Compaction and Compressibility

[Okiryu]

Hi, how can I relate a specific grade of compaction to its compressibility? For example, if soils are compacted to 80% modified proctor, how can I relate it for settlement calculations? I guess I should start looking at the void ratio for the 80% compaction. Soils are silts with high water contents. We are stabilizing them with cement for roads subgrades, but looking at the Proctor results, soils may only be compacted to ~80%. I was worried about settlement issues due to low compaction degree. I think that other issue is that the subbase/base may not be able to be compacted properly if the subgrade gets compacted only to 80%.

 

[Ron]

Yes, comparing void ratios is the proper approach; however, it might not get you very far depending on the soil character.

I'm surprised that you can only achieve 80% of the modified Proctor. Why? Can you post the Proctor curve and a couple of in-place density test results? What method are you using for density tests? What is the character of the silt? Is it calcareous? Organic?

 

[Okiryu]

Hi Ron, I do not have field density tests. My assumption for the 80% compaction is because of the water content of the stabilized soils. I will post the proctor later. The silts are volcanic silts.

 

[Okiryu]

Ron, see attached my proctor curve. Based on the natural water content, per my proctor curve, I can get a dry density of approx. 8.4 kN/m3 which is ~80% of the MDD.

Also, double checking using this equation:
dry density = wet density / (1+W%), I got similar dry density values to what I got from the proctor curve. BTW, these silts without cement stabilization have very high water contents (about 100%).

My question is, what may be the implications when allowing only 80% compaction of the subgrade. I feel that it may be a problem when compacting the subbase to say 95% modified proctor. Perhaps also some settlement issues. But, I am still figuring out how to quantify/proof with numbers this issue...

 

[EireChch]

HI Okiryu,

So the main reason for not achieving say 95% DD is because the soils are too wet of optimum MC. Am i correct in saying that after cement stabilization the MC is at 75%? That seems very high.

We generally just undertake CBR testing for subgrades and dont specify a DD percentage. Regardless, what difference does the DD matter if its cement stabilized? The increase in strength through cement binding may not be readily picked up in a DD test but should increase strength dramatically with the right mix. I would assume that a cement stabilised subgrade would easily achieve a CBR of 25%. This should allow a standard pavement build up.

In the event that you are getting a low DD% or a low CBR say less than 3%, you may need to look at geogrids. Geogrids would be a good (but expensive) option as they can bridge over softer soils.

Or revisit the stabilisation method and develop a cement/lime combination that achieves a good CBR.

Interested to see what other more experienced engineers think.

 

[Okiryu]

Hi EireChch, thanks for the input. CBR is not an issue, I can get CBR values greater than 3% for the subgrade after stabilization. MC was also about 75% during CBR test, so the stabilized soils is achieving acceptable CBR values at low compaction degree. I am getting acceptable compressive strengths and elastic modulus at 7 and 28 days. However, my only concern was about the compaction degree. I was thinking that about 80% is too low.

 

[EireChch]

If the material is too wet then it is going to be very hard to achieve 95% or close to for you subgrade. The only way to increase the DD would be to somehow dry out the material before compacting. Discing could be an option.

If the material wasnt cement stabilised i would be concerned with 80% as you may not get good compation in your subbase due to cushioning effects from soft subgrade. However in relation to your quote below. I dont think the 80% is an issue as you are getting good CBR values. You should easily achieve 95% compaction in your subbase when compacting on a subgrade of CBR 3+.

ps - I would have thought you would have gotten better CBRs for cement stablilisation. Did you do some test cube mixes? Say 3,5,7, 10% of cement and crush them? Sorry bit off topic but just interested.

 

[fattdad]

There is no reason you can't achieve 90 to 100 percent compaction of the subgrade or of the fill being placed. You have to control the moisture content to be compatible with the compaction process. The laboratory moisture-density relations help with that process.

You need to develop the moisture-density relations for the silt. In that regard, you'd need to develop the Standard Proctor (or Modified Proctor) moisture-density relations. That'd be ASTM D-698 or D-1557, respectively. Either of these would be important if you are looking to compact the native soil without any cement/lime stabilization (i.e., prepare the site for subsequent fill placement). However, if you have plans to improve the subgrade AND also use cement/lime in the subsequent lifts of new fill, you'd need to develop the moisture-density relations using the cement/lime stabilized blend. There are separate ASTM standards for developing moisture-density relations on lime/cement stabilized soils. Use those!

To actually gauge the field compaction, you have to take the field density tests immediately after compaction (i.e., before the cement or lime is fully hydrated). If you are doing the tests after several hours (or the next day), you are not following protocol and you'll get incorrect results.

I trust that you will actually seek out opinions from the geotechnical community and am surprised that few geotechnical engineers are providing opinions. Sure you are now up to two, hopefully something helps. I must say to the OP 80 percent relative compaction is almost like not compacting at all. Under the worst of all possible conditions, you'd be hard pressed to get below 75 percent relative compaction. Then again, you may be talking about relative density, which is completely different. 80 percent relative compaction is akin to 0 percent relative density, so that doesn't even make sense for a silt!

f-d

ípapß gordo ainÆt no madre flaca!

 

[BigH]

I am not sure of anyone who would compact a material so much higher than the optimum moisture content. Almost every spec I've seen requires that the fill be "conditioned" to an acceptable moisture content - in your case in the order of 50% IF the compaction curve is correct. Why wouldn't you dry it out first to approximate better your optimum moisture content?

Also - I am not sure that I have ever seen anything other than organic silt that would have a dry density in the order of 10 kN/m3 - this is the unit weight of water . . . so my question would be - what is the specific gravity of your silt? I've had volcanic residual clays with standard MDD of 13+ kN/m3.

