masonry wall on slab on grad

huangyhg

masonry wall on slab on grad
any recommendations for the capacity of a 4" concrete slab on grade to support a masonry (or any other material for that matter) wall?
i have a client that would like to install a new 8" masonry mechanical chase in a commercial building.  the wall is 10' tall.  i would like to avoid saw cutting the slab, and would like to build a masonry wall with #4 verts at 48" o.c. directly on the slab, epoxying the verts into the slab just for fun...
in the past, i have justified light concentrated loads to slabs on grade by assuming the load spreads out at a 45 degree angle in the slab, and using a decreased allowable bearing capacity.  therefore, if i use an allowable load of say 800 psf, and have a 4" concrete slab, i would end up with an allowable line load of 1066 plf for an 8" wall, which would accomodate a 21' tall masonry wall, assuming it weighs approx 50 psf.
any thoughts?
if the chase wall takes axial, structural load (which in your case it doesn't sound like it does) then the code would restrict "footings" to a minimum depth of (i think) 8".
since i'm guessing its just a "partition" and not supporting axial, structural loads then it may be permissible.
however, by adding some weight to the slab, there is the potential for the slab subgrade to give a bit, and this would initiate cracking around the chase wall, just outside the perimeter of the chase due to downward movement and the subsequent negative bending in the 4" slab.  that may be ok with you and the owner but that is the full risk you are taking as i see it....and after typing this - its probably a pretty small risk.
consider light weight block just for kicks?
add thickness to the top of the 4" slab. bush hammer, wire brush, roll on "moose milk" or other bonding agent and place another 6" of thickness to give a suitable footing for your new wall.
in order to design this "suitable" footing, it seems i would need to know why the slab, as is, is not suitable...  would you guys agree with my proposed analysis?
page 54 of the 2nd edition of designing floor slabs on grade has an example of this situation.  i created a spreadsheet to crunch the numbers.  this example assumes it is plain concrete, which you could reasonably assume since it is only 4in thick and most likely has some temp/shrink reinforcement.
assuming:
f'c=3000psi
t=4in (slab thickness)
fs=2.5 (factor of safety)
k=200pci (subgrade modulus)
i come up with the following wall loads:
interior wall - 1050plf
exterior wall - 814plf
hope this helps.  i don't have this book in front of me, and i can't re  
to make the solution work, you should minimize the weight of the walls
use lightweight block conforming the astm c90 for loadbearing masonry units. you can use 8" thick units. some engineers would argue that 6" would be better, but the weight savings are not that much. make sure the block are less than 105 pcf and 95 pcf may be common in your area. this should give you a wall weight of less than 40 psf.
you can periodically use short (8"-16") epoxied dowels in the slab.
a hidden bonus may be the increased fire resistance and insulating value of the lightweight block.
use empirical design criteria from aci 530. this should not be a problem if you can provide some lateral support at the top of the wall and do not have some unusual industrial loads (unprotected from fork lifts, etc.). the h/t requirement can even work for 6" block if it is used.
do not use grout and vertical reinforcement except at openings. use joint reinforcement at 16" o.c.
dick
maybe i'm off base, but, doesn't az have seismic activity?  shouldn't that be incorporated into the design?  i know i did a masonry partition wall for a building in asheville, nc and if i re  
chipb -
good catch! -
it could be possible if the az refers to a location. then, it is up to the code requirements for that area. empirical design is not for all areas, but it is very good where the detailed analysis and "simplifying" assumptions are beyond most engineers ability if you try to design/justify what does and has worked.
industrial situations can be even more difficult than the normal site specific wind and seismic, so often these are the cntrolling factors.
i designed my first reinforced masonry building for both seismic and blast. unfortunately, the assumptions did not include the improbable blast 6 months later from a "bubble" of vented hydrogen gas 100' directly above the building that exploded from static electricity. - the walls held, but most of the over-designed roof was on top of the instrumentation and recording equipment in the block house.
dick
you should check the slab on grade as a beam or plate on an elastic foundation.  checking the bearing on the soil is all well and good but you need to check the concrete also, flexure and shear like anything else.  you'll a subgrade modulus for the soil, or you'll need to assume a value.  
the army manual has a solution for this.  google tm 5-809-12.  
rgerk has the right answer.  a cmu wall (load bearing or non-load bearing) on a slab-on-grade does not load the slab like it would a strip footing.  the slab bends under the wall (tension in the bottom of the slab), and bends either side of the wall in the reverse direction (tension in the top of the slab).  i use the pca book, like rgerk.
daveatkins