Structural Arching

Posted in: , on 27. Jan. 2011 - 13:35

The blockage of a hopper outlet or other flow channel by lumps

By Lyn Bates

Ajax Equipment Limited, U.K.

This danger not as common a problem as cohesive arching but design guidance on how to avoid this hazard tends to suggest simple multiples of particle size for cones and slots, without regard to the many other factors that can influence this danger. It is not practical to lay down a specific formula or hard and fast rules; theoretically because the process involved is stochastic and there is no definitive safe dimension, but in practice this is because the phenomenon is affected by many interacting factors whose effects can be awkward to quantify. However, extended tests and experience allows a slightly more refined assessment of a reasonably ‘safe’ orifice size that takes account of the main influential variables that bear on the prospect of a stable arch structure forming within a flow channel. These figures should be taken as a guide only, with further investigation needed for applications that are particularly sensitive to flow stoppages and those where the size distribution is highly irregular.

Flow blockages become increasingly probable for cohesionless bulk solids when the ratio of particle diameter to discharge orifice diameter exceeds a critical value, the size of which depends on various physical and operational factors. Obviously, the probability is certain if a particle is larger than the orifice but even orifices that are several multiples of a particle size can be vulnerable to blockage. When the particles in the flow stream converge towards the discharge outlet, jamming occurs if the size of the particles is large enough to come together to sustain a stable, interlocking arch. Jamming phenomenon as structural blockages of this kind cannot be predicted from current theories, (Jenike and others), that apply for cohesive materials and can only be assessed in an approximate manner. However, a review of the factors and experience provide a useful basis for practical application.

Jamming is statistical in nature. The probability of jamming to occur at the outlet depends on the nature of packing as the particles move towards the discharge outlet. The jamming probability increases with particle size. Since the phenomenon is stochastic, there is no absolutely ‘safe’ value of orifice, although the probability falls off rapidly above a certain value of orifice size to particle size ratio. The problem is further complicated if the bulk material has a wide particle size distribution or multi-modal size distribution and the particles are non-spherical.

Designers of bulk solids equipment generally use an experience-based approach with crude rule of thumb guidance since experimental data tends to be scattered, and there are no usable theoretical models. Published figures do not discriminate between differing operating circumstances, which do influence the blockage prospects, so the following figures provide an improved starting point for design purposes. These suggested values do not apply to soft, sticky, elastic or plastically deformable particles, for which special investigations should be performed. Also, note that rounded or cube-shaped granules are less likely to adopt a stable contact relationship than those with significant flat surfaces or have an extended, elongated form.

The prospects of particles coming together in a converging flow channel to form a stable structure and block the flow depends on many factors. The three major factors are the span and shape of the flow channel and the ‘effective size’ of the particles in the media. . Flow channel span is key a design choice. Planar flow through a slot opening offers less prospect of a blockage than radial flow where the particles converge in two dimensions. ‘Effective particle size’, in this context, must take account of size distribution, shape, surface roughness and other features of the particle. Many bulk materials have a particle size distribution so an ‘effective’ dimension must be allocated to form the basis of an assessment of what may be considered a practical size of opening that has a very low probability of blocking. Account must be taken of any accumulative effect of segregation concentrating the proportion of larger fractions at some stage of the flow route.

Two other main features are the shape of the particles, as those with flat, rough or interlocking surfaces are more prone to resists slip against each other, and whether the flow boundary is smooth or rough enough to provide a stable foundation for an arch. A hopper designed for mass flow offers a wall surface on which the particles will slide so the boundary of the flow channel does not offer robust support to oppose flow and jamming is less probable than an equivalent size of opening which is surrounded by static product. A third feature is the velocity of flow as fast moving particles tend to have a more dilate condition due to the flow dynamics and greater inertia to overcome weak structures that may occur from time to time as the particles are jostled into ever changing conditions of packing in converging flow.

