roof truss

huangyhg

roof truss
a roof truss with a span of 35 meters ,it will be supported  on two concrete columns , the design was carried out assuming one support is pin and the other is roller ,but in the construction drawings 'the attached drawing' both ends are bolted down ,,is that ok , i have some concerns about this as the temperature changes are high in my place ,,moreover, the members near the support are not concentric, is that ok.
finally is there any standard procedure to test the truss.

if you are designing it as pin-roller, then you have to detail it that way.  if you detail it as pin-pin, then you have to design it that way.  i might try designing it both ways and use the worst case for each member, unless you can model the concrete beams directly in the analysis.  the beams will provide more of a spring support (laterally) than a pin.  this means that your situation will fall somewhere between the two analyses.  do both and use the worst case for each   
add differential temperature load case(s) on the truss, go back to the model, and change the support to pin-pin, then re-analysis the model with proper load combinations.
i would make sure the extended top chord (at end) can resist a moment from the support to the first upper pannel joint.  
also, you might want to explore the benefit of using slot holes at one end to accomodate the lateral movement.
depending upon column stiffness, and how much shortening of the truss is anticipated due to deflection, and how much thermal cycling you expect.
typically, you have to restrain the ends to prevent uplift and displacement.  it is standard practice that they be bolted and/or welded to supports, however the support conditions frequently allow rotation and some axial length changes.  be sure the bolted connections do not fix the ends from rotation due to deflection of the truss.
-on a typical span of 35 meters, with l/240 deflection, you should expect 14 cm deflection, with a corresponding shortening of truss span of a few millimeters.
-the thermal expansion due to a 20 c change of a mild steel on a   
there may not be a problem.  determine how much horizontal movement is required at the roller, and divide by 2.  this is how much the top of each concrete column must move (horizontally).
determine what moment is induced in the concrete column due to this horizontal movement.  if it is a tall, slender, flexible column, it may be ok.  if it is a short, fat, stiff column, it may not be able to accomodate the movement.
daveatkins
hi alladdin: try to use sap 2000 to design your roof truss
it appears that the end diagonal intersects the top chord a considerable distance from the reaction point.  make sure that the top chord is adequate in flexural strength to transfer the reaction to the truss   
the original poster mentioned the eccentric connection, but ba was the first to address it.
out of all the nodes to have an eccentric connection, the first diagonal away from bearing is probably the least desirable place to have one.  please pay special attention to the design of that truss bearing seat.