Geomechanics- direct shear test

[Dell]

i am taking a geomechanics course, as a part of the course we are required to do labs, one of which is to test the shearing stress of a sand using the direct shear method.

we repeated the experiment 3 times changing the compressive stress each time, 0.75, 1 and 2 kg/cm2

from each compressive stress we found the appropriate shearing stress and then plotted the shear- compressive stress graph,

from the 3 points i did not get a straight line, and after using MATLAB to calculate a linear approximation i found that at a compressive stress of 0, the shearing stress was larger than 0, (i got 0.4) from what i understand this means that the sand has cohesion of 0.4 which as far as i understand it is not meant to have. C is meant to be 0

what could i be doing wrong?

 

[PhanthomJay]

Theoretically, c should be zero, but if the sand is dense, the graph is slightly curved, and if this is an undrained test, then the curvature is even more noticeable, with the origin (s,p) at (0,0). You might want to try it using a low value for the compressive stress. The value of c = .4 seems higher than you can account for due to experimental errors.

 

[Dell]

thanks jay, unfortunately i don't have another opportunity to use the lab before i have to hand my report in, so i won't be able to use a lower compressive value,
as far as the density goes, the sand i used was 1.55 [g/cm2]
the 3 results were as follows

compression=0.75 ; shear= 0.6263
compression=1 ; shear=0.9648
compression=2 ; shear=1.2368

i don't really have anything to compare these results to either(they may all be totally wrong) so any further advice would be appreciated
thanks again

 

[PhanthomJay]

Your sand appears more loose than dense, where the angle of internal friction is somewhere around 25 to 30 degrees.,. The tangent of the internal friction angle should approximate the slope of the (assumed) straight line curve from the test. Your slope gives a higher angle for the internal friction angle, especially in consideration of point 2. I guess you can explain it in your error analysis.

 

[Dell]

thanks, do these results seem at all probable or have i maybe processed the results incorrectly? i don't see where i could have made a mistake, yet i don't see any reason for there to be such errors in measurements, the entire experiment was done in a lab and all the readings were done electronically and sent automatically to ms excel. from there i graphed the horizontal and veritcal displacement and shear force divided by changing area. from the graphs i took the maximum shear stress (from each experiment) and made the shear/ compressive graph.

 

[PhanthomJay]

What value do you get for the slope of the graph when you fit the points to a straight line?

 

[Dell]

it gives around 40 degrees

 

[PhanthomJay]

Seems a bit high. I think since you were using a cohesionless soil like sand that you could start your graph at (0,0) and draw the interpolated straight line using the other values. It's been many a year when I last did such a test, so I can't recall. Where did you get the values to use for the confining compressive stresses?