pt design in ram concept - interesting phenomena

huangyhg

pt design in ram concept - interesting phenomena
i have been using ram concept for about nine months now, and one thing that i have run into is that sometimes, when i have lots of walls constraining my slab, i end up getting less precompression on my section.  say i put 20 tendons into a beam, with an expected precompression of 20*26.8 kips per tendon = 536 kips effective prestressing force.  after doing preliminary checks, 20 tendons appear to be the correct number, plug it into ram concept and the section is nowhere close to working.  when i do an audit, i see the precompression on the section is a lot less than the 536 kips . . . sometimes it is more like 200 kips.
i am not very experienced at pt design, but am learning and working under an experienced person.  anyways, this experienced person is used to the pt data, equivalent frame type of analysis, and he generally scoffs when concept does something like lowering the precompression down on a section.
anyways, i am reviewing a design on an underground parking structure done by another firm . . . a very experienced firm in the way of pt design . . . and this "bleeding" of the precompression that the section should see "shows" that there are a lot of "overstressed" areas.  does this phenomena actually happen, or am i modeling something wrong in the ram concept finite element analysis program?  has anybody else ran into this problem?
thanks for any help that you can provide.     
  
check out our whitepaper library.
i'm not sure if this is answering your question...assuming you are talking about post-tensioned, unbonded tendons here...but if you have your anchorage or your stressing end located close to a stiff vertical element, like a shearwall, a big portion of your prestressing force indeed will go to your shearwall instead of your beam or slab.  we were taught this in our pt class in grad school...forces flow to stiff supports if the anchor end or the stressing end is located too close to them.
here's a rule that i always found helpful when trying to envision pt behavior:
unbonded post tensioned strands impose forces upon slabs/beams as point loads at anchor or stressing ends, and as uniform vertical forces @ changes in curvature.
if you put a big horizontal point load near a wall, it'll flow to the wall before your slab or beam because the wall is so much stiffer....does that help?
yes, stiff walls will steal the prestress forces.
a phenomenon you must be familiar with for continuous pt beams over supports, is a secondary moment.  this secondary moment will change the physical eccentricity of the p/t force and will take away or add to your precompression.
i am not sure if it helps.
i guess i am familiar with all of these ideas (secondary moment.  i guess what i am questioning is how this pretty knowledgeable p/t design firm (my boss said that these guys are "pt gurus") made the slab work, but when i plug it into my ram concept program, it shows beams being overstressed because of losing this precompression.
are there tricks to use when making models in ram concept to allow more of the precompression to hit the slab?  and are any of these tricks "realistic," and model the true behavior of the slab?  i guess i am for a loss on how the engineer of record was able to make the slab work . . . the only thing i can think of is that i am modeling it as a flat slab while in actually it is sloped (concept can't model slabs as sloping) . . . should i avoid using concept at all on a sloped slab?      
for sloped slabs (you mean top surface sloped, bottom surface flat, like for drainage @ a balcony, right?) i think you just have to model a thinner element, maybe the average of the thickest and thinnest portions...have to use judgment.
if you are using unbonded tendons, the forces really will go into the stiff walls first, like your model says.  i would say try to move your tendon ends away from any walls.  
put a picture of your model up here...i think there's some way to do that
pour strips or slip joints are often used at stiff elements that would cause significant restraint forces.  perhaps the "pt gurus" are using something like this to make the slab work.  or perhaps they were intending to, but forgot to include the detail.  do you have the original designer's entire set of drawings?
the slab is sloped in ramps, as it is a below grade parking garage.  it is constant thickness.  
but, no, they don't have any pour strips in the below grade portion of the structure.  i have the entire set of drawings, and they also don't have any slip connections at the walls (such as an epoxy injection sleeve system) as near as i can tell.  all stressing is done by blockouts, they have 30 inch shearwalls in the center core and being below grade, has a 12" wall all around the perimeter.  some of these levels have been poured all ready, but i need to figure out the slabs capacity as i need to use it for my work . . . and i just can't replicate it with my model.
if you and your pt design mentors think this is a serious issue, then it's probably worth a call to the other engineer to resolve your questions.  they may have a legitimate answer that you haven't thought of yet.  or they may be in store for one of those "oh $#!%" moments.  i have made such calls a couple times in my career.  if the original engineer has made an error, they usually appreciate the heads up so they have a chance to fix it (or cover their behinds) before the lawyers get involved.  sometimes this kind of thing is best handled between principals of the firms.