WR1

Good question. Seismic forces are actually inertial forces generated within a body in response to seismic acceleration at ground.
When seismic ground waves (PGA) accelerate a building (Sa; spectral acceleration), the inertia of a building resist these accelerations generating inertial forces in the building.
Inertial forces are nothing strange but the usual F=ma i.e. the product of mass and acceleration.
In softwares, the joints or nodes are loaded with this F based on the mass attributed to the tributary area of that particular node x accelration of that floor.
Like this, these forces are applied at each level like static procedures.

You need to size the slabs for long term deflections. Forget these tables.

Depending on the plan, you can try strap beams, connecting beams etc.

i am using SAFE 12.3.1 and i think its the latest version

to draw a design strip select the layer A or B and then put the values of start and end of strip widths

For example you want to draw strips of 1m width then click on draw design strips. Select layer A and put 0.5 in all 4 boxes of left start width, left end width, right start width and right end width. This 0.5m or 0.5ft is above and top of center line of design strip so the total width is 1m. Then just replicate this line to cover all your slab.

Similary draw design strips for B layer.

You can draw A layer parallel to X direction and B layer parallel to y direction each 1m width. If you click on "SHOW STRIP WIDTH" in view options dialogue box you can see the widths on screen

Usually it doesnt matter you select column strip or middle stip because you are designing your for unit meter width. So you get reinforcement areas per m. If your strip width is more than 1m then you have to divide the resulting moments or reinforcement areas over that width to get moment per meter or reinforcement per meter.

The main difference between SAFE 8 and 12 is that

SAFE 8: was analysis 2D plate behaviour only

SAFE12: by default it takes full analysis rather than 2D plate

I dont my self aware of the full analysis which SAFE 12 does but the answer to the question that why NON LINEAR UPLIFT analysis in SAFE 8 takes no time and why in 12 it takes hours.

We faced this problem and asked CSI guyz! because personally i dont like the SAFE 8 interface but there were people saying that if you have to do the non linear analysis export it to SAFE 12 which i really dont like.

SAFE 8 was doing only 2D plate analysis (so you dont have to put restraint in x and y) only vertical springs. But in SAFE 12 if you run the analysis and check the log you will find that there is instability error because of no restraint in x and y.

in SAFE 12 the options to select non linear uplift analysis are also very complex. Its not clear to select NON LINEAR UPLIFT in individual load cases or to select the option CONVERT COMBO TO NON LINEAR COMBOS

well dont select the first option to check UPLIFT in each load case (results will be very strange)

rather convert all load combos to non linear for uplift (results will be fine)

but remember here that if you are doing 3d analysis without restraining in x and y your simplest model could take upto 3 hrs to run bcz of instabilities.

to compare the results with SAFE 8 and to save run time goto RUN>Advanced options> and select 2d plate analysis option

so by selecting just option you convert safe 12 to 8. no need to use the older version of SAFE

and also note that in

SAFE 12: Tension is + and Compression is -
SAFE 8: Tension is - and Compression is +

uplift non linear analysis is used so that we get zero tension soil! it can be done by iteration process by manually applying soil springs and removing tension springs...or more easily by program non linear uplift analysis option.

there is a big conceptual difference between safe 8 and 12. i bet most of the guys are still confused to see the difference between 2 programs of same example..

pls continue this topic..im willing to go in details but step by step

No, scaling is for dynamic base shear. Scaling will not increase mass participation. If 90% is not achieved you have to include more mass, that means increase the number of modes.

How to: PARTIAL FIXITY in ETABS


PARTIAL FIXITY IN ETABS
Download article in PDF format

Let’s take the case PARTIAL FIXITY in ETABS. There is no simple option to just release a specific percentage of moments and shear at the supports. The only way is to provide the reduced stiffness of the members. Let’s have a look at different options in ETABS for releases.
You can access these options by clicking: ASSIGN>FRAME LINES>Frame Releases/Partial Fixity



· There are many sets of combinations possible. You can get the details in ETABS help menu. For example it will not allow you to release torsion at both ends.
· The various checkboxes you see in this form are for releasing (making 100% pin connection). One for start point of the section and the other for end point. Important point here is when you select the option either START or END the boxes for spring values will be enabled. By default the values in those boxes is zero which means the stiffness is reduced to ZERO so making it PIN connection.
· To make partial frame releases (say only 50% of the moment), you need to put the“FRAME PARTIAL FIXITY SPRINGS” values in the START and END boxes.


