Pinned or fixed foundation on a 3 story building. 2

[smcp4blo]

Hello,

This question may sound a little bit odd (for not saying silly), but I had a discussion with a colleague and I would like to have other opinions. I'm designing a 3 level steel building with HSS columns and W beams. Since the land is not very good I made the decision of modeling it in Staad with a pinned base, and so designing the foundation with only axial and shear forces. However, my colleague (who has more experience) says that buildings can't be modeled as Pinned in their bases for buildings more than 3 levels. I find this wrong, as for my structural criteria states that I can model a building as pinned since it isn't unstable, and the only difference with a fixed model would be that my displacements are greater and so the stiffness of my columns must be greater. Right?

So, pinned = less foundation, bigger columns.
Fixed = more foundation, smaller columns.

Anyway, my colleague refuses to accept this design as acceptable, and he tells me that according to his experience he has never seen a building constructed as pinned foundation. However he hasn't presented 'technical' data to support this, just his experience.

Can I have any opinions?

Thank you!

 

[smcp4blo]

By the way, the only "real" technical data I could think of to support his point of view, is that base plates for HSS can't really be "pinned" when putting the 4 anchors that are needed as minimum.

 

[slickdeals]

You can analyze it as pinned, but you should detail it such that it acts as pinned.

(or) conservatively analyze it as pinned, calculate moment that can be transferred by your base plate, depending on your bolt pattern and use that moment to design the foundation.

 

[SteelPE]

I went to an AISC seminar once a number of years ago that gave the argument that pinned base connections were not really pinned because they could take some moment in them not matter how hard you try (well unless you actually fabricated a pin at the base). They basically stated that a fixed base develops 4EI/L of fixity and that inside of the column stability charts (K) they allow you to use G=10 therefore you can use .4EI/L inside of your computer models as a spring at the base of your column. However, this was from the approach of trying to economize steel buildings (get people away from assuming a fully pinned base).

People who I have talked to will pin bases of columns because they believe that fixing the base will lead to larger moments. I find it sometimes difficult to fix the base of a columns because the moments that get driven through the system are difficult to develop.

 

[paddingtongreen]

For starters, there is no such thing as a fixed base, at best, they are partially fixed. I know of no restrictions, this should be based on circumstances, not on a rule of thumb.

We have seen on here, various practices that are common in one location but that most of us around the world haven't seen before.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[amec2004]

My reply as the followings

1. There is no specific code wise requirements defining the building base shall be designed as PIN or FIX. It’s all engineer’s own call and judgment.

2. Most of the structures and buildings base are designed as PIN base. Both PIN base braced frame and moment frame are stable structure and there is no stability issue.

3. When your lateral resistance system is braced frame, you shall design all the bases along this braced frame line as PIN, as the FIX base won’t take any lateral load and moment due to the much much strong bracing and is just waste of $$$

4. We only consider FIX base when the serviceability requirement cannot be achieved. One of the example is the crane top of rail lateral deflection < 2” requirement in building moment frame case. Again, in braced frame case FIX base will help little but in moment frame it will help to bring down the lateral deflection a little bit

5. FIX base with moment will create difficulties in anchor bolt design. There is a significant issue when your structure is in seismic design criteria SDC >=C and the code requires a ductile anchor bolt design as per ACI 318-11 D.3.3.4.3. You may need huge pedestal and 8d stretch length to meet the code requirements, so, try to use PIN base as much as you can to simplify the base and anchor bolt design, especially when in SDC>=C case

6. It would be difficult for you to achieve the PIN base assumption when your column is HSS as the anchor bolts can only be arranged around the perimeter. One of the solution would be using knife connection by cutting slots in HSS and welding two cross gusset plates to connect the HSS to base plate.

7. There is strict PIN or FIX in reality. The actual condition is something in between. What you can do is try to detail your connection as close to you design assumption as possible, as per item 6 stated above.

