The cuttability of rock using a high pressure water jet

1. Introduction

One of the fundamental processes in mining is the liberation of minerals from the in situ rock mass. This can be achieved by a number of methods including abrasion and fracture. The causation of fracture in rock has long been associated with some form of mechanical indentation or the cyclic application of large impact forces. This is exemplified by the old ‘hammer and tap’ method of rock drilling but it also underlies the more modern techniques of mechanical rock breakage with cutting tools such as picks and rolling disc cutters. This principle has not precluded other sometimes more subtle techniques from being used in the past such as the gentle knock in the correct orientation by a stonemason.

Apart from mechanical indentation, non-contact mechanisms to initiate fracture in rock have also been developed. Probably the most significant being the detonation of an explosive within a confined space. Before the use of black-powder, however, the ancient Chinese observed and adapted to their advantage the natural weathering process of exfoliation by accelerating the rapid changes in rock temperature with fire and water. A modern variation of this technique is the thermal jet lance (Fleming and Calaman, 1951). More recently Sellar (1991) has reported radiating rock with a pulsing laser to induce internal stress variations causing fracture wherein it was found that the pulse frequency was critical and should equal the resonance frequency of the rock. Common to all these techniques is an alteration in the internal state of stress such that bonds are broken and free surfaces created.

Another potential method of rock breakage reported to have significant potential is the application of high pressure water jets. High pressure water jets in this sense normally refer to pressures between 10 and 400 MPa with a nozzle aperture of less than 1 mm. Harris and Mellor (1974) have reported that a rock surface can be significantly damaged or cut by a water jet at high pressure. As shown in Figure 1, this damage is normally in the form of a narrow slot of varying depth. Various theories have been developed to account for this damage including a theory on cavitational drag (Crow, 1973), energy balance in brittle fracture (Mohaupt and Burns, 1974) and material erosion (Rehbinder, 1976).

Figure 1. View of several slots, or kerfs, cut in rock by a high pressure water jet.

During the early development work, it became apparent that water jets could not compete with conventional forms of rock fragmentation. A relatively large quantity of specific energy, in the order of 1000 MJ/m³, is required for this form of rock breakage which is several orders of magnitude greater than that required in conventional mechanical rock cutting. But it was found that water jets were useful when combined with conventional mechanical systems especially when cutting hard rock where tool life can be short (Hood, 1975). The most significant benefits derived from a hybrid cutting system being:

· a reduction in cutting forces. Fairhurst and Deliac (1986) reported an average force reduction of 30% in the cutting direction and somewhat greater reduction in the normal, or thrust, direction.

· increased tool life. Taylor and Thimmons (1989) reported a doubling in tool life. Hood et al (1991) reported an appreciable tool life was obtained when cutting a hard rock with a UCS of more than 200 MPa, where tool life was otherwise non-existent. Morris and MacAndrew (1986) suggested that water jets reduced the rate of tool wear by cooling the highly stressed rock-carbide interface.

· greater machine advance rates. Because of the reduction in thrust force, larger penetrations per revolution can be attained for a machine with a given thrust capacity. Also the reduction in thrust requirements with water jets can compensate for the loss in cutting efficiency of worn cutter tools.

· lower machine vibration. Fowell et al (1988) reported significantly reduced vibration levels on a roadheader cutter boom.

· increased product size. It has been reported that the proportion of fines can be reduced and coarser rock debris produced (Wang and Wolgamont, 1978).

· lower levels of respirable dust. Taylor et al (1989) found that dust make was 80% less during cutting with jet assistance then in cutting with conventional water spray systems.

· reduced occurrence of incendiary ignitions. Water in the cutting groove can dissipate frictional heat, hence lowering the possibility of gaseous ignition.

More recent work by Lin, Hagan and Roxborough (1990) has shown that greater efficiencies can be attained in water jet cutting of rock by focusing two or more jets below a rock surface. They reported large fragments were produced and that specific energy was reduced by nearly an order of magnitude. This work confirms the behaviour predicted by Mazurkiewicz et al (1978) which they termed the ‘jet accumulation phenomena’ and is an adaptation of the earlier work on shaped charges by Walsh et al (1953).

