addnl s re surcharge load on retaining wall

huangyhg

addn'l ?'s re: surcharge load on retaining wall
first, thanks for all the help on the previous ret wall/surcharge question. aggman's response included a method for which i have an another question. in addition to determining the pressure distribution vertically on a wall due to a point load it also determines the distribution laterally along the wall length (i.e. each direction horizontally from a perpendicular line from load to wall). naturally, the further you get from the perpendicular the lower the pressure, therefore the less impact on overturning. if you only consider the 1'-0" strip at the perpendicular then overturning is overstated because that pressure doesn't exist everywhere. my question is, what wall length should be considered for resistance to overturning? two times the perpendicular dimension?
i think for a reasonably lengthed wall you need to look at the global stability of the wall.  the key to this though is to make sure that your wall is stiff enough (lateraly and in the footing) to transfer the non-uniform loads and engage enough soil to maintain stability.  my typical application would be say a front end loader feeding a surge bin for aggregates.  so i might have a wall that is 25' tall and 50' or 75' long.  i have two concentrated loads and the pressure distributions overlap.  the limits of my loads are pretty well defined and i can easily check the wall so that it can globally handle the surcharge loads.
as far as distributing the wall pressures to the wall itself i again think that for a reasonable wall you can look at the stem as a cantilevered plate.  at least this is what i have done in the past.  you could use fea for this but what i have done is by inspection of the bell area of the pressure curve i distribute the "bell" pressures over an area spreading out on a 30 degree distribution.  this is probably conservative and you could get away with 45 degree also but that is for each to estimate on their own.  but essentially you can break the bell into a resultant load and you will get a series of loads making a pressure distribution which can be used to generate and cantilever moment.  i then distribute the cantilever moment over the width of the bell curve plus a 30 degree distribution on each side.  this would be in addition to the normal lateral earth pressures due to no surcharge.