The following design features are suggested as ways to avoid the prospects of blockages occurring as a result of the formation of structural arches in hoppers

1Select a design for mass flow in the outlet region. If the hopper is not of mass flow design there is a static bed of product around the discharge orifice. Invariably, some particles will overlap the opening to reduce its effective size and flow has to pass through the rough boundary surface of a static bed of particles. By contrast, mass flow takes place with slip on a smooth boundary that offers very poor support for a stable, structural arch to form. The following link to a video demonstrates the reduced jamming tendency in mass flow configuration versus non-mass flow. http://www.ajax.co.uk/model1.htm

2It is well established that a slot outlet with plane flow is less prone to jamming than circular outlet with three-dimensional convergence. If the outlet is square or circular, the preceding shape can still converge in one direction only to secure the benefits of a plane flow channel.

3Avoid use of a butterfly valve if at all possible as the central blade reduces the flow cross section to two semicircles that are approximately equivalent to openings half the diameter of the valve.

4Minimize segregation, as the accumulation of larger fractions increases the prospect of jamming. Segregation most commonly occurs during the filling of a container at a single point. Dispersing the feed stream by way of spreading the feed over an area or diverting to a multipoint feed is effective in reducing segregation. Other tips are given in my book, 'User Guide to Segregation', published by The British Materials Handling Board

5Do not restrict the hopper discharge as a slow moving bed of material has less inertia to disturb an instantaneous blockage. The bulk is also in a denser state when flow is restricted because a flow stream is more dilated when moving at higher flow velocities, therefore, there is less room for slow moving particles to escape from a structural load path forming and arch as they touch in a dynamic state.

6In marginal cases, the provision of a vibrator will reduce the potential for particle jamming an opening by structural blockage and aid its failure, should one tend to occur. . However, sustained vibration on a blockage that does not break down quickly will tend to increase the local packing density and create a more stable array that is more difficult to destroy. Vibration should not be applied when the material is static The optimal location for application of vibration, specific energy input, amplitude and frequency will depend on the bulk material and system configuration but should essential be directed as close to the outlet as practical. The effect of vibrations on jamming probability is an open area for further research.

7Whilst research is handicapped by the numerous factors that influence particle jamming, it would be useful to fit individual research contributions into a framework that distinguishes Plane Flow from Radial Flow, Mass Flow from Non-Mass Flow, Dynamic Flow from Incipient Flow, rounded particles from angular particles & so on, and include polydispersity.

8Fitting one or more bars across the flow channel, at a distance and spacing above the outlet that offers a larger flow channel than the final outlet, will dilate the flow underneath and allow more latitude for particle re-arrangement in passage through the outlet.

9If there is a prospect of an occasional lump of uncertain size forming, such as an agglomerate, than the fitting of an inverted cone in a conical hopper, that has a slot opening similar to the outlet size, can also serve as a flow dilator and capture significant oversize particles without blocking the whole flow channel.

10For ultimate protection, a grid may be fitted with openings fractionally smaller than the outlet and access or facilities provided for removing or breaking down the lumps.

For the reasons outlined, the following multiple of maximum particle size are suggested as the minimum allowable orifice size for round or irregular shaped particles, based on the form of flow channel, whether mass flow or not and if dynamic or starting from a static condition. A separate assessment must be made in circumstances where the extent or distribution of particle sizes within the bulk is irregular, as with agglomerates or segregation effects.

Suggested minimum orifice dimension as a multiple of maximum Particle Size, form of Flow Channel and Particle shape



Radial non-mass flow pattern, (from static condition) Round 8 Irregular 10

Plane non-mass flow pattern, (from static condition) Round 6 Irregular 8

Radial mass flow pattern, (from static condition) Round 6 Irregular 8

Plane mass flow pattern, (from static condition) Round 5 Irregular 7

Radial non-mass flow pattern, (from dynamic condition)Round 7 Irregular 8

Plane non-mass flow pattern, (from dynamic condition) Round 5 Irregular 6

Radial mass flow pattern, (from dynamic condition) Round 6 Irregular 7

Plane mass flow pattern, (from dynamic condition) Round 4 Irregular 5

If it is important that flow stoppages do not occur, then further allowances should be made on these guidelines, or the flow behaviour verified in representative trials.

