Ground Improvement 6

[Jastan]

I am examining a site for development which has loose sands at depth (say 6 metres). These loose sediments are overlain by dense stiff sands. The water table is quite high at about 2 metres from the surface

My question to the forum is with respect to dynamic compaction (in this instance the triangular and square wheeled rollers). Does anyone have experience with the dynamic compaction of these ground conditions, in particular below the water table?

My immediate thoughts would be that the compactive effort will be largely lost once it "hits" the water table.

Thanks

 

[BigHarvey]

The dense sands of the top 6 meters will not allow compactaion of the deep loose sand layer even with traditional deep dynamic compaction. The other methods you are mentioning can't improve sands deeper than say 3 m.
I would suggest you to use compaction grouting which is a mass treatment which can be selective in terms of thickness to treat. Obviously it will be more expensive than the methods you have been contemplating

 

[rchanke]

BigHarvey hit the nail on the head, compaction grouting can easily target the depth where soil improvement is required and can be quite economical when compared to the alternatives. Reasons why compaction grouting will work; 1. Loose sands are permeable enough to allow an economical grout injection rate. So your grouting production rate is not limited by pore pressure dissipation and thus can be performed more quickly ie "time is money". 2. The overlying dense stiff sands will assist in limiting surface heave from grouting operations. Therefore more energy will be focused on the targeted area creating a better product. The work can be easily verfied with pre and post grouting CPTs. Your more costly alternatives would be another form of ground improvement such as jet grouting or the most costly of all, remove and replace.

 

[jdmm]

Before you pose a solution you have to have a problem. Why are you concerned about a loose layer under a dense layer?

Is it a bearing capacity problem, liquefaction, settlement?

I would think that for most relatively lightly loaded foundations 6 metres of dense sand would be pretty good at supporting the loads.

Do you really need to densify the layer?

If you do then traditional DC (dynamic compaction) can do it.

 

[Myoho]

How do you know the sand is loose at 6.0m? Are you estimating density from Standard Penetration Tests (SPT's).

Be very carefull when interpreting the borehole logs. If SPT's are carried out without balancing the water pressure in the casing then instability will occur at the base of the borehole and hence the SPT N value would be artificially lower. This would suggest the sand is loose but it is actually dense.

 

[Jastan]

Thanks a lot people for your responses. All them have been helpful.

The sands were defined as loose by EFCPT probes, and the concern was for liquefaction during possible earthquake conditions. I realise that at depth, the dense sand overlying the loose sand will provide a good foundation under normal circumstances.

If it is decided that the loose sediments are a problem at depth, I understand that dynamic impact rolling is not the best method of ground improvement.

Thanks for the feedback. I hope I can be of assistance to others in the future.

 

[jdmm]

Jastan,

Just a note about a dense layer over a loose layer. From a geological perspective it is hard to imagine the geologic process that would allow the densification (compaction) of a surficial layer that would not also affect the underlying layers. I am very suspicious of loose under dense sand.

Our methods of analysis measure strength directly and density indirectly. SPT does not measure relative density it measures the strength of the soil to resist penetration. Frequently we see lower resistances from layers that are more related to the grain size distribution and the grain size and angularity (ie friction angle) than relative density. If you are using a CPT with pore pressure measurements then look for pore pressure increases in the "lower density" layers. This suggest that the soils have higher silt contents and frequently lower strength. If there are interbedded sand and silt layers the results will tend to indicate lower density layers than in reality.

regards

 

[Focht3]

I agree with [blue]jdmm[/blue] - look real hard at your data before proceeding.

Were any conventional borings done? What type of equipment - wash rotary or hollow stem auger? Did you get SPT results with depth? Grain size / -#200 results with depth?

As a rule, I won't rely on remote sensing data alone (i.e. CPT) without at least a few conventional borings unless I have A LOT of experience all around the site, and I have a thorough understanding of the site history and geology. I'm aware of too many cases where a CPT-only approach got a project - and the geotechnical engineer - in very hot water. If possible, get at least one conventional boring close to the "loosest" profile. Be sure to use an experienced driller, and wash rotary techniques. NO HSA -



Please see FAQ731-376 by [blue]VPL[/blue] for tips on how to make the best use of Eng-Tips Fora.

