Abstract: The paper gives a comprehensive comprehensive overview overview about the application and design of mat foundations. foundations. SAFE software software has been used extensively extensively throughout throughout the paper for analysis of mat foundation. Rightful importance has been given to modulus of subgrade reaction, its measurement and relation to various soil properties have been discussed. The Transcona grain elevator has been a subect of extensive case study. The foundation is treated in speci!c as a mat foundation and not ust ust as a shallow foundation. The foundation failure has been discussed from not only the bearing capacity point of view, but ta"ing into account settlement considerations as well. #arious #arious limitations of SAFE S AFE and possible ways of overcoming overcoming them have been discussed. The numerous methods of averaging "s to get a single representative representative value has also been discussed.
TRANSCONA GRAIN ELEVATOR : CASE STUDY Background: The Transcona Transcona elevator construction construction began in $%$$, by the the &anadian 'aci!c Railway Railway company to provide easy means for grain shipment. The grain elevator was designed to carry a million bushels of grain. (t basically consisted of a reinforced wor" house and a bin house for storing grain. Site investigation was conducted before construction to ensure that the soil could withstand the load of the grain elevator. elevator. This included some small diameter plate bearing tests at foundation level. The tests indicated that the soil should be able to handle a pressure of )*)+-% "'a "'a +/ 1 ton0ft 2. (t was calculated that the total pressure pressure with the elevator !lled with grain would be a maximum of )$3 "'a).) "'a).) ton0ft12 4owever, these tests only stress the ground to a small depth and would have only tested the upper !rmer clay. Subse5uent boreholes showed the soil in that area comprised of an upper layer of highly plastic clay,overlying a layer of soft gray clay.
6or" house7 The wor" house was was 1$.) by 1%.13 m -8 by by %3 ft2 in plan, and /.*3 /.*3 m $*8 ft2 high, with a raft foundation ).33 m $1 ft2 below the surface.
9in house7 The bin house comprised of 3/ bins !ve rows of $) bins2, each approximately 1*.8 m %1 ft2 high and .1- m $ ft2 in diameter. The bins were supported by a reinforced concrete raft foundation, 8.3$m 1 ft2 thic", 1).- m -- ft2 wide and /%./ m $%/ ft2 long and at a depth ).33 m $1 ft2 below ground level.
&onstruction of the Transcona :rain Elevator was completed in September of $%$). (t was the bin house that underwent failure. (t began to be !lled with grain as uniformly as possible. ;n ;ctober $*, $%$), the bins contained )$,/88 cubic meters *-/,888 bushels2 of grain and settlement was observed, increasing within an hour to a uniform )8/ mm $ ft2. The foundation of the building was supported by a line of boulders on its east side which then permitted the grain elevator to sin" more so on the west side.;ver the next 1 hours, the structure tilted to the west until it made an angle 13< /)= to the vertical. The earth on the west side bulged up as the structure moved towards it and this slowed down the movement. The east side of the bin house moved away from the soil surrounding it, leaving a gap to the depth of the raft foundation. The clay below the foundation was *.*m 1% feet2 below its starting level on the west side and on the east side it had risen $./ m / feet2 above.
Figure $ 7Transcona grain elevator failure &ourtesy 7 >?A Engineering @td., ?anitoba, &anada2
Several washborings were made immediately after the failure. The elevator was underlain by layers of uniform deposits of clay.
Figure 1a2 7 6ash boring of site after failure &ourtesy 7 'ec" and 9ryant, $%/)2
Figure 1b2 7 6ash boring of site after failure &ourtesy 7 'ec" and 9ryant, $%/)2
Testng o! so" at ste: (n $%/$, relatively undisturbed samples of subsoil, were obtained and bearing capacity calculations were made. The materials were obtained far enough from the failure Bone, so that the soil is undisturbed by displacements and settlements. From the borings, samples 1 inch thic" and 3 inches in length were chosen. The natural water content and index properties such as li5uid limit and plastic limit were determined from these samples. For computation of uncon!ned compressive strength, the 3 inch specimens were trimmed to a length of )./ inches and were tested until failure in uncon!ned compression. The material was then remoulded at unaltered water content. This time the dimensions of the specimen were $-0* inch diameter and )./ inches length. The uncon!ned compression test was repeated to these specimens. The index properties are represented in the !gures )a2 and )b2 below. The ground level is ta"en at an elevation around --1 feet. >pto an elevation of -/,ie, upto a depth of *.1) m, the soil is tan and grayslic"ensided clay. The average water content is this layer is about /C and the average uncon!ned compressive strength is about $8/.) "'a$.$ ton0s5uare foot2, with a sensitivity of about 1.The values of li5uid limit and plastic limit are around $8/C and )/C respectively. According to &asagrande plasticity chart,thesevBlues correspond to inorganic clay of high plasticity. 9etween the elevations -/ and -)- in boring $, there is graysilty clay layer of thic"ness 1. m. (t has an average water content of /-C, an uncon!ned compressive strength of about 31.1 "'a 8.3/ ton0s5uare foot2 and a sensitivity of 1. The Atterberg limits are approximately the same as before. 9elow this layer,lies tan silty gravel, containing limestone chips and clay poc"ets. >ndrainedtriaxial tests were also conducted on specimens and they were exposed to lateral pressure of 1--.- "'a 1.% tons0s5uare foot2. These are mar"ed by D on fig. 1b2. The confining pressure has almost no effect on the compressive strength. So we treat the soil as having angle of internal friction, 8.
