STP Borings in compacted fill

[geo1976]

Is there anyway to evaluate the compaction of fills using SPTs? The engineer of record has done multiple SPT borings, the material ranges from a SM to ML with anywhere between 23% to 65% fines. The engineer of record is now saying the material was not compacted properly based on the N values. There was a foundation failure after a large rain event 6.5 inches in two days and almost 8.8 in 9 days (which supposedly had nothing to do with the failure). Fills ranged from 5-11 feet.

My feeling is the N-values are an index test, and the values listed for sands are for clean sands not silty sands and sandy silts, the material definitely has cohesion as a 10 foot shear cut held up for 5 days with compaction equipment ran up against it for 5 days.

Any ideas???

Thanks in advance

 

[BigH]

Welcome to Eng-Tips. Hope that the site will be useful to you. I notice that you posted this in both this forum and the soil testing Engineering forum. By doing so, you will get disjointed responses. I've red-flagged the one in the soil testing forum and asked that it be removed leaving this one for responses. Double posting is not appropriate to the site. Most geotechs review all the geotechnical forums. I'll give your situation some thought and other will too. Do a search - for there has been a similar question recently. That was for confirming, or not, compliance with the specification. Why there wasn't any testing done during the construction phase is a mystery - as with yours.

 

[dgillette]

Aw dangit, Big H, I just replied to the other one (which happened to be directly above the one you were referring to). I said the same thing as before: test pits and sand cones.

 

[Ron]

me too, dgillette...I copied my reply before it get the inevitable can....



"I agree with dgillette. The "near surface" soils can be better tested directly with appropriate methods...SPT is ok for deeper stuff by correlation, but when other methods are readily available that will give more accurate data...use them."

 

[BigH]

Sorry guys - I had to pick one . . . Maybe when a thread is double posted and one is removed, the replies could be transferred?

 

[aeoliantexan]

In the early 70's an article by Bazarra in the ASCE geotechnical journal presented new correlations between SPT and relative density using tests in compacted fill. After some discussion, i believe the geotech community generally agreed that Bazarra's correlations appied to compacted fill and the old Gibbs and Holtz correlations still applied to sedimentary sands. Compacted fill seemed to produce higher SPTs for a given relative density than natural deposits because compaction locked in high lateral stresses. There has been a lot more published on SPT since then.

You can develop a correlation between SPT and relative density for a given material; it will depend on depth and compaction method as well. It will take a lot of work. Others can argue with your results.

If a dispute over money is involved, I suggest using test methods that all parties will agree to accept. Otherwise, a resolution will be held up by disagreements over interpretation of data.

 

[msucog]

what does a "foundation failure" consist of? was there excessive settlement or sinkholes or did the building heave over to one side like a sinking ship or what?

also, what sort of blowcounts were found? and you meantioned that relative to the blowcounts, "the values listed for sands are for clean sands not silty sands and sandy silts". what values are you referring to and wouldn't a blowcount be a blowcount? the blowcount doesn't care what the material is...only we care what the material is. so in other words, a 5 is s 5 although our interpreted consistency would be different if it were a silt versus a sand. even the topic of a "silt" or "sand" is fuzzy, varies dramatically with the available test data and depending on which way you want the result to fall. for instance, i've seen a wash #200 go from 35% to 60% simply because dispersing agent wasn't used by the testing firm...so that supposed SM then become a ML.

i'm rambling a bit to make a point...
1. there's not enough info to make any sort of reasonable guestimate about what's going on
2. a simple data point or two likely doesn't tell you anything unless it happens to fall on the indiviual reader's point of view
3. a data is useless if it wasn't obtained correctly to begin with
4. if you are the geotechnical engineer, who is this EOR? are they the geotechnical EOR or a structural or what? are you stamping the project or working for the contractor or another client?
5. what degree of compact was desired?
6. loads?
7. area of loading?
8. etc
9. etc

 

[DRC1]

SPT are an index tests of sorts that yield relative density. They are applicable to Silty sands as well as clean sands. They are not useful for cohesive soils. The chief use of SPT's is to estimate the internal friction angle of the soil, from which an estimate bearing capacity can be computed. Unfortunately, most compaction jobs are controlled by specifying a dry density Proctor or Modified Proctor value. These values do not correlate to SPT tests (unless you do the site specific testing aeoliontexan suggested.
I would point out that the SPT test is highly variable and is not precise. Laboratory testing of field samples would be the best way to determine soil properties or bearing capacity. fills with fine contents ranging from 23-65% wouldn't be expected to perform well in 9 inches of rain. Could you provide some more details on the nature of the failure?
Also note that unless you really now you have a solid clay running compaction equipment along the base of a 10 foot shear cut is not a good idea. These faces can colapse without warning

