anchor bolts between wood sill and foundations wall

huangyhg

anchor bolts between wood sill and foundations wall
i have a question regarding sill plate anchorage requirements.  the 2000irc requires that the wood sill plate be anchored to the foundation with minimum 1/2" anchor bolts at maximum 6' on-center.  i am concerned about basement walls that retain earth, but are not cantilever retaining walls.  these walls are braced at the top by the floor framing, and at the bottom by the footing or floor slab.  the sill plate is typically pressure treated southern pine, which is loaded perpendicular to the grain.  for a permanent load, the durattion factor is0.9.  therefore, according to the nds the strength of 1/2" bolt embedded in concrete with 1 1/2" sill plate is 400*0.9=360 pounds.  if you have an 8' wall retaining about 6 feet of good soil, the force at the foundation to sill interface is approximately 230 plf assuming at-rest soil conditions.  therefore, a 1/2" bolt is required every 1.57 feet.  a 5/8" bolt would be required every 2.2 feet, and a 1" bolt is theoretically required every 3'.  the force on a bolt at 6' spacing is 1380 pounds.  
what am i missing here?  i know that basement walls are going in with 1/2" anchor bolts at 6' spacing, but they seem to be overstressed.

i get, generally, your numbers.  the at-rest condition is appropriate, but wonder if the numbers we are used to using in commercial, etc. projects are a bit higher than used for residential.
that wouldn't explain the 6' spacing vs. 1.5' required by your calcs.   just thinking of our geotech friends and their 3.0 and 4.0 safety factors vs. what really occurs over the life of a home.
also, i see lots of basement walls with cracks and bowing vs. bolts shearing off.  perhaps at-rest pressures aren't entirely appropriate.
i would like to emphasize that my above numbers are for a good granular soil.  if the backfill is silty/clay, the applied load will increase.  i think the problem with basement wall failures is that people apply the code minimum requirements for anchorage and wall type without considering if those minimums will really be adequate for the actual site conditions.  i also find it troubling that i can't get the anchor bolts to calculate for even good soil conditions.  some claim that friction between the sill plate and the foundation make up the difference.
i think another problem is that reinforcing is not a requirement in residential basement walls.  at my last firm we did a lot of residential type work, and bowing/cracked basement walls were a relatively frequent problem. i'm not talking about a vertical shrinkage crack in the concrete, but horizontal cracks.  once a plain concrete wall gets a horizontal crack, it is incapable of spanning vertically.  even a small amount of vertical rebar gives a great benefit in flexural strength and ductility.  i think this could also be done at minimal cost.  
i also find it disturbing that the code does not appear to address specifically the anchorage of basement walls to wood sill plates.  i wonder if their rationale for the 1/2" bolts at 6' is based on past construction practice (without engineering backup)which hasn't resulted in an unacceptable number of failures.
i always believed that bolting the sill to the foundation was in response to horizontal and uplift forces that may arise from  storm events, e.g. hurricane, tornado, etc.
bjb,
the book "residential design guide" by nahb (2000) deals with this specific topic in example 7.7. they say "this conflict between analysis and experience creates a dilemma for the designer that may only be reconciled by making judmental use of the "extensive experience clause" in nds 7.1.1.4". i highly recommend this book if you do much wood design. to quote the forward, "this text strives to "bridge the gap" between conventional engineering practices and conventional residential construction practices".
mrengineer,
i looked for the above referenced design guide on their web-site with no luck.  can you pin point it for me?
thanks,
chip
often, small basements can function as open top boxes. the wall then acts as a panel that is pinned on the bottom, fixed at the sides and either free at the top or supported by the floor (pinned).
if the walls are too long then they start to behave like you describe (pinned at the top and bottom).
i think this anchor bolt requirement is a holdover from the days when structures were relatively small and the walls were short. the prescriptive codes do not get updated as often as the engineered codes do and there is alot of momentum not to change from the contractors.
i live in nc and when i first moved here there were many houses with basements with no  (yes no) anchor bolts whatsoever.  now most have 1/2" bolts at 6'-0"  and i have yet to see a single failure.
when the wall is 8' or less  i just let the code govern but if i am asked to design a wall that is over the 8' in height i call for larger bolts and shorter (way shorter) spacing.  but when i do so,  i feel like a fool.  but i go by the numbers.   so, is there a flaw in our engineering or what?
mrengineer- that book sounds great but as an engineer i feel i have to rely on sound engineering and if the book can't back up it's ideas with engineering principles and i rely on those ideas and a wall collapses, i know who will get the blame and it wouldn't be the book.
it appears a lot of people are confused over this issue and i know one thing for sure and that is  i am confused.
by
chipb,
i have seen failures where inadequate anchor bolts were part of the problem.  i am currently working on a project where the anchor bolt failed the wood sill plate.  most of the failures that i have seen in residential basement walls are ususally due to inadequate flexural strength because reinforcement is not provided.  these failures may not be news worthy, but in my experience i have seen them.  when the backfill is not suitable, such as silty clay soils, the probabilty of a problem goes way up.  the code prescribed connection does not take into account the nature of the backfill, yet it does take this into account when giving wall thickness requirements.  i think that full scale testing should be done to try to understand the actual behavior, because what calculations using acccepted engineering practice show to be completely inadequate often but not always work in the field.  i also think that flexural reinforcement in all basement walls should be required by the code.