To understand the mechanisms of water jet assistance in a hybrid cutting system, it is necessary to study the behaviour of a water jet acting in isolation to break rock. Experiments were undertaken to assess the effects of a high pressure water jet in rock cutting. Such knowledge can be used to optimise the total extraction system so that the failure mechanism of the mechanical tool is most effectively complimented by a water jet.

2. LABORATORY APPARATUS

The test program made use of a commercially available high pressure, low volume pump shown in Figure 2. Filtered town water was feed through one of two hydraulically actuated double-ended, reciprocating cylinders. Each cylinder had a 20:1 pressure intensification factor. Changes in water pressure were made by adjusting the outlet pressure of a variable displacement, pressure-compensated axial piston hydraulic pump. The unit was capable of delivering 4.7 L/min at pressures of up to 380 MPa.

Standard industrial sapphire nozzles with a conical outlet were used to form the water jet. The discharge co-efficient of the nozzles was 0.65. A range of nozzle aperture diameters was used in the experiments varying from 0.15 to 0.36 mm.

Figure 2. View of the water intensifier unit and linear cutting table used in the testwork.

A linear cutting machine was used to move the rock samples with respect to a stationary water jet. This modified planer could accommodate rock samples with plan dimensions of 450 ^x 450 mm on a horizontal bed at velocities between 50 and 300 mm/s. A variable frequency controller was used to regulate the speed of the electric drive motor.

Table 1

Material properties of test rocks

property	Woodlawn Shale	Gosford Sandstone
UCS (MPa)	145 ± 27	41.8 ± 4.4
UTS (Brazilian) (MPa)	11.7 ± 3.7	2.95 ± 0.53
E_S (GPa)	36 ± 3	6.3 ± 2.7
E_D (GPa)	26 ± 6	9.2 ± 0.6
G_D (GPa)	10.0 ± 2.6	-
Poisson’s ratio	0.32 ± 0.06	0.13 ± 0.05
Density - bulk (t/m³)	2.73 ± 0.01	2.21 ± 0.05
- grain (t/m³)	2.77 ± 0.01	-
Shore Hardness	62 ± 3	-
Schmidt Rebound No.	69 ± 1	47 ± 2
Hacksaw Abrasiveness	3.11 ± 0.56	-
Porosity - apparent (%)	0.5 ± 0.1	9.4 ± 1
- true (%)	1.8 ± 0.3	-

3. MATERIAL PROPERTIES OF ROCK

Two rock types were used in the study; these were Woodlawn Shale and Gosford Sandstone. A summary of the material properties of these rocks is given in Table 1.

Where applicable, the material properties were evaluated according to the suggested methods prescribed by the International Society for Rock Mechanics.

4. TEST PROCEDURE

Cutting of a rock mass by a water jet is termed water jet slotting or kerf formation. A typical configuration involves a water jet traversing across the surface of a rock as is shown in Figure 3. The principal variables in cutting with a water jet include:

· jet variables: nozzle diameter, water pressure, nozzle discharge coefficient and water density all of which effect the water flow rate and jet velocity, and

· operational variables: standoff distance, nozzle traverse speed, jet attack angle and number of multiple passes.

Figure 3. Main variables in water jet cutting.

Other variables include those of the rock (for example compressive strength, fracture toughness, porosity, grain size and surface roughness) and of the rock mass (for example structure). These variables are, however, site dependent and tend to be over-ridden by the jet and operational variables.

The study was undertaken with a continuous water jet on the linear cutting machine described in §2. The principal goal was to study the basic aspects of a water jet in cutting rock. Only those variables that effect the hydraulic energy of a water jet and hence the energy available to initiate fracture or erode the rock were considered. It can be shown that the energy of a water jet acting per unit distance along a rock surface, or the specific hydraulic energy, can be calculated from Equation 1.

