beams shear splices vs. moment splices

huangyhg

beams: shear splices vs. moment splices
hi...i work in an office (difficult to get clear answers from people, so that's why i come to this website) where we have 3 different basic steel beam splice details:
1.  bolted/bolted moment splice (flange plates bolted, web  
    clip angles bolted)
2.  welded/bolted moment splice (bot flange plate bolted,   
    top flange has no plate-it's full pen welded so
    as to create an even bearing surface, web clip angles
    bolted)
3.  shear splice (web clip angles only-no flange attachment)
i have been told to keep moment splices near pts of inflection/zero-moment areas...but then when would you ever use a shear splice, which has little moment capacity?  thanks!....biz

the moment connections are more expensive and for roofs where you do not get unbalanced live loads your inflection point would not move around.  in fact, by placing a "pinned" connection at the splice, you are forcing the structure to behave with an inflection point always at the splice.  this affects the moments that occur throughout the length of the beam.
for floors, where you have the requirement for alternating live loads, you can still use the non-moment shear splice.  again, you are forcing a pin into the structure and affecting the resulting moments.
by using moment splices on a floor you are adding to the rigidity of the continuous beam and perhaps reducing the beam size required....but this is offset by a larger cost of splice.
there is no point in putting a moment splice at a point of zero moment. i think beam splices always have to be designed conservatively, therefore, i would tend to specify the position of the splice on the beam and design it for the relevant moment and shear at that point.
one word of caution, dont forget to take into account the deflection. there will always tend to be a very slight 'give' in the connection due to the bolt hole tolerance.
i usually proportion roof framing beams using plastic design and use bolted end plate connections for combined shear and moment.  using plastic design nearly always requires splices for both moment and shear.  welding end plates to beams is fairly economical and with the sequence of erection, it is not normally required to provide for shims.  the connections are fairly simple to design; the end plate is generally thicker (and more costly for material only) to reduce prying action forces caused by the moment connection.  i try to stretch the beam lengths to be maximum (fewer pieces to handle); this can be a bit interesting when it comes to providing a slope for roof drainage and locating splice locations.
the additional connection cost is minimal and usually 'disappears' with a competitive bid.  the economy in the reduced section size is, generally, far greater.  this is partly because i have a personal quirk about only having 4 high strength bolts for an 18" deep beam if that's all that's required for shear.  with the additional bolts that i would spec, i can 'make' the cost difference disappear <g>.
also with any continuous construction in steel, elastic or plastic design, i use web stiffeners (even if not needed; another personal quirk).
the following for simple supported continuous beam design is somewhat simplified:
the alternate loading scenario virtually disappears with plastic design proportioning and the member is essentially proportioned for factored dead load and live load based on it's failure mechanism.  the inflection point for total load is a start for the splice if required.  on the total load moment diagram, the minimum loading moment diagram is superimposed.  this is only to help determine where the splice should really be located and to provide a magnitude of the moment connection required.  the splice point is 'juggled' to a location between the original point of contraflexure to equalize the original moment and the reduced load moment.  the new moment diagram uses the reduced loading, but, keeps the original plastic design support moments.  the splice is proportioned on the maximum of either the moment for the original loading or the moment for the reduced loading and maximum shear at the connection.
the canadian code requires that plastic design connections connect for 25% of the moment capacity; this is usually fairly easy to attain with high strength bolts and often provides the required splice moment.  in addition class 1 sections must be utilized (less prone to local buckling).  this may compel the designer to select a less cost effective section.
deflection calculations are a bit of a nuisance; as a first guess for udl loading i usually use 1/3 of the simple span deflection (1/3 * 0.00624ml^2/i; m in 'k, l in ft and i in in^4) for udl.
an example using a 2 span continuous condition of say 24' and 30' and a factored udl of 2klf.  combined length is 54' ok.  the elastic moment would be approximately 2*27^2/9 = 180 'k, the max simple span moment 2*30^2/8 = 225 'k and the plastic design moment 2*30^2 * 0.0858 = 155'k.  no splices for either elastic or plastic and stiff pl over support for both, just to illustrate the economy of the section required.  (i haven't checked these for the actual section required and it may be that some of these converge based on sections available.)