Single Bay Frame

[LSPSCAT]

Imagine a single bay portal frame designed with rigid girder to column connections and base considered pinned. In reality the base is composed of a standard 4-bolt base plate connection to a concrete pedestal.

No leaner columns or additional frames for support. If the base plate is considered flexible and the frame is designed as pinned then the anchorage does not need to be designed for any moment?

In practice it seems like it must take some moment?

I understand this is a commomn assumption in multi-bay frames I am just trying to resolve some issues with invidual frames in the field.

 

[rb1957]

no, i think you can have a pinned frame. the moment connection at the other end of the upright is enough to make a free body.

i think you can frame a frame with fixed foundations and pinned connections, or pinned foundations and fixed connections, or fixed and fixed; but not pinned and pinned.

 

[frv]

Correct- you do not need to design the base plate for moment. You DO have to design it for shear, though.

In practice, it is very difficult to get drift limits to work with realistic frames in realistic situations without fixing the base.

 

[MiketheEngineer]

rb and frv are correct.

That is why these are designed as such - so we don't have moment connections at base - pretty hard(expensive) to do.

 

[LSPSCAT]

My stumbling on this particular situation is that the single bay frame is an equipment support and must resist out of plane loading. Therefore to resist these loads we have detailed and designed the base plates as "fixed" for transverese loads applied to the frame.

Additionally the frames are heavily loaded in the vertical direction. Now, the frame design and connections will work with pinned bases for vertical loading. However, if I analyzed the bases as fixed for the frame analysis I have very large moments that I would prefer not to design the anchorage for.

Dilemna is possibly academic but something I want to understand. I need to design the bases as fixed to take out of plane loads applied to the frame. At the same time, I need to perform a frame analysis with the frame designed with pinned bases to take vertical loads. My base plate is detailed to account for this.

 

[rb1957]

an out-of-plane load can be reacted with pinned foundations and a fixed connection to the cross-beam ... the cross-beam would be in torsion.

mind you, if you have 4 good bolts attaching at the foundation, assuming the fixed end moment is conservative (the moment can't be bigger).

 

[SteelPE]

Years ago I went to a seminar given by the ASIC. They stated that since there are no true "pinned" bases and that you could realistically use a fixity of the base equal to 0.4EI/L. With the 0.4 being 10% of the fully fixed case and coming from using the old "G" charts for columns.

I've always tried to create my models the same way they would be constructed in the field. I can't really see you trying to fix your bases for one load combination and not the others.

 

[rb1957]

i would have thought there were no truly fixed bases. a single point interface (with a spherical bearing) is truly pinned.

the conpromise partially fixed joint can be easily modelled ... model the pinned structure, and then the fixed structure, and combine the two together as a linear combination = superposition (eg 40% fixed + 60% pinned).

 

[LSPSCAT]

Thanks SteelPE,

That is essentially where I am headding. Modeling the rotational restraint and moving on. The frame is similiar to a pipe support rack or large gas station sign. (Flagpole situation) No additional frames or support out-of-plane.

I am just having an academic dilemna in my head...all things being equal how does the base plate know how we expect it to behave?

 

[rb1957]

oh, the plate knows how to behave ... it's only our approximation as to how it behaves.

you have a pattern of bolts. clearly they can react some moment, but how significant is it ? when the bolts start to yield, as they would before failure, they allow the joint to rotate and approach a pinned connection.

assuming a fully fixed foundation is conservative for the foundation and representative of a joint that is tightly held together (bolts under preload); assuming a pinned foundation is conservative for the frame structure and representative of a joint free to rotate ('cause bolts have yielded, and relaxed their preload, and the joint is gapping).

 

[LSPSCAT]

Thats the last piece that I wanted to hear explained.

If the base plate / anchorage is not designed to resist the full moment we begin to form a hinge, allowing rotation - hence our assumed pinned behavior.

Thanks for you help and input.

 

[rowingengineer]

you could model it as a semi-rigid joint.

