pinned or fixed connection

huangyhg

pinned or fixed connection?
consider a multi-span, steel, wide-flange beam, continuous over all supports.  if the bottom flange is simply bolted to the column cap plate, is this intermediate support pinned or fixed?  
two of my "mentors" disagree.  i need a third (fourth, fifth...) opinion.
thank you!
if it is an intermediate support and if the beam is continuous over it, then it is something between hinged and fixed.  basically it is called a continuous beam and you can use different techniqes to calculate shear, moment, deflection for a continuous beam (e.g. moment distribution, slope deflection, computer software, etc.).
of course depending on ratio of the span to moment of inertia of the beam you can approximate its behavior.
depends on the nature of the bolting to the cap plate as to whether i would consider it "fixed" to the plate. what do you mean by "simply bolted"?
if you are looking at the moment in a uniformly loaded beam, consider the two extreme bounding conditions:
1. span assumed to have two simple supports: m(max)=w(l^2)/8
2. span assumed to have two fixed supports: m(max)=w(l^2)/12
the result is that the denominator of the equation can be assigned any value between 8 and 12, depending on the judgement of the engineer. for example, when cofferdams were designed with hand calculations the equation m(max)=w(l^2)/10 was almost universally used for lower spans to determine bending stress in the sheet piling.
it is entirely possible that both your mentors are right (although one is probably more "right" than the other).
as stephened points out, it is a matter of the connection type - is it a full moment connection between the beam and the column? secondly, if the columns are long and slender they will not provide much rotational fixity, but if they are short and fat they will.
if you assume the supports to be fixed against rotation, you will get a maximum hogging moment over the supports of wl^2/12. if you assume they are pinned, you get a higher hogging moment in the end spans, roughly wl^2/9. -> it is conservative to assume pinned supports if moment capacity is your critical factor.
but you bet that thousands of these things have been designed as mere continuous beams, even ignoring any frame action.
in fact till 25 years ago was common to see in spain steel structures -most of them today still standing- where the columns where battened columns and the girders merely stood even in the same battens. eccentricity if some was that of standing on a batten, and the girders were continuous ones, point. i even re  
i am not sure why you want to fix the joint at the columns. for your information, a three span continuous beam is equivalent to a bent with two columns and one beam and the second and third joints are fixed (meaning they can carry positive and negative moments).
when i design continuous beams, i assume my supports to be pinned and i design columns (assuming the support is a column, it can be a wall, another beam, etc.) for the vertical reaction.
keep in mind, you are the designer, it is your call and judgment. you can design it pinned and or fixed. however, you must provide a practical detail for that joint that will deliver and behave as you assumed and can be constructible!
good luck
if sway is prevented, it's conservative to consider it pinned.  in actual performance, it's likely very nearly fixed with a rotation (caused by the beam) having an input on the effective length; this assumes that the stiffness of the column is considerably less than the beam...
imho, you will need to consider the nature of the bolted connection.  if the bolting is such that the wf beam cannot move without imposing a rotation into the column and does so with some high degree of reliability then it may be treated as fixed.  if it behaves otherwise, it is pinned.
of course, this assumes that a yield condition in the columns doesn't exist.  if that is the case, you could have a rigid connection that will produce a rotation within the column section and will ultimately behave as a pinned connection anyway.
i hope this helps.
i appreciate this question , many young engineers have confusion over type of connection to be assumed for a particulae situation. they all go by the literal meaning of the term.
to answer this querry, my view point is that the moment connection is the one when the top & bottom flange of the beam is well built into column flanges, in case one wants to design as framed connection. a case like the one where  only the bottom flange is simply bolted to the cap plate of column, does not imply that the beam is built into column. hence the most of of the moments(99%) are carried into beams itself for which the beam shall be designed.
hope i have clarified  to eit2

if your spans are  more or less of same length and the loads on the adjacent spans (right and left of the column of the interest) are more or less the same you can assume it pin connected in the vertical loads point of view.but if the above condition is not the case, it depends the rotational rigidity of the column and the transmission degree of the moment at the joint to the column by the bolts.you should observe and judge it.
in case of earthquake( if it is of concern), if i were you, i would want my connection to be fixed.as lutfi says you are the designer, make it fixed/pinned and assume nothing about the joint.
uk2211