OPEN-PIT MINING Evolution of sandwich belt high-angle conveyors

JOSEPH A. DOS SANTOS
Design Project Engineer
Dravo Corporation

EARL M. FRIZZELL
Mining Engineer
Spokane Research Center
U.S. Bureau of Mines

Abstract

Conventional belt conveyors offer an economical method for transporting materials at inclines approaching 18 degrees. Sandwich belt conveyors may be used to transport materials at high angles, to 90 degrees, while maintaining the most positive features of conventional conveyors, that is, (1) proven conventional components, (2) high speed operation, (3) high-single lifts, (4) continuous cleaning of the smooth surfaced belts, and (5) easy, quick repair by hot or cold vulcanizing. The present discussion deals with the evolution and development of sandwich belt conveyors to satisfy the operational requires of the open-pit mining and materials-handling industries.

A look at the limitations of conventional shallow and angle conveyors reveals the conveying angle at which a cover belt is needed. The relationship between conveying angle and hugging pressure by the cover belt is developed mathematically. A study of past and current developments reveals that the Loop Belt and Beltavator are significant advances in the state-of-the-art. The vertical radius of curvature constraints are discussed in detail and constraint equations are developed. Criteria for future development is established as a basis for evolutionary developments in sandwich high-angle conveyors. The use of intermediate drives in sandwich belt high-angle conveyors is investigated.

Introduction

In connection with U.S. Bureau of Mines funded-study relating to continuous haulage from deep open-pit mines, the authors had occasion to study extensively, the use of high-angle conveyors as a means of hauling mine products continuously from the pit while maintaining optimum mine slopes to minimize total excavation. The many possible high-angle conveying methods investigated included skip hoists, bucket ladders, pocket belts, fin belts, cleated belts, sandwich belts, pipe belts, screw conveyors, slurry pipe lines, and other. The sandwich belt high-angle conveyor appeared to be the most operationally-appropriate and economical solution for the mining industry. The present paper presents the theory, historical development, state of-the-art, and evolutionary development in sandwich belt high-angle conveyors.

History and Theory of Sandwich Belt Conveyor
Limitations of Conventional Conveyors

Conventional belt conveyors offer an economical method for transporting bulk materials at recommended inclination angles ranging from a low of 7 degrees for soda ash briquettes, to a high of 30 degrees for cinder concrete and ground phosphate fertilizer().

Typical recommended inclination angles for open-pit mine products such as excavated earth and blasted, primary crushed rock vary from 15 to 22 degrees with respective angles of repose from 29 to 44 degrees⁽¹⁾.

The conventional conveyor is often the most economical, reliable, and safe means if transporting bulk material⁽¹⁾. There are however, many cases which strongly warrant an increase in the conveying angle. The West Germans used high-angle sandwich belt conveyors on some bucket wheel excavators to increase the depth of cut without increasing the boom length(2-4). A substantial cost savings resulted from reduced structural and mechanical components associated with the shorter boom length required to accommodate a high-angle boom conveyor. In open-pit mines with steep mine face angles, or in any case where the surface incline angle in the path of the conveyor substantially exceeds that of the recommended conventional conveyor angle, much excavation or elevated support structure is needed in order to accommodate the conventional conveyor.

In order to logically develop the theory for high-angle sandwich belt conveyors, it is important to understand the nature of the conveying angle limitation for conventional conveyors.

In a static case, a cohesionless material on a rubber belt will begin to slide back when the incline angle of the belt surface just exceeds the angle of internal friction of the material or the friction angle at the material to belt surface interface, whichever is smaller. The angle if internal friction is equal to the angle of response for such materials. Both the angle of repose and the friction angle for bulk materials on rubber will vary from one material to the next and will be affected, even for the same type of material, by the maximum lump size, the lump-size distribution and orientation on the conveyor cross section, and the shape that the particles or lumps take as a result of the reduction process, that is, blasting and the varying degrees and methods or crushing. Although there is much published information on angles of repose for various bulk materials⁽¹⁾, very little is available in the friction angles for various bulk materials on a rubber surface. H. Colijn⁽⁵⁾ lists coefficients of friction of material on rubber for six different fine-grain materials ( = tan , where is the coefficient of friction, and , is the friction angle). They vary from 0.45 for Ottawa sand 2 pct moisture to 0.624 for bituminous coal fines with 15 pct moisture. The respective sliding friction angles are 24.2 and 32.0 degrees. The corresponding angles of internal friction are not listed. An investigation by the U.S. Bureau of Mines⁽⁶⁾ revealed much higher friction angles when coarse material was literally displaced over rubber belting material in a Large Direct Shear Tester. In one set of tests, river run gravel of size 0.25 by 0.185, 0.375 by 0.25 and 0.5 by 0.375 in. resulted in friction angles at the material to belt interface of 387.5, 42,0, and 38.0 degrees respectively. The corresponding angles of internal friction are 43.5, 48.4, and 52 degrees. A second set of identical test of 1 by 0 in. Upper Freeport coal, 1 by 0in. Middle Kittanning coal, by in. dolomite, and by in. limestone yielded respective friction angles, between the material and rubber belt, of 34.8 to 38.1, 40.6 and 33.7 degrees. The correspondence internal friction angles are not listed.

The recommended conveying angles are, in general, far below either of the friction angles mentioned. This is due to the dynamics induced in a moving belt conveyor, which result in relative motion between adjacent particles or lumps of the bulk material and between the material and carrying surface of the conveyor belt. The three major sources of dynamic effects are:

1. The agitation of the material on the conveyor belt as it approaches and is carried over each successive carrying idler.
   This effect is amplified when belt tension is low and idler spacing is high, thus resulting in large belt sag between idlers.

2. The acceleration of the material at loading or transfer points. This results in relative slip and turbulence because velocity and direction cannot change instantaneously.

3. The vertical impact at transfer points. Such impact is absorbed by the resilience of the belt and impact idlers and results in bouncing of the material. This adds to the turbulence at the loading area.

These effects are increased when the belt speed is high and even more so when the material handed is loose and contains large rounded lumps.

These effects are increased when the belt speed is high and even more so when the material handled is loose and contains large rounded lumps.

