Valid Pin Boundary Condition? 1

[duckhawk]

For a square moment frame, do you think this is a valid pin boundary condition at the base?

Where does the moment go?

 

[SE2607]

It will develop some moment.{/quote]

With tension in the anchors which may or may not have been designed for tension.

But to answer your question:
bending of the toe and stretching of the bolt

You are describing post yield deformations which exceed ASD code values, if, in fact, the OP was designing based on ASD. It doesn't sound to me like the OP was doing a collapse prevention or an ASCE 41 analysis.

Any moment developed by the base plate will go into the footing.If the footing was not designed as fixed, it will rotate and relieve the moment.

Not before likely exceeding allowable soil bearing pressures.

Once the rotation of the column end equals what it would be as a pinned base (per your analysis assumption), then it is effectively hinged. That anount of rotaion is very small. It should be readily available. A fixed base has ZERO rotation,

You are making a distinction between "very small" and ZERO rotation. Even a footing designed as fixed base will rotate. Anything with a load will deform.

so it is safe.

I do not agree, but I do acknowledge "standard practice" where based plates with anchors located within the flanges have been considered pinned for a long time. My contention is that this is a business decision in which there have been few failures and not one founded on structural analysis.

 

[SE2607]

Really? Surely that could readily be unconservative. The bolt positioning and the plate thickness doesn't lend itself to a high stiffness connection, it would fall in the semi-rigid category, hence my earlier nuanced replies.

The only place where I can see that this could be unconservative is in the moment connection between the beam and column. "Back in the day" before I had software that would do the work for me, I would use the Portal Method to design frames. I would design the moment connection for F x 0.6h and the base connection also for F x 0.6h. A modern day analogy would be to model it both ways and design the moment connection from the pinned based model and the base plate and anchorage from the fixed based model.

 

[271828]

Fixed.

The critical force effect in most frames like the OP's would be the moment at the top of the column, and a fixed base assumption just made that a lot smaller. Thus, you need to show the base plate is thick enough to justify fixed. What's the method for that?

Note that for flush moment end plates -- a cousin to the OP's connection -- there's a specific procedure for making sure the plate is thick enough to justify a fully-restrained (FR) moment connection. Otherwise, it would be a partially restrained moment connection.

I anticipate it being tough to make the OP's base plate thick enough to get anywhere near a fixed condition. I'm pretty sure the fixed assumption would result in an unconservative frame design.

 

[Ron247]

Most people I know would design this as pinned but they would consider that the baseplate and foundation design can affect where along that transition from pinned to fix occurs. Really thick baseplate coupled with a wide and thick foundation moves the needle closer to fixed while the thinnest baseplate that meets the needs and a more narrow and thin foundation moves the needle closer to pinned.

You already have wide steel section which moves the needle towards fixed. I have seen a design where the baseplate was not full length of the column. They did not let it get too close to the flanges. I have no idea how it faired over the years. The column depth is your biggest problem in this design.

I would design pinned, check it for fixed but not get alarmed if it was slightly over stressed under fixed.

 

[SE2607]

Really? Surely that could readily be unconservative. The bolt positioning and the plate thickness doesn't lend itself to a high stiffness connection, it would fall in the semi-rigid category, hence my earlier nuanced replies.

It would only be unconservative for the moment connection between the beam to column.

And for the record, to bluntly answer the original question I'd model that connection as pinned. But if I had any inkling that assumption would be unconservative I'd model it as both to ensure missing some important failure mode.

Modeling it both ways would be best. "Back in the day", I would design moment frames using the Portal Method using F x 0.6h to design the moment connection between the beam and column and F x 0.6h to design the connection (and footing) at the base.