Keys for Successful Analysis – Representative Sampling & Estimation of Standard Error Calculation

Log in to watch this webinar

Not registered yet? Create an account
00:00:00 Keys for Successful Analysis -Sampling for particle size analysis Estimation of standard error
00:02:22 The presenters
00:02:30 Overview
00:04:13 Obligatory Opening Quotation (O2Q)
00:05:21 Why? Mining examples
00:07:50 Sampling: The impact on costs and decision making
00:08:39 Sampling in the 1890’s
00:10:12 LinkedIn - Mining – GEOSTATISTICS group
00:10:40 Published Abstract
00:11:15 (ALL) My University Notes on Sampling
00:11:48 University Notes (continued)
00:12:22 This is not the time to decide on Ms……
00:13:10 Adapted from: F J Flanagan “Reference Samples in Geology And Geochemistry” U.S. Geological Survey Bulletin # 1582, (1986)
00:14:24 Hofman (1901) – units?
00:15:06 Math will be based up that of Pierre Gy
00:16:15 Assumptions of the listener
00:17:07 What do I need to think about and define before I even attempt a particle size measurement?
00:20:27 What do we mean by representative sampling?
00:22:16 The 3 R’s
00:23:23 Francis Pitard
00:24:08 Cliché time
00:24:38 Fundamental Sampling Error (FSE)
00:25:27 Sampling
00:26:04 Sampling - segregation
00:26:36 T R Woodridge ‘Ore-Sampling Conditions in the West’
00:27:25 The taking of a representative sample
00:27:56 Esbensen – 2 SlideData
00:28:21 NASA lunar regolith
00:28:40 ‘How much sample do we need to take for any required degree of precision?’
00:29:00 Standard Error (Gy: Fundamental Sampling Error, FSE)
00:31:03 Fundamental Sampling Error (FSE)
00:32:01 Fundamental Sampling Error (FSE)
00:32:22 ‘I wish to detect a small amount of agglomeration in my system’
00:33:39 Minimum Mass
00:33:58 ‘I wish to detect a small amount of agglomeration in my system’
00:34:47 Minimum Mass
00:35:58 Calculation of minimum mass
00:36:28 Break out – back to the course – x95
00:36:44 [(1/aL) – 2]
00:36:54 Vezin (yes he!) 1866 (?)
00:37:09 So compare Rawle and Vezin
00:37:42 ‘Gy's Formula: Conclusion of a New Phase of Research’
00:38:05 Professor Robert Richards, MIT
00:38:39 We can turn the calculation ‘around’…
00:39:14 OK - up to 200 mm
00:40:02 The big stuff
00:40:36 So you’re taking only 20 mg……….
00:40:56 So how do you take a sample/specimen?
00:41:21 The x100 represents the single largest particle in the (sampled) distribution
00:41:50 The largest particle….x100
00:41:59 The largest particle….x 100
00:42:18 x100 – ISO 13320: 2009
00:42:31 Sampling - summary
00:42:56 Cliché time
00:43:09 Charles Babbage
00:43:30 References
00:43:49 Sampling personality webinars
00:44:08 Contact Information and Q&A
00:48:28 Thanks for listening

In any analytical technique you get out what you put in – that is, the instrument and technique measure what is given to them. 

In this webinar we’ll explore (from a theoretical aspect, the Theory of Sampling (TOS), as first formulated by Pierre Gy) the variation seen when different samples are extracted from the bulk lot. In many instances analytical techniques get superb results on tiny amounts of sample but this has repercussions from a representative sampling perspective (think fruit cake and just sampling the raisins).  

The variation caused by the inherent heterogeneity of the material represents the best variation that can be achieved and is easily calculable.  

Our examples will be based on particle size distribution analysis but the conclusions are transferable to all other metrologies including x-ray fluorescence where elemental content variation is repeated samplings is the norm.  The sampling formulae also have implications for count length in x-ray diffraction.

Speakers

Alan Rawle -C.Sci., B.Tech., Ph.D, C.Chem., FRSC
Applications Manager/CoChair E56.02 Characterization SubCommittee of ASTM E56 Committee on Nanotechnology

More information

Who will benefit: anyone taking measurements on an instrument and wants to understand why each sample taken from a bulk lot is different. This allows achievable specifications to be set based on the heterogeneity of the material.