This application is a real bag of worms. Geoff rightly points out that the residence time distribution in a non-mass flow silo is indeterminate. Also, the extraction pattern from an air slide is certainly not progressively related to the square of the length from the discharge point, as it would have to be to take uniformly from the silo diameter were this to be the main extraction device. The discharge slides only cover a fraction of the silo base and are sequentially activated, so mass flow is virtually impossible. At best, the air slides may be considered to be discharge aids, and an assessment made on the basis of non-mass flow.

The inlet rate of 240 T/hr and outlet rate of 170 T/hr implies that the infeed ceases from time to time, so the level in the silo changes. In practice, this is expected to remove some material that has built up over the prior period of differential feed rates, as most of the fresh material passes straight through the system within a few hours. The level will build at about one third meter per hour during filling and drop by about 0.8 metre per hour when filling stops. The residence time for 70% of the incoming material will be about two hourse, through a growing static bed of approx 1500 Te. when the silo is virtually full. The residence time for the static product will depend on the duration of sucessive fill periods.

The length of the inlet slides will determine the fill points, which interact with the draw pattern to establish the change of surface profile. The cycle time of one hour for sequencing the extraction confines the main action to the top metre of so of the surface contents. Ten minutes between changes, and consecutive operation of adjacent air slides, hardly give time to develop deep repose contours that would collate product from multiple filling layers. The blending action of the silo may therefore be expected to be relatively inefficient with most of the benefit arising from the split fill, even though much of each discharge will draw from the sector filled.

To develop a study of the behaviour of the silo I would suggest that you prepare sectional sketches of the fill and emptying sequences of the cross sections based upon poured repose fill contours, preferential draw from the air slides with core flow and drained repose profiles. As an example, you may consider an article I published in ‘Powder Handling and Processing’ April/June 98, pp 135 – 138 on ‘Re-mixing of a segregated product during discharge from a mass flow silo’.

As an interim measure, I would suggest that improved blending would be secured by drawing the material off over 20 min periods in the circumferential air slide sequence of 1,3,5,2,4,6. This would develop deeper drained repose cones that cut across many different sequences of fill layers.

Lyn Bates ■

We are pleased to introduce ourself as one of leading engineering company having expertise in Fly ash silos, Cement silos and Raw meal blending systems.Please send us the details of your existing blending silo.We presume from your letter that you might have Inverted cone silo. Various factors such as residence time, consistant rate of extraction,Material discharge from different time zone etc effects the blending of raw meal.We also undertake modification of existing blending silo to achive better blending efficiency.

thanking you,

SANDEEP SANGHAVI

SANGHAVI ENGINEERING,

121, C.I.E, BALANAGAR EXPN,

GANDHINAGAR,

HYDERABAD-500037,

ANDHRA PRADESH,

INDIA.

sanghavi@hd2.dot.net.in ■