In 1934, A. Vierling⁽⁷⁾ performed tests on a belt conveyor of V-shaped or wedged-shaped cross section and showed that he could increase the conveying angle simply by increasing the troughing angle. The conveyed materials were brown coal, household briquettes, briquette fragments, and overburden. It was theorized that it did not matter whether the material to be transported is carried in a single wedge-shaped trough, using two-roll idlers, or is carried on a belt with parallel wedge-shaped grooves, as long as the size of these grooves is in a prescribed ratio to the size of the pieces of transported material.

His tests showed that the conveying angle could be substantially increased only when the troughing angle exceeds 45 degrees. Tests conducted with grooved belts conveying over-burden with a natural angle of repose of 40 degrees showed that a conveying angle of 38 to 40 degrees could be achieved. It is not known at what belt speed, belt tension, idler spacing, and material size distribution these test were performed. It was announced however that the grooved belt could convey material with mixed grain-size composition at conveying angles 4 to 5 degrees less than the natural angle of repose.

Extrapolation of these results would lead to the conclusion that a grooved belt is effective in suppressing the dynamic effects due to belt travel over the idlers and in possibly creating an added normal pressure, due to wedging action, at the material to belt surface interface. The conveying angle cannot, however, exceed the angle of internal friction of the material at the free surface.

Sandwich Belt Conveyors

Conveying angles greater than the angle of internal friction can be achieved by a cover which, when pressed against the material, will create a hugging action to prevent sliding of the control surfaces.

For a cohesionless material one can idealise the situation as shown in Figure 1. The material is idealised as parallel layers spaced infinitely close.

For the case where the cover surface is free to follow the material as it slides back, sliding will occur when the tangential component of the material weight exceeds the frictional forces which resist it or:

Wm sin α > (N + Wm cos α)μ ....................... (1)

where:

= _m or _b whichever is smaller

Wm, α, N _m, _b are defined in Figure 1.

To achieve an inclination angle α, a normal lineal hugging load, N, must be exerted by the top surface such that:

N ≥ Wm (Sin α - cos α) ....................... (2)
               

If the cover surface is fixed and it resists the motion of the material at the interface, then material will begin to slide back when:

Wm sin α>N ( + ) + (Wm cos α) ....................... (3)

where:

= _m or _T whichever is smaller.

_T is as defined in Figure 1.

To achieve an inclination angle α;

N ≥ Wm         (sin α - cos α) ....................... (4)
             +    

If = , substitution into the above equation shows that the normal hugging load, N, needed to prevent backsliding, is only half of that required in the previous case (see Equation 2).

When the above principles are related to sandwich belt conveyors, the bottom surface represents in carrying belt, while the top surface represents any cover surface such as a pressed belt, a weighted belt, a link chain matt, etc. For a covering surface which depends on its self-weight, the normal hugging load N becomes the normal component of the lineal weight of the cover surface. The second set of equations, 3 and 4, clearly show that the required hugging load is much less if both surfaces are driven at the same speed.

In his 1958 review of patented high-angle conveying methods⁽⁴⁾, P. Rasper breaks down the sandwich belt approach into three different categories:

1. Cover belt acting by its own weight
   

2. Cover belt with additional pressure
   

3. Cover belt with carriers.

Figure 2 illustrates a category (a) solution which was patented in West Germany in 1953. It consists of a carrying belt conveyor on three-roll, 30-degree troughing idlers with special spring-suspended impact idlers at the loading area, a cover surface which consists of a link chain matt, and a cleated belt to drive the cover surface. Special grooved return pulleys are used to carry the link cover matt.

During operation the material loads onto the bottom horizontal portion of the carrying belt, and is carried into the sandwich where the chain matt cover hugs the material virtue of its self-weight, and prevents it from rolling or sliding back. The material is then elevated in the sandwich and is discharged at the top where the two surfaces separate.

The chain matt could conform easily to the irregularities in the material profile and this was thought to be an attractive feature. It was noted that a cover surface which depends on its self-weight can be uneconomical at higher conveying angles because the required lineal weight of the covering surface increased drastically. Some elevating conveyors were built according to these ideas.

Figure 3 illustrates a category (c) solution which was also patented in West Germany in 1953⁽⁴⁾. It consists of a carrying belt conveyor on three-roll troughing idlers and a cover belt with flexible carriers of flexible chain or rope secured to the belt at prescribed intervals along the length. It requires a loading belt in order to place the material into the sandwich.

During operation, the material is placed into the sandwich by the feed conveyor and it is immediately covered by the carrier. The cover belt and carriers hug the material, by virtue of their self-weight, while it is elevated to the discharge point at the top. The main selling feature of this method was claim that the cover surface was self-cleaning because of the flexibility of the carriers. No elevation conveyors were ever built according to this design.

In his book The Bucket Wheel Excavator⁽²⁾, L. Rasper discussed the elevating conveyor illustrated in Figure 4. It is a category (b) solution which was patented in West Germany in 1954. It consists of a carrying belt conveyor on three-roll troughing idlers and a cover belt which is pressed onto the material by rubber tires distributed across the belt and spaced at large intervals along the conveyor length. A feeder belt is used to load the conveyor.

During operation, the material is fed onto a low-angle, uncovered portion of the carrying belt at the bottom of the elevating conveyor, and is carried into the belt sandwich. The cover belt, which is pressed by the rubber tires, hugs the material as it is elevated to the discharge point at the top. According to L. Rasper⁽²⁾, this solution was used in several occasions on bucket wheel excavators of the 1950s.

As a result of his survey, P. Rasper⁽⁴⁾ lists three important operational requirements of high-angle conveyors. The first pertains only to their use in bucket wheel excavators, while the second and third are important in any application.

1. The high-angle conveyor must provide for easy removal or lifting of the cover belt when the conveyor is to be operated at shallow angles.
   

2. The system must lend itself to easy and quick repair of the belts.
   

3. Any high-angle conveyor must lend itself to easy cleaning. Requirement 1 illustrates a recognition of additional wear on the belt and other components when operating with the extra pressure from the cover surface. On a bucket wheel excavator boom, such a conveyor will operate at high-angles only when it makes the extreme high and low cuts.

