Posted by Em Prof Peter Arnold, Univ of Wollongong, Australia on May 02, 1999 at 19:14:04:

I have been involved with the flow property testing of a wide range of (principally Australian) coals for 30 years. Our laboratory has both Jenike Direct Shear Testers plus a Johanson Hang-up Indicizer and a Johanson Flowrate Indicizer. Which procedure is best for 'coal' depends very much on the purpose for which the information is required and the condition required of the coal when tested.

Generally for coal the moisture condition where it displays maximum strength (worst handleability) is of interest along with its time consolidation characteristics and wall friction characteristics (a feature that is vital for mass-flow design and which does not seem to have been mentioned in the current 'debate'). The Johanson Hang-up Indicizer does not provide any information on wall friction - the Hopper Indicizer does. However, the concept of wall friction measurement is so simple (even if the results are not - see Haaker's paper in the latest issue of 'powder handling and processing' Vol 11, No 1, 1999 pp 19-25 and the paper by Prescott et al. in the same issue pp 27-36) that such complicated equipment is not necessary for reasonable measurements.

It is our experience that for the flow property testing of coals at elevated moisture contents and for use in the design of large coal storages (bins and bottom reclaim stockpiles) that the Johanson Hang-up Indicizer has some severe limitations. We use a top size for test samples of 4mm which can present problems in the Indicizer cell. At elevated moistures the Indicizer does not completely fail the test sample (push a complete plug in the one-dimensional failure mode) which casts doubts on the relevance of the Arching or Ratholing Index calculated by the machine's 'black box'. The capacity of the Indicizers is often exceeded which is limiting when large bins and stockpiles are required to be designed. We have also found that the machine is not particularly robust when used with bulk solids such as ROM coals. As indicated above it give no wall friction information. Generally with above ambient moisture coals wall friction is variable and depends not only on the wall lining material but also on the consolidation level. From a design point of view it is necessary to see how the hopper half angle for mass-flow can be varied with hopper outlet dimension or span to ensure that a practical geometry is attained. The Hang-up Indicizer only provides one piece of time storage data over the time storage interval of interest. If verification or more than one piece of time storage information is required then this can be a time consuming exercise as well as tying up expensive equipment. The moisture loss problem referred to by Brian Cox is also a consideration. The Hang-up Indicizer does have the potential to provide, fairly quickly, some 'relativities' between samples based on arching and/or ratholing capabilities. Our experience is that the repeatabilty obtained is not without some 'noise'.

The Jenike is not without its problems either. As Tom Troxel points out a skilled operator is required. I would put it more strongly, a very skilled operator is required! This is especially the case when determining the family of instantaneous yield loci. This is a often rather difficult task when dealing with coals at their maximum strength moisture content. Not only is skill required to undertake the testwork but care and some considerable skill is required to interpret the data collected. The main advantages of the method are: it's wide acceptance and associated well tried design procedures for storage bins, feeders, chutes etc; its provision of a wide range of data to allow both large and small storage facilities to be designed; its provision of considerable time storage information, collected 'in parallel' ie over the time storage interval of interest; provision of a simple procedure to obtain vital wall friction data. While procedures based on the Jenike data have been developed to provide relative 'handleability indices' to compare various bulk solid samples, the time taken to collect the relevant data is not trivial.