Plasticity Index Survey 1

[BigH]

Statement of problem: Contractor has a residual soil to be used for, say, subgrade - subbase, base and bitumen layers to follow. He samples the soil, and determines that the soil consists of 80% coarser than #40 sieve; 20% finer than #40 sieve. He determines the PI of the material finer than #40 sieve to be 20. He has a requirement that his PI for use as subgrade be less than 12.
Question: What is the material's PI? Can he use the material as subgrade?

Reason for asking? - a simple survey of the "experts".

 

[Focht3]

[grin]

I've dealt with this one before!

As far as I'm concerned, the material is fine for use as select fill under most circumstances. It's essentially inert and has enough clay binder to hold together moderately well. It's not a very good waterproofing material, though - too much granular material.

This material can be subject to more post-construction 'shrinkage' if it isn't placed on the wet side of optimum. And it can be very sensitive to moisture changes.



Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora. See faq158-922 for recommendations regarding the question, "How Do You Evaluate Fill Settlement Beneath Structures?"

 

[BigH]

So, Focht3 - what is the material's PI???

 

[Focht3]

Well, according to ASTM - 20. The Atterberg limits tests only apply to the -40 material. While I understand the reason for the way the test is specified/reported, I have long thought that a 'rock correction factor' should be applied to account for the presence of the +40 material in the "whole" sample. Of course, this concept isn't 'important' enough to academic-types, so no research has been done (to my knowledge) to address this issue.

For me, the material is essentially a non-plastic to very low plasticity (0<PL<10) material, but I can't point you to a study to support my opinion. That's 'just' what my experience says -

Bean counters and pencil pushers giving you fits again?!

Don't - just 'em!





Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora. See faq158-922 for recommendations regarding the question, "How Do You Evaluate Fill Settlement Beneath Structures?"

 

[BigH]

As the Scots would say "I".

 

[dirtsqueezer]

I have a question. I am copacetic with discussion thus far, but isn't the PI used to evaluate liquifaction potential? I thought liquifaction was a relatively well-studied phenomenon. Is the percentage of material over the #40 not part the body of knowledge correlating the two? I've never seen any study on this, I just thought sand was usually present in cases of liquifaction, and figured there had been research linking it to atterburgs. Please feel free to disregard if I am way off base.
Thanks.

 

[BigH]

dirtsqueezer - it might be but the percentages of fines is one that is more common. This question isn't about liquefaction, though, so . . . can you give your two cents worth to the question posed????

 

[VAD]

Hello Big H:

I agree in principle with Focht3 re the ASTM test method that the PI should be 20. However, the Atterberg limits are done on part medium sand to smaller sizes.

An examination of the gradation curve is necessary since the medium to coarse size fraction, if significant, can indicate that the material is of lesser plasticity in a field assessment.

On the other hand, if the fraction retained on the 40 mesh is of gravel or larger sizes then this blending to reduce the plasticity would not occur and one has fines as distint from coarser material. This is not an aggregate that is usually wanted.

I have noted that such aggregates if used for crushing one has to be careful about the plasticity characteristics.

The material should be suitable for subgrade use.

Curious about LL and Pl of material.

One can also do an Atterberg Limit test with all the sand sizes included. This is sometimes done on sand clay materials despite the test method.

Regards

 

[dirtdoctor]

Big H -

This specification has reared its ugly head on SCDOT jobs for years. To get around it (when samples are only marginally above the PI spec of 13), we suggest to grading contractors that they blend fine sand with the material, thereby lowering the PI. SCDOT has also toyeed with the idea of specifying AASHTO A-5 or better soils, but slightly clayey ML soils almost always classify as A-6 or A-7-5 soils. So, upper stratas of Piedmont residual soils have become useless for fill on SCDOT jobs. This is quite distressful to local graders who were always instructed to save the 'red clay for capping out roadways (it's actually not a clay, but an ML/MH).

I am still waiting on some geotech genius at the FHWA or AASHTO to come up with a modified PI that considers both the PI and %fines. Just use a specified constant to multiply the PI by the %fines, i.e,

LL - 48
PI - 19
%fines - 45%
% +40 material - 25%
USCS - SM
AASHTO - A-7-5
MDD - 103.5 pcf
OMC - 18.5%
% swell (95% compacted CBR) - 0.3%
CBR value (20 lb surcharge) - 7

Here's a stab at it:

Modified PI = (.45/.75)*19*designated constant
= 11 (that's what I would guess it to be)

I ask each of you, would you be hesitant to use this material as embankment fill?? Well, we can't use it on roadways in this state. As most of my graders say, 'all the good dirt's long gone around here'.

 

[SirAl]

Hi BigH

A good question and good responses.

In my opinion, the PI is a way of characterizing the fines component of the material you are studying. I can't really see any practical way of projecting this result into an 'equivalent PI' to account for the amount of +40 material. I suppose the PI of the fines component of a product can be a useful parameter in qualifying say an ACP aggregate where an excess of clay particles could coat larger aggregate particles and thus be detrimental to the final product. As for a subgrade application, I'm not sure why this would be specified. Perhaps drainage is an issue. I really don't see a problem with this material used as subgrade in most situations accepting the limitations that Focht3 has brought forward.