During operation, a bucket wheel excavator can pick up large boulders of different sizes and shapes, and tramp iron such as beams, plates, etc., from previous underground mines. These can cause tears in the belt. Requirement 2 recognises that it is impossible to guard against this so the effort must be to minimise the consequential downtime. For a regulated and more predictable material and size distribution, that is, the discharged material after primary crushing, good design can reduce the frequency of damage. The shape of the lumps cannot, however, be controlled. P. Rasper⁽⁴⁾ suggests that the repair can be made by hot or cold vulcanising and, on this basis, rules out the use of fins or cleats.

Requirements 3 is based on a realisation that belts get dirty and must be continuously cleared if they are to maintain their alignment. Build-up of material on the belt and other surfaces also leads to additional drag and premature wear of all support and drive components. He states that the most popular way to clean belts on bucket wheel excavators is by screw-type return idlers, and this requires the use of smooth surface belts. This requirement still holds true for the many belt cleaning methods available today.

On this basis he concludes that only solution of categories (a) and (b) lend themselves to the operation requirements and thus warrant further study and development.

L. Rasper also points out that most of the development and use of high-angle conveyors on bucket wheel excavators occurred in the 1950s and that the bucket wheel excavators commissioned in the 1960s had abandoned them in favour of higher tonnage rates and reduced belt wear of conventional conveyor⁽²⁾. This reduced the cutting depth of the machine, but operational experience showed that deep-cut cohesive soils and excavated lignite permitted conveying angles of up to 23 and 27 degrees, respectively, with conventional conveyors.

More recent developments in the category (a) solutions include the Retainer Belt by Stephens-Adamson⁽⁸⁾ and the Overlay Conveyor by R.A. Hansen Company⁽¹¹⁾. Both claim the capability to convey material at angles up to 45 degrees.

The Retainer Belt, as illustrated in Figure 5, consists of a conventional carrying conveyor on three-roll troughing idlers and a heavy cover belt with smooth rubber surface. The cover belt is shot with lead for the extra weight needed for the normal force component.

During operation, the material loads onto the uncovered conventional conveyor, at a nearly horizontal conveying angle, and enters the belt sandwich prior to a gradual increase in conveying angle. The sandwiched material is conveyed at the high angle to the discharge where the belts separate.

The Retainer Belt conforms well to the second and third operational requirements listed by P. Rasper⁽⁴⁾. Other than the special cover belt, it uses only conventional conveyor components. The higher cost of the special cover belt, however makes the Retainer Belt expensive at high angles. There is not apparent capacity limitation other than those which govern conventional conveyors.

The Overlay Conveyor is illustrated in Figure 6. It also conforms well to the second and third operational requirements set by P. Rasper, but it does not address the need for hugging pressure to keep the material from sliding back. It uses an ordinary (non-weighted) cover belt which is driven along with the carrying belt by mechanical coupling. The carrying belt is supported on three-roll troughing idlers. A specially suspended single-roll idler presses the cover belt at the entrance to the belt sandwich. A speed-up belt is used to feed the high-angle conveyor.

During operation, the speedup belt throws the material into the loading area and its momentum carries it into the sandwich entrance. From that point it is carried in the belt sandwich until it is discharged at the top where the belts separate.

It is argued⁽¹¹⁾ that the ordinary non-weighted cover belt is sufficient to keep the large lump from rolling back and the rest of the material will not slide back. It is also argued that the speedup belt will throw the material into the loading are at a prescribed speed and angle of incidence so that the momentum of the material will carry it into the belt sandwich with a minimum amount of turbulence. This justifies an uncovered conveyor at the high angle.

Two test units were built to verify the theory. The first used a 20-inch cleated carrying belt with an 18-inch smooth cover belt and a conveyed wet sand with 10 pct rocks at 45 degrees while the second used two 60 inches smooth belts and conveyed wet sand with 10 pct rocks at a conveying angle of 40 degrees. These test units were successful in conveying wet sand at high conveying rates and wet sand with 10 pct rocks at reduced rates. When the rocks exceeded 10 pct, they would roll and bounce in the loading area without entering the sandwich. The ability to convey a limited type of material, without additional weight on the cover belt, illustrates the advantage of driving both belts and indicates that the effective friction angles within the sandwich may be quite high.

In light of the state-of-the-art and current developments in sandwich belt-type high-angle conveyors and the requirements of the present development, two new categories (d) and (e) are added to the possible solution types as listed by P. Rasper⁽⁴⁾.

The current solution categories are then as follows:

1. Cover belt acting by its own weight
   

2. Cover belt with additional pressure

3. Cover belt with carriers
   

4. Belt sandwich with prying resistance by virtue of the transverse stiffness of the two belts.
   

5. Belt sandwich with radial pressure by virtue of the belt tension and the conveyor profile geometry.

The Loop Belt by Stephen-Adamson(9) represents a category (e) solution. It is illustrated in Figure 7.

With this concept, a non-weight inner belt loop is pressed against closely-spaced troughing idlers by an outer belt loop. The outer belt is straight at the loading region and supported on troughing impact idlers. When the conveyed enters the belt sandwich, the carrying strand (outer belt loop) is no longer supported by idlers. As the belt sandwich enters the vertical curve, the belt tension and the radius of curvature are such that sufficient radial pressure is produced to overcome the weight of the belt and the conveyed material. The material is thus conveyed in a sandwich made up of the suspended carrying belt and the idler-supported inner belt loop. This sandwich loops around, approximately 155 degrees, to the discharge point. The lineal hugging load produced by the outer belt is determined by equation 5. As the conveying angle increases, the vertical radius of curvature and the outer loop-belt tension must provide sufficient radial pressure to support and hug the suspended material in order to conveyor it up to a 90-degree angle. Above the 90-degree angle, the material begins to sin on the supported inner belt loop; and the outer belt loop becomes the over belt which supplies the hugging force. The material is conveyed in this manner to the discharge point. Typically, the outer belt loop is driven while the inner belt loop is forced to follow by the frictional drag of the conveyed material. There has been occasion to drive both belts by a load-sharing drive arrangements which utilises the combined tension rating of both belts in order to maximise the elevating height.

Pr =  T/R  ....................... (5)

where:

T = Outer belt tension at the point in question on the conveyor profile.

R = Radius of curvature corresponding to the belt tension. T.

Pr = The corresponding lineal load applied by the outer belt.