For more information, please visit:

https://edir.bulk-online.com/profile...-equipment.htm

http://www.ajax.co.uk/

Re: Structural Arching

Posted on 29. Jan. 2011 - 10:26

Thanks for this posting, It is a very good general overview to what is a vexatious issue at times

Colin Benjamin

Gulf Conveyor Systems P/L

www,conveyorsystemstechnology.com

Yup

Posted on 31. Jan. 2011 - 06:42

Thanks Lynn. Professional as ever! A true Boltonian. from an estranged pie eater.

Thanks Colin for VEXATIOUS. If you don't mind I'll borrow it a few times tomorrow and see how it performs.

Re: Structural Arching

Posted on 31. Mar. 2011 - 02:31

Thank you for a very interesting and informative post.

I was particularly interested in your comments regarding the inability of conventional theory such as the Jenike and Johanson approach to predict structural arching.

This opens up the interesting question of what constitutes the limits of the JJ approach.

In the case of powders, shear cell test results, both from one of the conventional types of shear cells, or your vertical shear cell version, provide some level of confidence and have a long history of successful application.

For coarse and very coarse bulk solids, one or two orders of magnitude larger in largest particle size than the typical shear cell test; we can perhaps have some doubts. A structural arching check against your data is one safeguard, but it is one that, if alluded to, is seldom if ever quantified.

In industry today, some extensive materials plant is being designed on the basis of shear cell tests on material that bears little resemblance to the material in the actual process. A sample of moist material with a lot of fines and a top size of a few millimetres is likely to indicate some cohesion when subjected to a shear test. The same material, at 10 times that size, with the fines removed by a screen, is, I believe, likely to exhibit a quite different behaviour, with little or no cohesion being present. And yet, it is not uncommon to see the same shear cell tests being applied to designs to handle that larger material. Perhaps the answer is that we are led to a conservative design by this approach. The downside of conservative design is increased cost.

Similar comments can be applied to wall friction tests.

So, I throw the question open to the forum. What are the practical limits of the powder shear cell test and wall friction tests when we are dealing with large particles and what alternative tests do we have that may provide better answers?

Peter Donecker Bulk Solids Modelling [url]www.bulksolidsmodelling.com.au[/url] [url]https://solidsflow.wordpress.com/[/url]
jag_minhas - Guttridge Ltd, U.K.
(not verified)

Structural Arching And Practical Limits Of Powder Shear Cell Te…

Posted on 4. May. 2011 - 10:34
Quote Originally Posted by doneckerView Post
Thank you for a very interesting and informative post.

So, I throw the question open to the forum. What are the practical limits of the powder shear cell test and wall friction tests when we are dealing with large particles and what alternative tests do we have that may provide better answers?

"Powders and Bulk Solids" by D. Schulze provides a good discussion on the various tests and testers that are used in bulk solids industry.

When we mention powders, firstly we need to establish what the particle size distribution is allowed for a granular substance to be classified as powder. I use a Brookfield Powder Flow Tester and what has been mentioned when using the tester is that 90% of the material needs to have a diameter less than 1mm and the maximum particle size allowed is 2mm (for a 230+33 cc trough and vane).

All different shear testers should have their specific size limits dependant on the material volume they can contain and the maximum stresses they can generate during the tests. The higher stresses should allow the engineers to interpret the forces involved when designing massive industrial hoppers.

Larger (capacity and stress generating) shear testers could be used to find out the arching dimensions and internal friction of materials with large particle sizes.

But the question would still remain, how much can you step up the values by? And are the current methods of calculating stressses from material head viable as we keep on increasing the hopper sizes.

