Grain Size Distribution by Hydrometer

[MRM]

The hydrometer test, according to the ASTM guidelines, says to air dry your specimen. After the test you dry and weigh the specimen, correcting the whole deal by applying a hygroscopic moisture correction factor and correcting for the additional mass of the defloc salts. Ok, that's fine. It'll work.

My question is why not dry the sample completely using a lower oven setting, use the actual dry weight so that no hygroscopic moisture correction factor and defloc correction factors are required, place the sample in a container with defloc solution and allow it to soak for several days prior to mixing, testing, etc. Does anyone see any problems with my proposed method?

Thanks for your input!

 

[Focht3]

That used to be how the test was performed; as I recall, the results were erratic and didn't agree with other tests/methods. (We would oven dry them, grind with mortar and pestle, weigh out enough material, then put in cylinder with dispersant and water.) I don't recall when the change was made - I suggest that you contact ASTM to see who to contact, or get papers describing why the change was made.

The same is true for Atterberg limits - we used to oven dry, then grind them up with a Waring blender. (Isn't that special!) But that method was shown to consistently result in Liquid limits that were 5 to 10 points lower than the current "don't dry it back" approach. Perhaps the hydrometer change was influenced by the Atterberg change; I don't know.

But I'd stick with the current method unless you have a very good reason to alter it...



Please see FAQ731-376 by [blue]VPL[/blue] for tips on how to make the best use of Eng-Tips Fora.

 

[davecooper]

MRM,

Even if you did dry the sample in the oven at a lower temp (60c) you would still have hygroscopic moisture in clay. Clay has bound water that makes up part of its structure. If you cook the bound water out of the clay at a higher temp (110c) the size of the clay particle will change, hence the air-drying procedure. The deflocculant, meniscus and temperature will always have to be corrected for. I think it is time to come in to the 21st century and start using laser diffraction. I’m not very knowledgeable in soil chemistry; I hope I got it right.

David Cooper
Cooper Testing Labs, Inc.

 

[Focht3]

Hmmm,

I saw a laser diffraction system demonstrated in the mid to late 1980's. I wasn't impressed: very small sample volume, didn't deal well with "coarse" particles (i.e silts), easy to clog, and incredibly expensive!

I'm already uncomfortable with the small sample size of the current hydrometer test method; and I have a real tough time calling only 2 to 5 grams of soil "representative."

Improvement is needed, but the results must be reasonable, reproducible and "representative."



Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.

 

[SirAl]

Regardless of sample prep, the presence defloculant (sodium hexametaphosphate)in the water will affect the fluid density thus affecting the hydrometer reading. Unfortunately, I don't think that you can get away from adding a defloc agent when using this test procedure.

 

[MRM]

Thanks for the responses, everyone. I agree with Focht3 regarding the laser diffraction system. I've seen some tests run on that equipment in the past, and the results did compare reasonably well to the grain size by hydrometer, but the machinery plugged easily, had very small specimen sizes (which may or may not represent the soil being tested), and it was indeed expensive to operate.

I have enough of a challenge in convincing clients that they should let me obtain Atterberg limits or get grain size by hydrometer for another $100. I can only imagine what they'd say if I suggested they give me another $1,000 or so to get grain size dist. by laser diffraction. As I've written in other threads, as more and more people look at geotech engineering as a commodity, and more and more geotechs are providing recommendations based only on gut feelings (without any testing whatsoever) and praying for a good outcome, it is getting extremely difficult to properly evaluate soils. That's a whole other thread though!

davecooper
I had a thought on your response and was wondering if you could help me out. If I were to air dry my specimen only and place the specimen in a defloc solution of some kind, what guarantees do I have that the cations (and maybe some anions too) that are locked in the bound water layer (double layer) would not interact, and possibly disassociate from that double layer, causing the clay particle to change size anyway? I would think it would depend on the actual makeup of the double layer, makeup of the defloc solution, exchangeable cations, temp, pH and everything else. But maybe this effect would be more pronounced if the specimen were completely oven dried and then subjected to a defloc solution. Any experts on clay chemistry or electric double layer interaction out there?! I think it's somewhat beyond me.

