Perhaps one of the NAVFAC manual tables could help you out. It will depend on the thickness of the fill, too.

Jeff
 

I have seen to fail a foundation of these characteristics (of some colleagues) in a pair of months of ending the works. The foundation sustained only weak doors between pillars of masonry and small mesh for fencing. The pillars were by then 2" tilted atop and some cracks nearing 1 inch wide were visible.

Water was a factor here, but everything else in the soil else similar.

On another works I saw about 2" inches settlement on pavement about a year later; however was localized, and due to junk produced by the works outrightly covered by the fill.

Substitution to sound soil was the solution in this, and may be the case in yours, or else make a very stiff foundation or whatever ensures no problem will appear.

 

If your name and reputation is involved, then why take a chance?  Right now you need to take a stand and do the right thing.

The usual routine for this sort of thing is to dig out the fill and replace it with compacted fill.  Or, lower footings to competent ground, or other way to avoid using the fill.

I seriously doubt anyone here will give a recommendation, short of assuming the bearing is very very low and using a very wide footing, but even then, the fill may settle of its own bringing the footing down with it.

I personally would not accept the liability of the uncontrolled fill,  I would consider over-excavation or pipe pile, or if the uncontrolled fill region is small, bridging the area with the foundation wall, with large pad footings at either end of the bridge.

Mike McCann
MMC Engineering

I know the answer! The long term settlement of uncontrolled fill is. . . unacceptable.

f-d

¡papá gordo ain't no madre flaca!

i wouldn't even hazard a guess other than it could be substantial and highly variable (not to mention all the other problems that could show up).

i would however feel quite comfortable in rejecting it and replacing with a controlled fill.

Thanks everyone for the responses.  All are as safe as I would be, however, I have clearly left out critical information.  This is forensic work.  The foundation (not of my design, I would have taken the fill, apx. 7', out) is already on the fill and it is performing badly.  Another engineer is of the opinion that any settlement that was going to occur would have taken place in the first few years after construction, 20 yrs ago.  I am of the opinion that the settlement from the foundation load would only have brought the foundation soil to equilibrium with the load and that migration of fines, change in moisture content, etc would continue to affect the equilibrium and therefore continued settlement.  But I can't find an accepted source to establish this specifically.

Thanks jdonville, I checked in the NavFac manuals and find some info to the effect that small movements, even in engineered fills, can be expected for 15 to 20 years.

Again, thanks to everyone, any source that you can point out would be appreciated.

GG

In my experience, long-term or delayed compression of loose fill is most often initiated by saturation or an increase in water content.  You might approach this by researching literature on collapsible soils, which include dirty sands and gravels as well as the better known silty clay loess.  Figure 5, page 7.1-40 of DM-7.1 shows a relationship between dry density and liquid limit to identify fine-grained soils with a collapse potential.  The farther the soil data plot above the curve, the greater the compression that may occur on saturation.  There are some good references on Page 7.1-39 of DM-7.1.  More recent literature can be found.

As you describe the fill material, I'd feel confident in saying settlement is long ago done.  Being quite granular is good, but even with some clay there, seems fine.

If there were cinders or wood products there, that would be different.

Are any of the cracks recent?  Dust in them? spiderwebs?  Painted?

Occupant reporting any recent noise or noticing change?

If you can track down the following publication: BR 424: Building on Fill, Geotechnical Aspects, 2nd edition this will provide you with exactly what you want. The document covers all aspects of building on fill, from controlled filling with a high degree of control and earthworks spec, to very deep un-compacted fill with buildins on top. There are numerous case studies in it, and as it was written by the Building Reseach Establishment, the monitoring goes on for many many years.
ISBN 1 86081 509 X, pubished in the uk by EMAP, and you can e-mail them at crc-at-construct-dot-emap dot-co-dot-uk.
For a similar situation, check out case study 8, where after 12 years a pump failed, water level rose by 15m and 0.50m of settlement occured near instantly (geotechnically very quickly, well over a couple of months at least).  

I went to the NavFac manual again for Fig. 5 and that may be useful, We will likely have another round of drilling in the future to get some additional data.

iandig, I see the publication for download at BRE Bookshop, do you know if it is available for download anywhere else.  I haven't had any luck emailing them regarding US funds.

GG

I believe fattdad answered this question for me in summer 2007 for a similar excercise.  Schmertmann's method does not work well for this situation nor other NAVFAC methods.  I think the answer was us a c-alpha value.

The BRE document should be available for download with online technical indices such as IHS dot com. Do an advanced search with BRE as publisher and 424 as document number.
I just checked and found it!

Perhaps another approach.  If the structure has already settled (and presumably caused distress to the structure), why not temporarily transfer the foundation loads to jacks  to support the structure, and add water.  Sure, you'll have to monitor the foundations and adjust the jacks to accommodate the settlement.  And do this until settlement stops.  Just a thought!

Unfortunately, the research I mentioned earlier in this post does include the use of inundation to induce settlement of fill. The problem is it takes a long time for it to occur which usually means it is not a viable option. My comment before about when the pump failed is an indication of this, there was 0.50m of settlement which occured over a year, very qucik if its your buildign settleing but too slow andmore importantly too unpredictable in most instances as a ground improvement technique

I would perform a density test on the soils as close as possible to the same level as the underside of the footing. The density test should also be performed as close as the footing as possible. The geotechnical technician will have to re-calibrate the density gauge because the test will be in a hole and next to concrete. That will be no problem and should only take 5 minutes. But it is critical this calibration of the density gauge is undertaken.

Once you have the density results back you can estimate the degree of compaction and the potential settlement left for the soils. I would suggest a compaction result of 95% will be equal to 0 future settlement. Use 95% as your reference settlement estimator. If you get say a result of 90% compaction I would suggest you could estimate for 5% of settlement is still left to occur. (ie 5% of total fill depth = settlement)

The advantage of a compaction test is it will give you a definite answer to the compaction  state of the fill near the footing and if the density ratio is less than 95% it will tell you the site is liable to still undergo future settlements.

The compaction/density test as a settlement calculator will get less reliable as the depth of fill increases and you get uneven consolidation through the fill. In this case you would do a number of density tests all the way down the fill depth to arrive at a more accurate settlement calculation.