The Loop Belt has had great success in self-uploading ship applications and several outdoor installations. It has achieved 10,000 tones per hour, in conveying iron ore pellets and coal, and lifts up to 150 feet. Typical belt speeds are 800 to 1,200 feet per minute. A special low-modus fabric belt is used to achieve tight vertical curves with a troughed belt.

The Beltavator, by Stephen-Adamson⁽¹⁰⁾, is a hybrid which represents a solution of categories (d) and (e). It is illustrated in Figure 8. The Beltavator is an extension of the Loop Belt. It is identical to the Loop Belt up to a conveying angle of 90 degrees. At this point, it conveys the material vertically for a desired elevating height and it again becomes identical to the Loop Belt above the 90-degree conveying angle. It is the straight vertical portion of the conveyor profile that qualifies as a category (d) solution. On the straight vertical portion, the belt sandwich is held together by closely-spaced, staggered-edge roll which press the belt edges to keep the sandwiched material sealed between. The hugging pressure required to convey the material at 9-degrees is provided by the prying resistance of the two belts as the material introduced into the sandwich. The prying resistance is by virtue of the transverse stiffness of the two belts. The required transverse stiffness imposes a capacity limitations in the Beltavator. As the belt width requirements become large, in order to achieve higher tonnage rates, increasingly thicker multi-ply fabric belts are required to provide adequate transverse stiffness. Such considerations limit the practical belt width to a maximum of approximately 36 inches, and maximum capacity is approximately 1 000 tons per hour when conveying dense material. Only the outer or receiving belt is typically driven as was the case in the Loop Belt. The Beltavator, too, has had great success within its capacity range. It is conveying profile can follow a C Belt Path a Z Belt Path, or an L Belt Path, as shown in Figure 8.

It is important to point out that solutions which fall into categories (d) and (e) represent a significant advance in the historical development of the sandwich-type high-angle conveyors. All of categories (a) through (e) address the need for hugging pressure which increased the frictional resistance to relative movement between adjacent particles of the conveyed material and

between material and belt surfaces. The significance is that solution categories (d) and (e) do not require added weight or a pressing mechanism.

In the category (d) solution, the hugging pressure results from prying resistance and the transverse stiffness of the two belts. This transverse stiffness is utilised to an advantage, but it exists whether or not it is utilised.

In the category (e) solutions, the hugging pressure results from radial force by virtue of the tangential belt tension and the curved geometry of the conveyor profile. Here the belt tension is used to an advantage by the selection of a particular conveyor profile geometry. The belt tension at any point on the profile is related to the tension at the tail pulley, the lineal material load, and the height above the tail pulley. As with the belt transverse stiffness, it exists whether or not it is used to an advantage by the selection of the appropriate conveyor profile geometry.

Additional implications in category (e) solutions are discussed when addressing the vertical radius of curvature requirements are specified by CEMA⁽¹⁾.

It may seem surprising at first that Stephens-Adamson typically chooses to drive only the carrying belt in the Retainer Belt, the Loop Belt, and the Beltavator when it is clear from equation 2 and 4 that the required hugging pressure is greatly reduced by exerting drag at both the carrying and cover surfaces. A closed and more realistic look at the belt sandwich model will reveal that both surfaces do indeed exert drag on the material, even though only one belt is driven.

The Belt Sandwich model, illustrated in Figure 1, is very instructive but it is not totally accurate. It assumes that the cover surface contacts only the material, but the edges do not touch the carrying belt. It is known that lateral movement of the cover surface during operation will cause the edges to bear, intermittently, on the carrying belt; thus losing a portion of the hugging load directly to the carrying belt and support idlers, while uncovering the material at the other edge. More realistically, a minimum distance from the belt edge to the material is required in order to assure that the material is always covered and does not spill out. On this basis, Stephen-Adamson chooses to derate the capacity of the sandwich type conveyors (as compared to CEMA and capacity recommendations for conventional conveyors⁽¹⁾). This assures large edge distances and, thus, a sealed envelope.

A new, more realistic model (Figure 9), illustrates the interplay of forces. The minimum. The minimum normal hugging load, Nm; which must be exerted on the material in order to prevent blacksliding, if both surfaces resist motion, is expressed by equation 6.

(min) Nm = Wm (            ( sin α  - cos α) ....................... (6)
                            +       

This follows from equation 4. The drag which must be exerted on the material by the top surface must be counteracted by the friction drag between the top and bottom surfaces at the edges, as expressed by equation 7.

(min) Ne e = (min) Nm ....................... (7)

The minimum required total normal load, N, can be expressed by combining equations 6 and 7 to obtain equations 8 and 9.

(min N = (min) Ne + (min) Nm = (    + 1) (min) Nm ....................... (8)
                                                   _е

 (Min) N = (  + 1) Wm           ( sin α - cos α) ....................... (9)
                _е                 +     

If both carrying and cover surfaces are of rubber conveyor belting, then = . If _e = = , equation 9 reduces to equation 2. It will be shown later that this is, in fact, a reasonable assumption, Equation 2 assumes that only one surface is driven. At first impression it would seem that nothing has been gained by using the more realistic model, because the required total normal load, N, is the same when only one belt is driven regardless of which model is used. What is important is the recognition of a more realistic behaviour and that only the portion of the normal load which bears directly on the conveyed material actually contributes to an increase in the impact on each successive idler as the material is carried along with the belt, and thus, only this portion contributes to any increase on belt wear. The required edge distance corresponding to the assumption that _e = = , is b/4, where be is the belt width.

What then is the advantage of driving both surfaces? If the type of material to be handled and the environment constraints do not require a large edge distance, then the total required additional normal load, N, can be reduced. If there are circumstances which would cause doubt that the required counteracting edge drag could be developed, that is _e , then it would be advantageous to drive both surfaces. In order to maximise total elevating height, it is advantageous to drive both surfaces by a load-sharing drive arrangement so that the combined tensile strength of both belts is utilised.

In light of the more current solutions in sandwich belt conveyors a new set of operational requirements is established. This set retains requirements 2 and 3 as originally listed by P. Rasper⁽⁴⁾ (now numbered 4 and 5), but omits his requirement 1 because it is applicable only to high-angle conveyors when used on bucket wheel excavators.

