Reliability of Nuk Density Testing in Exploration Test Pits

[rockchip]

My question for the honorable forum is this: Does anyone have an opinion regarding the reliability of using a nuclear density probe device to obtain indications of density within a soil exploration test pit? We typically use them in ours within the upper 4 to 5 feet but have not heard of any liability issues associated with making judgements on density using this approach as opposed to other methods such as visual (caving=loose)or driving resistance like a Dynamic Penetrometer. Any thoughts or comments?

 

[rchanke]

What is your test pit, excavated pit or formed concrete laboratory testing pit? Give dimensions and wall height. Have you considered the limitations of the nuc when used in deep trenches or deep shored pits etc..?

 

[BigH]

Why not do some "conventional" density determinations (e.g., sand cone, rubber balloon) to obtain correlations with the nuke values - then you can make your own judgment as to the level of reliability you would place on the nuke test results. There is little better than direct correlations at a particular site and a particular site condition.

 

[Focht3]

You can adjust for the pit with the newer nukes by taking a "temporary" standard count while in the trench. This standard count will only apply to a single density test; I haven't worked with all the machines, but understand that this option is available from the major manufacturers.

I have used this several times - as early as about 1990. (I'm pretty sure that it's been around for more than 14 years, though.) I have compared the results to lab moistures and sand cone densities - they compare as well as "level" ground tests. At least in my experience.

If you are concerned, run sand cones on 25% of the in trench densities for comparison and QA/QC. This will also give you confidence in the machine results -



Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora. See faq158-922 for recommendations regarding the question, "How Do You Evaluate Fill Settlement Beneath Structures?"

 

[rockchip]

Thank you for your responses, rchanke, BigH, and Focht3. Yes, we are taking standard counts in the test pits and moisture samples from the test and keeping track of the correlations with moisture. The typical test pit we use this in is about 4 to 5 feet deep, and about 3 feet wide. We have not done sand cone yet but just started correlating the nuk testing with a Wildcat Penetrometer (dynamic cone penetrometer). The majority of sites we use this in are sand or silt soil, but some are in gravel. We have also found we have to specify the backhoe sub use a smooth edged bucket to reduce disturbance from bucket teeth, although this is not always practical and we often end up digging our test sites below the teeth marks. Thanks

 

[dirtsqueezer]

A fellow user of the the Wildcat DCP! That's another good way to impress a date, once you pound enough holes with that 40lb weight. :]

I remember over the penetration depth of the DCP, we were able to correlate a bearing capacity directly from the blow counts, given a weighted average over the entire depth. It seems to me there was a some science other than the spreadsheet provided, that I might be able to track down if you'd like.

It seems to me that if you were to encounter muliple soil layers of multiple resistances, a single proctor/nuke test may be impractical given the 5 meter range of the DCP. On the other hand, I don't think more information is ever a bad thing. I don't know if the forum would agree, but for shallow consolidated sediments, I just use a 1/2" probe rod, which speaks for over 10,000 PCF of bearing, and has basically the same depth-range of a nuke.

Let me know if you'd like the reference. I just changed companies, but I'm sure I could still get in contact with the PE we subcontracted for the work.

 

[rockchip]

Thanks dirtsqueezer. I would appreciate the reference you mentioned if you can dig it up.

 

[Ron]

George Sowers was one of the DCP pioneers. Check his textbooks (particularly the third edition) for appropriate references.

 

[lallatin]

The main advantage of using a nuc in a trench is the fact that the gage works just fine on the SIDES of the trench, meaning that the surface is smooth, requiring little preparation before the guide-spike is driven, and any number of tests, at the exact soil conditions you need to evaluate, can be tested in a single trench. The moisture readings will also be more accurate, without having to do a standard count inside the trench. THE GAGE DOES NOT CARE WHAT ORIENTATION THE TEST IS TAKEN IN. Use a maximum source depth, 12" if the gage has a long probe, and hold the gage fully seated. I've done this hundreds of times. Just e=mail me if you have further questions.

 

[rockchip]

That approach (sidewall testing) would certainly allow tests in "undisturbed" conditions. It seems like it would work best for silty sands and finer grained soils. Gravel soils may be more problematic. We will try using this method. Thanks for the idea.

 

[lallatin]

It's true that granular material could require extra care, and in some cases [dry, coarse-grained, etc.] might be impractical. The longer the probe, the less surface loss will affect the accuracy of the data. Safety must be your primary concern: wide or well-shored, strict OSHA compliance. This is a method that certainly reduces equipment costs if a backhoe is being rented. Pits can be dug wherever needed at once, instead of having to hand-trim the bottom of a pit at each depth for each test.

Glad to help. I do geotchnical/constructability services and seminars for a living, and don't usually give this advice away.