So, WHAT value I put for Fixity Springs

1- First you need to calculate the stiffness of the FULLY FIXED support and this calculated as k=4EI/L
Where k = Fully fixed stiffness of the connection, E=Modulus of elasticity of the member, I=Moment of inertia in the direction of analysis, L=length of the member between supports.
Here the important point is that L is the member length between supports in that particular direction (unsupported length). If the member is divided in let’ say 10 parts, you will not put the length of one part, rather the full unsupported length.
2-After calculating the actual stiffness value of the connection, you need to multiply it by the reduction factor by which you need to reduce the moment, shear etc. The reduction factor is:-


REDUCTION FACTOR = n/(1-n)

Where n is the percentage you want to reduce. For example if you want to reduce by 25% you will get REDUCTION FACTOR = 0.25/(1-0.25) = 0.33
You need to multiply 0.33 with 4EI/L to get the final spring stiffness value and put it in ETABS.
There are following 2 cases involved:-

I) Simple one frame analysis.
Suppose I have a fixed end beam of 6m length. A load of 10kN/m is applied. The fixed end moments are wl²/12 = 30kN.m. E=2E+8 kPa, I=4.787E-3 m^4
Now I want to make the ends partially fixed/pinned, so that I get only 30% of the moment I’m getting now. (30% of 30kN.m is 9kN.m.).
The REDUCTION FACTOR = 0.3/(1-0.3) = 0.43
Full fixity stiffness = 4EI/L = 638266.67 kN.m
Reduced stiffness = 0.43 x 638266.67 = 274454.67 kN.m
If I put this value in frame releases option in ONE END only, I will get 30% of 30kN.m moment that’s 9kN.m. Now the rest of the 21kN.m will be distributed in the beam and at the support on the other end.
II) Second case, when we have a full 3D integrated structure. We may not get the moment values reduced by that percentage by which we applied the reduction factor, meaning to say we wanted 50% reduction in moment values but after analysis we got only 35%. So this process is iterative. You have to change the stiffness values based on many iterations unless you get the desired results. This is because the remaining moment should be redistributed to the other elements of the structure. Please see the picture below for the iteration process.

Just to correct you here (and my self too) that mass source command is used for calculating Center of Mass of the diaphgram. So this option is used even for wind load or any other load where we need these two centers. Pardon for my old post which was a mistake

Depending on what you are trying to achieve with tie beams? Reducing bearing pressure? Or reducing forces in pile cap? Modelling it as line element or shell? 

In my understanding these are NOT 2 different methods;
 
This is just a differentiation;
 
There are two torsions; 
 
1. Compatibility torsion (where redistribution of moments take place) like slab on beams
 
2. Equilibrium torsion (where there is no path available for redistribution of moments, like a cantilever slab resting on a beam)
 
These are not two different methods of analysis in ACI or ETABS. This is just to distinguish the cases.
 
That is why it does not matter in ETABS because in ETABS loads will follow the paths that is available.
 
So does not matter if it is case 1 or 2, apply J modifiers but watch for slab moments.
 
Also make sure your detailing handles all these issues.
 
For example if the beam is torsionally too stiff as compared to slab, it will take more moment as compared to slab, and if you are applying less J modifier to beam then make sure the detailing also follows the same approach. (try to increase bottom reinforcement of slab).

They are doing it right. It depends on you. It is the beauty of the structures that they will behave the way you designed them. 
 
When reducing the torsion modifier for beams that are failing to a value approx equal to 0 then watch for the increased moments in slabs. If you put the reinforcement in slab for additional moment then it is ok!
 
It depends upon the relative stiffness of beam and slab that how much load beam will take (Torsion, moment etc).

