Settlement of sand under fill

[mad2323]

I have a site that will have 6 ft of fill placed over 6-feet of loose fine-grained sand with N value of 3 with water at the ground surface. I have looked at a schmertmann approach and came up with less than 1 inch of settlement but I'm not buying it. Any thoughts out there?

 

[SlideRuleEra]

I'm not a geotechnical engineer, but I have spent a career working on projects with poor soils and high (ground surface) water table. Agree with you, significant settlement may be expected even with negligible loading. Vibrating loads (or seismic events) will make matters worse... quickly. One example that comes to mind was a 1970's project that called for 5 feet of select, well compacted fill placed on top of local waterfront soil. Got about 6" of settlement in 2 months, and still moving. Geotech investigation revealed the 5 feet of properly compacted fill was settling intact into the underlying local soil. This happened with no applied load, only self-weight of the fill.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[mad2323]

slideruleera: Yes fill is scary. Only about 2-feet is equal to most building loads. Thank-you for your comments.

 

[geomane]

The settlement of the sand layer with likely happen as the fill is placed and compacted. I would be concerned with what is beneath the 6 feet of in-situ sand. Compacted fill is an aerial load and therefore, your vertical stress increase will be felt at depths deeper than 6 feet.

 

[oldestguy]

jmcc is right. Many times low blow count on saturated sand is due to the light submerged unit density, along with the gradation likely as uniform. Uniform sand may show low blow count, but its compressibility is low. However, depending on what will sit on the fill and whats below the sand, you may or may not want to monitor settlements. This is simply done by placing one or two settlement platforms on existing ground and monitor the elevation of these plates during and after fill is placed. I'd expect that the subsequent (after fill is compete) contribution from sand compressing probably will be slight. Depending on what eventually sits on top, you may or may not want to do some more work, such as placing a surcharge on that fill for a while to overload deeper soil and minimize or eliminate any settlement problem for what will sit on top eventually.

 

[mad2323]

After reviewing the boring logs I forgot about a soft (N=3) 3-foot layer of clay above the sand. My calcs came up with about an inch of total settlement for both layers with a 300psf building load on top the fill.

 

[oldestguy]

The job may take some re-thinking especially if getting that fill in place and compacting it. Any chance of soil rupture type failures while doing so, messing up that fill? Displacing or excavation of the soft clay may be a practical way to deal with it. Can the stuff be wasted on site where no building will sit?

 

[mad2323]

oldest - Yes I see what you mean. The clay may have to go or maybe we could use geogrid? It may be worth a test patch to see how it works.

 

[geomane]

The geogrid/geosynthetic is something I would like to hear more feedback on. I do not have much experience with it. I think of a geogrid as being used to prevent shear failures such as in pavements. Can it be used to prevent elastic settlements under aerial loads such as a fill, or even a large mat foundation?

 

[oldestguy]

mad2323 Test patches or other tests have proven to me to help a lot with establishing details for major cost portions of jobs, such as thickness of base course or amount of undercut. As to grids a small scale test may have to be looked at with mathematical ratios of various stresses and dimensions. Once a grid fails, there is no repair that is cheap. The stuff then is in the way of the fix. Regardless any buried utility trenches later will run into it and cause the contractor considerable grief.

 

[BigH]

My two cents worth: As I understand, you have 3 ft of soft clay overlying 6 ft of very loose (N=3) sand. Presumably, although you didn't say it, underlying the sand would be competent material. You have not indicated the "size" of the construction - i.e., a "house" size area, a industrial bldg sized area or . . . Nor have you indicated if this is for a raft type structure or individual footings or . . .

Assuming you would be developing an area, while three ft of clay seems like a lot, what I would suggest, and again this depends on the nature of the job, is to remove the 3 ft of clay. You have indicated that the groundwater is at the surface of the ground - is this by standpipe or ??? If the water level is at the surface of the sand when the clay is removed, put in some strategically located sump pumps and pump. When the water level is drawn down, put a heavy vibratory steel-drum roller onto the sand which should densify the sand considerably.

Given the soft clay, I don't think that geogrids would be appropriate - they certainly have been used building roads in swamps, etc. but even so, the fills of the embankments above will have flexibility and settlements of the road surface of a few inches would not be detrimental especially with the ability to repave the road from time to time for smoothness.

 

[Ron]

....and to compound things a bit.....you are putting a sand layer on top of a clay layer that is on top of another sand layer. That gives two potential drainage paths, thus potentially increasing your rate of settlement.

You might consider surcharging or using a compacted sand column approach such as using a 30-36 inch diameter pipe vibrated about 10 feet into the soil on a 5 foot grid.

 

[fattdad]

back to the OP. . .

consider what the modulus value of the loose sand may be? Let's say, 40 tsf. Consider the unit weight of the 6 ft of new fill - let's say 125. Let's ignore Boussinesq and just claim the entire sand layer experiences a 750 psf stress increase.

Integrate the stress change with depth - 750x6=4,500 #/ft
convert units of modulus - 80,000 #/sf

Long division time!

4,500/80,000=0.056 ft - 0.67 in

The modulus could be a bit lower, but not by too much? So, is it 20 tsf? That'd then return 1.5 inches.

After such modulus-based compression, there is some likelihood of latent effects. These would be unlikely to exceed 30 percent in the two to three decades to follow.

I agree that the majority of compression occurs as fill is placed. Post placement settlement would then be the 30 percent component, which in my example would be 30 percent of 1.5 inches (my conservative assessment) or 0.3 inches.

Having fun with numbers. . .

f-d

ípapß gordo ainÆt no madre flaca!

 

[Okiryu]

I agree with f-d numbers. Recommended Japanese correlations between modulus and SPT N-value for loose sands are E=1.4N (MPa), with your SPT N-value of 3 it becomes 4.2 MPa (87,700 sf). For large loaded areas, I also consider the simple Hooke law for settlements (delta = [sigma x H]/E).

Also, with saturated loose fine-grained sand and if the OP is in a seismic area, I would also be liquefaction concerns...