W' = 1)

where:

W' = specific hydraulic energy, MJ/m

C_d = discharge co-efficient

p_n = nozzle pressure, MPa

d_n = nozzle diameter, mm

u_t = nozzle traverse speed, mm/s

r_f = fluid density, kg/m³

As indicated by Equation 1, water pressure and nozzle diameter are the two main jet variables. Other variables such as discharge coefficient, water temperature, addition of polymers or abrasive substances and pulsing of a jet were not studied. These alter the structure of a water jet and would only tend to further enhance performance.

The range of water pressure values in the study was selected on the basis of the mechanical compressive strength of the rock. It has been observed that the minimum pressure required to initiate fracture, commonly referred to as the threshold pressure, is typically of the same order as the rock compressive strength. A spread of water pressure values about the equivalent compressive strength was selected to test this observation as well as to evaluate the effect of changes in energy of a water jet on slot depth.

Of the operational variables, nozzle traverse speed and the number of multiple passes were examined. In each experiment the water jet standoff distance (shown in Figure 3 as the distance between the nozzle outlet and rock surface) was fixed at 21 mm. This meant an effective clearance between rock and nozzle holder of approximately 14 mm. The standoff distance was equivalent to 70, 90 and 140 nozzle diameters for the three nozzles sizes used. Even though this was within the range of maximum effective jetting distance, the constant absolute standoff distance tended to give some advantage (in terms of additional effective jet energy) to the largest nozzle diameter.

The parameters used to assess the effectiveness of changes in variables during the slotting of rock are slot depth and to a lesser degree, specific energy. Slot depth is the mean measured depth cut below the rock surface.

The study was conducted in a mode which simulated a water jet preconditioning the rock surface by forming a slot in front of and in line with a cutter tool. This mode differs from a synergistic arrangement where the actions of a water jet and tool are combined.

The study involved three series of experiments. These are outlined in the following sub-sections. The test program was arranged to cover as wide a range of values as possible. Unless otherwise stated, five levels were selected for each variable which increased approximately in arithmetical progression. The test schedule was randomised to minimise any errors that may have arisen due to the heterogeneity of the rock. In each case, the rock was tested after being air dried for at least 48 hours.

4.1 Slot depth variation

The purpose of this test series was to determine the variability in depth along the length of a slot.

The series involved measurement of slot depth in Gosford Sandstone. Three slots were formed in the sandstone at pressures of 100 MPa, 140 MPa and 210 MPa, at a fixed traverse speed of 50 mm/s and a nozzle diameter of 0.23 mm.

4.2 Principal variables in jet slotting

This series concerned three variables of a water jet in the slotting of Woodlawn Shale. Although slot depth was the main parameter, measurements were initially made also of slot width. Details of the experimental program are contained in Table 2.

Table 2

Variables in jet slotting test program
rock type: Woodlawn Shale

variable	unit	level
traverse speed	mm/s	50	100	150	200	250
water pressure	MPa	100	140	175	210
nozzle diameter	mm	0.15	0.23	0.30

The series was at first organised as a partial factorial program in which one variable was held constant while the other two were varied. Nozzle diameter was fixed at 0.23 mm while water pressure was varied at each level of traverse speed. Each combination of variables was replicated at least four times.

The test series was later extended to a full factorial program where nozzle diameter was also varied for each combination of traverse speed and water pressure. This was done to verify the trends established at the original fixed nozzle diameter of 0.23 mm. The number of replications at the other two nozzle diameters of 0.15 mm and 0.30 mm was reduced from four to two.

4.3 Multiple passes of a jet

This series involved an examination of the effects of successive slot deepening in rock by multiple passes of a water jet. The tests were conducted in Gosford Sandstone.

The series involved two parts. The first part involved the progressive deepening of a slot and measurement after each successive pass of a water jet. A total of twelve consecutive passes were made over the same slot. Each test was performed at a fixed nozzle diameter of 0.23 mm, traverse speed of 150 mm/s and water pressure of 210 MPa. The relatively high level of water pressure was chosen to ensure that even the strongest regions within the rock would be cut by a water jet.

