precast pile connection to pile cap

huangyhg

precast pile connection to pile cap
i am curious to know how others have detailed pre-cast piles to the "poured in place" cap?
i am looking for an equivalent shear connection to (4) #9 hook bars which are epoxy grouted into pre formed holes.  possibly a more economical approach.  the piles are 20"x20".
thanks so much,
mark  
check out our whitepaper library.
mjohan,
the standard detail of joining a precast pile with an in-situ cap is to "socket" the pile between 75mm to 100mm (3 to 4") into the cap.  the precast concrete above that level is broken up leaving all vertical pile reinforcemet exposed to a length equal to or longer than the development of the rebar.  the cap is cast as normal but incorporating the pile's exposed reinforcement and socketed section of the concrete pile.
it is done throughout the world and gives you a full moment connection as well.  there are even hydraulic tools specially developed to break quickly the unwanted top section of the precast pile, otherwise it is a straight forward job for the jack hammer.
thanks bbird,
what would you say the most common connection of a round pre-cast pile to a pre-cast pile cap is? a square pre-cast pile to a pre-cast pile cap?
i appreciate any comments.
seimic would be the governing force.  piles are both battered and vertical.  i am investigating alternatives to a current design, however, and am considering eliminating the battered piles.  in any event, can these connections be performed efficiently as pre-cast?
no comment on the moment connection details - not my area.  but i strongly urge you to eliminate the battered piles since you are dealing with seismic loads.  keep all the piles as plumb as possible -
focht3,
i am very appreciative of your response to my thread!  i have read many of your threads, regarding the elimination of battered piles.  
the job i am looking at is located in an inner harbor, baltimore.  a combination of battered piles and vertical piles have been designed.  i believe the piles will be approximately 50' long and are typically 20"x20" prestressed, precast piles.  i believe they extend through approximately 30' of water.  (at this point i have no geotechnical report.)  i am interested in the concept of eliminating the battered piles and allowing a more flexible structure.  it seems to me that the structure was designed in two parts... first, vertical piles were specified according to allowable spans for the cap beam... then, battered piles were located to withstand the lateral loads.  i certainly could be wrong, but that is my hunch.  at particular areas, many battered piles are specified, @ expansion joints, etc...
my question is, how do vertical piles typically behave in the redundency of small waves?
and... if the geo-tech recommended battered piles for the lateral stability, what could i say to initiate a possibly different recommendation?  ...i believe i read a p-y analysis??
it seems as if there might be approximately 1/3 more piles than required!  
mjohan,
i haven't done any pre-cast pile cap before and cannot comment your first question properly.  i do feel confused though.  
a precast pile is usually driven to a set and cannot always end up with the top of the pile at the exact level in the design.  by the time you cut off part of the pile the integration with a pre-cast cap must be problematic because a joint has to be achieved at the underside of the pile cap.  is it possible that you really mean cast in-situ cap?  
as far as driven piles are concerned the square section is more common because they are mass produced in a casting yard and a square formwork is a lot easier to fabricate than a round one, bearing in mind the piles can be prestressed in series during manufacture requiring the moulds easily installed in the correct position relative to the reinforcement.
i shall refrain from offering any comment on precast pile cap as i haven't done one before.  structurally i doubt one can achieve a full moment connection with a precast pile cap.  a pile cap exists for providing a foundation for a column and the setting out position of the pile cap msut be dead accurate.  people generally allow a pile, either driven or cast in-situ, to deviate from its design position by about to 3" or 75mm.  i cannot imagine in my little mind how by stacking another big lump of precast concrete on top of them will give you the accuracy required for the column which by any standard must be within about an inch or 25mm from the designed position.  i also have no idea how are you going to compensate the verticality of the column if the piles and cap are not 100% plumb.  
i must admit that battered piles are rare and few in between although they are fairly common for jetties and offshore work.  
i am surprised by focht3 that they are not good for the seismic.  does a battered pile not offer a higher lateral resistance than a vertical pile?  may be i should start reading the geotechnical section of eng-tips
the issue is relative stiffness.  battered piles are so stiff - in comparison to plumb piles of the same design - that they can lead to a catastrophic failure of the pile cap.  it's a pretty big deal, yet i find a lot of engineers (geotechs and structurals) have no idea that battered piles can be a real problem -
i doubt the pile cap will be precast.  never seen one, and can imagine lots of construction difficulties.  reminds me of the, "maybe you can draw it, but you can't wish it into place" warning i received as a young engineer.
increasing the pcc piles to 24 inch widths will have a big impact on stiffness.  they are much heavier and harder to handle, but considerably stiffer - you will get much more than a 33% increase in stiffness with the same number of piles.  and i would definitely use a p-y analysis to evaluate the piles.  don't forget to consider group action!
greatly appreciate the advice!
in your opinion, does 30 kips sound about right for the lateral load applied at the top of each pile?  not factored, no seismic reduction taken.
i'm sorry - too many unknowns.  thirty kips may or may not be a reasonable load.  i just don't know.