is a calculator. The designer is responsible for generating models and making design decisions. The computer simply does the calculations and produces a set of results. It is the users job to then interpret those results and make design decisions as to its appropriatness and to iterate the design to get to a final solution.
Safe, RAPT and many programs give users control over columns stiffness relative to floor member stiffness and many other areas of design. None that I know of actually recommend that punching shear problems can or should be allievated by reducing column stiffness. It is the users responsibility to use the software appropriately for their project.
Another common example of bad use of software by engineers happens with drop panels in PT flat slab design by designers following the ACI/PTI logic for flat slab design. No matter how many times I tell people they are grossly underdesigning using this methodology, they continue to do it, because everyone in Asia and India does, so they have to to compete!!
The problem with software arises when an incorrect option is recommended by the software or argued by the software to be appropriate or offered as the only possibility. This happened in a case that has been covered earlier this year and covered in some posts on this site regarding the use of Mxy moments in design. At least 2 FEM concrete programs are/were actively recommending that Mxy moments be ignored in design and their default oiption is/was to ignore them. This is blatently wrong and very unconservative. That is when Software developers can be blamed. It is encouraging designers to underdesign and produce substandard designs.

But the software cannot be blamed for incomptent engineers. That is the fault of the system, and it is progressively getting worse.

 

[hokie66]

Hear, hear!!

 

[ash060]

I have used a pinned column support in flat plates when I what to represent a steel column supporting the slab.

 

[JedClampett]

I'm sure there are software developers out there that have extensive design experience and their software reflects that. BUT, for every good software package, I can give you examples of poorly thought out, end user unfriendly cases. Both of these examples are from the same, very common software.

Example One: Why does retaining wall software insist on asking for the phi angle of the soil retained? Every Geotechnical Report I get gives equivalent fluid pressures of the soil. To use the software, I have to convert a easy to understand fluid pressure to a mysterious phi angle. So I do retaining walls by hand.

Example Two: Concrete Beam or slab software. "d" for a single layer of bar can be calculated automatically using clearance, bar diameter and stirrup size. The software makes me calculate it, every time. I have to do the calculation. It's a pain, something I would hope the software would automate.

Both of the examples above give examples where computers could make my life easier by removing a tedious, mistake prone task (retaining walls; "d" for concrete) and throwing it back on me. I suspect an actual designer would make better choices in writing this software.

 

[hokie66]

Jed,
Example One: I don't want someone to tell me what "equivalent fluid pressure" is, because I want to be able to control the assumption as to where the water table will be. I do them by hand as well, but want to start with phi if in a granular soil.

Example Two: I want to be able to control "d", particularly for top bars in slabs. The effective depth depends on the size of the bar chair used, not on the specified cover.

 

[JedClampett]

Hokie, you're not going to like the Geotechnical Reports we get. We never get the phi value. Ever. And I've been through a couple of hundred of them. We get an active pressure (in pcf), an at-rest pressure (in pcf) and pressures below the water table for both of those.
As far as the "d" issue, you can control it by changing cover. All I'm saying is the "d" is thickness minus cover minus stirrup size minus one half the bar diameter. Why can't I just input the clearance (2 inches or 1.5 inches or whatever) and let the computer calculate the "d"? If I say I want 2 inches of cover, they better use a bar chair that accomodates that. Otherwise, I have to add a little hand calc in the margin to show where "d" came from.

 

[hokie66]

That does make your hand calculations easier, but I like to know how the geotech got his recommendations. I do the "d" calc based on available height of bar chairs (because they come in standard increments), not on the specified cover. Admittedly, I'm probably the only one who does that.

 

[rowingengineer]

Hokie,
You’re not that idiosyncratic. Not only do I like to be able to adjust d for chairs/blocks (have to say I prefer the concrete blocks to chairs for bottom steel), but I also like to be able to adjust it for external/internal applications. Then there is out of sequence cantilevers, and steps to account and I’m not even going to start with beams/PT…….It really annoys me when for a program all you can input is one cover and you have 5 different covers.

As for punching shear, in flat plates/slabs, I not only like to account for column stiffness, but I also like to account for column projection, every single project has a projection these days and if you don't catch it early in the formwork laying, I can guarantee you it is a pain in the royal ass after the steel has been laid.

The column stiffness I like to select is based on axial load there are some good articles around, however I have found the NZS code a good guide for this purpose. It would suit my purposes if software allowed you to spec different stiffness for different applications, ie I would like to be able to set the column stiffness conservatively for slab bending and conservatively for punching shear. As these are complete opposites in return value, and column stiffness from all the reports I have read are all over the place, best estimate is +/- 30%.

Rapt,
While software developers cannot be hold fully responsible for the degradation of engineering in regards to use of software. It can be held responsible for Poor manuals (I will exclude Rapt Manual from this statement which is about 60% better than most other manuals). Most software developer these days fail to convey what theories or fundamentals their software is using, mainly only suppling manuals with directions on how to input the parameters, which is normally only about 2% of the parameters used in the process. I want to be able to control the parameters used in every step of the process, from column stiffness to code controlled values.

I also dislike software’s that select shear steel for punching instead of highlighting the problem, for using arrangement of shear steel in beams that is unrealistic. I say this from a checking engineer’s point of view, for junior engineers to learn they need to be able to do their own projects, and will get lost from time to time in the local practices. In north qld it is rare to use shear steel in thin slabs for punching shear, with drop panels being the norm. It would be good if the software made them select an input or direction, as so that this problem is highlighted to the supervising engineer, instead of the poor supervising engineer having to sort thru 28 pages of program output to find the error.

ok time to get off the soap box and have a beer.


An expert is a man who has made all the mistakes which can be made in a very narrow field