 

[BigH]

Focht3 has a good point - in any site investigation it is necessary to get some recovered sample so you can feel it, smell it and taste it!! Roll the material about on your hand - get dirty. For many of us, seeing is believing .

In my company - the "new" soils engineer was required to be in the field doing drilling logging almost 5 years (off and on - mainly on). I strongly urge all new geotechs to go out and get their hands dirty - or, in Canada, freeze your bottoms!!

 

[ashjun]

Hi Jastan,
Yes your observation that dense sands overlie loose sands does not surprise me. I myself have observed such a phenomenon. There was a case when the geotechnical engineer was susceptible of the kind of cone penetration values observed and merely "corrected" (I hate to use the word manipulated) the values based on his "judgement".

You are right if concerns of liquefaction potential of the lower sands are raised (especially in earthquake prone areas). Have you thought of considering sand compaction piles to improve the ground?
Regards

 

[22222222]

Hi Jastan,

What is under the 6m loose sand layer? What is the area of your site for development?

 

[jdmm]

With much respect to BigH and Focht3,

Your feelings about CPT testing are noted.

Eventually the dropping of a weight by a harry fellow pulling a rope to cause a rod to penetrate the soil will be abandoned for a computer controlled strain gauge measurement of soil resistance combined with dynamic pore pressure measurements.

Correlations with the "old" established techniques are well documented.

With respect to Focht3 comment about "remote sensing" I would consider a junior engineers description "remote sensing" and CPT testing direct sensing.

With respect to BigH wanting to taste the soil I have no comment.

 

[BigH]

jmdd - you never did the chew test for silt/clay? (I never have either!!) I'd rather down the lager with .

 

[Focht3]

I prefer to think of it as the "grit test" - and I use it from time to time...on selected sites, of course!



Please see FAQ731-376 by [blue]VPL[/blue] for tips on how to make the best use of Eng-Tips Fora.

 

[insitu1]

If liquefaction is a big concern I'd advise doing some siesmic CPT before you embark on any mitigation. I've seen more than one ground mod job go bye bye after we had some Vs data.

 

[doc11]

Jastan,
As far as Loose sands underlying dense sands, that is a relatively common phenomenon, and it can happen with any soil or rock type, especially if there is groundwater or faulting invovled.
I would have to agree with Focht3 and BigH, you do need to get some good samples of the suspect zone (mud rotary no HSA). Cpt's are very useful and can dramatically reduce the cost of a site invesigation but I would be leary of useing them as a sole source of information. There is still alot they can't do, namely give you give you a sample that you can look at (and taste if you want too).

jdmm you are right something will eventually replace the cathead and rope and maybe even the drive sampler but I will still refuse to shave.

 

[BigH]

Perchance Harry is hairy??

 

[doc11]

Not as hairy as I once was, time and too many hardhats have taken most of it off the top

 

[AZdinak]

If you do in fact have a loose condition that cannot be reached with conventional DDC, I suggest two other methods that could be more economical that compaction grouting: vibrofloat/densification and explosive compaction. Both of these can treat relatively large areas fairly quickly and be designed to treat the relative desity of your sands at the site if they do indeed need improvement. Not saying to use these, but at least consider them, as the technologies have been used quite widely.

Zdinak

 

[dirtsqueezer]

jdmm,

I know this may be talked out, but if you find that your deeper layer is in question to the point where you proceed to improvement, I've seen swampland improved to the point of structural building quality using stone columns. You may need an underlying competent layer, but the ballast injection process also serves to densify the surrounding material, in your case the elusive loose soil. It also and may be slightly cheaper and practical than grout fill, and require less testing. I don't know if this method is exclusive to our local companies, but for what it's worth. Best of luck.