Good luck !

anchor bolt design per ACI 318-11 crane beam design

http://www.civilbay.com

 

[smcp4blo]

amec2004,

What do you mean in item 2 with "PIN base braced frame"? Are you saying that if it is pinned then it must have braces (e.g. conncentric o excentric)? What if I have a model with PIN base and my lateral resistance is just resisted by my columns stiffness and cables (mainly for lateral displacement due to wind)?

 

[amec2004]

>> What do you mean in item 2 with "PIN base braced frame"?

It means the lateral reistance system is braced frame and all column bases are designed as PIN

>> Are you saying that if it is pinned then it must have braces (e.g. conncentric o excentric)?

No

>>What if I have a model with PIN base and my lateral resistance is just resisted by my columns stiffness and cables (mainly for lateral displacement due to wind)?

Can your cable be treated as a brace ?

Correct my typo above
7. There is NO strict PIN or FIX in reality.

anchor bolt design per ACI 318-11 crane beam design

http://www.civilbay.com

 

[smcp4blo]

Well yes, it is a brace that works only in tension, and since I need resistance in both ways I arrange them in "X" so it has resistance in both ways.

About the connection of the base plate with HSS columns, I find that quite tricky but its true, I cannot have a real "pinned" base plate with HSS since my anchors must be arranged in the perimeter and it takes moment. What other solution, more practical, can be made with these pinned columns?

 

[delagina]

The AISC Specification does not define what assumptions the responsible design professional should make in regards to base fixity considerations. There is probably no such thing as a completely fixed or completely pinned base. It is usually a matter of it being considered as closer to fixed, than pinned or vice-versa. The specific details of the base play a large part as to what goes into making these assumptions. But the type of foundation system also may be a major consideration. Bases with the anchor rods located outside of the column shaft are often assumed an 'pinned' due to flexibility of the base detail, as well and rotational characteristics of the foundation itself. These are really engineering judgment calls that must be made by the design professional.

Kurt Gustafson, S.E., P.E., F.ASCE
Director of Technical Assistance
American Institute of Steel Construction
866.ASK.AISC

 

[TXStructural]

As another data point, every steel building I have designed and seen designed was considered pinned at the base of columns UNLESS the base plate was detailed as a moment-resisting connection. Use AISC Design Guide 1 "Base Plate and Anchor Rod Design" to guide that decision. Where I designed base plates to transfer moments to foundations, I usually attached the anchor bolts to the column rather than the base plate (using stools or sleeves welded to the flanges or walls of the column.)

 

[amec2004]

>> There is probably no such thing as a completely fixed or completely pinned base

Agreed

>> Bases with the anchor rods located outside of the column shaft are often assumed an 'pinned' due to flexibility of the base detail

Disagreed.

For normal base plate design practice we control the base plate thickness to be min. 1/4 of plate overhang to maintanin the rigidity of base plate matching our design assumption of concrete stress distribution as linear rectangular or triangle.

If you consider the flexibiltty of base plate, that might affect the design assumption most enginers used, and will be more accurate to go for FEM modeling which is overkilled in a daily design routine.

Plus, ACI 318-11 D.3.1 limit the redistribution of stress utilizing the yielding or plastic theory when the anchor bolt is non-ductile design.


>>as well and rotational characteristics of the foundation itself

Disagreed.

The rotating stiffness of concrete pedestal plus footing or pilecap is MUCH MUCH larger than the steel colum and base plate. It's not realistic that you rely on the contribution from the concrte part to make the base as PIN

anchor bolt design per ACI 318-11 crane beam design

http://www.civilbay.com

 

[paddingtongreen]

@amec2004,
">>as well and rotational characteristics of the foundation itself

Disagreed.

The rotating stiffness of concrete pedestal plus footing or pilecap is MUCH MUCH larger than the steel colum and base plate. It's not realistic that you rely on the contribution from the concrte part to make the base as PIN"

I don't know where this came from, concrete is much more (about 9 times more) flexible than steel and soil is much more flexible than concrete, that is why we state there is no such thing as a fixed base.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[amec2004]

>>concrete is much more (about 9 times more) flexible than steel and

I am talking about EI as the stiffness. You can also say it's EI/L. The steel column normally has much larger L over concrte pedestal's L as pedestal is a short column.

