Column Bases Restraint in ETABS

[Ali Shan]

Hi everyone.

I have seen many structural engineers using Pinned column bases for RCC buildings in ETABS and many others are using FIXED ones (soil being of compressible nature in both cases). I want experts to comment on the right approach and also please explain what are the detailing requirements for both cases as I have seen that people using both approaches use same detailing pattern in which column bars go straightly down to the footing.

[Simple Structures]

A good question. No base is fully pinned or fully fixed!

1. Engineering judgements are needed for base fixity, and experienced engineer use this judgement to analyse, design AND detail the column/base connection.

2. Fixed based require large footings - uneconomical in most instances; also, very difficult to get full fixity with a pad bases once the base rotates under moment. Only piled foundations with thick pile caps may offer anything close to full fixity!

3. Sometimes (not common) column flexural stiffness (EI/L, 4EI/L, 3EI/L etc) is used to derive "partial fixity" for analysis purposes. Majority of column/base connections are inherently partially fixed - but in analysis this is ignored.

4. I have analysed buildings with pinned bases, and then re-analysed the same frame for serviceability wind deflection assuming 10-20% fixity! All an engineering judgement call.

5. Pinned bases are used as a matter of course for most designs of concrete frames. There are exceptions to this rule: (a) for portal frame type crane buildings require fixed bases to limit lateral sway of frame (H/400), and make sure crane does not come off the crane rails; (b) Multi storey steel frames often use ‘fixed bases’ to limit frame sway at bottom, and design foundations accordingly.

6. In Pakistan I would always assume pinned bases for analysis purposes (although 10-20% fixity will always be inherent e connection of column/base, which helps but should not be assumed in analysis). Although this leads to higher moment on columns between first and second floors, and hence lower column needs to be sized the same as column above.

7. With a ‘fixed base’ frame the lowest column will attract greater damage to lowest column/beams during an earthquake. seismic regions. With ‘pinned base’ less damage to the column would occur and more to the beam ...beams are easier to strengthen later then columns! 

8. My Conclusion: Use pinned bases in rc frame upto say 5-6 storey high for analysis purposes. Note design is half the story, make sure you sketch the critical reinforcement connection details for the steel detailer & fixer too.

Hope this helps?

[Ali Shan]

Very nicely explained. 

But I have never seen any practicing engineer in Pakistan who provides detailing of column bases as HINGED CONNECTION in case of RCC buildings. Most of them simply design the footing for AXIAL LOAD ONLY and let the column bars go straight into the footing pad. Could you please reveal that myth? 

[Simple Structures]

Design: You assume "pinned" for analysis.

Detailing & Construction: You still detail the column for reinforcement to anchor fully into the base....with compression anchorage provision of rebar into base....hence partial fixity is always there.

 

 

 

[Ayesha]

Simple Structures has explained it really well. Kudos.

Another statement that you will hear in this conversation is, that what you will assume is the structure will behave. This is actually true mostly.

[SMAQ]

Simple Structures explained the topic very nicely. But if we analyse the building with fully pinned supports, the reactions we are going to get for foundation design are without moments i.e Mx=My=0. If a structure involves large moments at base and we are not considering it in our design, because practically structure will not behave as 100% pinned or fixed but indeed there will be partial fixity, so we are making it fully pinned in analysis and finally we are neglecting those partial moments that are going to act on the foundations.

[Simple Structures]

@SMAQ 

1. Let’s look at a simple single storey frame, with single bay, sitting on isolated pad foundations. The frame will be subject to VERTICAL gravity loading + HORIZONTAL wind loading.

2. At the top of the base there will be horizontal force from wind (as well as vertical load). For bearing capacity check you would convert that horizontal force to a moment (H x base depth) and then check ground bearing capacity is less than allowable. 

3. In analysis, you put a pin joint at this location. (V, H but no moment). Vertical loads are fine; The sway (lateral displacement) of the frame will be determined by the column stiffness / inertia in that direction. As the base node is input as "pin", deflection will be higher. This will then be reduced, or brought into to allowable, by two things:

         a. by increasing the inertia (size) of the column.

        b. by providing a moment connection at the eaves, column/beam, joint location (zero M at base).

The above two will enable you to bring the sway of the frame within allowable limits (say H/300 or H/400 etc). Where H is height of the frame. 

4. In concrete frames base size and base connection are often determined by vertical load (& compression length of rebar into the base). The floors act as diaphragm and take all sway/lateral loading to the shear wall, stair or lift core walls, and transfer them into the ground. In these bases the forces are higher, there can even be uplift on one end of the base, and base design and reinforcement anchorage is more onerous.

5. You will now design/detail the column to pad base connection.

6. For a steel frame you would use 4-holding down bolts as a minimum for safe erection of column and design them for the compression load, and any tension generated by the horizontal case.