The analysis is Linear. Automatic redistribution would have been possible if analysis were Non-Linear. In order to approximate redistribution (during cracking), the USER has to put some kind of stiffness modification. This applies to shear, torsion, moments, axial...everything.

the idea is like a pin support..slabs supported on masonry walls has max moment at mid span wl²/8
no moment (ideal) at the wall support...so in that case for gravity loads..yes no moments at joitns
but if you talk about lateral loads...they do take somehow out of plane moments but they are very weak in tension as concrete. So in that case reinforcement is added and moment can also be transffered to base (footing) and by inbuilt concrete cols or beams...
remember masonry system can be used as lateral resistant system

Building Drift in ETABS
Drift is a very complex topic in structural engineering. It involves too many factors to arrive at a suitable decision. It involves engineering judgment, the phenomenon fresh engineers might not feel. In this article, I have tried to explain what is building drift, allowable limits, ways and means to check in ETABS models and to control the excessive drift. Please keep in mind, this article is not about the building drift as far as structural science is concerned, rather this topic of drift is related to ETABS software.
First of all you must be familiar with the term story drift. For convenience, I am quoting here the definitions from UBC-97 code:-
STORY DRIFT is the lateral displacement of one level relative
to the level above or below.
STORY DRIFT RATIO is the story drift divided by the story
height.
1) Maximum Limits
Now what for story drift limits? What is the maximum permissible value? Well it depends upon the type of drift. Is it seismic or wind?
For seismic, I will refer to UBC-97 code which in section 1630.10.2 talks about drift limits for earthquake.
 

 
Now in simple words, the maximum limit for seismic drift is:-
delta M shall not exceed 0.025 x story ht (if building seismic period is less than 0.7)
delta M shall not exceed 0.020 x story ht (if building seismic period is equal or greater than 0.7)
Important to note here is that it talks about SEISMIC drift so SEISMIC building period not the WIND period.
Now delta M = Max inelastic response displacement = 0.7R delta S
where R = from Table 16-N
delta S = displacement from static, elastic analysis
this value is read from ETABS.
you multiply this value by 0.7R to get delta M

This was all about seismic drift, but for wind drift code is mute. I will refer you to ASCE 2005 commentary CC.1.2

So we can understand that the limit for wind drift is "on the order of l/600 to l/400" for "common usage". This is common thing, however, in reality this figure can be up or down depending upon the ductility of cladding material and finishes. However for common usage value of l/400 is thought to be well satisfactory. Here l means story ht.
The concept of drift limits is same throughout all the governing codes, and the typical limits of story height by some number is same, but obviously you have to take care of the process of calculating the wind force or seismic forces. You should not calculate wind force from one code and apply limits of another code.
2) Load Combinations
Once the drift limit has been determined separately for seismic and wind forces, now is the need to check the actual drift vs the limit. Determination of actual drift depends on the load combination and the period of recurrence. If not properly calculated, this may dramatically increase or decrease the accepted drift values in model.
Seismic force E is always already factored so that's the reason its factor is always 1.0 in load combinations of ACI/ASCE code. The recurrence period for seismic force is 50 years. In seismic drift we do not convert it into service seismic force. Seismic drift is checked against the direct load case of EQx, EQy etc in ETABS.
For wind drift, we need to convert 50 year wind to service wind force. It has been recommended by ASCE commentary CC.1.2

To convert 50 year service wind force to 10 year service wind force it is multiplied by 0.7, as the equation says, and other gravity loads; D and 0.5L are also added.
So in a nutshell we create following load combinations in ETABS to check our drift:-
DRIFTWx1 = D+0.5L+0.7Wx
DRIFTWx2 = D+0.5L-0.7Wx
DRIFTWy1 = D+0.5L+0.7Wy
DRIFTWy2 = D+0.5L-0.7Wy
For seismic drift, as discussed earlier, we do not need any combination, drift will be checked just on EQx and EQy load cases only.
3) How to check in ETABS
Now we have obtained both the actual drift and the drift limit, but how can we do this in ETABS easily?
Well, after creating the drift combinations as discussed in step 2, we need to do as below:-
For seismic drift goto File>Print Tables>Summary Report

Select the file name

Scroll down to the end of the page, you will find out a section about drifts, similar to this one:-