The combined operation requirements are:

1. The receiving end of the steep-angled conveyor must insure that the turbulent material at the load point is settled quickly and experience no backsliding prior to entering the sandwich.
   

2. The hugging pressure exerted on the conveyed material and, thus, the additional pressure on the cover surface must be applied by a soft loading system which minimises load concentrations.
   

3. The cover surface must be a floating which will not obstruct the flow of lumps larger than anticipated or oriented unfavourably.
   

4. The system must lend itself to easy and quick repair of the belts.
   

5. Any high-angle conveying system must lend itself to easy cleaning.

Requirements 4 and 5, which have already been discussed, preclude the use of carriers, and makes it impossible to cover the material immediately as it is fed into the high-angled conveyor. The Overlay Conveyor by R.A. Hansen Company⁽¹¹⁾ was loaded at a 40-degree angle by using a speedup to throw the material into the loading area and allowing the momentum to carry it into the sandwich. This was apparently successful in conveying wet sand, but many difficulties arose when attempting to convey wet sand with 10 pct rocks. It was found that the orientation of the speedup belt, with respect to the loading area, was very critical and the optimum orientation varied with variations in the mix of the bulk material. Rocks alone could not be transferred in this manner because they would bounce and roll back But even more basically, such a system, which depends on the momentum and angle of incidence of the fed material, could not tolerate an emergency stop of the conveyor. It would be impossible to re-accelerate any material remaining at the loading area after the shutdown. Requirement 1 recognised the problems associated with loading a sandwich belt conveyor at a high angle. On this basis, it is recommended that the conveying angle onto which material is loaded should not exceed 10 degrees, and the material should enter the sandwich at a maximum conveying angle of 18 degrees.

Compliance with requirement 2 will minimise the additional wear on the belts and other components due to the additional hugging load. This is very important because L. Rasper⁽²⁾ cites the reduced belt wear of conventional conveyors as one reason why the bucket wheel excavators of the sixties has abandoned the use of high-angle boom conveyors.

The design must convey material which is discharged from a primary crusher. The crusher setting only assures that in dimension will not exceed the prescribed value. If the feed into the crusher consists of slabby material, then a minus 8 inch setting could result in slabs of dimensions of up to 8 by 12 by 24 inches. The loading area should be designed to assure that slabs are oriented favourably. Requirement 3 assures that oversized or unfavourable orientated material will not encounter pinch points and will result only in local lifting of the cover and possible spillage of the material. It should be emphasised that oversized lumps and slabs of the dimensions mentioned are not to be considered part of the normal operation. If slabby material is to be crushed, it may become necessary to reduce the crusher setting.

Constraints On Vertical Radius of Curvature

Operational requirement 1 carries with it severe implications. After the material is loaded, the conveying angle must be increased from 10 degrees to a much higher angle. In open-pit-mine application, the ultimate conveying angle is dictated by the slope of the mine face. Such an angle increase cannot be instantaneous with a troughed belt. The angle change must be sufficiently gradual so that no part of the troughed belt is subject to buckling or overstress. Within these constraints we must strive to increase the conveying angle within the shortest possible distance. Figures 10, 11 and 12 illustrate why this is so.

If α represents the highest mine face angle which is attainable because of slope stability or other considerations, then a multi-run high-angle conveyor system may be incorporated in to the mine, as illustrated in Figures 10 and 11.

Figure 10 illustrates the adaptation of the conveyor system without additional excavation. The consequences of a long transition radius of curvature are two-fold. First, the ends of each module divert farther from the mine face, thus increasing the structural support requirements. Secondly, the ultimate conveying and α, corresponding to the module with long transition radius, is greater than α, which corresponds to a module of short transition radius.

Thus, a higher ultimate hugging force is required to prevent the conveyed material from backsliding.

If additional local excavation is favoured in order to reduce the structural support requirements and to keep the conveyor profile within easy access from the mine benches, then the conveyor system is adapted to the mine as shown in Figure 11. A long transition radius of curvature results in increased excavation requirements.

If a single run conveyor is used to carry the material out of the mine, as in Figure 12, then a long transition radius of curvature requires that the conveyor loading area extends further into the mine. In many cases it is more efficient to locate the loading area several bench depths above pit bottom. In such a case, the imposition is the conveyor loading area makes it difficult to recover the material below. Depending on the slope stability requirements, the amount of irrecoverable material can be substantial.

The equation to determine the allowable vertical radius of curvature are listed by CEMA⁽¹⁾ and derived here. The final form of the equations as derived are equivalent to the CEMA equations but are presented in a more instructive form.

The vertical belt curve must be designed so that it will not cause buckling nor over stress at any part of the belt cross section. CEMA, in fact, requires that the minimum allowed tensions anywhere on the belt is 30 pounds per inch of width.

When any section of linearly elastic material is subject to simultaneous tension, Tc, and bending, M, which is induced by the curvature 1/r as shown in Figure 13, the stresses due to the independent forces may be superimposed. The model assumes that the curve is smooth and that plane sections remain plane. When applied to a conveyor belt section, this model predicts the stresses accurately at points sufficiently far from the ends of the curve. If the trough depth and width are much less than the length of the transition curve and the support idlers are very closely spaced when compared to the radius of curvature. Having made these assumptions, the bending moment can be related to the radius of curvature by the following equation:

M =  EI/r  ....................... (10)

where:

E = modules of elasticity (lbs / in.²)

I = sectional moment of inertia (in.⁴)

and M and r are as defined in Figure 13.

Superposition of stresses may then be expressed as follows:

f = fa + fm =  Tc  +   EI  ....................... (11)
                      A        rS

where:

F, fa, fm, Tc, A, r, S are as defined in Figure 13.

Recognising that I/S = y and substituting into equation 11 yields:

f = fa + fm =  Tc  +  E   y ....................... (12)
                      A       r

Where y is the distance from the neutral axis to the extreme outer fibers (see Fig. 13).

This equation can be applied to a multi-ply conveyor belt of arbitrary geometry by making the following substitutions:

• Replace the cross-sectional area A (in.²) in equation 12 by the product of the belt width and the number of piles, (b) by (p) (in.-ply).
  

• Replace the elastic modulus E (lb / in.2) by the commercially listed belt modulus Bm (lbs / in.-ply).

