Self settlement of engineered fill

[TW88]

Hi, I have a site where we have a variable thickness of fill (Class 2C), generally placed to >95% MDD (2.5kg proctor) and generally <5%. Due to some unforeseen circumstances we have had to excavate much deeper in areas to the natural formation (which is a stiff clay). For reference the fill was removed from a backfilled quarry and re-used insitu, compacted in 250mm layers with a vibratory roller (approximately 13T). The fill varies from approximately 4m thick to 16m deep. Whilst most of the self-settlement of the fill would have occurred during compaction and construction (built over approximately 6 months), I need some calculations to back up time to 95% consolidation settlement and calculations for total anticipated self settlement of the different thicknesses of fill (i.e. 4m vs 16m). To put this is context, depending on the final development plan houses may be built over different thickness of fill, so we need to ensure that differential settlement is not going to be an issue (for reinforced strip foundations) and the time it may take, so a construction plan can be developed. Any help/ideas very welcome.

 

[oldestguy]

Sounds like you are worrying about nothing. Likely any settlement will come from that lower natural clay, but you already have applied a pre-load there and any compression settlement (if any) is completed. Compacting the fill likely has applied pressures far above an future pressures and any "self settlement" is very unlikely. That's why we use compacted fill in place of compressible material. It also tends to give a form of "beam action" to even out any differential settlement that may occur.

 

[Ron]

Agree with OG....I'd be more concerned about the underlying clay layer being affected by the additional overburden.

If you want calculations, you can compare the void ratio of the pre-fill material to the post fill material and show how much more compact the material is now as compared to the natural state.

I would anticipate no change from self settlement in controlled, engineered fill....that's the primary purpose of engineered fill.

 

[hokie66]

It's a rare occurrence, but I find myself disagreeing with both oldestguy and Ron. My experience is different than theirs, but as I am not a soils specialist, I can't offer advice as to time dependent self consolidation of fill, other than to say I would be concerned as well.

 

[oldestguy]

Hokie66: Interesting post. Perhaps explain, there may be some experience to show why the comment.

From my point, I'd have to say the numbers of jobs I've had where building on compacted fill was done, including over very deep marshes (after surcharging), easily runs in the thousands. Never a problem of consequence because of the fill.

 

[aeoliantexan]

I am not sure what you mean by Class 2C material or generally <5% (water content?), But I have a lot of experience with buildings on compacted fill, mostly lean silty clay or sand. I've seen very few problems attributed to the fill unless it was poorly placed or got wet after construction. I believe it takes both a competent contractor and a diligent inspector to build a good quality fill. And thick fills, let's say >20 feet thick, can compress 1-2% or more when wetted, even though compacted to 95% Standard Proctor. The company I used to work for learned this the hard way on properties developed on hilly ground, especially after automatic lawn sprinkling became popular. I have been searching the literature lately, and this behavior has been well documented since at least 1953, but the general practice still seems to be unaware. With the differential thickness you described, I would be concerned if there will be irrigation associated with the houses.

 

[ATSE]

Check out the paper "Hydrocompression Settlement of Deep Fills" by Brandon Duncan and Gardner, ASCE, 1990 (7 pages).
Dr. Noorany (retired professor at SDSU) also wrote and ranted a lot about this. Southern California has many case examples of bad outcomes from non-uniform deep files and hydrocompression.
1/2% of 50ft is a big number when it comes to differential settlement.

 

[oldestguy]

From what I can tell of the post with 2C soil, it is low plasticity glacial till per this reference.

https://www.engineersireland.ie/Eng...Glacial-Tills-by-Conor-O-Donnell.pdf?ext=.pdf

With it compacted as indicated, this would not appear to be susceptible to further desnsification even if "saturated".

 

[EireChch]

aeoliantexan - in may last work place my boss always used the 1-2% of the thickness for self weight settlement. He said it was a rule of thumb but never had a reference for it. Would you by any chance?

 

[aeoliantexan]

That R.O.T was told to me by Doug Moorhouse, president of Woodward-Clyde at the time. I believe it came out of research and experience related to some of the southern California cases that ATSE mentioned.

Since moving to Texas, I have seen several cases where the settlement was much larger than that. 9 inches and counting in a 30-foot-deep fat clay backfill. 15 inches in a 35-foot-deep lean clay backfill. Both had reportedly been compacted in lifts and tested. I admit that both the quality of compaction and the testing can be questioned, but unfortunately, they were consistent with the state of the practice.

One other thing learned from those cases is that pushing Shelby tubes into a poorly-compacted fill can make it look dense. We had the 35-foot fill drilled and sampled, and most of the results would plot close to the zero air voids curve. In other words, the sampling compressed the sample to its maximum density at that water content. I do not know the details of the tubes, except that they were 3-inch diameter. A lot of sampling is done with relatively thick-walled tubes, often with blunt ends.

 

[fattdad]

If the soil is not saturated and above the water table, I would not be too concerned with, "Consolidation," which is limited by the soil's ability to relieve excess pore pressure. I'd expect any mechanics of settlement to be related to compression - i.e., related to modulus.

You may not know the modulus value? I'd get that. You may be able to divine some value though? Don't know. Let's say you have a modulus value of 200 tsf within the fill itself. Let's also say the fill is 20 ft thick with a unit density of 125 psf. That'd mean the average stress change in the 20-ft thick layer is 1,250 psf (0.62 tsf). If a 20-ft thick layer with a modulus of 200 tsf receives 0.62 tsf if stress change would return 0.74 inches of settlement (i.e., (20*0.62/200)*12)

That's nothing! Also, as the OP points out, it'd all happen during fill placement.

That's not the end of the story though. . .

Modulus-based settlements are also influenced by time. The parameter, "C-sub-t" is used to anticipate additional, "Modulus-based" settlement that may develop after the immediate settlement is complete. For a forecast of 1 year beyond initial placement the value of Ct is 1.2 (i.e., 20 percent of originally-calculated settlement will occur during the year that follows. For a period of 30 years, that value is 1.5. That means for the example above, you'd forecast 0.15 additional inches of settlement in the year that follows and 0.37 inches of additional settlement in the 30 years that follows.

Class 2C is a cohesive granular fill. It certainly seems reasonable that settlement could, "Lag" from the more typical response of sand. I doubt there'd be a performance issue though.

I'd agree the greater concern being whether the seat of settlement extends into the native soil.

f-d

ípapß gordo ainÆt no madre flaca!

 

[dik]

My experience is similar to OG... never had a problem, even where fill has been in the order of 20'.

Dik

 

[Okiryu]

The Navy DM 7-02 manual has some interesting information for compressibility and other parameters of compacted soils. See page 62 of this link:

http://www.geotechnicaldirectory.com/publications/Dm702.pdf