 

[BigH]

There is, as Focht3, knows a bit more to this problem - can't say for now - but we have specs and specs are mute on mixing (so take that as mixing is not allowed) -

dirtdoctor - there is something out there called the Plasticity Product. This is used in TRL Overseas Research Note 31 for crushed rock crushed gravel or crushed boulder base courses where the PI is limited to 6. If the PI is approaching 6, then they want to limit the fines - so they define PP as PI x % passing #200 sieve. They recommend that this be maxed at 45. So if PI = 5, then % passing permissible is 45/5 or 9%.

As SirAl likely knows, the TRL is a great wealth of information and hasn't really been mentioned much (if at all) in the threads to date. They have a nice book on designing roads in mountainous areas as well as another on surface dressing (TRL ORN3).

 

[Focht3]

I haven't read TRL OR Note 31, but suspect that the PP limitation has to do with maintaining a freely draining material. Under some circumstances this may not be a critical design parameter; in others it likely is of great importance.

[blue]BigH[/blue]:

Does India have any potassium chloride (KCl) mines? What would the Indian authorities think about using a soil treated with KCl so that PI [&asymp;] 0? KCl is better than lime in that it is fully soluble in water, so much less mechanical mixing is needed for it to be effective. Anyway, a thought -



Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora. See faq158-922 for recommendations regarding the question, "How Do You Evaluate Fill Settlement Beneath Structures?"

 

[steve6076]

These sorts of soils are common in South Africa. The South African design standard (TRH 14) provides subgrade to a maximum PI = 10, EXCEPT for materials with a large coarse fraction for which a higher PI of the soil fines may be acceptable.

As a guide, the acceptable PI can be calculated from:
Max PI = 3 GM + 10
where the grading modulus GM is given by

Metric: GM = (P2 + P0.425 + P075)/100
where P2 is the percentage RETAINED on the 2mm sieve, P0.425 = 0.425mm sieve, etc (note its percent retained and not percent passing as normal grading analysis is expressed. So percent passing of 20% means a percent retained of 80%. Typically GM = 2.1 to 2.9 for these sorts of soils).

American GM = (P#10 + P#40 + P#200)/100

EXAMPLE
The original problem had 20% passing #40 sieve. Assume the rest of the grading is 10% pass #200 sieve and 35% pass a #10 sieve (or in metric, 35% pass P2.00mm, 20% pass P0.425mm, and 10% pass P0.075mm). Then the:

GM = [(100-35) + (100–20)+ (100–10)]/100 = 2.35

Maximum allowable PI = 3 * 2.35 + 10 = 17.

So the contractor should not use the soil for a subgrade. In South Africa, he would probably mix it with sand as has been suggested in posts above.

 

[BigH]

Just to follow up - the specs didn't specifically allow for mixing. In some countries, I am finding out, they stick to the word! Most specs will say "residual soil" natural or mixed meeting . . . For interest, there is also something called the plasiticity modulus - this is the %passing #40 sieve times the PI in % (i.e., in my case 20x20=400). This is then used as a given range - sort of like dirtdoctor's modified PI. For low traffic gravelled roads, I've seen the PM to range from 200 to 1200 but PI must be less than 20 in rainy areas or 15 in now monsoon zones. This gives enough of Focht3's binder to keep the coarser particles together.

Thanks for the great discussion - it is good to see the different practices about the world.

 

[steve6076]

BigH

I think that the " plasticity modulus" might derive from Phil Paige-Green's (and especially Frank Netterberg and also Sampson) work on unsealed (unpaved or gravel) roads. The TRL people also picked up on it from joint research in Botswana. Performance-related specifications have now been developed for southern African conditions (TRH 20, various conferences, or CSIR Transportek) which are pretty exciting.

Phil took it further. He found specs to pick the wearing course (basecourse or top layer) gravel materials so as to avoid those that could: ravel or corrugate / be slippery / erodible / dusty. He plotted the shrinkage and grading properties of a potential unpaved wearing course gravel on his graph of Shrinkage Product (linear shrinkage times percent passing 0,425mm sieve) versus Grading coefficient (Per cent passing 26,5 mm - per cent passing 2,0 mm) x per cent passing 4,75 mm)/100, and the materials could be chosen.

He had a general spec of:
Shrinkage product = 100-365 (< 240 preferable)
Grading coefficient = 16 – 34
CBR soaked greater than 12
plus limits on the oversize.

He also provided for engineering judgement to be used in his graph. In flat, dry areas, materials falling into some zones on the graph may be acceptable if the site-specific potential to erode or become slippery is not excessive. Similarly if there is low traffic and a high maintenance capability, the use of materials prone to corrugation may be acceptable.

 

[BigH]

Steve6076 - you are probably right - the only place I actually saw the PM was in a Laos spec - written by a Brit. Do you have any downloadable references that you've quoted? Where could I find out more. We also have a problem with "earth shoulders" of similar ilk. What is acceptable, etc. Thanks. I'm giving you my favourite (the closest I can find for "snaps" or Parliamentary "table pounding") for your great info.

 

[steve6076]

I got lucky and found the following:

It's on page 9 of:

http://www.ilo.org/public/english/employment/recon/eiip/asist/publ/techbrf/techbrf9.pdf

I reckon there might be things on earth shoulders as well, but I haven't read it thoroughly yet.

 

[BigH]

Terimah Khash! Muchas Gracias!