Equation 13 then follows from equation 12.

f = fa + fm =  Tc  +  Bm y (lbs / in.-ply) ....................... (13)
                      bp        r

The units of stress f, fa, fm are now expressed in lbs / in.-ply.

To prevent overstress of the bottom fibers, the combined stresses must not exceed the working tension rating, Tr, of the belt. This means that the following equation must be satisfied.

  Tr   >  Tc   + Bm y ....................... (14)
 B p      b p        r

where:

Tr = Working tension rating of the entire belt (lbs)

y = distance from the neutral axis N.A. to the extreme bottom fibers (in.)

Solving for r yields the following:

r ≥ b Bm p y ....................... (15)
       Tr - Tc

To maintain a minimum tension of 30 lbs / in. on the top fibers, the following equation must be satisfied:

 30  ≤  Tc   +  Bm y ....................... (16)
 p       b p        r

where:

y = distance from the neutral axis N.A. to the extreme top fibers (in.)

Solving for r yields:

r ≥   b Bm p y ....................... (17)
       Tc - 30 b

Next, consider a belt which is carried on three-equal-roll idlers of troughing angle . It is assumed that the belt is divided into three equal parts, as illustrated in Figure 14.

If the vertical curve under consideration is concave (that is, belt edges are at inner side of curve). Then y = (b / 9) sin and y = (2b / 9) sin . Equations 18 and 19 follow when these identities are substituted into equations 15 and 17, respectively, and the equations are divided by 12 so that R is the radius of curvature in feet rather than inches.

R ≥  b²   Bm p sin (to prevent overstress of middle when curve is concave) ....................... (18)
     108     (Tr - Tc) 

R ≥  b²   Bm p sin (to prevent edge buckling when curve is concave) ....................... (19)
      54     (Tc - 30b)

If a convex vertical curve is considered (that is, belt edges are at the outer side of the curve), then y = (2b / 9) sin , and y = (b / 9) sin . Equations 20 and 21 follow when these identities are substituted into equations 15 and 17, respectively, and the equations are divided by 12 so that R is in feet rather than inches.

R ≥  b²   Bm p sin (to prevent edge overstress when curve is convex) ....................... (20)
      54     (Tr - Tc)

R ≥  b²   Bm p sin (to prevent buckling of center when curve is convex) ....................... (21)
     108   (Tc - 30b)

Equations 18,19, 20 and 21 are equivalent to the CEMA equations(1) and apply to multi-ply fabric belts. For steel cord belts, equations 18, 20, and 21 are applicable if p = 1 and the elastic modulus Bm is in (lbs / in.). When the curve is concave, the belt edges of a steel cord belt are allowed to buckle to the extent that it is not detrimental to the operation. To reflect this, the right-hand side of equation 19 is divided by 2.5, thus yielding equation 22.

R ≥  b²      Bm p sin    (allow edge buckling in steel cord belt of concave curve) ....................... (22)
      108   (Tc - 30b) 2.5

where

p = 1 and Bm is in (lbs / in.)

Equations 18 through 22 reveal the governing parameters as they relate to the allowable radius of curvature. The belt width, b, and the troughing angle, , must be established ahead of time so that they are compatible with the capacity requirements and maximum lump size of the material to be handled. For any solution of category (e), it is established that the trough depth shall not be less than the maximum lump size (that is, for belt on three-equal-roll idlers, then (b/3) sin > maximum lump size). With belt width and troughing angles established, the remaining variables are the composite elastic modulus of the belt, Bm p (lb/in.), the rated working tension, Tr (lbs), and the operating belt tension at the point under consideration, Tc (lbs). For a concave, curve, equations 18 and 19, applicable to multi-ply fabric belts, or 18 and 22, applicable to steel cord belts, must be satisfied simultaneously. Tc may be increased by an increase in tail tension in order to counteract the buckling tendency but not to the point where the working tension rating. Tr, is exceeding. The same arguments apply to equations 20 and 21 for convex curves. In order to minimise the radius of curvature without violating the governing equations we must seek a belt which maximizes the ration of the rated tension to the elastic modulus (maximum Tr/Bm p)). However, very low values of Tr cannot be acceptable because this would severely limit the elevating height of any single run. The nylon-by-nylon reinforced fabric belts seem to offer the best solution of the commercially-available multi-ply fabric belts. In addition, it is possible to order a special nylon-by-nylon belt of reduced elastic modulus, Bm. The cost of such a special belt is higher, but installations in limited space many require a very short vertical radius of curvature and, thus, warrant the extra cost. The Loop Belts installed by Stephens-Adamson on self-unloading bulk carrier lake ships incorporate a special low-modulus belt. An added benefit is that the tail tension on the carrying belt, required to suspend and hug the material at the sandwich entrance of the Loop Belt, is reduced by the reduction of the radius of curvature by equation 5.

If we consider a standard nylon-by-nylon fabric belt, 60 inches wide, carried on equal-roll troughing idlers at 30-degree troughing and subject to a convex curve, we could expect a minimum allowed radius of curvature in the range between 50 and 80 feet. The variation is due to the different standard elastic moduli listed by the various manufacturers. One could expect radii of curvature to one half of this value if a special low-modulus belt is used. With a steel cord, belt, the minimum allowed vertical radius of curvature would typically exceed 10 times that of the standard nylon-by-nylon fabric belt. In open-pit mine applications, the steel cord belt is feasible only in a single-run application with the substantial penalty of making it difficult to recover the material below the transition curve, as illustrated in Figure 12.

When large transition curves, such as required for a troughed steel cord belt, are used to increase the conveying angle from 10 degrees at the loading area, to the ultimate angle of the mine face; a category (e) solution is not possible along the transition profile (see Fig. 16) because the corresponding belt tensions, as dictated by equation 5, are not practical. If category (a) or (b) solutions are employed, as shown in Figure 15, then an additional constraint applies to the transition curve. Because the carrying belt is supported on troughing idlers and is uncovered from the 10-degree angles at the loading area to the 18-degrees angle at the sandwich entrance, uplift of the belt in this region must not occur when starting or running empty. Typical operating tensions in this region will probably not cause uplift the empty belt but, if it is necessary, the added load of the covering system may be applied at angles lower than 18 degrees or prior to the start of the transition curve in order to insure that uplift does not occur.