It displays the max drift for each lateral load case for each story. As we want the drift for wind to be on drift load combinations and not on wind load cases, so we will not compare this wind drift without limits. In this table we are going to check just the drift values of our ETABS model for individual seismic load cases; EQx and EQy.
As you noticed, this table shows us values in fraction format. For example 1/105 that becomes 0.009523809524. This 1/105 value is story drift divided by story ht. It means delta S / story ht.
Now this value is delta S. First we need to convert it to delta M by multiplying it with 0.7R. Assume R here is 3.5 so
delta M = 0.7 x 3.5 x 1/105 = 7/300 = 0.023333 which is less than 0.025 so safe ( if T<0.7).
So instead of calculating every time by 0.7R we can check these limits in other way.
If our limit is 0.025 then the limit we get is 0.025/R/0.7. Assume R=3.5. Now the values in ETABS are inverse so our limit is 0.7x3.5/0.025 = 98.
In ETABS the drift is reported as 1/x where x is some number. Now as long as x (some number) is greater than 98 our limit of 0.025 x story ht is being satisfied. This way you can quickly check and compare seismic drifts.

Now for the wind drifts, goto Display>Show tables, select Point displacements>Story drifts and then select only drift combinations for results. Click on and then copy the table to EXCEL.




To save time you can right click on EXCEL taskbar and select maximum and minimum. Then just select the column H or I and see the maximum value that should be less than H/400 to H600 limit (0.0025 t0 0.00167). Again the values reported in ETABS are divided by story ht.
http://4.bp.blogspot.com/-9qv8XKHgL8Q/UALNKflmVsI/AAAAAAAAAEQ/AwKBYWt2iys/s320/image022-773193.jpg
4) Controlling Excessive Drift Values
sometimes you may face problem of excessively large values in drift tables in ETABS. Well we are not going to talk about different measures and modeling techniques to control the drift values. We are going to talk about large numbers in drift tables. Sometimes it happens that a point or node is free in the model or is connected to a NULL line or very flexible section. Drift tables for example the story drift table in wind captures the maximum displaced points. Obviously the displacement of several meters in tables is not what we are looking for. Drift values (relative) may be still okay for these points, but it requires you to check the displacement values too before checking directly the drift. Unlock the model and remove all free points, check for any discontinuity and modify your models to remove all the errors.
 
 

I guess you are talking about modal super-position, but that is not my question.
 
Yes and according to SK Gosh (http://skghoshassociates.com/SKGAblog/viewpost.php?id=5), we need to amplify dynamic torsion because;
"....  accidental torsion is not determined as part of the dynamic analysis, but as the result of a separate static load applied at an eccentricity. The only way we can use the exception [to ignore amplification in dynamic] is to incorporate the accidental torsion effects into the building model itself by defining a floor mass distribution that is not uniform so that the center of mass has a 5% offset from the centroid of the floor area."
Agreed. Now, the next step is to how to do it practically. We are not just talking about the concept but to actually do it.
Now, see CSi ETABS Wiki (https://wiki.csiamerica.com/display/etabs/Accidental+eccentricity)
1. In the first method, accidental torsion is included in dynamic analysis by actually shifting CM as SK Gosh suggests that would change the dynamic properties, natural characteristics and stiffness matrices for each eccentricity and so we do not need to amplify further, because accidental torsion has been calculated through dynamic analysis.
Con: The main drawback is that as the properties of model change for each eccentricity, hence 4 separate eigenvalues analyses must be performed for each eccentricity and then finding a way to envelope the maximum response of these 4, which is not possible in majority of software. What is the simple solution? See point #3.
2. Second way is to model a static torsional load [or a static force applied at an eccentricity] at each story for each eccentricity to approximate these effects. Then static + dynamic response is combined. This is where we need to amplify accidental torsion. Because accidental torsion has not been calculated dynamically.
3. An "efficient and practical" approach is adopted in ETABS. After the analysis of MRSA cases;
a. Acceleration at each node is multiplied by tributary mass and given eccentricity so the result is a torsional force = m.a.e = F.e
b. A static response is generated under these torsional loads and added to MRSA dynamic results.
Now, strictly speaking, eccentricity was not directly analyzed in dynamic analysis in method 3 but atleast there is a satisfying globally used practice. So we dont need to increase accidental torsion by this method.
 
4. There is another method developed by  Fahjan et al. and quoted by CSi wiki. It also contains good background information. You can view the paper from csi wiki page.

