Cobbles in PSD graph

[Deacy]

I need some advice on reporting cobbles as part of a particle size distribution analysis.
I test by BS1377 with help from BS5930. This standard tells me to describe mixed coarse soil as if it had no cobbles and boulders first and then to add them in to the description as occasional, some or many. That's fine but should they appear on a log %passing graph? If not it appears that they are not a significant part of the test sample and if they are included it throws out the relative % of all other particle sizes. This has cause some confusion since the ammended standard came out in 1999!

 

[slowoldgit]

Why not present both, with relevant comments?

Also, in my experience (marine SI), the weight of sample for testing required by BS can be beyond the recovery in shelby and drop-core samples. Something for you to check.

 

[iandig]

You should include them on the graph IF they appeared in the representitive sample. From BS 1377-2:1990 your sample mass could be very large if you have cobbles - was it? and did the representitive sample include them?
The descriptive terminology from BS 5930 is just that, a method and means to produce a representitive and repeatable description. The advice in BS 5930 is that you are consistant for all samples for a specifc job, so don't do one thing for one sample and one thing for another.
Although BS 1377 deals mainly with soils <63mm, there is no reason why you should not include Cobbles and Boulders, IF they were in the representive sample.
Why keep stressing representitive? because its damn hard to do with very coarse samples. I have had experience of grading Rip-Rap, and they don't make sieves for that size of material, so choose your method well and keep it repeatable.

 

[Robert168]

With reference to the ASTM specification on testing the extremely coarse soils, the test sample size needs to be about 100 times(with an accuracy of within 1%)the weight of the largest particle existing to be representitive.