design of shear key for steel column

huangyhg

design of shear key for steel column
hi,
i am to design a shear key for a steel column of a braced bay in a pipe rack. since there is a significant amount of shear at column base which is difficult to be borne by anchor bolts ,shear key has been envisaged.
my problem is where ever i seen calculation for the design of a typical shear key element (rolled section or plates in + shape welded at the bottom the base plate )flexural stress of this element (shear key depth is 200mm + 25grout-10mm gap = 215mm)has been calculated .
my question is why we need to check the flexural stress of this element when there is no place for this element to bend (shear key will be filled with non-shrink grout)?
shbh - it may be theory vs. reality and sometimes reality is harder to calculate.  in this case, the "theory" of the bending is the conservative and safe route to go as the grout may crack, crumble, be flawed somehow, and some bending actually occur.
one other method that you could consider is to embed horizontal steel plates in the top surface of the concrete, on either side of the base plate.  then, after the base plate is installed, bolted down, and grouted, you can place twin vertical plates that weld to the embed and extend up the sides of the base plate - with a welded connection there.  this would then take horizontal shear out of the base plate and deliver it to the concrete support.
i use a method similar to what jae describes with a setting plate (supporting anchor bolts) with headed studs embedded into the concrete. i use side plates then field welded to the setting plate and base plate.
actually my concern is, for high shear if you check shear key for bending then you land up with bigger size of this element and ultimately you are forced to place your anchor bolts outside the column flange. so even being a pinned connection modelled in analysis this actually becomes a moment connection. moreover bigger size base plate and further even bigger size plinth.
it seems like you are saying that if your anchor bolts are outside the column flange then automatically it is a moment connection. in order for column bases to resist moment you must reduce the rotational contribution of these elements:
anchor bolt elongation
base plate bending
footing rotation
ok, you are saying that even if i have bolts located outside the column flange, this base still can behave like a pinned one if i don;t introduce stiffener plates....agreed.
well coming back to my original question, do i still  need to check my shear key element for bending. even if you have grout/vent holes in the base plate do you still see any possibilities of pockets etc in the shear key.

you need to account for the tab bending. every publication i've seen says to calculate it, so unless you have specific test data to say otherwise i wouldn't neglect it.  i dont think it's as simple as saying the tab is totally supported on both sides, because once you have 10 or 20 kip acting on the face of the tab you need to take the stiffness of the grout vs stiffness of the baseplate/tab into account.  i'd suspect the baseplate would allow enough rotation that you'll get full bending.  one reason why they suggest matching the baseplate and tab thicknesses.
you have to check bending in the shear key plate, and it does become thick and bolts will probably have to move out.
even though widely spaced bolts will tend to act like a fixed connection this is probably not going to be significant in a braced frame.
you should consider concrete bearing and breakout. bearing will be important for sizing the shear key but if you have small pedestals then breakout could govern design. for breakout i would use aci318 app d with fictitious bolt centers at the edges of the shear key. otherwise i don't know of any published methods.
thanks all of you for you valuable suggestions. i got some design aspect given in aisc steel construction manual which are quite useful.
thanks again for you comments guys.