Collapse on Wetting, Perhaps

[dgillette]

Got a weird one here.

I'm dealing with an old dam that has shown some minor cracking recently, for which there are a couple of plausible explanations. One of them is collapse on wetting.

There are three main kinds of fill in this embankment, hydraulic fill, "rolled" fill (primarily CL), and rolled fill. The distinction between the last two is pretty big. The "rolled" was probably dumped off of wagons, and spread with horsedrawn Fresno scrapers, with little or none of what we would consider rolling nowadays. There is also an overlay of dense, for-real rolled fill put on 10-15 years ago for seismic reasons.

1. They had the wettest spring and early summer in anyone's memory.

2. When we drilled it ~15 years ago, unadjusted SPT N was mostly 2 to 5! I don't have dry density, but it can't be very high.

3. Moisture contents in the "rolled" were consistently lower than the PL; MC 13-16, PL 20-22. (The material is practically all -#40, so I'm not getting tricked by the presence of +#40 in the MC sample and not in the Atterberg sample. Already checked that. In fact it is mostly 90-95% -#200.)

While I've never heard of collapse potential in manmade fill, with such low N and MC here, it would seem possible. Ever seen anything like that happen? We don't have much info beyond water content, N, grain-size distribution, and Atterbergs. Unfortunately, unless we decide this is a serious dam-safety problem (which I doubt, because of the overlay and the unusual geometry of the dam), I won't have any money for drilling or testing, dangit.

The underlying hydraulic fill is similar, but for having much higher moisture contents (a little below LL), and slightly higher unadjusted N. (Yes, higher N in the hydraulic fill than in the dry fill.)

Settlement in the area of the crack and at one other place accelerated quite a bit this summer, something like 0.3 feet, whereas previous settlement since the overlay was put on was more like 0.2. (No info before the overlay in the late 90s.) The geometry of the crack would fit with collapse of the dry fill or consolidation in the underlying hydraulic fill, although there is no reason I can see for the hyd fill to suddenly start to consolidate again in wet weather.

I have experience with only a couple of collapsing soils, one of which, for fattdad's interest, was the slopewash in Tim S's dissertation. The other was a collapsing silt we were having to pre-wet before we built a canal in it / on it, >20 years ago.

Cheers!
DRG

 

[BigH]

David - do you have any phreatic surface data in the fills? At first glance, I find it hard to believe that this material will collapse on wetting - I've dealt with loess before and I know that it has . . . Being an old dam, it must have had some wet seasons before - the only way that I could see it collapsing is the original fill being "hard clay balls" - and now the wetting softening them to a degree that they are starting to squish under the upper weight. But might this not have happened earlier? Interesting - I'll keep thinking but now - its time for bed - need to get up in 5 to watch USA play Italy in World Cup Rugby. Game has already been played but I don't know the score.

 

[dgillette]

We do have piezos in the hydraulic fill, but the "rolled" is high and dry.

They certainly have had plenty of wet years there before, but apparently that did not get the fill wet prior to the drilling in 1991. I think it's unlikely that the fill had been previously wetted then dried deep within the embankment. One reason I say that is work we did in the 1990s in Phoenix area. It's a long story, but in digging up an old embankment, they found that the moisture content was pretty close to the original placement moisture, once they were 2-3 feet below the surface. This was after decades of Arizona sun and heat, not Idaho. This part of Idaho only gets 12" of precip/year, but that's almost twice as much as PHX, and the year-round average temp is 20 F lower than PHX.

 

[emmgjld]

Sounds like some problems I have had in Western Colorado, with fine grained soils experiencing piping & settlement.
Going back to some older books dealing with embankment seepage and problems similar to Dispersive Soils.

My best article is Relevance of Atterberg Limits in Evaluating Piping and Breaching Potential, Daniel Resendiz, pp341-353, within Dispersive Clays, Related Piping and Erosion in Geotechnical Projects, Sherard/Decker, editors, ASTM STP 623, 1977.

The point of interest to me has been the article dealt with soils/embankments in drier environments, the soils were not necessarily Dispersive but seemed to behave as such and he came up with a criteria for initial identification. My experience indicates his criteria of identification is valid.

Skempton's Activity is determined with values of 0.3 to 1.0 indicate soils which may be troublesome. Piping, formation of 'sinkholes' and areas of settlement may be associated with migrating moisture.

I know these STPs are hard to get a hold of. I am fortunate that my Father believed in lots of literature so I have a better than average library. If you cannot get to a copy, post here & I will try my hand at uploading a pdf, don't tell ASTM.

 

[dgillette]

Thanks, emmgjld. I should be able to find most of the relevant ASTM pubs here in the Front Range. (The canal I mentioned above was in the Four Corners area, and I've also seen it or felt it underfoot in a friend's yard in Glenwood, when he was sawcutting another door for his walkout basement and the water ran onto some freshly exposed soil.)

I am 98 or 99 percent sure that this crack is not a result of piping, because of the fact that it is roughly parallel to the crest for a couple hundred feet (no localized sink holes), and the improbable amount of material that would need to have moved out through the filtered overlay. That, and the odd geometry of the dam that puts the crack so far from the reservoir.

I'll probably dig that out of the library next week when I get time.

 

[emmgjld]

Cracking along the crest of a dam/embankment..... Check other articles in that STP and note articles by SCS, now NRCS personnel in similar publication. The SCS/NRCS has been blessed with dam crest cracking. Usually not piping ... initially but can progress to piping.
Most dam crest cracking I have dealt with appear to be a result of drying shrinkage. Possibly a recent period of somewhat lower precipitation of slightly warmer temps? Once the cracks open, a lot of changes begin.