Differences in Sand Cone Side by Side Tests

[BobR549]

I do independent assurance testing in our department. One of the issues I face is when doing a side by side sand cone test with another technician we are at times 4 - 6 pounds different on our wet density. These tests are done no more than a foot away from each other in the same material. It is my job to determine where the error is and correct it. Other than the sand and cone calibration, what would be some other aspects to look at that may cause this much difference in wet density? This occurs at times in both granular material and soils.

 

[oldestguy]

Have you checked your sand for variations in gradation? Possible variable moisture content? How is it stored? How about how the hole is dug? How big a hole? Disturbance? Is moisture test done on a sufficiently large sample? Almost every step can have variations. I hope you are not using those small cones and small jugs.

 

[BobR549]

We use the larger 6" cone with the gallon jug. moisture content is usually within 1 - 2%. Holes are close to the same in volume. 500g moisture sample size.

 

[L0k]

how about balance calibration? it will measure the same weight for 1 pound?

 

[BobR549]

Balances are all calibrated. (That's my winter job)

 

[moe333]

Unless it is a very consistent, homogeneous material, some density differences due to variability in gradation and compaction effort should be expected even side by side.

 

[Ron]

Have you tried doing both side by side tests yourself to take out the "operator" differences?....then have the other person do two side by side tests to see if it is an operator issue or material anomaly.

 

[bb29510]

its going to be in the calculation, somebody rounding off, somebody not, not carrying enough decimal points, on the hole size i like five decimal point, unit weight of sand rounded off, there a lot of variable but most mistake are on the calculator. if you write up an excel sheet to calculate, that take care of those variables.

master ICC inspector, AWS CWI

 

[Ron]

bb29510....sorry to disagree, but there is no need for 5 decimal places of perceived accuracy. The test itself cannot carry that level of precision.

While it is good to check the computational procedure, if something is off in the 4th decimal place, it would not yield a significant difference in the end result.

 

[BigH]

Agree with Ron - the sensitivity to that degree shouldn't be there in a test like this - especially with a 4 to 6 pound difference. Can I ask if anyone has plotted the dry density and moisture content against the Zero Air Voids (ZAV) curve? Just to make sure that the result is on the correct side of the ZAV? While the OP said the "wet density" - I would wonder what the difference is with the dry density.

 

[mikeofBelAir]

Have you eliminated All the variables except the operators? Are you using the same balances? are the calculations being done by one person? I like to set up a spreadsheet to do the calcs--that way I know all the numbers are being crunched in the same way. You need to eliminate all sources of differences except the operators.

What is the material? If its some crushed rock product, i can see some variation in the digging of the hole and the transferring of the material. ( I once had to do a SCD in CR6 for a second Lt of the Army Corps. I told him the disturbance of the material while digging would make the result inaccurate. He didn't care--the spec said to do a SCD. I did the SCD and made sure that I disturbed the hole sides while digging so that the result was about 115% of the ASTM D1557 density. The office was happy and then went on his way.) The point is that is possible to dig the SCD hole in a way to slant the results---if that is desired.

 

[darthsoilsguy2]

balances level and away from wind, especially the gram scale....
have the techs switch equipment and redo...
take the remainder of sample to moisture test in the lab oven, does it match field burns?....
push both spots with your thumb before you test, do they seem similar...
Google "Hawthorne Effect"...