In my understanding these are NOT 2 different methods;
 
This is just a differentiation;
 
There are two torsions; 
 
1. Compatibility torsion (where redistribution of moments take place) like slab on beams
 
2. Equilibrium torsion (where there is no path available for redistribution of moments, like a cantilever slab resting on a beam)
 
These are not two different methods of analysis in ACI or ETABS. This is just to distinguish the cases.
 
That is why it does not matter in ETABS because in ETABS loads will follow the paths that is available.
 
So does not matter if it is case 1 or 2, apply J modifiers but watch for slab moments.
 
Also make sure your detailing handles all these issues.
 
For example if the beam is torsionally too stiff as compared to slab, it will take more moment as compared to slab, and if you are applying less J modifier to beam then make sure the detailing also follows the same approach. (try to increase bottom reinforcement of slab).

I guess you are talking about modal super-position, but that is not my question.
 
Yes and according to SK Gosh (http://skghoshassociates.com/SKGAblog/viewpost.php?id=5), we need to amplify dynamic torsion because;
"....  accidental torsion is not determined as part of the dynamic analysis, but as the result of a separate static load applied at an eccentricity. The only way we can use the exception [to ignore amplification in dynamic] is to incorporate the accidental torsion effects into the building model itself by defining a floor mass distribution that is not uniform so that the center of mass has a 5% offset from the centroid of the floor area."
Agreed. Now, the next step is to how to do it practically. We are not just talking about the concept but to actually do it.
Now, see CSi ETABS Wiki (https://wiki.csiamerica.com/display/etabs/Accidental+eccentricity)
1. In the first method, accidental torsion is included in dynamic analysis by actually shifting CM as SK Gosh suggests that would change the dynamic properties, natural characteristics and stiffness matrices for each eccentricity and so we do not need to amplify further, because accidental torsion has been calculated through dynamic analysis.
Con: The main drawback is that as the properties of model change for each eccentricity, hence 4 separate eigenvalues analyses must be performed for each eccentricity and then finding a way to envelope the maximum response of these 4, which is not possible in majority of software. What is the simple solution? See point #3.
2. Second way is to model a static torsional load [or a static force applied at an eccentricity] at each story for each eccentricity to approximate these effects. Then static + dynamic response is combined. This is where we need to amplify accidental torsion. Because accidental torsion has not been calculated dynamically.
3. An "efficient and practical" approach is adopted in ETABS. After the analysis of MRSA cases;
a. Acceleration at each node is multiplied by tributary mass and given eccentricity so the result is a torsional force = m.a.e = F.e
b. A static response is generated under these torsional loads and added to MRSA dynamic results.
Now, strictly speaking, eccentricity was not directly analyzed in dynamic analysis in method 3 but atleast there is a satisfying globally used practice. So we dont need to increase accidental torsion by this method.
 
4. There is another method developed by  Fahjan et al. and quoted by CSi wiki. It also contains good background information. You can view the paper from csi wiki page.

Yes..According to Taranath,
 
Total lateral deflection of a rigid frame is the combination of following;
 
1. Axial deformation of columns (15-20%)
2. Shear racking due to bending of beams (50-60%)
3. Bending of columns (15-20%)
 
So the most economical and effective way to reduce drift is to increase beam depths (and slab thicknesses as well).

In my understanding these are NOT 2 different methods;
 
This is just a differentiation;
 
There are two torsions; 
 
1. Compatibility torsion (where redistribution of moments take place) like slab on beams
 
2. Equilibrium torsion (where there is no path available for redistribution of moments, like a cantilever slab resting on a beam)
 
These are not two different methods of analysis in ACI or ETABS. This is just to distinguish the cases.
 
That is why it does not matter in ETABS because in ETABS loads will follow the paths that is available.
 
So does not matter if it is case 1 or 2, apply J modifiers but watch for slab moments.
 
Also make sure your detailing handles all these issues.
 
For example if the beam is torsionally too stiff as compared to slab, it will take more moment as compared to slab, and if you are applying less J modifier to beam then make sure the detailing also follows the same approach. (try to increase bottom reinforcement of slab).

They are doing it right. It depends on you. It is the beauty of the structures that they will behave the way you designed them. 
 