Let's talk about EI only.

Concrete E=2.5x10^7 kPa, Steel E=2X10^8 kPa, consider concrete cracking 50%EI --> Steel E / Concrete E = 16 times <-- you are right at this point, BUT

Say column W310X97 (W12X65) Ix=2.2 x 10^8mm^4

Concrete pedestal 500mmx500mm (20"x20") Ix=500/12x500^3=52.1 x 10^8 mm^4 around 24 times more


Overall EI Concrete/Steel=24/16=1.5 times

Consider L, Say column storey height=10', pedestal height=3'

Overall EI/L Concrete/Steel=1.5 times x 10/3 = 5 times

anchor bolt design per ACI 318-11 crane beam design

http://www.civilbay.com

 

[paddingtongreen]

The column and the pedestal are continuations of the same member, the end of the member and approximate center of rotation is at the footing.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[amec2004]

>> The column and the pedestal are continuations of the same member

If you design your (base plate + anchor bolt) as PIN assumption and bolt pattern, it would be very difficult to maintain the continuity at the column base to pedestal top point. You need the strong rigid link between column base to pedestal top point, which means a full MC connection. It’s very difficult to achieve that given the big difference of rotation stiffness between these two elements. In reality there would be no continuity between these two and there will be relative rotation between column base and pedestal top to create a hinge at that point.

I believe most engineers model the frame's pin support at base plate elevation, not the center of footing elevation.

anchor bolt design per ACI 318-11 crane beam design

http://www.civilbay.com

 

[wannabeSE]

smc4blo,

Often times pinned bases are used because it is more economical to add steel to the frame than provide a baseplate and foundation with fixity. I have even seen steel beams with moment connections encased in concrete below grade to provide a fixed base.

 

[SAIL3]

A wise engineer on a previous post on this subject mentioned not to design for a fixed condition or you are going to regret it, ofcourse, if you have the option for pin-based design....I use pinned, but, where I do have a concern is with post-installed AB's...in this case, I design the AB's for a semi-fixed condition or in other words, overdesign the AB's to ensure a capacity beyound the pinned-base assumption...as others have stated, there is no case of a truly pinned or fixed condition, unless, one goes to extra length to design it as such....this also applies to typical steel connections that are assumed to be pinned....

 

[JAE]

If you have spread footings - don't forget about the flexibility of the soil underneath - that adds to the "pin" type behavior.
If you have drilled piers - flexure in the piers adds to the "pin" side of the behavior.
Flat, non-stiffened, base plates also flex under moment and add to the pin type behavior.

 

[smcp4blo]

So basically, to summarize the information, there is no problem in modeling my building as PIN based (as long as it is stable).

In reality of course, PIN and FIXED are never completely true, since this are idealized representations of what we are trying to accomplish, however if the correct way to do it would be to take into consideration soil interaction, stiffness of the structure and represent this as a spring, that would take a lot of time and effort and I definitely can't do it if I'm running optimization into my model and have to design approximately 2 o 3 buildings per week.

So, the best approach to this PIN situation and HSS columns can be to model it that way, and then design the anchors on base plate for tensile loads (or minimum for OSHA reguirements) and if I may, consider the moment that could take this arrangement and transfer ir into my foundation (spread footings).

I agree with amec2004 that creating a perfect rigid connection between base plate, anchors and pedestals is quite difficult.

And.. yes, a fixed connection on the base plate can lead to big costs in foundatios, for example some of my daily designs are warehouses, sometimes with a height of 20m (65ft) and spans of 30m (100ft) or bigger, and due to wind loads my base plates can get to 1.75" and 12 anchor bolts of 1.25"; which if you ask me, is quite a lot of steel.

Comments?