I have a question regarding the moisture content value that is dispayed on nuclear densimeters.

In our laboratory a discussion was brought up as to whether or not the value displayed for moisture is already corrected for the amount of rock that also is within the tested area.

I have perfomred sand cone correlation tests and speedie companion tests and I have shown to our Materials Engineer that the gauge DOES NOT make this correction, it doesn't discern what is rock and what is soil, it gives you a total moisture content. The correction is only applied once you screen your material over the applicable screen dependent upon proctor method once you excavate underneath you gauge.

The general argument is that you should not be correcting the nuclear gauge moisture value, and should instead, utilize a speedie moisture tester and correct its displayed value in leiu of the gauge value. I explained to the engineer, that both methods are acceptable, and you can correct either value, they are both giving you the same number. He wants to change Line 13 to read, (speedie reading).

Attached is our worksheet that we utilize, we use Proctor Methods A primarily and C proctors occasionally.

Firstly, ASTM D2922 and D3017 are discontinued standards. The newer standards should be followed.

Second, there is no requirement in ASTM D5220 to adjust the moisture content obtained from the guage for oversize particles. Nowhere in the standard does it say to correct for oversize particles by ASTM D 4718.

The optimum moisture is corrected on the proctor to correct for oversize particles that were removed from the sample. The nuclear guage considers the whole soil mass including all the rock and gravel so there is nothing to correct for.

It is sometimes useful to compare the guage moisture to moisture obtained from another source (usually direct heating method). You probably need to correct for gravel removed from the speedy sample since you can't fit much sample into a speedy (another reason to lose the speedy).

I would also mention that there is no rock correction to density from the gauge as shown on the form. It looks like whoever created the form does not entirely understand what they are doing. Just show them the relevant standards and go through line by line.

Its been discussed before but the use of the nuclear density meter should, if at all possible, be correlated to the sand cone - you can do this as "end result" dry density or you can do moisture to moisture, etc. Back in the earlier days - 1978, I found that the nuclear density meter was readings for moisture were high - water content x 0.7 = oven moisture content .. . almost all big jobs I have been on a correlation was/is a must. For small jobs, you are likely using with fairly known materials so you should have a general region-area correlation.

Agreed with the sand cone correlation, that is a very good practice when utilizing oven dry samples. Your experience with a +0.7% solidifies what I have been saying, that the gauge gives relativly the same value as an oven/dry or speedie, give or take a %.

When we utilize our sand cone form or nuclear form to perform rock corrections for Method A or C, you end up with a rock corrected h20 content by which you back calculate the moisture out of the wet density. If the gauge back calculated for say 50% rock, it would read a value of basically half of the oven/dry value, which it doesn't, but that is what is being purported.

This is the heart of the debate, what is being said is that the nuclear gauge automatically does this already, which from my studies, I have shown that it is usually closer to an oven dry or speedie total moisture, but never close to the rock corrected moisture value, line #13 on the sheet. The only reason why the form says nuke or speedie, is in case you didn't have a speedie with you, or you were doing lots of testing and wanted to rapidly obtain a moisture value. The materials engineer believes that if I use the nuclear gauge moisture value, and correct the h20 content with a rock correction, that I am double correcting the value, which isn't true, especially if I have already obtained oven dry samples and made the appropriate offset corrections within the gauge programming.

If the gauge automatically corrected the moisture value based on the rock content, I would never even need to excavate beneath the gauge to begin with, but we all know that nuclear gauges don't do that yet.


To the first poster,

I will make some updates to this post tommorow, I believe you are referencing a different test method, both ASTM and AASHTO methods explicitly point out making corrections for oversize. I will update soon with the correct methods/designations.


Hopefully I didnt rant, and hopefully I got my point across well enough for people to understand what I am saying.

@itzobi - it isn't 0.7% - it was 0.7x; the nuclear would say 9%; the oven dry was 6.5% or so. Machines were calibrated - but perhaps the newer machines are a bit more refined . . .

I stand corrected. I see now the paragraph you are talking about. It's not part of the new standard. Looks like the old standard lets you average a few tests as an alternate.

The oversize correction certainly opens a can of worms. Are you performing bulk SG tests on the oversize particles in the field according to C 127? I hope you have the correct sample size, calibrated measures, etc. Do a one point first though before you do the nuke test. Meanwhile, the contractor is standing by to know if he can spread the next lift.