7. For a concrete frame you will have the column reinforcement going into the base, with full compression laps ...or tension laps if say it was braced bay in longitudinal direction. The rebar is always L-shaped and sits on the bottom reinforcement mat of the base. Say if the base was 750mm thick then the rebar will be 650mm+ben of say 250mm .... 900mm into the base. 

8. now re-read my earlier post again about partial fixity always being present, but ignored in analysis, for a safer frame design etc.

Does this help in understanding the practicalities of the column base connection?

 

 

[Simple Structures]

Yes, you can use "fixed bases" in the analysis. There the bases have to be designed, ensuring no rotation, for "full fixty". The cost of such foundations in many such cases can be prohibitive! 

As said, I have only designed fixed bases for crane supporting industrial buildings.

With fixed base performance, one is also relying a lot on "soil-structure" interaction, which is new debate and topic in itself .... not all structural engineers have sufficiently versed in geotechnical knowledge.

Design safe, be safe.

 

[Ali Shan]

@Simple Structures Thanks for nicely explaining the concept of base fixity and I have understood all that. You are right that even in case of hinged base, the column bars are required to go straight down to the footing for bar anchorage in compression. But I have seen in many books which present the attached  reinforcement detailing (please see the attached image) of hinged column bases in which it seems that the column bars have been terminated at the face of footing and have not been developed into the footing. Your expert views on this are awaited please.

 Thanks

[Simple Structures]

Good question - I will respond using the same numbering as previous post – read that again - thus saving me repeating the same points again.

It’s the "Hinge Detail" that you are asking about.

9. I have yet to meet an engineer who has detailed this connection in practice for concrete column in a building structure! Only place I have cross across true "pins" or "hinges" is in wooden and concrete arches. The proposed detail is not practicable for concrete fame buildings.

10. This is a very unpractical detail with very little, I would say zero, benefit. Could you construct a column like this, remove the formwork and leave it standing vertically as a cantilever, till the slab is cast? No. Therefore it’s not a buildable detail either.

11. To get a true "pin" or "hinge" with the illustration you have given to allow rotation, the connection would also require the dowel bars to be "de-bonded" and to work as true pin or hinge would require "crushing" of the reduced area of concrete, to activate the "hinge" and "pin". This then begs the question how the column compression load is transferred through the interface concrete.

12. There are no stirrups at the hinge in the detail. As hinge is supposed to move, the concrete cracks. Therefore, any water in the ground would penetrate the crack and corrode the reinforcement - inherent durability problem with the detail.

13. Let me think, the only location you could adopt the detail is when two cantilever beams meet, and the "hinge" detail is adopted to stop any relative (vertical stepping) at the detail.

Ignore the book detail, indeed cross out the detail yourself from it, if you don’t want to bin the book...it’s not a really a practical detail. Forget it.

 

 

[ANStructs]

I agree with the point made by SMAQ.

Although, analysing the frame with pinned supports will be on the safer side as far as the design of frame is concerned, but the same could not be said for the foundation.

The design of foundation will most definitely govern for a "fixed base support" model.

Not to mention, the hinged base support detail at column-foundation joint, is not a preferrible joint detail for "code" desired sway mechanism for a moment frame.

Edited March 29, 2020 by ANStructs

[Ali Shan]

@ANStructs

So does it mean that we should prepare two models of building one with hinged base and other with fixed one? The hinged should be used for frame design and fixed for footing design? 

[Simple Structures]

 

...carrying on this debate....

14. The foundation is there to provide safe bearing pressure onto the soil below. This will have Horizontal and Vertical reactions - with Moment released.

15. With a concrete frame you transfer later loads from the sides of the building, into floor plate diaphragm and then down the shear wall into the ground. This later load converts to push-pull onto the foundation (resolving the moment). The foundation at this shear wall location sees maximum forces and is designed appropriately with the increased forces.

16. The stiffness of a column is very little compared to a wall - hence load is attracted to stiffness, the wall. 

17. As we are all locked in our homes due to Covid 19 isolation, I suggest for your own peace of mind - run both analysis - with varying base restraints, check results, design foundations to suit and satisfy yourself.

18. Don't forget computer software is only an analysis tool, and practical engineering judgement is needed for safe and economic structures. 

19. I do always carry out a quick design by hand, get a feel for the answers to be expected, and then move onto computer analysis. 

I am sharing my thoughts here to pass on my experience for the benefit of young structural engineers, the final decision is always the Design Engineer's!

 

 

[Simple Structures]

*later should read *lateral

[ANStructs]

@Ali Shan

Short Answer is, Yes. This way you will definitely be on the safer side as far as the design forces for foundation and structure are concerned.

