Controlled fill QC 8

[amazing azza]

Starting on a new project. A small sugar cane processing plant in sunny Indonesia. We are scheduled to begin earthworks soon. Going to strip the vegetative layer (20 cm) and do a controlled fill with clay up to grade level for the main building. Total height of fill - 1 to 1.5m. I need help with specifying a QC procedure for this from the owner's side. Our spec is 95% MDD, 20cm lifts.

* Is it realistic to do sand cones on every lift? Several samples?

* I am planning to require a minimum size for the sand cone - 200mm cone with 4L jug.

* What is the proper equipment for compacting each lift? I have seen them simply rolling over the incoming fill with a dozer before. Dozer spreads and drives over, spreads and drives over.

 

[BigH]

Talk to a geotechnical engineer - unlike years ago, there a quite a few good ones around these days. Where in Indonesia? I worked before in Padang, Kudus area, and Sorowako.

 

[GeoEnvGuy]

Old school sand cones probably not the best. In canada the standard is nuclear densometer but they also sell a electric gauge for non nuclear applications. Compacting clay is done with a sheepsfoot or pad foot roller.

 

[BigH]

GeoEnvGuy - nuclear gauges, while they are there - the use of sand cones would be "standard" . . . for a lot of reasons - money and headache of nuclear. In India, no one wanted to use nuclear guages - they thought you'd become sterile (same back in the late 90s in Northwest China). Once in India, the lab guy came out for a nuclear guage reading on a big job using a full lead head to toes apron. Big issue will likely be quality control supervision of whatever they do.

 

[amazing azza]

Thank you guys for the replies.

Well whadayaknow? We are in Semarang, very close to your old stomping ground BigH, if you have anyone you would recommend, I'd love a contact. For geotech, we have been going through the local university, but they are not very talkative. Its hard enough to nego and schedule the testing with them as it is. I guess there isn't a shortage of customers.

I was hoping for some pointers. Some dirty contractor tricks to look out for... How likely is it that they'd be willing to accept testing of every lift?

 

[BigH]

amazing_azza - know Semarang well - Simpang Lima! I used to stay at the Dibya Puri when in Semarang (which, if you are too young, your parents would know). Torn down now; supposedly was inhabitated by a number of Dutch officer ghosts from WW2. My wife is actually from Kudus.

How big is your area? How many m3 is expected to be? I am not familiar with geotech firm there but you might want to try Dr. Paulus Rahardjo of the Catholic University of Bandung. I will contact him and let him know someone may be calling. GIve me a day or two.

First, you need to establish the borrow/fill material. Characterize it by Atterbergs, Standard Proctor, typical in situ moisture content. As GeoEnvGuy says, sheep or pad foot rollers are usually used. A pneumatic roller has also been used. There are also small tag-along type padfoots. Please provide some details - like the fill areal size, anticipated fill layer thickness. In building a dam in Malaysia and using 22 ton vibratory pad foot rollers, we used 300 mm loose lifts. If you would be using a tag-along padfoot, the lift thickness will have to be less - something like 150 mm. For a large area, this will take time given the tag-along's small size.

There are several ways to determine the in situ density - first off, sand cone. In Asia, labour is cheap and it is typically, in my experience, used. You could also do drive tubes - push in or gently tap in a 100 mm thin walled tube - and then get the density from the core taken - nuclear can be used; I personally have less faith in them - they are typically "off" on the moisture content.

I would suggest, especially given the likely quality of the local contractors to do several density tests per layer. In India, the highway requirements are a minimum of 6 tests (seldom is this done on other jobs but it was a government requirement). You would determine the average and ensure that the average was greater than 1.65x the standard deviation plus the unit weight required at 95%. (or whatever you are using) This is to ensure you have uniformity. Again, the number of tests would be determined on the areal extent of the fill layer.

 

[amazing azza]

BigH, thank you very much for the detailed info! I have never been to Dibya Puri, indeed it was before my time. But, I have been to Lawang Sewu (old Dutch railroad depo). I was instructed to only walk on a certain side of the stairs and to touch the railing, or else risk being haunted... What a small world, I have some relatives (wife side) in Kudus as well!

The area to be filled is about 2400 m2, estimated fill depth 1-1.5m, giving a total fill of 2400-3600 m3. The contractor we are talking to right now said he would do that in 3 lifts. Given your info, seems a bit thick. I will check on the roller weight/size they are planning on using.

Definitely planning on the sand cone, radiation is like voodoo magic around here. For a food safety x-ray, the operator has to complete a *month* long training in Jakarta. Oddly, the contractors around here seem to frequently use CBR (California Bearing Ratio) for quantifying fills. I know it is inappropriate, but what is the mechanism why?