 

[jrm73]

If we think about the reason for specifiying soil density as a control mechanism for compacted fills, we might find that the SPT can be useful. The compacted fill will probably have to support a load even if it is its own weight. To support a load, the soil will have to possess a minimum shear strength. It can be expensive to test the shear strength of a fill with a frequency that would provide reasonable confidence. We know that there is a rather dependable correlation between strength and density. And density tests are less expensive to perform in the field. Over many years and many projects we have found that relative density values of say 95% density (Standard or Modified depending on the project requirements)typically provide adequate soil strength. Therefore, I think that performing SPT would be useful if you are convinced of this arguement and you can convince the Engr of Record of the same. I have done this with both SPT and Shelby Tube sampling on certain projects with acceptable outcomes.

 

[BigH]

That's fine from a pragmatic and practical point of view. But if this is for judging that a contractor has lived up to his stipulated requirements in the contract - it doesn't wash. It is now contractual vs practice.

 

[geo1976]

To shed some more light on the situation. The geotech of record recommended the removal of soils within the building pad. The soils removed consisted of organic clays, fill materials, tree roots, old pipes, etc... The report recommended the removal down to bedrock (sandstone and shale) with the perimeters being cut a further 10 feet beyond the footing lines. The materials from one end of the site were approve by the geotech for use as fill (SM-ML). The fills were placed and tested and compaction was attained. However the material went in slightly drier than +-2% and in lifts of roughly 12-15 inches, which the contractor was advised outside of the specs. Density testing revealed that the recommended compaction was attained.
During footing excavation the footings were tested with a drop bar and favorable blow counts were attained. After a 6.5 inches of rain in two days the footers in one corner of the building had dropped 3 inches, there was no load on the footers. It should be noted that the site sat under 2 feet of water for at least 2 days after the rainfall. Test pits were dug in the failure area and the moisture contents revealed that fills were then above optimum moisture, and that the contractor had not removed the required 10 foot perimeter excavation. The geotech of record then performed SPT borings, which revealed blow counts ranging from woh to in excess of 50. The area in which the failure occurred was removed and backfilled with recycled material.Subsequently the eor has submitted an opinion that the materials were not impacted by the rainfall, improperly compacted, and that the whole building should be pressure grouted.
To me discounting the impact of the rainfall is way off the mark, as pictures of the fill going in and afterward is like night and day. He has discounted the washing of fines, but how would you tell if the material already contained varying amounts of fines? Is it possible that the rainfall event combined with the material being placed dry of optimum caused the grains to go through some type of reorganization and caused settlement or a loose or soft condition? Also is it possible that the rainfall event sufficiently raised the water table and decreased the bearing capacity of the soils?

 

[DRC1]

1. Somebody should have records of construction control, such as density testing and weather or not the contractor achieved density during construction.
2. Obviously the lack of excavation at the perimeter will be an issue.
3. Optimum moisture only applies during the compaction operation. The water content can increase after that with no increase in DRY density.
4. getting to the meat of the matter, the problem appears to be two fold: a) the fill soils seem to be highly variable, and probably some have high fine contents and low sand contents and others are just the opposite and b) there does not appear to be any mention of drainage provisions, which based on the flooding are non existent or inadequate.
Soils with high fine contents and granular soils do better with different types of compaction, so that would explain why some parts of the site were good and others are not. The other problem is that fine soils do not drain quickly. as the weight of the fill rises due the saturation by the rain, the water trapped by the clay particles can not bleed off fast enough, the internal pressure rises and the clay or silt soil structure can fail.
*** Obviously I have not seen the site or any data beyond what you have mentioned. This is just my feeling based on your notes. You may want to bring in an independent geotech to review the site and data to make recommendations. If the soils are not coarse and drainage is an issue, compaction grouting may cause more problems than it cures.
Hope this helps

 

[msucog]

the fact that the contractor put the fill in 12-15 inches thick will kill any argument. running compaction at the top of that tells you nothing. with the rain you mention, it sounds like someone is putting in a loose rocky-soil fill that then washes down in to the voids since it was put in way too thick/loose. i don't know the geology but just assuming the problem is in the fill, then the EOR has my vote. if it went in compacted in the first place, it doesn't magically become uncompacted with a rainfall. and raising the groundwater doesn't necessarily lower the bearing pressure...in fact, it could increase the "bearing pressure". bearing pressure is not a simple test...it is much more and has multiple inputs. again, the EOR has my vote.

good luck.

