dowells to connect a new concrete

huangyhg

dowells to connect a new concrete
i need to extend a culvert (6' each side), for that need to design a new culvert without bottom (frame) and join to the existing culvert approx. 28' long.  i want to know the best way to connect the new concrete section to the existing concrete frame.  the beam or slab and the walls of the frame are 8" thick.  the new section will be 6' wide (each side of the existing culvert) and 28' long.  should i use dowels or some other method to connect the new concrete? how should i specify the join?
check out our whitepaper library.
use of dowels is the time-tested way to do this.
slideruleera, thank you very much for your response. could you please tell me how you would estimate the length of the dowells??. i can estimate the number and the diameter using the shear but have no to much idea about the length.
on the other hand, could you give me your opinion about placing the new portion of curlvert independently of the old.  is it necessary to join the old with the new frame (culvert without bottom and had no fill over the top)??
why no make it work separetly? are the settlements very important (i know it depend on the soil but i have no data for the soil)?.  what do you think about the option of making work the culverts separetly and providing some sealants for the joints. do you know what are the advantages of using dowells?
i am sorry if i am asking obvios things for some people, but for me this is new.
thanks!!
ggc69889 - in my opinion, a culvert should be a continuous structure for at least two reasons:
1. when water flows through it, you want to minimize the chance that the water can get into the cracks between the the old & new concrete.
2. if a external load is applied to the culvert you don't want differential settlement.
length of dowels is limited by the length of bits that are available to drill into the the existing concrete. twelve inches is probably the minimum embedment that i would consider... but that depends on the details.
you may want to go to my website (link below) and download the out-of-print document "concrete culverts & conduits" on the portland cement association page. probably not a help on this problem, but a good reference if you work with these structures.
ggc-  usual practice to install dowels is to embed them in slightly oversize diamater holes into the existing concrete and attach them with either a structural adhesive or grout.
the minimum dowel embedment depth and hole diameter are usually governed by the adhesive manufacturer's recommendations, and a consideration of how far away the dowels are from a free edge and how often spaced.
the actual embedment depth that is specified will also depend on the load that the dowel has to carry.  adhesive manufacturers such as hilti and powers/rawl publish test data that indicate pullout and shear strengths for various types of dowels, base materials and substrate geometries.  suggest you get your hands on one of their catalogs.
the minimum projected length of the dowel can be determined using typical aci data for rebar development length.  length will depend on whether you plan to resist shear or tension or both.  in your case it sounds like no specific force can be calculated, you just need to tie the new and old sections well together.  
consider using a concrete bonding agent at the joint between new and old, as well as dowels.  the b.a. is not expensive and can't hurt.
  
if you go with hilti products, they have the hy-150 product that probably work pretty well.  they have very specific requirements for size and depth of holes, follow their instructions to the letter for those items.  you can go to the hilti website and download a program to figure the proper embedment length for their products.  other vendors may have similar products; follow their recommendations for embedment length.  
regards,
chichuck
slideruleera and samdamon, thank you very much for your answers, they are very helpful. i am also looking at the information you recommended.
if possible, i just want to know something else: why in my case, no specific force can be calculated in order to determine the forces the dowells are going to take ?? isn't it correct to determine the size and number of the dowells considering the shear produced at the edges of the span, at least for dawells along the leg of the frame (my frame is 9' long (span) by 6' wide, very small)?? probably for that frame or culvert size, typical size of dawells are enough, however, i would like to know if my assumption to size the dowells is correct.
thank you very much for all your help and time. ggc
chichuck, i just read your message. thanks a lot for your response. i just got a magazine from hilti and i am going to know more about dowells.  again thanks!!!
ggc- i haven't ever designed a culvert, but assume they work as some kind of arch.  not sure exactly what the load would be at the joint between two identical arches coupled together sideways.  might be some shrinkage forces as the new concrete in the added arch cures, might be some shear due to differential settlement as the new arch compresses the subgrade beneath its haunches and thus wants to move down or possibly rotate relative to the old one.  
anyway good luck with your culvert .
ggc69889 - in your case the size (diameter) of the dowels will be limited by the edge distance. for 8" thick concrete the best that can be accomplished is a 4" inch edge distance (maybe less, depending on where it is possible to drill holes in the existing concrete). once you have "backfigured" dowel diameter this way, then hilti, or similar catalogs, will recommend a minimum anchor spacing (for that diameter).
these calculations will determine the maximum number of dowels that can be used... that is just the best that you can do under the circumstances. the nice thing is that it is usually "good enough".
i want to thank everybody who responded, everything was helpful. i finally desigend the new culvert as a box curvert without bottom slab (a frame). to tie the new and the old section i used dowells #7 x 1'6" long @ 12" c.o.
i read most of the referenced you recommended.
thanks !!! ggc