Floor Slabs on fill soils

[Okiryu]

Just wanted to ask what will be your approach when providing recommendations for floor slabs for a site that has already an area filled with cohesive fill (silty clay) soils. The fill does not have debris. Fill soils are consistent in the type of soil material (mostly silty clay soils) and ocassionally are mixed with gravels. Average SPT N-value is 5. Fill soil layers vary from 2 to 4 meters thick. The fill was placed 25 years ago and records for the grade of compaction of the fill are not available. No groundwater was observed within the fill layer. Based on information from past projects, water content is assumed to be around 25%-30%, LL=45-55% and PL=20-25%. Under this fill is an overconsolidated medium stiff clay (3.8m thick) with Su=40 Kpa and then very stiff clay soils (mudstone).

The building will be a 2-story school building and part of the building will be on this old unrecorded fill area.

Also, our site will receive 1 meter of new fill. I checked consolidation settlements of the medium stiff clay due to this new fill and settlements are on the order of 10 mm. Will not be a problem for the footings since we are planning to do soil improvements.

My question is about the floor slabs:

1. I have not done any tests for the unrecorded former fill (we do not normally test fill soils) and thinking that since I do not have information for the former fill, I may have to recommend a structural slab in that area rather than typical slab-on-grades.

2. Or, is it reasonable to assume an elastic modulus for the former fill based on SPT (in my area we normally E=700xN (KPa)) and calculate elastic settlements due to the placement of the new fill (I am assuming that the former fill will not be saturated)? I have calculated this already and I got 20mm. So,assuming that the elastic settlement will occur fast, a slab-on-grade will be acceptable.

What do you think? Thanks for your assistance in advance...

 

[GeoPaveTraffic]

I see it, the biggest risk is that there are pockets of debris, wood, or soft material in the fill. We would normally use test pits to evaluate the fill. In that way you see much more of the material and get a better feel for it.

If the existing fill is consistent and does not contain soft area, debris, or wood, then the proposed plan should work.

Mike Lambert

 

[Ron]

OK....let me make sure I have this right....
You are going to use public money to build a school on silty clay and you have done no testing and do not contemplate any?
What is the standard of care for engineering in your area?

Further, with 1m of similar fill over fill with N=5, you are only concerned about elastic settlement? Will the slab carry any of the building load or is the structure isolated and the structural bearing soils will be improved?

Reconsider your elastic settlement and don't consider that the material will react elastically and fast....it will bridge over because of the lower moisture content, but moisture conditions change so give that some thought.

I would certainly consider some lab testing on the materials, particularly consolidation and shear testing.

 

[Okiryu]

Thank you for your replies.

Actually, I am concern about the former fill and although in my area the fill is not usually tested, I will do additional soil borings and am planning to run some testing for the fill as well.

The floor slab will carry a floor live load of approx. 5 KPa. The floor slab will be connected to the walls and footings. Under the footings we are planning to do some soil improvements with soil-cement slurry. Considering 1m of new fill, I checked elastic settlements based on a total load of approx. 25 KPa. Once I get the consolidation test results I will check for consolidation settlement as well.

 

[aeoliantexan]

I practiced for many years in an area where the dominant material was silty clay. We were very suspicious of fill that did not have good documentation of its placement and compaction. Remember that samp;ing by pushing thin-walled tubes in fill can greatly increase it density and make it look well compacted when it is not. Test pits are useful to examine the fill for debris, topsoil, lift thickness, and low-density zones. Density test can be taken in the test pits using nuclear gages, sand cone, or hand-carved samples. I would take the approach that the fill is not to be trusted to support even the floor slab unless I can prove otherwise with a thorough investigation.

If it was well compacted, it should compress little and quickly under 1 meter of fill. If not, there is potential for several inches of settlement if the water content increases, perhaps due to changing the site drainage or a future plumbing leak.

 

[Okiryu]

aeoliantexan, thanks for your reply. Since your dominant soils were CL-ML, did you use them as engineered fill underneath slabs and to support footings? If so, what was the typical specification for the installation of that type of engineered fill? For instance, I see that 95% compaction is difficult to achieve for that type of soils. Also, what was the typical W% and LL for that soils?

 

[Okiryu]

I went to the site today and somebody was doing some test pits. The depth of the pits was around 2.0~3.0m. Attached are some photos. The upper granular fill might be compacted above 90% but I feel that the clayey soils below are less than 90% (based on experience in the area, even try to achieve 92% is difficult for these type of soils). Based on our soil borings, I was thinking that the soil material for the fill was consistent in detph, however the test pits show the contrary. I could see construction debris as well. The test pits were done in a playground area where there are no structures around, so I imagine that the Contractor did not put to much effort on compaction and also did not pay much attention on the soil materials for the fill. I feel now, that a slab-on-grade is not the correct approach unless the uncontrolled fill is removed. Finally, I will be doing some additional borings to double check the thickness of the uncontrolled fill.

 

[Okiryu]

another photo..BTW, is it possible to upload more than 1 file?

