Sieve analysis

(not verified)
Posted in: , on 19. Jun. 2007 - 16:36


We are QC testing a powder by sieve analysis and the material is failing to meet specification(pass through on 40, 70 and 50 mesh). It complied when tested by the manufacturer. The manufacturers method shakes the sample for 30 mins at 80000 cycles, whereas we have an older shaker which shakes at 50000 cycles, so we set the run time to 48 minutes to give an equivalent number of shakes.

Damn Statistics

Posted on 19. Jun. 2007 - 08:10

Hi mjohnson,

You are right, and the manufacturer is right.

There are statistics, and there are damned statistics. You are screening powder at

40 mesh = 420 micron

50 mesh = 297 um - Product 1 > 300 um

70 mesh = 210 um - Product 2 > 210 um

Product 3 - powder <200 um

That gives 3 products. What you have found is that powder does not behave like a set of billiard balls. The flow of product from one screen cloth to another is not related to the number of cycles, or the number of shakes, you give the stuff.

Powder at less than 200 um, particularly the very fine component, is very suseptable to air entrapment, and hence changes void ratio, or bulk density.

Now - what exatly are you measuring, and what do you want to know?

Life can be fun, if you don't take it that seriously,

Regards - Sgt John.Rz

(not verified)

Re: Sieve Analysis

Posted on 19. Jun. 2007 - 09:49

Hi John,

We are measuring the particle size distribution of an API. When tested by the manufacturer, it just about met the specification for % pass through of not less than (NLT) 95.0% for 40 mesh, NLT 90.0% for the 50 mesh and NLT 70.0% for the 70 mesh. When we tested the material it failed at 80% for the 50 mesh and about 50% for the 70 mesh. The material is known to be hygroscopic, so we suspect that it may have absorbed moisture which has resulted in the failure to meet specification.

Do you think that the increased shake time (i.e. 48 min's instead of 30 min's) is likely to have contributed to the failure?

thanks in advance,


Shake, Shake, Shake Yo Buudy

Posted on 20. Jun. 2007 - 02:27

Hi mjohnson,

Not likely. It is easy enough to test if the sieve time has any practical effect. Do a batch at 30, 48 and 60 minutes. Generally the shake duration is chosen to be long enough so that any increase in duration will only make a very marginal difference in the test results.

Get a Lab to do a microphotograph of your oversize, and mids grains - to identify if there is any moisture bonding or adhesion effects.

We have similar moisture degradation problems with TiO2. But in our case the moisture forms a "rust" film on the particle surface, reducing its ability to react chemically.

All the best - John.Rz

(not verified)

Re: Sieve Analysis

Posted on 18. Jul. 2007 - 10:16

Try a Microtrac particle size instrument, They are diffraction instruments, but they can display and print sieve data. The unit is very fast and easy to operate, typical analysis times are 10-30 seconds.

You can send a sample for a no-charge sample analysis.

Re: Sieve Analysis

Posted on 17. Mar. 2010 - 01:59

The advice given was spot on. I would also make sure that the sampling procedure of the API is representative of the entire sample.

Re: Particle Analysis

Posted on 24. Mar. 2010 - 09:52


WipWare particle analysis systems offer fast sieve results with pinpoint accuracy on virtually any material. Please visit for more information, and be sure to contact me about the systems and services WipWare sells.

Best regards,

Mark Wagner

WipWare Inc.

Tel: 1-705-472-2664

Fax: 1-705-472-2645


(not verified)

Re: Sieve Analysis

Posted on 26. Feb. 2013 - 06:54

A sieve analysis can be performed on any type of non-organic or organic granular materials including sands, crushed rock, clays, granite, feldspars, coal, soil, a wide range of manufactured powders, grain and seeds, down to a minimum size depending on the exact method.