 

[fattdad]

Regarding the initial inquiry: Nothing wrong with sand cone data. I don't like 0.5m lifts; however.

A watchful eye is important. During a work effort, they will spread and fill various areas - perhaps simultaneously. Each of these work areas warrant testing, as these work areas are unique. How big is the work area? How consistent are the soils? That will inform a reasonable testing frequency.

How to address your reference density? I mean is there going to be gray and brown soil? Will that be your basis for which MDD to use? Will it all be red clay with varying sand content and plasticity? Hard to eyeball the correct reference density.

My advice? Have in the field a concrete block and a proctor mold. Carry a frying pan (or Speedy moisture tester) and capture your own proctor density. Perform the test at less than optimum water content and walk the field proctor point onto your lab curve. Now you'll know.

If you check your reference density and focus your tests on the specific work area(s), a handful of sand cone tests may be all that you need on a daily basis.

Bear in mind that the sand cone hole is only a few inches deep, so a 0.5m lift will not be properly gauged.

f-d

ípapß gordo ainÆt no madre flaca!

 

[cvg]

Is it realistic to do sand cones on every lift? Several samples? - no, typically nuclear which is verified on a daily basis with sand cone

* What is the proper equipment for compacting each lift? I have seen them simply rolling over the incoming fill with a dozer before. Dozer spreads and drives over, spreads and drives over. this depends on the type of soil, but generally one pass with a dozer is not adequate. you might consider a test fill to determine an approved procedure (type of equipment, necessary moisture conditioning and number of passes/thickness of lifts) and then just forgo most testing after that point as long as they follow the approved procedure.

 

[BigH]

Mates - I think to remember what country we are looking at and what Contractors will be involved. It is very doubtful that a project of this size will see any level of sophistication - I'm hoping that they will have a padfoot and not a smooth drum. Agree as I indicated before, 300 with a 22 tonne vibratory padfoot should work quite well. Fattdad made a good point in piecemeal placement. I would hope that with an area of only 2400 m2 (i.e.,50 m by 50 m or 40x60) that it would be wise to do the whole area at one time per each lift. Check out the borrow area for "uniformity". Take several samples and do lab proctors on them. Note colour, texture, feel of plasticity. As most of the likely fill is from a flood plain - the Semarang area except for Candi Baru is flat. So it is quite likely a fairly uniform borrow deposit. I would suggest in an area of 50x50 to do at least 3 sand cones per lift. Ideally 6 could be done (have a couple of crews) - then you can get a good handle on average and uniformity of the results. Again, labour is cheap.

300 mm lifts give say 5 lift. In lieu of a "trial pad" which would likely not make sense for this site, go with a minimum of 6 single passes of the 22 tonne padfoot. This should give a good result - on the first lift - you can check with the sand cones. For an area of this size, a lift a day would not be unreasonable. Can study the results at night and give okay or not for next morning. Also, observe the behaviour of loaded trucks on the fill that is placed after compaction (when putting on the next lift). Any "sponginess" or rutting? On the material we used in Malaysia (PI of about 20, LL of about 45 to 50), we ran heavy articulated loaded trucks with hardly any deformations under the wheel load (8 passes). When I worked in the area, I doubt there was any such testing on fills like these (road from Semarang to Kudus - 1988).

Spent a lot of time in Kudus - my "home" as we were doing feasibility study for coal-fired power plant - ended up at Tanjung Jati - near Jepara. Played a lot of basketball at the Kudus military base.

 

[fattdad]

I'm in complete agreement with BigH. I have also worked in non-US countries and know that what we do as customary is not so much in other places.

f-d

ípapß gordo ainÆt no madre flaca!

 

[amazing azza]

So much great info. Thank you all! It took me a whole day to digest it

fattdad, if I understand correctly, you are suggesting to do the lab proctor curve on the fill sample. Then, say, at the end of the day, do a single-point proctor measurement on-site, followed by some sand cones. Then compare the sand cones against this value? The reason for doing this is to control variability of the fill soil? The shape of the proctor curve would be assumed the same as previously determined in the lab, but it will be transposed onto the nightly single point measurement?

cvg, the project is small in the grand scheme of things. It will be hard to justify a trial pad...

BigH, I like the nightly testing approach. Combined with fattdad's single point proctor, I think that will be my method. The machine will be a 10-ton, vibrating, flat drum in the front. I guess that means I should go with a bit less than 300mm per lift. The fill soil will be large lumps of sandy clay. They call it "tanah padas" around here. I am attaching a picture below. Lab tests will be done shortly. This material comes in boulder-sized chunks (other choice is red dirt "tanah merah", fertile soil that is fine and crumbly, more expensive).