How could you do anything so vicious? It was easy my dear, don't forget I spent two years as a building contractor. - Priscilla Presley & Ricardo Montalban

 

[SteelPE]

In conjunction with rb1957. I went to another seminar where the speaker said "steel will do what you tell it to do". That is, as long as you have a positive load path for every force in the structure the system will work. You may end up with some cracks here and there....... but it will not fall down.

 

[ToadJones]

For a typical baseplate of thickness determined by vertical loading, and four bolts, I typically model as pinned. The plate with deflect and relieve moment.
If I need to use a fixed base, I typically go with some variation of the base assemblies you see on old (or new) mill columns...as shown in Blodgett.
If I want the base to behave more like a pinned joint, I may only use two anchors and place them "inside" the wideflange shape (if a WF is being used) of place 4 bolts "inside".

I was called to inspect a cantilevered canopy for a hospital once. There was a single row of cantilevered columns supporting cantilevered beams (inverted "L" shape frames).
The columns were massive rectangular tubes as were the beams.
Inexplicably, the base plates were 3/4" thick plates with a square four bolt pattern with the columns full pen welded to the plates. The plates were large in plan....say 36" square. A simple survey of the structured showed the columns leaning with no roofing material or any finishes on the structure. The beams were 90 degrees to the columns as they should have been. An elevation survey of the base plates showed some significant plate bending. The A/E firm designed this bases as fixed.

What a total nightmare to fix.

 

[MiketheEngineer]

Can you put "frames" in the other direction??

 

[LSPSCAT]

Frames are existing and similiar setup is proposed for future work so I need to validate the design and prepare new designs.

I do not have any of the original calcs or analysis except 40 years of service. I was just mainly hung up on the concept that for the same base plate I need require to modes of behavior. Out of plane requires fixity otherwise it is not stable and tips over. Now, say you apply the fixed base condition to the frame for normal gravity and vertical loads. A quick look at the anchors and base plates reveals significant stresses are developed.

Next, say the base is pinned and this problem seemingly goes away when analysing the base plate for vertical loads applied to the frame. My contension was the ability to categorize the base plate to suit the required results.

To add, the base plate is somewhat detailed this way. Base plate is gussetted to the column to transfer the loads in the out of plane direction. Anchor pattern is rectangular with short side in the plane of the bay.

Morning project turned into an all day affair! Loads are high so it has me checking - 200 kip.

Thanks for all your thoughts.

 

[asixth]

Model the base with some minor fixity (0.1*4*E*I/h). This will create a minor moment at the base which a 4-bolt connection should easily have capacity for.

 

[Teguci]

A further concern:
Consider plastic deformation at the base plate due to frame action (the theoretical pin to foundation). The anchor bolts or the base plate either fail in tension or yield. For the frame action, this doesn't matter since what little moment there is at the base can be redistributed to the moment connection. However, what happens for the out of plane direction where a full moment is required on the now deformed anchor bolts/base plate?

Clearly, if the anchor bolts fail in tension the structure is in jeopardy. If the base plate deforms, we are probably still OK due to yielding but our final deflections may be considerably more than calculated for the out of plane condition.

I think the answer for me is to configure the anchor bolt properly and eliminate tension failure of the anchor bolts except by yielding.

For anchor bolt configuration where frame action is up and down, the bolts should be close together as follows:
x x
C
x x

 

[TXStructural]

The methods of analysis we use is strictly notional. Unless you go to the difficulty to analyze the precise flexibility of each connection and member, by convention we model a standard baseplate as pinned. A 3/4" thick baseplate with widely spaced anchor bolts allows the columns to rotate significantly (in structural terms.) While there is some restraint, it is generally not adequate for building structures.

In an existing open frame, where deflections can be tolerated, you can imagine that there is some fixity, but I would be loathe to use that since anchor bolts are probably inadequate. Anchor bolts with bent ends (rather than headed ends) will slip and fail under the forces involved. Also, the bases may well be under- or unreinforced, and likely were not designed to resist rotation of the column bases.

I'm always hesitant to use "successful history of use" as a basis for any future use assessment, since few structures see full design loads during normal use, and seldom are they exposed to full lateral loading.