 

[geo1976]

Not arguing that the material becomes "uncompacted". When the water table is raised if you calculate the theoretical acceptable bearing pressures you essentially cut the calculated pressures by half.
I have had situations where material went into trenches or underneath pavements, the material was fine grained in nature, as it went it it was below the optimum moisture and compaction requirements were met, after rainfalls the material lost its strength. That is the argument, its no different then if you have a clay that is dry and strong, then its saturated and it looses all its strength. The contractor has to answer for its errors, pertaining to lift thickness and not removing the necessary perimeter cut. However, the EOR needs to accept responsibility for approving a fine grained soil as structural fill for a building pad.

 

[BigH]

With respect to the last post's "theoretical acceptable bearing pressures" - you are not actually describing the acceptable bearing pressure but the acceptable (allowable) bearing capacity based on shear. The "acceptable" bearing pressure is not the same as the "acceptable" bearing capacity - in most cases it is less due to the service limits on movements set by the structure design criteria.
I doubt that the footing settled because of the lack of "strength" - but it is possible that the compaction, being drier than optimum caused a metastable state of the grains to be developed and that the subsequent inundation of the material gave "lubrication" to allow the metastable state to rearrange to a more dense state - and this was significant. Others have postulated, correctly, that for most fills, the dry density doesn't change on inundation - I agree but obviously with the settlement, the volume is now smaller for the same mass of soil - so the dry density would be higher.
Do you have records of where the compaction testing was carried out? Was it done on a random basis or was the testing selected in a biased fashion? Who selected the locations? Was there any actual tests run in the area where the footings experienced the distress? Perhaps, as the contractor did not "extend" the excavation out beyond the footprint as stipulated, perhaps he didn't extend it to the perimeter of the footing either? In this case, the footing might not be founded partly or wholly on the placed fill.
I find it difficult to believe that any compacted fill would have a WOH "blow count" This suggests the postulation noted above. Just a thought.

 

[BigH]

A caveat on my last statement about WOH - I have seen this (or a situation that would be so) but this was when compacting a loess soil and it was allowed to dry out completely - the upper 300 to 500 mm was "powder".

 

[muuddfun]

Sounds like a classic case of "hydrocollapse".
Your lift hights were to much, and the soil was not completly compacted. Once it was inundated it collapsed.
Also not completing the removal as specified probably contributed as well. There is no reason to blame the EOR for using a finer grained soil for structural fill. If there was then a lot of dams would not exist.
If I was the EOR I would make the contractor remove the fill entirely, to the proper limits of the removal as specified, and start over, that would be cheaper anyway than compaction grouting. (unless you let the contractor continue with the construction of the building?)


Collapse looks like this on the consol diagram.
__________________ Log P
|\
| \
| \
| | <- H2O added
| |
| |
| \
| \
|
Percent
Consolidation

 

[aeoliantexan]

I agree with msucog and muddfun. It looks like a fairly simple case of a very inadequate job of overexcavation and replacement. A contractor who doesn't excavate outside the perimeter as required and places 15-inch lifts probably doesn't do a very uniform job of placement and compaction, so there will be hard and soft zones after water gets to it. There is nothing wrong with using ML and SM for a building pad if the fill is properly compacted. Yes, it will absorb a little water and soften somewhat if inundated, but it won't become "woh"; it won't settle 3% of its thickness under its own weight; and the bearing capacity won't become so low that the footings settle under their weight, no matter where the water table is.

I agree, the contractor should remove everthing and start over. And the testing lab should watch every lift being placed and test it with judgement applied.

My comments are based on 25 years of geotech investigations and inspection where thousands of buildings were founded on spread footings on structural fill, usually CL, but also ML, SM, and SP.

 

[jrm73]

I agree with Aeoliantexan on several of his points. Silts are OK if you can compact them properly. Thick lifts in silt are really a problem because of the narrow moisture band at the proper density range. Plus, there can be some amount of bridging that does not show up with thicker lifts. The lower portion of the bridged soil can still be loose and settle with the introduction of water. I read two important items in geo1976 information. First, only one corner of the site settled indicating to me perhaps a local problem with different compactive effort, soil type, etc. Second,SPT values in the problem area ranged form WOH to 50? Has anyone ever seen that large of a range in properly compacted soil in a small area of the site? If the rest of the site has performed well, does that need to be removed? Just asking?

 

[TDAA]

{quote]Second,SPT values in the problem area ranged form WOH to 50? Has anyone ever seen that large of a range in properly compacted soil in a small area of the site? [/quote]

No, and that is why I agree with aeoliantexan as far as removing and starting over. There seems to be a lack of control for the fill, and therefor, I would consider all of it to be suspect.