You guys hit the nail right on the head about the standards abroad. I realize that I cannot do things truly 100% properly simply because the techniques/equipment are either non-existent or too expensive, but I still try to get as close as I can with the resources given. In the land of the blind, the one-eyed man is king

PS. Kudus KODIM basketball/tennis court is still there and is as popular as ever. I recently played some tennis there Where were the other choices for building Tanjung Jati?

 

[BigH]

First - the use of a smooth drum roller in clayey soil - even like you have above is not ideal. Do they not have pneumatic rollers which in my view would be better. There is a need to knead the material, not pound it.

My point on the general construction approach - I do not know the level and efficacy of the contractors and you may have one that isn't as well qualified as on "bigger" jobs. I hope I am wrong and that the contractor will do the work according to good practice.

Played many a basketball games a the court. There was a missionary from Erie PA when I was there and every Thursday night he would come from Pati and he and I would pick three girls and play any 5 boys who wanted to try their luck. We usually won - the girls played great when against the boys - but when they played with girls against girls, they were quite sloppy in their play. They were kinda cute - but I was nearly engaged at the time! When I was in Kudus - I stayed at the Asri Jaya until I married, then we stayed for a few months at the Notosari before heading off to Pakistan and then Canada. I really liked the satay ayam we could get at the old stall complex near Simpang Tujuh. Take care - and remember the most important words for a guy when he is with his gal . . . walking down the street or in the mall . . . "Liku-liku laki-laki"!!

 

[fattdad]

a lab proctor is accurate for that sample only. Small changes in clay content, grain-size, etc. can affect the ultimate MDD. If you take a sample from the field and run one point on the field sample, you can walk that proctor point onto your lab curve and see if it falls reasonably onto that lab curve. You MUST do this on the dry side of optimum, 'cause they all merge together on the dry side.

Let's say the field point plots below the lab curve. Just translate the lab curve to the new point and get a new MDD. You must translate along the line of optimums (parallel the ZAV).

f-d

ípapß gordo ainÆt no madre flaca!

 

[amazing azza]

fattdad, I think I understand. If you look at the graph below, let the black line be the multipoint lab proctor. The field sample is done, and is represented by the orange circle. The proctor curve then shifts along the ZAV line until it meets the field sample (orange curve). The new maximum dry density becomes 103.2 lbf/ft3. The new optimum moisture becomes 18.4%. Sandcone test would then be compared to this new MDD, and not the lab MDD.

BigH, pneumatic rollers, you mean like the kind used for laying asphalt? Heh, I like your basketball approach. Did you get a chance to try sate kerbau or tahu telur (some more Kudus favorites)?

 

[amazing azza]

Found an excellent video on the subject:
CEEN 341 - Lecture 6

In the video, he mentions that the compacting spec is, say, 95% modified proctor, +-3% moisture of optimal.

Question: how does one control the moisture? Spray the area with water? Compact more/less such that the proctor peak is in the right location? What if the soil dries out in the sun afterwards?

 

[BigH]

amazing_azzA
Where were the other choices for building Tanjung Jati? One site was near Rembasng - east past the city - beautiful white beach - could wade out 500 m and still have the water just above your waist. The other site was near a rubber plantation west of Semarang.

 

[amazing azza]

One site was near Rembasng - east past the city - beautiful white beach - could wade out 500 m and still have the water just above your waist

Yeah, that area has some nice beaches. I have to yet to explore there...

The other site was near a rubber plantation west of Semarang.

I think I know this one

 

[fattdad]

Hi Azza, Yes, you're example is my approach. The shift is along the Line of Optimums, which is how you drew it.

There was a provisional ASTM on this topic. My former boss, Steven Poulous drafted it. He was a founder of a US firm, GEI.

f-d

ípapß gordo ainÆt no madre flaca!

 

[amazing azza]

An update and a question.

Stripped the vegetative layer with a dozer, then 30cm more using dozer + excavator. Placed the borrowed soil elsewhere on the property. Underneath removed material could observe traces of a previous fill with gravelly clay and some natural soft black silt. Proceeded to compact this surface with a vibrating roller (this is the only kind of compactor I was able to source, given the limited supply of heavy equipment).

Just as we were preparing to start the fill... it rained. A lot. The whole field turned to mud. It's been 48 hours since, but still not dry due to cloudy weather.

Question: what is the best way to proceed?

- wait until dry? how dry? to 95% (modified) proctor levels?
- wait until top surface is just dry enough for trucks to pass?
- place the dirt on top of the muddy field as is?