Issue combining sieve analysis and hydrometer data using ASTM D7928

[ahgeo11]

I'm wondering if the sieve and hydrometer performed on the same sample should meet up smoothly when combined on a curve. Say my sieve ends at 0.075mm = 28.5% passing, but my hydrometer starts at 0.0505mm = 33.54% passing, that is quite a jump in the graph if they were combined. Am I missing something in my understanding that these should theoretically match up with a smooth curve? Or is the percent passing in the hydrometer referring to something different and they need to be corrected to line up.

Hope this makes sense, thank you.

 

[ahgeo11]

If I adjust the calculated values of N[sub]m[/sub] using Adjusted=N[sub]m[/sub]*(percent passing 0.075mm of sieve)/100, the values come out quite low with a large drop from the sieve results to the hydrometer. I guess I'm wondering if they're supposed to make a smooth looking curve or not.

 

[zelgar]

Per ASTM D7928:

1.3 When combining the results of the sedimentation and sieve tests, the procedure for obtaining the material for the sedimentation analysis and calculations for combining the results will be provided by the more general test method, such as Test Methods D6913 (Note 2).

Note 2: Subcommittee D18.03 is currently developing a new test method “Test Method for Particle-Size Analysis of Soils Combining the Sieve and Sedimentation Techniques.”

 

[TigerGuy]

I've never seen them link smoothly from the sieve to the hydrometer. Most samples I've processed contained considerable clay, which does not separate from the coarse grains very easily.

If my rusty old memory is correct, D422 had guidance on how to reconcile these differences.

 

[DanielSgeo]

Ugh ever since D422 went away this has been an annoying issue. The standards do not adequately address this in my opinion.

One thing is you should always do a no 200 wash after running the hydrometer. Then you can do the correction method I discuss below.

I make the adjustment by adding or subtracting a percentage to the hydrometer data based on my sieve as follows: if -200(Sieve) = 30%, and -200(hyd) =25%, then I add 5 percent to all the hyd data. This in effect shifts the hyd curve vertically on the plot and preserves the shape.

When this doesn’t work is when you have a significant amount of material retained on the split sieve (default No 10) such as at GM or GC.

 

[TigerGuy]

Shouldn't the material retained on the split sieve be washed, prior to sieving? Everything that washes through the split sieve is then presumed to have the same gradation as the material tested in the hydrometer.

 

[DanielSgeo]

The split sieve is typically the no. 10, and the first hydrometer reading typically corresponds to a particle size of about 0.05 to 0.06 mm. So you would have a gap in data unless you also do a sieve down to the no. 200, at which point I think you end up at OP’s original question.

 

[morecoffeeplz]

There is most definitely an error somewhere. Ensure you are using the correct specific gravity for this material and also double check the dry mass of the specimen.

 

[TigerGuy]

It's been a few years, so maybe my comments are out of touch with the procedure. After every hydrometer I ran, the cylinder was poured out and washed over a #200. Once dried, the retained material was sieved over a #20, #40, #60, #100, and #200 (I think) to fill in the gap between the #10 and the hydrometer data.

 

[ryan7crew]

This is fairly common when combining hydrometer analysis with sieve analysis. You'll need to dive into the theory behind hydrometer analysis to understand why. Hydrometer analysis is based off of Stokes Law, which is assuming the soil particles are spheres (which they are not). Stokes law only reports a diameter. Sieve analysis only reports the minimum sieve size the particle passes through. Basically, its not quite an apples to apples comparison, although it appears that way.

 

[geotechguy1]

A hydrometer is giving you grain size distribution approximating flat-shaped clay particles (maybe still surrounded by ions etc) as a sphere.

A sieve is giving you a grain size distribution approximating a wide range of circly-angular type shapes based on how they fit through square sieve openings

They wont match - and you shouldn't expect them to. However I have a theory that lab managers everywhere long ago got tired of answering this type of question from engineers and they just started fudging the results so the curves line up and the engineers don't ask so many questions

 

[ahgeo11]

Would you use a hydrometer result to classify silt vs clay? Or keep your analysis to strict particle size results and not infer whether the material is silt or clay?

 

[DanielSgeo]

There are several correlations between index properties and other soil properties that use "Clay Fraction" which is based on particle size, i think usually 40 or 4 microns (i can't remember exactly). For example Stark's residual and fully softened clay strenght parameters.