restrained vs. unrestrained

huangyhg

restrained vs. unrestrained
folks,
recently there has been considerable debate in our office regarding whether the end spans/corner spans of post-tensioned flat slabs are classified as restrained or unrestrained.
refer a similar thread
in researching i find that there is considerable redistribution of moments in continuous systems due to the fact that negative steel remains cool in comparison to positive steel. also, there is a mention of tension membrane action. intuitively, i feel that the corner/ exterior columns provide axial and rotational restraint to a pt slab regardless of the presence of spandrel beams, but how does one quantify it and present it to a building official or a fire marshall?
does anyone have any experience in this regard for concrete buildings.
i note that a recent research (by troup et al.) concludes that most typically framed (with simple end connections) steel buildings are considered as thermally restrained.
thanks
does anyone have any opinions?
slickdeals - as you have found there is considerable literature on both side of the fence on this issue.  you can find reputable resources that say the end bay is restrained as well as unrestrained. as far as i am aware there is no definitive guidance and i would follow what is considered the standard of care in your part of the country.
personally, i treat it as restrained just like the rest of the slab.  reason being, it does not seem to make sense to classify an interior bay as restrained and the exterior bay as not restrained, when a fire at the interior bay (with less cover) could pop the same tendons that extend through the exterior bay (with more cover if designed as unrestrained) thus ruining the integrity of that bay.     
slickdeals,
i think it gets down to the mechanics of how a slab works under fire.
for an internal panel, the bottom reinforcement capacity under fire is not as critical as, with as lot of redsitribution, the slab can work with no positive moment capacity and basically act as 2 cantilevers spanning to the middle of the span.
for a simply supported span, there is no negative capacity possible so the cover has to be increased on the bottom to keep the temperature down and allow the bottom steel to still work.
for an end span, you cannot develop much capacity or redistribute much moment to and end column, so the positive steel needs to provide some capacity so more cover is needed than the internal case. if the end column was very stiff, then you could develop moment there and so it could be treated like an internal span.
so it really depends how stiff the end column is.
in most cases i would think it would be ignored.
you then have the problem hinted at by willsv with unbonded pt that if the strand loses capacity in the internal span, what capacity does it have anywhere? i would say that this one has been missed by the code committees. everyone else in the world using bonded pt does not have this problem. maybe you should change to the better product!!