As it's already pointed out in previous posts that, the actual behaviour is something between a pinned and fixed support condition. The actual rotational stiffness of the support depends on the interaction of two main factors.

1. Flexibility of the soil.

2. Rigidness of foundation.

The higher the flexibility of the soil is, the more it will allow the foundation to rotate (act more like a pinned support) no matter how rigid is the foundation. Similarly, a flexible foundation will not resist any meaningful rotations no matter how rigid is the supporting soil.

If you want to learn more about this, try googling "Soil-Structure Interaction".

 

 

 

[ANStructs]

Finally finished reading the debate whose sole participent was SimpleS Structure(just joking. No offense intended)..

I would like to add a few comments on what you've said so far.

Regarding hinge detail.. All your points seems valid "if" the construction of a hinge detail is carried out like a normal column. However, this is not how a hinge detail is designed and constructed in an rcc member. I'll give you my 2 cents on those of your points which i don't agree with (using your reference points numbering).

On 3/27/2020 at 9:45 PM, Simple Structures said:

Ignore the book detail, indeed cross out the detail yourself from it, if you don’t want to bin the book...it’s not a really a practical detail. Forget it.

Although, I don't think this action will be necessary.

On 3/27/2020 at 9:45 PM, Simple Structures said:

10. This is a very unpractical detail with very little, I would say zero, benefit. Could you construct a column like this, remove the formwork and leave it standing vertically as a cantilever, till the slab is cast? No. Therefore it’s not a buildable detail either.

What's more beneficial than this that this detail will serve it's intended purpose and will not transfer any significant moment? Offcourse, there'll be some limitations such as how much load this can bear or whether this should be provided where large lateral loads are expected.

Assuming I'm providing a hinge detail in an RCC column, such as shown in the attached figure, then I'll make sure that the contractor knows about the requirements for vigilant inspection and potential stability issues, and will take preventive measure by providing adequate braces untill the upper floor/beam shuttering is removed and desired load path is achieved. Also, as far as I know, these sections are designed for a minimum accidental eccentric moment due to axial loads. Still, the tensile stresses produced due to these moments are not allowed to exceed the tensile strength of concrete.

On 3/27/2020 at 9:45 PM, Simple Structures said:

11. To get a true "pin" or "hinge" with the illustration you have given to allow rotation, the connection would also require the dowel bars to be "de-bonded" and to work as true pin or hinge would require "crushing" of the reduced area of concrete, to activate the "hinge" and "pin". This then begs the question how the column compression load is transferred through the interface concrete.

12. There are no stirrups at the hinge in the detail. As hinge is supposed to move, the concrete cracks. Therefore, any water in the ground would penetrate the crack and corrode the reinforcement - inherent durability problem with the detail.

Well actually, most of the hinges are engineered by locally reducing the size of column. This effectively reduces the moment of inertia of the cross-section which results in very small rotational stiffness and zero (or close to zero) moment from analysis.

Yes, the reduced section will cause a very high compressive stress, due to which, a very high compressive strength of concrete and bursting reinforcement may be required. The gaps might or might not  be filled later on with an elastic compressible material.

As I can see, all these provisions are considered in the picture provided by AliShan.

4 hours ago, Simple Structures said:

15. With a concrete frame you transfer later loads from the sides of the building, into floor plate diaphragm and then down the shear wall into the ground. This later load converts to push-pull onto the foundation (resolving the moment). The foundation at this shear wall location sees maximum forces and is designed appropriately with the increased forces.

I think the point you're making here is that column will see very little to no moment. And i agree, that's what dual frames are for. But, what if only moment frames are provided to resist lateral loads? By designing for partial fixity condition, you're ensuring that plastic the hinge will form at supports first. Which is not a good design practice and might produce very high P delta moments.

[Jorge Andrade]

On 3/27/2020 at 8:22 AM, Ali Shan said:

@Ali Shan could you tell me what book is this?, please.

[Shahzad Khan]

And how to assign partial fixity to the support in ETABS?

[Simple Structures]

I am not a ETABS user but "base spring stiffness" follow the same rules for all structural analysis programmes, inc ETABS I would think.

Base fixity is a determined by column stiffness : 4EI/L 

As column length (L) changes, the base stiffness changes.

For a "nominally pinned base spring stiffness" take 10% of 4EI/L.

For a "nominally fixed base spring stiffness" take 100% of 4EI/L.

 

Nominal base stiffness explained.pdf

[UmarMakhzumi]

On 10/22/2020 at 8:36 PM, Tufail Mabood said:

And how to assign partial fixity to the support in ETABS?

This might give you some clues:

Thanks.

[Omer Ahmed]

@Simple Structures Nicely explained. Pardon my ignorance but could you please explain how moments are resisted from foundation/footing if we assume base to be behaving like a pinned/hinge support?