When reducing the torsion modifier for beams that are failing to a value approx equal to 0 then watch for the increased moments in slabs. If you put the reinforcement in slab for additional moment then it is ok!
 
It depends upon the relative stiffness of beam and slab that how much load beam will take (Torsion, moment etc).

First of all my apologies, if my question seems stupid because of my unfamiliarity with the topic.
When defining response spectrum load case in ETABS, we have three directions; u1, u2 and u3.
Lets assume the excitation angle = 0 and building principle direction is u1
So I will apply response spectrum function (per IBC 2006) in u1 direction by a scale factor.
My question is why we don't apply u2 and u3 in the SAME LOAD CASE.
My understanding is, according to WILSON in his book on STATIC & DYNAMIC ANALYSIS, CHAPTER 15. The response is calculated for each direction separately. Then this response is combined by different methods (CQC, SRSS etc) which is called MODEL COMBINATION.
In CQC method response in 2 directions is assumed to be a portion of 1 direction means S2=a.S1
Where a = 0 to 1.0, recommended value is 0.50 to 0.85
I know in EQ building has only one principle direction. Although EQ can hit from any direction but there is always one major direction. or you can say major acceleration in 1 direction. Lets say u1.
As in CQC method the response in other direction will be some portion of response in major direction. I also know that If we have equal spectra CQC is reduced to SRSS method which is independent from the excitation angle theta.
My confusion is we will do MODAL COMBINATION (cqc or srss) only when we want to combine u1, u2 and u3 at the same time. But if we are applying only u1 in one response spectrum load case, then we DONT have to do modal combination. It is the same thing as in CQC S2=a. S1. Now in this case a=0. so S2=0. All we have is only response in u1 direction.
Then in my building models what i do? include u1 and u2 both in one load case or in separate load cases?

Read chapter 10...Actually concrete cracks just after it reaches its tensile strength (about 7.5 to 10 or 12% of its compressive strength). before that the reinforcement is not taking action. So once concrete starts cracking...(when concrete cannot take tension) reinfrocement starts taking the tension. Now obviously when concrete cracks...we cannot analyze our section based on full inertia...so we have to reduce the stiffness, because its in reality. like this..
Now how we can reduce its properties? there are many methods...like the one given in chapter 10 of the code (ACI). like we reduce beam stiffness by 35% and column by 70% (uncracked) ... this is all relative.. I mean if you have only all the beams. even if you dont reduce stiffness it will be ok! but why we reduce stiffness is to redistribute internal forces from less stiff to more stiff areas like from beams (35%) to columns or walls (70% reduction).. So this way it ensures proper load path and distribution of loads based on stiffness

Read chapter 10...Actually concrete cracks just after it reaches its tensile strength (about 7.5 to 10 or 12% of its compressive strength). before that the reinforcement is not taking action. So once concrete starts cracking...(when concrete cannot take tension) reinfrocement starts taking the tension. Now obviously when concrete cracks...we cannot analyze our section based on full inertia...so we have to reduce the stiffness, because its in reality. like this..
Now how we can reduce its properties? there are many methods...like the one given in chapter 10 of the code (ACI). like we reduce beam stiffness by 35% and column by 70% (uncracked) ... this is all relative.. I mean if you have only all the beams. even if you dont reduce stiffness it will be ok! but why we reduce stiffness is to redistribute internal forces from less stiff to more stiff areas like from beams (35%) to columns or walls (70% reduction).. So this way it ensures proper load path and distribution of loads based on stiffness

Solutions;
1. Why dont you import the cad drawing (as shown in image) into SAFE and draw footings with proper orientation?
2. Like you got reactions under so many combinations from ETABS for each column, in the same way, you could select all columns and export to EXCEL the local forces at and then filter for zero location and proceed with manual design.
3. Get global reactions (like you did already) but one group at a time. By one group i mean all the columns on one radial grid line. All columns on that grid will have same orientation and angle right. Export to EXCEL, transform forces in XY to that angle. New rotated forces will be;
Fx' = Fx Cos theta + Fy Sin theta
Fy' = -Fx Sin theta + Fy Cos theta
there you go, you now have the new rotated forces. Repeat it for each radial line and then design footings manually.