Transition profiles of short radii of curvature, such as may be obtained with nylon-by-nylon multi-ply fabric belts, lend themselves to a category (e) solution. This is illustrated in Figure 16. The belt tension required at the tail end to suspend the material and carrying belt, as dictated by equation 5 is low enough to make this solution very attractive.

Utilising 2 nylon-by-nylon fabric multi-ply belts, driven by a load-sharing drive system, net elevating heights of up to 350 feet could be obtained without oversizing the belts specifically to maximise lift. A single-run conveyor, utilising two steel cord belts driven by a load-sharing drive system, could achieve practical elevating heights of up to 900 feet. In mines with depths exceeding 900 feet, the high-angle conveyor system could consist of fabric

belt conveyor modules of short transition radius, arranged as shown in Figure 10 or Figure 11; or a single-run approach, as in Figure 12, utilising steel cord belts, or a combination of both. The single-run approach is attractive because it eliminates the need for additional intermediate transfer points and reduces the number of pulleys and drive components. For multi-level operation with conveyors of varying lift required, any single-run lift exceeding 350 feet would require steel cord belts, and this a long transition profile, causing an imposition into the mine pit which would limit material recovery.

A strong desire to achieve lifts greatly exceeding 350 feet, without the penalty of a long transition radius, let to the investigation of a method using a steel cord carrying belt and a fabric cover belt with spring-mounted pivoting wing rolls at the idlers of the transition curve. The operational characteristics of this method proved unsatisfactory and it was not pursued further.

Evolutionary Concept for Future Development

Before introducing new concepts in sandwich belt being high-angle conveyors, it is appropriate to establish criteria which recognise the positive and negative features of past developments, and serve as the basis for evolution of these into operationally superior new concepts.

Criteria for Development of Sandwich Belt Conveyors

All new developments in sandwich belt high-angle conveyors must address the theory and constraints discussed and should judiciously select and incorporate the desirable features of past developments while omitting, where possible, those features which are not considered desirable.

In addition to the five operational requirements already listed, new developments in sandwich belt high-angled conveyors should conform to the following specific criteria. Criteria 4, 5 and 6 apply specifically to hard-rock material discharged from a primary crusher, set at 8 inches maximum lump size.

1. Design coefficients of friction at the belt surface to material and belt surface to belt surface interfaces are taken as .5 which corresponds to a friction angle of 26.6 degrees.
   

2. The belt width is sized so that the cross sectional area of material, corresponding to a uniformly loaded conveyor operating at capacity, is less than the cross-sectional area recommended by CEMA(1) for a conventional conveyor with 0-degree surcharge angle.
   

3. The trough depth of the idler-supported belt in a category (e) solution is equal to or exceeds the maximum lump size.
   

4. Belt speed is approximately 700 feet per minute and does not exceed 750 feet per minute.
   

5. All carrying idlers are of three-equal-roll troughing type and all idlers conform to requirements of CEMA Class E7.
   

6. Minimum wear cover thicknesses are:

3/8 inch carrying belt.
inch cover belt

Grade of Rubber: RMA 1 (Rubber Manufacturers Association)

The design coefficient of 0.5 between conveyed material and rubber surface, as established in Item 1, reflects the authors1 best estimate in light of the limited amount of data available in this area.

Because the material is totally bounded in the sandwich, local impact as the material is carried into and over successive carrying idlers is, to some extent, cancelled by the resisting reaction of the cover belt. The present of the hugging force and elimination of the rolling tendency of the larger lumps, which rise to the top surface, assure a nearly static frictional resistance to movement at the material to belt surface interface. This value is thought to be conservative; but because of a lack of data in this regard, it provides a reasonable design criterion.

A design friction coefficient of 0.5 for the interface between two belt surfaces, as established in Item 1, is consistent with the values used in calculating frictional development in belt-on-belt-type intermediate drives. It takes into account surface wear and the introduction of dirt and moisture.

It is very convenient to have equal values for these friction coefficients, because the additional pressure to transmit relative drag between the two belts, in addition to that required to hug the material, does not depend on the material-free edge distance. This means the prediction of an exact configuration of the sandwich cross section during operation is not necessary. This may be verified by considering equation 9 and reviewing the discussion which precedes and follows it. This becomes important when considering a load-sharing drive arrangement.

Item 2 is an artificial way of achieving the desired cross sectional configuration for a belt sandwich at the rated capacity. A scaled representation is shown in Figure 17. It can be seen that such a configuration yields ample edge distance to assure a tight, leak-proof seal and a margin for overload. It should be noted that Item 2 represents a derating of the belt cross section by approximately 50 pct when compared to a conventional conveyor with a surcharge angle of 25 degrees (as recommended by CEMA for conveying hard rock, lumpy material.

Item 3 reflects the experience of Stephen- Adamson with the operation of Loop Belts. It assures a relatively-smooth hugging pressure distribution without excessive local stretching anywhere on the belt. As noted, this criterion applies specifically to those methods which incorporate category (e) solutions. It turns out that all methods which use a short transition radius to increase the conveying angle from the loading area to the mine face angle do, indeed employ a category (e) solution.

Item 4 sets a conveyor belt speed which will assure smooth operation and extended life of the components. It reflects CEMA recommendations and the experience of industry, domestic and foreign, in conveying hard lumpy material.

Item 5 establishes that the most rugged idlers (the highest CEMA Class) shall be used in accordance with the requirements of continuously conveying hard and lumpy material in a harsh outdoor environment.

The RMA grade and minimum rubber cover thickness listed in Item 6 are as recommended by belt suppliers.

The Mechanical-Pressed Sandwich Conveyor

The Mechanically-Pressed sandwich conveyor is illustrated in Figure 18 and 19. As noted in previous discussion, conveyors which relied on a self weighted cover belt did not prove to be economical for conveying angles approaching 45 degrees. The category (b) Mechanically-Pressed conveyor, illustrated in Figure 4, had limited success on bucket wheel excavators but was eventually abandoned, for reasons which included the accelerated wear of the belt and other components. This is not surprising because the total required hugging load is applied as large discrete loads spaced at large intervals along the conveyor. Operation requirement 2 address the problem by calling for a soft loading system which minimises load concentrations.