 

[EireChch]

When we encounter fill in a geotechnical investigation, unless it can be certified by completion documentation, it is always considered non-engineered. We also never specify cohesive soils for engineered fill given the compaction requirements (i.e. Sheeps foot rollers etc).

I would not consider the fill to be suitable to support a new structure. I would investigate piling to the stiff clay layer at apporximatly 7m depth. This may require a machine bore hole with shear vane testing every 0.5-1m when in the stiff clay layer.

Piles should potentially be designed for negative skin friction due to the on-going settlememt of 4m of fill and the concrete slab should be designed as fully suspended to span between piles.

Over-excavating the non-engineered fill to say 2m could be considered. The soils should be replaced by engineered fill (I.e a sandy gravel). Tri axial geogrid reinforcing could be placed in the base of the hardfill to aid the load distribution. Using a stiffened slab could also be considered.

Also who is doing the test pits? Is there another geotech involved?

My two cents anyways.
Cheers

 

[Okiryu]

EireChch, we are planning to do some soil improvements using cement slurry under the footings. The soil improvements will reach the stiff clay. These soil improvements will allow to support the structure. My concern was regarding recommendations for the floor slab: slab-on-grade vs. suspended slab. After saw those test pits, as you suggested, a suspended slab should be recommended. However, we might need to explore other options for a possible slab-on-grade recommendation. As you mentioned, these options could be over-excavation and replacement with engineered granular fill, geogrids, etc. The test pits are for cultural assets (archaeological) survey.

About specifying cohesive soils as engineered fill, I would like to see aeoliantexan opinion since appears that he usually had to deal with silty clays.

 

[EireChch]

When you say the cement slurry will reach the stiff clay (~7m depth,are you proposing to do a soil mixed column?

Cohesive soils are our main soil type, so for bulk filling (i.e. for subdivision filling) we utilise these as they are readibly available.
Our specification requirements are as follows: Undrained Shear Strength 120kPa (typically measured with a hand held shear vane), Air Voids 7%, Air Voids over 10 Consecutive Average 5% (typically measured using NDM testing, Nuke testing).

Generally for residential/light commercial buidlings we recommend hardfill for compaction reasons as above. I would be intersted to hear others opinions.

Cheers

 

[aeoliantexan]

The dominant soil in the area was loess, mostly CL, some ML, rarely expansive. We used it often for structural fill to support floor slabs and footings for one-to-three-story buildings, placed at optimum +/- 3% and compacted to at least 95% of Standard Proctor. For pavement subgrades and heavier footings, optimum +/-2% and 98% of Standard. For very heavy structures, 100%. The ML has a very narrow moisture range for successful compaction, but the CL is quite useable. Because the underlying native soil was potentialy collapsible, we avoided granular fills, unless the loess was already saturated. We often overexcavated and recompacted 2 feet or more of the soil under footings and floors to increase the allowable bearing pressure and reduce infiltration of surface water. If the new site fill was to be more than about 4 feet thick, we preloaded with a surcharge to reduce delayed settlement. If the loess was drier than about 2% below optimum, it would not respond to preloading, and we might resort to deep foundations.

 

[EireChch]

That's interesting, what type of compaction equipment do you use?

I have worked with loess before and like you said it has to be close to optimum mc or it it's very difficult to compact. I have seen 4m high unretained cuts standing vertical in dry weather however once there is any rain it fails, shows how sensitive loess is to mc

I also agree with the increased standar compaction above fixation of 95%. There has been many discussions on this topic on this forum. We typically specify 98% standard compaction.

In relation to your comment about avoiding granular fill due collapsible soils. Why would structural loess fill be any better? Also what is the reason for native soils collapsing? Tunnel gullys and under runners are common on our loess soils.

 

[Okiryu]

Yes, we are proposing soil-cement columns. And very interesting, my experience with ML-CL soils is that even 95% compaction is difficult to be achieved...

 

[aeoliantexan]

EireChch,
Sorry, I have been busy.

Loess fill was compacted mostly with sheepsfoot rollers or vibratory sheepsfoot rollers. Tamping plates were used in small areas.

Compacting the loess eliminates its collapse potential, so you have stable soil in the highly-stressed zone immediately beneath the footings. It also reduces the permeability, say to 10E-6 cm/sec, which limits infiltration to a rate that can seep down thru the underlying loess without saturating it.

Collapse potential depends primarily on density and clay content, but also, I believe, on moisture content history and stress history. In eastern Nebraska, which is fairly humid, the loess is stable at water contents moderately higher than the present water content, because there have been natural fluctuations of the water content over the past 15,000+/- years that it has been there. One can add a few feet of fill with little effect, but a few years later, during the next wet cycle, it settles. We had a rule of thumb that up to 4 feet of fill was OK, but more would cause settlement someday. Preloading could reduce that risk if the loess was wet enough to compress under the preload, but dry loess had a higher apparent preconsolidation pressure, and would not respond to preloading. In those cases, if we couldn't change the grading plan, we resorted to deep foundations.

Tell me about the underrunners.