proving compaction in granular fill embankment post-construction 1

[Tiarnan]

Hi,

Looking for any thoughts on proving compaction in granular fill material post construction. The site is an approach embankment, roughly 7m high, and due to insufficient testing during construction additional data is needed to ensure compaction targets have been achieved. The material is graded AP65, so too coarse for CPT testing, and NDM testing will only give answers for the first metre or so, without running into issues with excavating out the material and reinstating it on a large scale. The contractors have suggested SPT testing, though I assume with this there would be the need for some benchmark testing correlated to an NDM or water-replacement test in the same material. Does this sound like it will result in reliable compaction information for a targeted 95% MDD in compacted fill, or would it have to be tailored to a penetration rate in mm/blow, rather than just aiming for N=50+where all 50+'s are assumed equal....
Open to suggestion for other techniques too, though the client probably isn't too keen on the whole 'tear it down and start again' approach.

 

[Mad Mike]

I would have also suggested SPTs through the material- so I agree with the Contractor's proposal. What it won't give you is a reliable means of correlating with a design compaction- keep in mind that even if you can correlate a material density to the SPT penetration rate, that could only be applied holistically to an isotropic material.

I would use continual SPTs to gauge for myself the relative compaction- I would supervise the SPTs full time to make sure the penetration rate is relatively uniform through the measured test intervals and lastly, I would want that information supplemented by some construction photographs and/or details on how the construction was carried out, plant that was used in compacting; validation of the material used from its source (borrow / supplier). The Contractor's reputation would be at the very front of my assessment- if he's known for shoddy work, that would be information enough.

Provided the granular fill is shown to be consistently very dense (N = 50+ or even 40+), I can't see internal settlement being an issue.

All the best,
Mike

 

[oldestguy]

What do you know about density variation from loose to dense? Degree of uniformity? If this range is relatively small,(loose to dense) as for uniform material, there ian'r much risk of problem settlement, even if not at 95%. With coarser material, the risk is even less. I often wonder why folks speck 95% but will accept natural deposits at say 80 -83%. Go over the surface with a heavy loaded scraper and see what it does. At least it will provide some benefit up where it really matters.
Edit: If this is really that coarse, with boulders, etc. percent compaction testing by the usual tools is not possible and meaningless if tried. Even penetration type tests won't mean much. Watching the contractor work probably is the best method I'd propose. Watching that scraper second best.

 

[MTNClimber]

SPT's for sure. Do a couple of continuous borings on top of the embankment and at least one half way down (or up) the slope to make sure they were't just pushing material over the edge. N>50 may be a little high though.

 

[Tiarnan]

At the moment, SPT testing and correlation with a water replacement test from a trial pit are on the cards. Looking into the possibility of adding downhole shear wave or MASW to confirm no soft spots.

 

[oldestguy]

This sort of a question comes up frequently in areas where there is plenty of larger sizes. You can do the group here a favor by coming back and describing just what you did and the outcome, if possible. If it turns into a court or other situation, we can understand.

 

[fattdad]

seems that the ap65 is a crushed rock sort of material. If it is a manufactured aggregate, I'd expect it to be generally uniform.

Nobody saw it going in. No clue on lift thickness, number of passes, or water content. Clearly all those factors will influence geomechanical properties. (to what extent do you know the underlying soil properties?)

How much can the behavior of the crushed rock vary? Clearly, the friction angle and modulus can be influenced.

How would friction angle affect the embankment stability? Well, maybe the sideslope safety factor? Are the embankment side slopes 2:1, 3:1?

How would modulus affect the design? Well, that would inform the settlement that primarily occurs during placement, so that's not a huge risk. What about settlement after construction? Well, I'd think that would also be a minor concern, well, unless the natural soils are prone to deformation.

If this was in my highway network, I'd go in there with in-situ testing. I'd try the cone, I'd try the dilatometer and I'd also try geophysics. I'd think together with the direct reckoning of the material used in construction I'd be able to showcase variability (or lack thereof) and get an indirect reference to modulus and strength.

I don't think I'd try to relate any of this to relative Proctor compaction; however.

f-d

ípapß gordo ainÆt no madre flaca!

 

[Tiarnan]

So the material was too dense and coarse for CPT, which ruled that out. We had a decent idea of the material and the variability of it as this was imported fill with lab testing. The main outcomes therefore were to attempt to prove the appropriate compaction lifts had been adhered to, there were no obvious soft spots or less compact zones, and that the material in the zones where we were lacking compaction measurements was similar density/compaction to the areas where we did have testing. The whole process was on a pretty tight timeframe, and we had nearby embankments which had fully compliant testing, so we settled on the following methodology:

Go to the embankment with full compaction testing and carry out continuous SPT testing to establish the penetration rate for the engineered fill throughout this embankment, as well as the range of results possible. As the material was very dense and coarse, we settled on a slightly modified SPT process, to make sure we got a reliable estimate of the penetration rate (as everything = N 50+ wasnt too helpful). This involved first carrying out a full 450mm SPT test (which resulted in about 220 blows for the measured 300mm test) to establish a penetration rate. Following this, the rest of the tests were carried out to either the full test penetration, or to 100 blows after the initial seating 150mm, whichever came first. The penetration rate was then extrapolated to give the number of blows required for 300mm. Penetration/blowcount was measured in 25mm increments rather than the normal 75mm so as to identify the possible effect of large clasts within the fill. Typical extrapolated blowcounts (N160) were in the range of 100-350 blows/300mm.

As well as the SPT test, MASW testing and GPR testing were carried out on the embankment. MASW showed the stiffness of the material in the embankment and the level of consistency in this, and GPR was able to identify the individual compaction lifts to ensure the correct lift height.
After this testing was carried out on the fully compliant embankment, the same testing was carried out on the embankment with missing compaction information, and compared to ensure similar levels of penetration resistance (SPT), stiffness (MASW) and compaction sequence/lifts thicknesses (GPR).

Results showed tests from the non-compliant embankment closely matched those from the compliant embankment and compaction lifts were in accordance with the spec. Alongside the consistency of the fill material as shown from lab testing, this was deemed to give sufficient confidence that the embankment had been constructed and compacted in accordance with the specifications and could be signed off.

 

[Mad Mike]

Tiarnin- thanks for the feedback- you guys did a great job out there of proving the embankment compliance...the chaps doing the SPT tests must have loved you!

All the best,
Mike

 

[EireChch]

Sounds like a reasonable approach to test an already certified embankment and compare the results to an un-certified embankment.

Id be interested to know what Vs values you were getting in the embankment and also how GPR showed the lift thickness. Not criticizing just interested.

 

[oldestguy]

A nice summary. The method used might wall serve as an example of how to do it in these situations. Thanks for the explanation.

 

[Tiarnan]

Yea they had an electronic trip hammer which counted the running total of blowcounts too, so they made sure to mention as we crossed every thousand. At least we were able to use a solid cone so we didnt wreck too much gear, as the material was relatively controlled as we didnt need to sample, and due to the problems with large clasts blocking the splitspoon and messing with the penetration.
Vs values were in the region of 200 m/s in the upper metre, increasing to ~400-500m/s at about 5m.
A bit of the GPR screenshot below. You can see shallower lifts in the upper material vs wider in the lower in the upper image; and the transition from compacted fill to in-situ gravel in the lower image.