compaction problem.

[bb29510]

Been doing this 26 years so no dummy here. I got this pink silty whatever, it not a sand, not a clay. It classified as a SM. its almost talcom powder. It wont get over 92% modified. I have double check nuke, sand cone ,drive tube all within a pound. double check PD. one points under nuke, still on curve.
Pd is 121 at 12 but I cant get over 110 lbs no matter what. Huge vibration steel wheel, rattle the building. 6 inch lifts. I cant figure this dirt out I can do a one point at the nuke moisture under the nuke and get 119 lb so that mean the roller should get close to 116 lbs, cant get over 110.

anybody have any ideas?

I,m flooding it right now hoping to use the water as a lubicant to get the final compaction 2 % Its not pumping and never had pump. the soil under it is white beach sand.I got about six feet of this pink

 

[cvg]

rock flour? wasn't there a geotechnical investigation?

 

[bb29510]

its borrow material

 

[cvg]

ok, so geotech of the borrow pit? sounds like unsuitable material. sounds like classification error - maybe it is ML, not SM

 

[DRC1]

How loose is the underlying sand? if that is loose, that can cause compaction problems, as the sand is densifying also. Also I know you said it was an SM, but when wet, rock flour can act clayey. Maybe a sheepfoot roller would help, although this is more of a stab than an educated guess.

 

[Ron]

Agree with DRC1...walk it out with a sheepsfoot, starting on wet side of optimum. If you can get that much compaction in the lab, you can get it in the field, but you just have to modify the procedures.

 

[cvg]

I would highly question a lab that says the material is SM. You are saying it is silt (ML), not sand (SM). And based on your description and experience, it is behaving like ML. Therefore I would also question any PD given by the lab.

 

[Drumchaser]

I am sure you have tried shooting through a lift to check the lift underneath.

 

[bb29510]

the testing lab is me, I beat my own proctors. I have done three this week on same material, still coming up the same. It a weird material, not a sand , not a clay but have enough, grainer material to be classified as a sm but behaves like a ml. it will compact tight as a brick but bump it and it falls apart.

I have check three nukes today, one being an asphalt nuke, sandcone, drive tube, all withing a pound of each other. Just cant get it in the field. thinking it might need a day of wobbly wheel

 

[BigH]

Have you done a sieve after compaction to see if the gradation is the same - perhaps it is breaking down and you are comparing apples to apple sauce . . .

 

[Ron]

brownbagg...it sounds like the material might be severely gap graded. Do what BigH suggested...run a grain size on it.

A wobble wheel won't get you as far as a sheepsfoot will with this type of material.

Post your gradation...we'd be interested in seeing it. Also post one of your Proctor curves.

 

[darthsoilsguy2]

as drumchaser said...
you checked the lift under the previous since you might have confinement issues with the compaction at surface.

 

[BigH]

darg2 agree --- which is what I think specs ought to say when compacting relatively uniform sand - the compaction would be based on checking the previous lift . . . interesting dilemma

 

[Drumchaser]

brownbag


Sounds like economics are coming into play.
What will the embankment support?
Is the pit/cut area at the contractor’s discretion/choice?
What is the avg fill height/CY to be placed?
Is stabilization possible? Or a change in borrow sources?

The experiences with silt I have encountered were not easy to work through. If I am remembering correctly...the borrow source was in close proximity to an old waterway that had been around a long long time.

I would be interested in knowing what the atterberg limits are. Also would like to know the pyramid of shots (roller/compactor passes) as this material is being placed and at what point it stops gaining.

 

[iandig]

I had a really similar problem a couple of years back, with an esturine silt. (Site was in the middle of the UK!).
The solution for the monitoring of the density of the fill was from a multiple approach, including looking at the air void content and the variation in MDD between the 2.5 (standard) and 4.5 (modified) rammers. I also used the DIN 18134 plate load test as a rapid means of assessing the abiltiy of the ground to be further compacted. If you are not familar with the test, the set-up is pretty similar to a normal plate test, but the manner of loading and the calcs are very different. The test uses a 300mm diameter plate and is loaded in 2 cycles. The first cycle has 6 loads and 3 unload stages, and the second just has the first 5 loading cycles again. For each load cycle you determine Ev, then do the Ev2/Ev1 ratio. German (DIN) guidance then provides data on the acceptable ratio for different materials and different degrees of compaction. Test only takes around 25mins as each load is kept on for a maximum time, not till all settlement stops so its very different from a lot of the other plate test methods out there.
Finally, in order to be able to compact the fill, each and every layer had to be dampled down. As the material was relatively free-draining, any wind or sun would dry the surface and you just could not compact the fill.
Of key note was that the minimum air void content you could achieve was ony around 5%, so we widended the envelope to less than 8% (actualy did the calcs on sphere packing and came up with very similar densities to what could be acheived on site, next step would have been to do a petroraghic analysis and check shape of silt fines). Any clay fines would increase the density which would be achieved. When the material was not confined (i.e. in a mould in the lab) the top 200-300mm always seemed to act loose, so we resorted to excavating the top 150mm and only testing the material below. The MDD between the 2.5kg and 4.5kg was often very similar. The compaction curves were very flat. All plate testing was done using multiple cycles (not just the DIN) method and it took a couple of load cycles to get the material to act elastically. Finally, once the slabs were down we did drilling and TRL DCP probing through them to check the stiffness and performance of the fill. Once it was confined under the slab and roads, the stiffness/CBR was what you would have expected, just very difficult to prove this when the upper surface is unconfined.

 

[BigH]

Thanks Ian. Would be interested sometime to see the information you had gotten - did you write a paper or technical note on it? Would be well received.

 

[iandig]

Hi BigH, No, but I am thinking I should have. There have been a couple of other 'brownfield' sites which I have done over the last couple of years where I have used a similar approach. One of these was the engineering of 120 seage lagood filled with human waste which we re-engineered a a general fill, and other was with a highly variable made ground containing ash, slag, partially burnt coal and other nasties which was also contaminated (ex chemical works which was still in partial operation). By doing the on-site assessment and re-engineering, we were able to limit exposure of these materials to workers and lab staff in particular.

 

[BigH]

Okay Ian - you admit you should have . . . so DO IT!!! Be a Nike star. I think that if you have the time, it would be very beneficial to the general geotech. Whatever happened to the old heal test????

 

[iandig]

We replaced the heal with the CBR! means about the same!