Figure )a27 (ndex properties of soil at site &ourtesy 7 'ec" and 9ryant, $%/)2
Figure )b2 7(ndex properties of soil at site &ourtesy 7 'ec" and 9ryant, $%/)2
GiHerential thermal analysis was performed to identify the clay minerals present. About twothirds of the material was illite and the remaining one third was montmorillonite. The nonclay portion was almost negligible. The speci!c gravity of this was found to be 1.-8.
Load at !a"ure: The load at failure, at the base of the elevator can be easily computed. The elevator held *-/,888 bushels of grain which weighed about 13,888 short tons. The dead load of the structure was calculated as 18,888 tons Allaire, $%$32. This load was distributed uniformly over the mat area 1).- mx/%./ m2. This essentially means that a uniform surface load of 1%)"'a).83 ton0s5uare foot2 acted over the mat. The mat was positioned ).33 m below ground level. So there will be a reduction of load acting on the mat due to this excavation, e5ual to IG, where I is the unit weight of soil e5ual to $*.*/ "J0m) $18 pounds0cubic foot2 and G is depth of excavation e5ual to ).33 m. So the reduction in load will be around 3% "'a and the net load acting on the mat is e5ual to11 "'a 1.) ton0s5uare foot2.
The ultimate bearing capacity of any soil is given by TerBaghiKs e5uation. 5u cJc L IGJ5 L 8./I9JI According to TerBaghi, when 8, JI 8, J5 $, Jc /.S"epton, in $%/$, proposed that when G09 is less than 1./, the value of Jc is given by,
Jc
(
5 1+
B
5 L
)(
1+
D
5B
)
(n case of the bin house, 9 1).- m, @ /%./ m and G ).33 m. This gives an Jc value of /./3. 6e will use this lower value of Jc for bearing capacity calculations. 5u can be written as, 5u cJc , since JI 8 and the value of IG is negligible comapred to that of cJc . c value is ta"en as half of the compressive strength since 8. The weighted average of compressive strength for both the layers is *% "'a. 6e must consider the entire depth of both the layers, $8.3- m, because it is lesser than 901 and the inMuence of the entire depth is felt on the raft.2 5u 8./N*%N/./3 1-.1 "'a This value is greater than the imposed pressure of 11 "'a on the mat. 9ut, however, this might be an optimistic value, as we do not consider uncertainities due to slic"ensides in the upper surface and due to diHerence in stiHness between two layers. Jo factor of safety is adopted. The worst case scenario would be to consider the lowest value of compressive strength,ie, of the second layer. This would give us, 5u 8./N31.1N/./3 $-) "'a 9ut it is improbable that bearing capacity would be this low.
Fig. shows a represntation of the bin and the underlying soil layers. The data available is used in SAFE to chec" for stability.
Figure 7 9in house and underlying soil layers
Est#aton o! #odu"us o! subgrade reacton !ro# t$e abo%e data: Since 8, undrained shear strength is e5ual to undrained cohesion, by ?ohr &oulomb e5uation s c L Otan 2 9owles relates undrained shear strength to modulus of elasticity as, Es $88 to /88 su when plasticity index is greater than )8. &onsidering the mid value, Es$ )88su )88 N /1.3- $/*8$ "J0m1 Es1 )88su )88 N )$.$1 %))3 "J0m1 From #esicKs relation, "s Es09$P12 "s$ $/*8$01).-$8.)12 -)%.*) "J0m) "s1 %))301).-$8.)12 )-.$) "J0m)
739.83 × 8.23 + 437.13 × 2.44
"sweighted2
10.67
3-8.3$ "J0m)
6hen this value is used in SAFE it estimates a settlement of )) mm which is close to the initially observed settlement of )8/ mm.
A!ter#at$ o! t$e !a"ure: The bin house itself suHered little damage, so it was subse5uently emptied and righted using ac"s. (t was underpinned with bored pile foundations ta"en to the limestone level.The !nal structure ended up with a basement at about $8 m below ground level. This failure shows the importance of proper site investigation and also that the entire Bone which is under the inMuence of the foundation should be studied.
