Concrete Grade/Tie Beams to Reduce Lateral Deflection of Steel Frame Structures

[newyorkranger]

Hello all,

Long time browser of these forums, first time poster.

I'm researching low rise (a few storeys) steel frame structures and looking at alternatives to cross bracing in order to reduce lateral deflection due to wind and seismic loads, with one method being the use of concrete tie beams located at the footings.

I've done some basic modelling and have generated some good results, however I'm trying to find some good papers, journals and research into these tie/grade beams.

Has anybody come across/worked with concrete tie/grade beams? A lot of the papers/documents I find reference combined footings which I understand work in a similar manner to adsorb moment due to eccentrically loaded footings often located on property boundaries.

Cheers.

 

[JAE]

To utilize the grade beams like that the key I think would be to successfully create a full columns-to-concrete moment connection - hard to do sometimes (especially with Appendix D).
Also - I would be extra careful with your analysis model/assumptions as the actual effective moment of inertia of the grade beam can vary greatly depending on the applied moments (see ACI 318 chapter 9).

 

[Ron]

Agree with JAE....

It is easier to develop fixity in the column with a large footing and an embedded column. That can present a variety of problems, not the least of which is corrosion of the column at the top of the embedment.

Using grade beams is moreso a way to prevent overturning of the footing subjected to a high moment. As JAE noted, you still have to achieve fixity at the baseplate...difficult unless you embed the column.

 

[msquared48]

I was thinking just that earlier Ron, but it seemed to be too much of a hassle - running the column continuous through the grade beam to a pinned footing connection below to attain the leverage needed. But it seems like a complicated solution for the contractor to make a mistake to me.

Mike McCann
MMC Engineering

 

[newyorkranger]

Thanks for raising some great questions guys, it seems it's not as clear cut as I made it out to be.

I was thinking of using the grade beams and pinned supports as an alternative to the costly fixed footings, which depending on the soil capacity may present further issues in itself.

 

[Ron]

Mike...I've used that approach many times for aluminum columns with high base moments. Works for steel as long at you coat the interface at the top of the concrete with asphalt mastic.

 

[KootK]

The latest edition of the AISC seismic manual covers a system of this nature in example 4.4.4 (SMF Embedded Column Base Design). Mechanically, it's pretty straight forward. The construction details do not seem very constructible to me however. That being said, I've done this once and didn't get so much as an RFI or a raised eyebrow in a construction meeting. I don't plan to test my luck again any time soon.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[jvvse]

Embedded columns are very expensive because it delays the pour schedule. You have to pour the column footing first and once it cures you have to come back and pour the tie beam/grade beam.

The ultimate goal is trying to create fixity at the base to reduce deflections. If you model it, you need to model the actual properties of the grade beam in the model and I would even use cracked properties of no more than 50% of I. Then you need to make sure your base connection can transfer those moments. Keep in mind that you need to account for the lengthening of the anchor bolts which will increase your rotation of the column and thus your deflections. I think AISC has provisions for that but I haven't gone that far myself.

Link below for a simple sketch on how to model a 1-bay, 1-story frame with a grade beam. Keep the supports pinned at the ends and model the actual properties of the grade beam (reduce I for cracked).
Link

 

[Ron]

jvvse...delays are not always necessary. If uplift on the column is minimal, a styrofoam blockout can be used for the concrete placement and then removed to set the columns later. The columns are then grouted in place.

 

[JAE]

Sometimes the cracked moment of inertia (full Icr) is about 17% of I_g...not 50%.

For pure bending you usually see the effective moment of inertia more like 35% of I_g(per ACI chapter 10 approximations).