In the second part, water pressure was varied with nozzle diameter and traverse speed fixed at 0.23 mm and 150 mm/s respectively. Measurements of slot depth were made after one and then five consecutive passes of a water jet for each level of pressure. Each test was replicated at least three times. Details of the experimental program are contained in Table 3.

Table 3

Level of variables in multiple pass tests
nozzle diameter: 0.23 mm
traverse speed: 150 mm/s
rock type: Gosford Sandstone

variable	unit	level
water pressure	MPa	70	140	210	275	345
no. of passes		1	5

5. RESULTS

5.1 Slot width

There were little discernible changes in slot width with either water pressure or traverse speed. This is in agreement with previous research. A casual observation made during the tests indicated that the incidence of surface spalling tended to decrease with increasing water pressure.

5.2 Slot depth

Figure 4 illustrates the extent of the variation in slot depth. The irregularity along the slot base when reported in absolute terms of standard deviation was found to increase with pressure from 0.66 through to 0.83 and finally 1.13 mm. The coefficient of variation in the mean slot depth decreased with water pressure from 50% to 33% and finally 24%; that is the relative variability in slot depth tended to decrease with water pressure and consequentially with slot depth.

Figure 4. Superimposition of the depth profiles for three slots formed by a water jet at three different
water pressures in Gosford Sandstone.

This indicates that within the rock matrix, there exist regions of high strength material that are highly resistant to fracture and erosion by a water jet. Although slot depth increased with water pressure, it competes against the variation in toughness within the rock where the latter can dominant the forces of a water jet. The effectiveness of the high water pressures diminishes with depth causing greater absolute variations in depth.

5.3 Effect of nozzle diameter

Slot depth was found to increase with nozzle diameter as shown in Figure 5. Evidently as water pressure increases, the effect of nozzle diameter on slot depth becomes more significant. Hence, nozzle diameter is not the insignificant variable that has sometimes been supposed.

Figure 5. Effect of nozzle diameter on slot depth
at different pressures in Woodlawn Shale.
Traverse speed was fixed at 150 mm/s.

Nikonov (1971) has reported that slot depth increases linearly with nozzle diameter. Based on these results, such a relation would imply a positive slot depth at zero nozzle diameter. It would appear more likely, however, that slot depth would vary as some power function of nozzle diameter such that:

h = (d_n - k)^m2)

where:

h = slot depth, mm

k = constant

and m is some value less than 1.

In this rock it appears there is a pressure, somewhat less than 100 MPa, at which there is little variation in slot depth with nozzle diameter and, where the water jet is ineffective in cutting rock. This may equate with the threshold pressure concept referred to earlier in §4.

As was shown in Equation 1 both nozzle diameter and water pressure influence the specific hydraulic jet energy and hence the level of energy available to cut rock. Figure 5 confirms also, as might be expected, that for a given amount of energy more benefit can be got from increasing pressure rather than by increasing nozzle diameter. For example, the two points A and B shown in Figure 5 are points of equal hydraulic energy but B is five times deeper than A. Point B has an effective nozzle diameter equal to only two-thirds of A and is about double the water pressure. Therefore, in terms of maximising slot depth for a given level of energy, greater benefits are gained from higher pressures than larger nozzle diameters (and hence flow rates).

It is worth noting that each curve for the different pressures appears to approach a limiting slot depth, indicating that there is an optimum nozzle diameter above which no useful increase in slot depth can be gained and that this diameter increases with water pressure. The four curves also indicate a common intercept which can be termed the critical diameter such that:

k = d_c

where:

d_c = critical nozzle diameter, mm

at h = 0

This critical diameter is the minimum diameter necessary to cause any significant failure in rock. It can be expected that the value of critical diameter will vary between rock types and be dependent on grain size, porosity and permeability.