As a bit of an aside, Dr. Jian Chu wrote an interesting paper for the 19th SE Asia Geotechnical Conference on "Innovation in Soil Improvement Methods". In it he describes, now that Indonesia is basically refusing to export sand to Singapore, the use of sludge or soft clays in land reclamation. To reduce the settlement, i.e., cause more settlement during construction times rather than normal consolidation after construction, they are looking at using horizontal drains (like PVD) between layers of the fills and then applying a vacuum to the drains - which causes quick consolidation of the fill. It is an interesting concept. I heartily recommend that my colleagues obtain a copy of the paper . . . a good read.

Getting back to the OP, he has not described if this is for a road, a land reclamation, or . . . but it might be a novel real test to use horizontally placed drains under vacuum to consolidate the fill quicker and reduce long term settlements.

 

[Okiryu]

Hi BigH, thanks for your comments. Typical Gs values are about 2.7, so the void ratios are quite high (about 2). These soils have unique engineering properties (high water contents, high sensitivity: that's why CBR values of untreated silts are very low, so treatment is needed). This is for a road with medium to high traffic. I am checking more mixing trials. I know that if we add more cement, the water contents will decrease and then can get better compaction degrees...

 

[Ron]

Okiryu....how are you determining the density? Are these trials based on a typical soil-cement approach or are you using the CBR as your only measure of stability? Also, it is difficult to get an accurate moisture content in a cement stabilized mix during the hydration process....you have two things going on at the same time...free moisture and the loss of that free moisture to chemical binding during hydration.

More commonly, an unconfined compression test for cement stabilization would be used as the strength/stability correlation rather than a CBR. For lime stabilization, the CBR would be ok.

EireChCh.....why specify a CBR value if you don't also specify a compaction percentage? They go hand-in-glove. The CBR value is some stability value AT some compaction value. For instance, do you want a CBR of 10 at 100 percent compaction or a CBR of 10 at 95 percent compaction? These can be vastly different in their moisture sensitivity.

 

[fattdad]

residual silt derived from volcanic ash in Panama with a maximum dry density under 60 pcf. I couldn't believe it and flew to Panama to check out the lab - like they were out of calibration or something. No, it was real and they were right. We built a big facility on that earthwork, but the foundations were drilled shafts through the silt fill.

f-d

ípapß gordo ainÆt no madre flaca!

 

[EireChch]

Ron - Ill be honest and say that I have never really thought about it like that. In my area the soils are typically residual silts and clays. If we were doing a road cut we would test the subgrade by undertaking a hand held shear vane test. This would give us an undrained Cu. We correlated this to a CBR by the old rule of tumb Cu/22 = CBR%. Anything below CBR of 3 would require some proof rolling with a sheeps foot or over excavation and replacement. Regardless of 95 or 100% wouldnt a CBR of 3 be adequate? (obviously depending on vehicle loading etc, the roads I have designed and overseen construction on where housing estates roadways and single lane roads)

 

[Ron]

EireChCh......it depends on what you put over it. I worked on a project where there were high wheel loads and the design used a CBR of 3 to 5. The overlying pavement failed. When we did our preliminary evaluation of the failure, we found that a CBR of 8 to 10 was necessary for the thickness of the overlying pavement section.

In my area (coastal plains soils, mostly poorly graded fine sands), we typically require stabilization to a CBR in the 20 to 30 range, depending on traffic loading. For unstabilized subgrade, we generally assume a CBR in the range of 10 or so.

Okiryu...are you sure about your specific gravity? It doesn't seem to fit with the zero air voids curve.

 

[Okiryu]

Ron, I plotted the ZAV in my proctor curve and it fit. See attached. As you mentioned, I tried to look for a mix that can give me a minimum CBR of 20. I tried different trials, plotted a graph of cement content vs. CBR and get the required cement amount. Then do a Proctor test for the stabilized soil with the required amount of cement. So, again, as you mentioned, I can say: I got a CBR of 20 at "XXX" compaction. In my case, it may be a CBR of 20 at ~80% compaction. I did also unconfined compression tests (based on water contents from these tests, the samples are assumed to be also compacted at 80%) and I got acceptable secant modulus values. Since I do not considered a 80% a high value for compaction, I was thinking that this low compaction degree may imply some settlement, but checking the secant modulus from the unconfined compression tests it looks to fit in an acceptable range. Perhaps I am overthinking this...As always, thanks for your input !!

 

[BigH]

Again - why are you not drying out the soil to a more reasonable moisture content in line with your lab density values . . . Most specs require you to do this. Of course, time and storage constraints come also into play. I am not sure of your magnitude of volume requirements but artificial heating might be considered in place of natural drying by aeration. You really need to get the moisture content to a more reasonable level, in my view. Ron points out that a soil cement structure will not be similar into behaviour as to a natural soil structure; hence, the concept of CBR may be totally meaningless.

If you are not familiar, I have attached a graph from a TTRL publication showing compaction levels and CBR for a sandy clay - of course the moisture contents are out to lunch compared to yours but, for reference.

 

[Okiryu]

Hi BigH, appreciate your input. I agree with you: I am planning to recommend that some soil conditioning will be required in order to get water contents to more reasonable levels.

Also, the link for your attached TTRL publication did not open. Could you upload it again, please?

 

[BigH]

They sometimes change links - try goggling TTRL ORN 3 and ORN 31.

http://www.trl.co.uk/reports-publications/overseas-road-notes/

http://www.transport-links.org/tran...rosoft Word - Overseas Road Note 31 edit1.pdf

(Don't know why they blacked out some tables and figures)

 

[BigH]

TTRL not blacked out after you save and open

 

[Okiryu]

BigH, thanks for taking the time to upload again those documents...the proctor curves with CBR values are interesting. Thanks again.