cast-in place bars versus epoxied

huangyhg

cast-in place bars versus epoxied
i have always wondered why there is such a large development length with rebar cast into fresh concrete versus a short one with rebar drilled and epoxied (hilti and powers) into existing concrete. after all, how can 1/16" of epoxy around the perimeter of the bar make such a huge difference in values without the surrounding concrete being affected?
i know the epoxy folks use safety factors on the order of 4 and the concrete folks use safety factors on the order of 1.7, but this would skew the differences the other way instead of making epoxy lengths so short.
can someone help me see what i am missing here?
check out our whitepaper library.
the rebar development lengths are based on mechanical anchorage (friction) between the bar deformations and the concrete.  i presume the epoxy anchors develop chemical bond between the bar, epoxy, and concrete.
i believe it is to do with the crical concrete surface in shear.
imagine that the critical failure surface is a cylinder around the bar.
for a cast in bar this cylinder is at the outside of the bars ribs. for an epoxied bar it is the outside surface of the drilled hole.
as the drilled hole is larger than the bar diameter, then the failure surface is larger and therefore it is stronger.
another factor that comes in is that the epoxy engages the surrounding concrete along the whole surface, whereas the rebar ribs give more localised stresses.
we called hilti one time about this very question, becasue it perplexed several of us in this office.
their response was that the values listed in their catalog are based on pull tests of rebars epoxied into concrete not on mathimatical formulas.
it is also the reason that one of the footnotes in their hva capsules table reads 'hilti does not recommend the use of alternate embedments other than those tested and listed above.' they do not have adequate mathimatical formulas to explain the behavior adequately, they just know it happens. there is no proper way to interpolate values between the embedment depths shown in their catalog.   
if i re  
my understanding is that the development length is based on a maximum achievable bond stress between the concrete and rebar.  as nutte mentions, the epoxy gets a much higher bond stress (to both the concrete and rebar), hence the shorter length required.
another thought comes to mind......is there a better or more efficient rib design that could relieve localized stresses and therefore shorten our lap lengths? is there any incentive for the industry or crsi to study this? if the rib design is improved, they would sell less rebar.
the adhesive anchors are very sensitive to edge distances; therefore if your concrete is not thick (i.e. thin slab) then the bar isn't developed as quickly, or maybe at all. just because one has the apparent embedment doesn't mean that the full bar strength can be developed into the concrete substrate.
the data that hilti or powers represents in their technical guide are based on rebar used as an "anchorage" in lieu of a splice.  so these shorter embedments shouldn't be used if you need to transfer load from an existing rebar into a new rebar, these embedment's should only be used if you would like to simply develop the rebar tensile or yield strength like you would with a standard adhesive anchor.  also, be aware that these embedments are typically ultimate values instead of allowable values for rebar anchorage.  if your desire is to lap the rebar into the existing slab and have load transfer from the new bar to the old bar in the concrete, the manufactures should recomend you to use the lap splice lengths from aci, not the values from the technical guide.  a description of this is on page 15 on this link from one manufacture.
interesting discussion ... i believe it has to do with the surface area of the cone formation which will have to fail, particulary in case of pull out for an anchor. and then, of course the properties of materials will come into play as well. pull out tests will normally result in a conical chunk of concrete breaking away from the parent body.
yogi anand, d.eng, p.e.
energy efficient building network llc
anand enterprises llc
packerfan raises an excellent point--even if the new bar is developed into the existing concrete, the existing concrete will "pull off" of the existing rebars, if they are not developed per aci.
daveatkins