 

[dirtsqueezer]

Here's what I think of ASTMs. Regardless of sample size, ASTM has produced a practically usable method on which to base recommendations in the real world. I think they can legally and practically, as a soils report or earthwork section. There is error and limitation inherent in a repeated test method just as there is in a geotechnical report. For example, how much of the total ground is explored out of an entire project site, compared to the amount of sand used from a sample? I have sometimes wondered just how damaging a large log between boreholes would be to a slab on grade. How much do you really know? I think the point is that we know all of these limitations, and that's why we make the recommendations we do, with the confidence we do, and the disclaimers we write along with them. If nothing else, you can base it on finance. Setting aside the empirical work and expertise that goes into the making of ASTM test methods, even if there were variables such as that cation/anion reaction, you would have the method, and therefore the company to hold accountable. On the other hand, if you were to deviate from the method, as in the hygroscopic question, it would be you that 'broke' the contract, and legally, your problem. I have heard it said that the business of geology is based on integrity. If one knows one's limitations and is legally and practically, I think these test methods are at present, at least less of an unknown than others you may have on a project. Follow the method and its limitations, and you're legally safe. Just as you expect designers to follow your recommendations according to the limitations you have included. I think ASTM at least makes it possible for some elements of the project to be predictable and acountable. I don't believe I've heard of an accreditating agency for contracters, for example...

Good thread!

 

[rochplayer]

I think part of the reason for air-dry versus oven dry is the problem of that oven drying can change the mechanical properties of certain soils, particularly residual soils in tropical and subtropical climates. Even air drying certain kinds of these soils may result in irreversible changes to plasticity. This may be evident in the hydrometer tests as colloids and clays agglomerate into silt-sized particles, skewing the distribution from the natural state.

 

[DRC1]

Dirtsqueezeer has a point. ASTM tests have their good points and not so good points but they are ASTM tests. ASTM tests are a comprimise between many points. Considerations include accuracy, operater judgement or skill, and repeatbility. However the test do change over time because people like you continually look for ways to get better results. I would split the samples. Do the test in accordance with ASTM and report those results. Then I would do the test as you suggest ( although I wouldn't oven dry. My understanding is that it can materially affect the clay properties which will adversely effect the the test). I would repeat your test method and report those results seperately if you are happy with the results and carefully explain how you used the results in any conclusions you drew.
Let us know how you make out.

 

[GTSLABS]

The simple solution is to oven dry the entire sample AFTER the hydrometer test is complete.
This will give you the wt of dry soil used in the test directly (after subtracting out 5 grams for the defloc.). Then there is no assuming the Hygroscopic MC correction.

We have run our Hydrometer tests on the -#200 fraction for years as permitted in ASTM D422 and have always had excellent proficiency test results.


Steven Grimaldi
Geotechnical Testing Services, Inc.
AASHTO and ACOE Certified Laboratory

http://home.comcast.net/~gtslabs

 

[Robert168]

If there is a difference in result between these two methods (air dry vs oven dry), is it negligible in most cases?

If the difference in results between the two methods is minor than that between two operators using the same method, then it will not make sense to be confined to one method.

 

[GTSLABS]

Let me clarify. I oven dry my 1L of sample AFTER the test. I start with -#10 air dried material but I don't record the weight because I get it after the test is complete. I wash my 16h soaked air dried sample over a #200 sieve and transfer that material, which includes the defloc. solution to the hydrometer beaker. Run the test and then in 24 hours oven dry the 1L of -#200 material to get my "W" in the formula.

So my test is still run on air dried not oven dried material.

Steve


Steven Grimaldi
Geotechnical Testing Services, Inc.
AASHTO and ACOE Certified Laboratory

http://home.comcast.net/~gtslabs

 

[iandig]

In accordance with the BS test for the same procedure, the sample is NOT dried before it is tested. The sample of soil is prepared, and two 'identical' sub-portions taken. One portion is dried at 105-110ºC, and the other portion used in the analysis. The moisture content from the oven dried portion is used to calculate the dry mass of sample used in the test. I have not dried the sample for analysis prior to testing for many years,[and unlikely to do so as I no longer get dorectly involved with the sils testing] and so I cannot comment on the repeatability of the two different procedures, but the reason for adopting this method was driven by the very concerns at the beginning of this thread.