reliability of SDG

[killswitchengage]

Hi
I will be using the soil density gauge to control soil compaction for the first time. Since i have no knowledge of this test beside the fact it uses dialectic properties of the soil and air and water , i want to know more about it . Especially , the reliability of this test and how deep it can probe , also what will happens if large boulders or cobbles are in the area of the test or when salty water is used in the compaction process ( i know salty water transmit electricity more efficiently )

 

[oldestguy]

I'd do a lot of research on it so as to be very familiar with what has been found already.

For instance, this is a quote from one paper I found.

Triplicate readings were made at each test po
int without moving the 6-inch probes. The EDG
measurements made at each test point were ad
justed based on calibration densities obtained
from sand cones or nuclear density readings th
at were suppose to cove
r the range of values
expected for the project.

I'd also have calibrated sand cone data for site, to calibrate all readings you get initially.
As to calibrating, be sure the sand-cone is calibrated also. Not all sands have the same density.

If the nuc is used, also that needs calibration

It ain't so easy is it?

 

[killswitchengage]

I am sorry but i am not getting it ! maybe its the language barrier .
If i am correct you said i should use either nuclear or sand cone tests and calibrate them with the SDG right ?
But my questions are more specific as you can see

 

[emmgjld]

Agree with OG.
Back in the mid to late 1960's, my task for my father's Test Lab was to field calibrate the Nuclear Gauges (Nuclear Chicago) and work out the individual gauge characteristics for the different soil conditions. This was before Troxler and all measurements were Back Scatter. Careful calibration with Sand Cone was required. In my semi-retirement, I am looking to go alternative density testing & do not look forward to the numbers of careful sand cones I will have to perform.

 

[oldestguy]

OG here. First I would assume the SDG is not correct and needs calibration against a known density that can be verified. What might that be? Well with soil density being measured there would be some difficulty, usually. See comment at end.

Instead you can take readings on a given spot with the SDG and then a reading with either the sand cone method or the nuc gags. However, either of their readings must be correct and having been previously calibrated by some other means.

For the sand cone calibration, a CBR mold (of known volume) can be used to calibrate the sand used. For the nuc one method is to have large boxes filled with different density of soil mixes and of known moisture content.

There after with either the nuc or the sand cone are calibrated, they can serve to calibrate the SDG gauge for the your site conditions of the individual job being tested. On another job site you may have to do the calibration for that site again with either nuc or sand cone due to resistance changes not related to density.

In summary, the order of calibration: the nuc or sand cone methods have to be calibrated against known densities first. Then calibrated nuc or sand cone is used to calibrate the SDG, quite probably necessary on every job due to chemical content that affects electrical resistance, not necessarily density. In addition, it likely is necessary to know the job site moisture contents by method other than nuc , which means lab or other separate moisture content tests are required.

There might be soil techs that by-pass the individual job site calibrations. I'd find it quite easy to give that tech a push down the road to another employer.

 

[Ron]

KSE....I assume from your terminology you are considering a Transtech SDG. The reliability of the gauge can only be assessed by calibration as OG and emmgjld have noted. My concern, as with other types of soil density gauges is, particularly, the accuracy of the moisture content measurement and how it is incorporated into the internal calculations. Here's an example:

Transtech states that its SDG has a moisture content determination accuracy of +/- 2%. Suppose your compaction specification says that you must compact within 1-1/2% of optimum moisture. Well, you can't check that accurately with the gauge.

Now assume you have a wet density of 110 pcf. If the moisture content varies by +/- 2%, then assume your measured moisture content is 12%. That means it could be from 10 to 14 percent in actuality. So your dry density could be from 96.5 pcf to 100 pcf. Let's assume your Max. Dry Density from the Proctor is 105 pcf and your specified compaction is 95% of that value. Based on the inaccuracy of the moisture content alone, you could either pass that test or fail it, depending on which end of the range your "measured" moisture content might be; however, at the measured moisture content of 12% you would actually tell the contractor he failed the test (94% compaction).

Bottom line....you must do a sufficient number of comparison (calibration) tests with a direct volumetric method such as the sand cone or drive sleeve, along with oven dried moisture contents to properly assess the reliability of the gauge.

 

[killswitchengage]

Thank you all
I have acquired the astm D7830 which talks about the SDG and predictably it talks about this calibration issue. Now i am about to control compaction of Tuff and i want to advice the client not to let the contractor use sea water ( we are in the shoreline). The soil brought in by the contractor has large boulders and stuff so i am looking forward to remove the material and change it with a much controlled one . But the thing is i see big agglomerations of sand in this material that can be broken apart with hands , so when it comes to sieve analysis and proctor test what can we do about it ?

 

[oldestguy]

I have seen cases where the on-site material has peculeary properties and conventional testing isn't working well. So we have used other methods to "calibrate" our conventional test. One method was to dig a large sample and measure the volume from where it came to compute actual density. One means of measuring the volume was to use calibrated sand (as from sand cone method) dropped on the area as if from a cone. Anther was to use thin plastic sheet and fill with known volume of water (top of hole had to be level). Lab compaction results had to be corrected for cobbles, etc. In your case you may have to make some other corrections to account for the difference between lab compaction and field compaction methods. Lab may break down material while field does not.

 

[Ron]

For the agglomerations, about the best you can do is use a tilling or reclaiming device to break up the material as best you can, then make sure the in-place moisture content is above optimum and drying back as you compact.