Regards

Jagteshwar Minhas

Limits Of Powder Testing

Posted on 17. May. 2011 - 02:38

Peter raises an important point about testing materials with a wide range of particle sizes. Wall friction measurements, which are both simple to take and extremely usful, are probably not so sensitive to particle size, and so no significant difference should register if fines or large particles are tested of the same bulk material.

There are two main issues with shear measurements, whether the effective size has cohesive properties and the proportions of the differing fractions. If sufficient fines are present to fill the voids of larger fractions, or they can segregate in sufficient quantity to be a bulk in effective scale, then fines measurement will give the correct design value. Likewise, if the combination of relatively small fines and powder are cohesive, it is not unreasonable to test the powder only.

The problem arises when the coarser fraction constitue the bulk of the material, but a proportion of fines are present or can be created by attrition. Judgement is then required as to whether the material is likely to reflect the flow condition of the fines or if the effect of the fines can be neglected. Once the behaviour is outside the cohesive condition, shear testing is hardly relevant but it has to be remambered that cohesion is not a matter of particle size alone, so there is a consideration of how to test large fractions that may be cohesive.

Large annular shear cells have been made, up at least one meter diameter, but they are not generally available. Constructing a large shear box is not difficult, but the forces involved may be considerable and bring out an often overlooked feature of shear expansion, as the overlap of large hard particles in confined circumstances can give rise to exceptional stresses. This is a major factor in many feeder start-up loads on crystalline and lumpy materials which has nothing to do with cohesion, but emphasises the combination of mechanical interferance with surface sliding to produce an effective angle of 'internal friction'. I suppore that that the answer rests in the need for training in the technology and reference to specialists as an alternative to blindly following a proceedure that may give ultra-conservative and expensive design consequences. You cannot put old heads on young shoulders, but the vast scale of educational needs in this field is a major industrial problem.

Jagteshwar's points about classification of 'powders' is covered by various bodies, but the behaviour of particles in bulk does not fall into similar clear boundaries. His query about the effective head values for larger hoppers is normally addressed by using Janssen's formula of pressures in bins.

Re

Posted on 2. Aug. 2012 - 08:29

Thanks for your sharing on the structural arching. Now, I understand what it is.

If I have an requirement, will be visit the website.

johan121
(not verified)

Structural Arching

Posted on 25. Dec. 2012 - 07:59

An arch is a structure that spans a space and supports a load.

Re: Structural Arching

Posted on 26. Dec. 2012 - 10:10

Johan 121’s comment that ‘An arch is a structure that spans a space and supports a load’ is correct, but does not exclude the situation that the arch may be in a state of collapse. Gravity flow through an orifice mobilises a dynamic arch that indeed support some of the weight of the flowing material and transfers it to the walls, in the case of a mass flow regime, or to the surrounding static product in funnel flow systems, but is constantly reforming with different members. This applies with any size of particle, but the vast numbers of contact points across the flow channel involved in the flow of fine particle materials preclude the prospect of a stable structural arch spanning a flow channel. Cohesive forces do develop arches of a similar basic structure by virtue of the bulk strength generated.

The optimum shape of this natural arch is a catenary, but the network of contact points between particles forms rapidly changing load paths that are only stable if they fall within the range of contact friction. The probability of this occurring falls off rapidly with increasing numbers of contacts in the chain across the span of the flow channel. Static product boundaries also tend to have projections overlapping the flow orifice, effectively reducing the opening and providing a rough wall condition for slip. Inertial effects of high flow rates also reduce stability prospects, hence the reason the variations are proposed for a ‘safe’ opening size based on the differing conditions of flow regime, particle shape etc.

Interestingly, the height of the dynamic arch depends on the boundary conditions as determined by the flow channel shape and the frictional conditions on the boundary, so the distance through contact points of particles in a structural arch is larger than the horizontal distance across the opening and there are discrete steps, according to the angle of stable forces between particles, at which such arches can form. Modelling is making great strides in assessing this problem, but the complication of particle size and shape variation limits this approach to relatively regular systems*