4. The radius of curvature must decrease along the driven curve. As the tension on the carrying belt is relieved by the drive belt, the radius of curvature must decrease if the prescribed radial load is to be maintained. The hugging pressure from the op belt increases inversely to the requirement, because its tension is constant ant the conveying angle decreases. The radial pressure from the drive belt increases exponentially.
   

5. Determination of the curved profile between the driven curves is largely dependent on the ability to predict the tension of the carrying belt at any point on the profile when, it alone, is driven. This means that a control system is necessary to assure that all drives contribute equally or nearly equally to the total driving effort. The complexity of a control system to co-ordinate load sharing would increase with the number of drives.

Such considerations are overwhelming and the incorporation of such drive units is ruled out.

The rubber-tire intermediate drive utilises a squeezing pressure which allows the development of traction on the belt. Effective coefficients of friction between the tire and belt surface were found to be from .44 to .55 for wet and dirty surfaces(12). The combined traction-to-squeezing-force ratio of .8, as indicated in Figure 27, is conservative. The inflection points on the snake sandwich conveyor are logical locations for this type of drive, as shown in Figure 29.

B.F. Goodrich and Continental Conveyor Company(13, 14) have developed and used rubber-tire intermediate drives. Drive modules of up to 200 hp are in existence. The system utilises a special belt which has extra wear cover at the edges and limits reinforcements, for steel cord belts, to these edges. Testing with such drives leads the Russians(12), to the conclusion that ordinary belts can be driven in this manner without prematurely wearing out the cover at the edges.

Incorporating 200 hp drive modules into the snake sandwich conveyor design, as shown in Figure 29, would require such units at height intervals of approximately 30 feet for a conveying rate of 6,000 tons per hour, and approximately 50 feet for a conveying rate of 3,500 tons per hour. A bigger unit would be needed for the higher conveying rate.

Considerations 1 and 5, listed for the drive-belt type intermediate drives, are equally applicable to the rubber-tire-type intermediate drive. This makes their incorporation impractical.

If a straight-run conveyor is used, as few as two intermediate drives, utilising steel cord belts could achieve elevating heights in excess of 900 feet. This is illustrated in Figure 30. However, the penalty is substantial because the required contact length between the carrying and drive belts is approximately 100 percent of the conveyor length (because = .5 for material on belt as well as belt-on-belt), and additional hugging pressure must be applied specifically to reduce this contact length. In addition, extra costs are incurred for drive belts, pulleys, and take-up arrangements. Applying the rubber-tire-type intermediate drives in a straight conveyor is no different from its application on the snake sandwich conveyor.

For the reasons discussed, intermediate drives are not considered practical at the present stage of development.

Summary

Conventional belt conveyors offer an economical method for transporting bulk materials at inclines approaching 18 degrees. Many applications warrant conveying at steeper angles to 90 degrees, to minimise the support structure and / or excavation. Of the many high-angle conveying concepts known, the sandwich belt conveyor appears to be the best solution for conveying at high rates, above 1,000 tonnes per hour, and achieving high lifts up to 350 feet per module. By studying the operational characteristics of past and present sandwich belt conveyors, the authors have established the design constraints and developed criteria for future development.

On this basis, new evolutionary methods in sandwich belt high-angle conveying are suggested fur future development. One class of solution is based on added mechanical or pneumatic pressure to the cover belt to develop the necessary friction between the material and the belt surfaces. The second class is based on hugging pressure derived by snaking the conveyor profile to exploit the inherent belt tension in any elevating conveyor. The use of intermediate drives in sandwich belt conveyors was investigated, but the many complications make such drives impractical with todays technology.

References

1. Conveyor Equipment Manufacturers Association (CEMA). Conveyors for Bulk Materials. CBI Publishing Co., Inc. Boston, 2nd edition 1979.
   

2. RASPER, E.L., The Bucket Wheel Excavator, Development, Design Application. Series on Bulk Materials Handling. Vol. 1. 1975, No. 2, Trans Tech Publications, Calusthal, Germany, 1st edition 1975.
   

3. GAERTNER, E., Entwieklungstenzen in der Geraete-und Foerdertechnik der rheinischen Braunk Braunkohlentagebaue (Development Trends of Machines and Material Handling Techniques in the Rheinbraum Open-Pit Mines). Braunkohle, Waeme Und Energie HEFT 11/12, 1955, pp. 22-241
   

4. RASPER P., Bank/Steilooerden in deutschen Braunkohlentugebau (Elevating Belt Conveyors for Lignite Open Mining in Germany). Deutsche Hebe-und Foerdertechnik in Diente der Transportration-isierung, Dec. 1958, pp. 22-29
   

5. COLIJN, H., Design Considerations for Belt-Conveyor Systems. Iron and Stell Engineer, August 1962.
   

6. WANCHECK, G.A., and FOWKES, R.S., Materials Handling Research: Shera Properties of Several Granular Materials. BuMines R1 7731, 1973.
   

7. VIERLING, A.,Die Keilband-Forderanlage, ein neues Mittel zum Steilen Foerden Foerdern von Massenguetern. Braunkhole Heft 11. March 1935, pp. 161-169.
   

8. Stephens-Adamson. The Stephen-Adamson Loop Belt Elevator. Belleville, Ontario, Advertisement.
   

9. Loop belt to be installed in Port Cartier. Engineering and Mining Journal, February 1977, P. 114.
   

10. Stephens-Adamson, Beltavator, Belleville, Ontario, Advertisement.
    

11. R.A. Hansen, Co., Inc. Development of Cross Pit Overburden and Waste Material Handling System. BuMines Contract No. H0252084, 1979.
    

12. SHARIKOW, V., KRAWEZ, N., and FRITSCH, F., Foerderbandantriebe durch Luftreifen (Belt Conveyors Drives with Pneumatic Tires). Foerden und Heben, v. 11 November 1979.
    

13. ARGALL, G.O., Belt Conveying Not Pumping, Used by Brewster P205 For Florida Matrix. World Mining, January 1978, pp. 48-51.
    

14. Coal Age, Continuous Haulage Methods Designed to Keep Continuous Mining Continuous, July 1976, pp. 119-136. (For Coal Age Reprint Dept., P.O. 690, Highstown, NJ 08520).