If a linear relation were assumed between slot depth and nozzle diameter then the critical diameter would tend to zero. If on the other hand the trend is as indicated in Figure 5, which is similar for all four water pressures, then for Woodlawn Shale:

d_c » 0.08 mm beware the changed sign

hence Equation 2 becomes

h = (d_n - d_c)^m

=(d_n - 0.08)^m

Figure 6 shows a graph of normalised slot depth against nozzle diameter. Normalised slot depth is a measure of slot depth expressed in relative units of nozzle diameters and is a parameter used in some Equations to predict slot depth, as for example the General Cutting Equation.

Figure 6. Effect of nozzle diameter on normalised
slot depth at different pressures.

The graph shows that over the range of nozzle diameters studied, normalised slot depth remains constant for a given water pressure, that is normalised slot depth is independent of nozzle diameter. Similar relations were also found for the other two traverse speeds of 50 and 250 mm/s. Based on the model of Equation 2 this would be expected if m = 1. Although a slight benefit in the form of an increase in slot depth is evident at a speed of 150 mm/s, it is not significant nor is it consistent with the other two traverse speeds.

5.4 Effect of water pressure

Figures 7a, 7b and 7c are graphs of water pressure against slot depth at different nozzle diameters for the three traverse speeds.

Figures 7a, 7b and 7c. Effect of water pressure on slot depth at different nozzle diameters and traverse speeds.

Over the range of water pressures studied, the results suggest a near linear relation between water pressure and slot depth. This linear relation has also been observed by Brook and Summers (1969) in sandstone at pressures up to 70 MPa and, by Harris and Mellor (1974) in granite at pressures up to 400 MPa. The graphs suggest also that slot depth is dependent on nozzle diameter and traverse speed, where the rate of increase in slot depth increases with nozzle diameter and inversely as the traverse speed.

There is clear evidence from each of the graphs of a threshold water pressure and that it appears to be independent of nozzle diameter. For this rock type, the value of the threshold water pressure is about 60 - 80 MPa. Although this water pressure is equivalent to about half of the mean uniaxial strength of this test rock, it is of a similar magnitude to the minimum measured value of compressive strength. It is also equivalent to about three times the mean uniaxial tensile strength or about twelve times the minimum measured value of tensile strength.

The value of the threshold pressure appears to be only marginally effected by traverse speed. Crow (1973) noted that the threshold pressure should increase with traverse speed. Considering the five-fold increase in speed, the small increase in threshold pressure shown in each of the graphs does not appear very significant.

The variation in normalised slot depth with water pressure appears to be linear as shown in Figure 8. This holds equally true for all nozzle diameters since it was earlier shown in Figure 6 that for a given pressure there is little variation in normalised slot depth with nozzle diameter.

Figure 8. Effect of water pressure on normalised slot depth at different traverse speeds.

Hence, normalised slot depth can be expressed in terms of water pressure such that:

where:

p_n = water pressure, MPa

assuming a linear relation this can be re-written as:

= a p_n - b beware symbol

which applies for the case when

£ 100

where:

s_d = standoff distance, mm

An intercept can be observed on the water pressure axis in Figure 8 where normalised slot depth equals zero. The value of this intercept is equal to the threshold pressure. This intercept is also shown to increase slightly with traverse speed. Hence, threshold pressure is equal to some function of nozzle traverse speed as was earlier predicted by Crow (1973) that is:

where:

p_t = threshold pressure, MPa

Figure 9 shows a plot of normalised slot depth against water pressure where pressure is expressed in terms of the intercept on the pressure axis from in Figure 8. Hence the plot should show an intercept on the normalised pressure axis equal to unity.

Figure 9. Effect of normalised water pressure on normalised slot depth.

Re-arranging Equation 3 in terms of normalised pressure:

substituting on the basis of the observed linear relation:

= a' - b' 6)

as the lines pass through the point (1,0) then

a' = b'

substituting this relation into Equation 6 and multiplying both sides by the constant k:

with

k = a'^-1

that is k is equal to the inverse of the gradient of normalised slot depth against normalised water pressure.

Table 4

Values of threshold pressure and pressure constant
at different traverse speeds.

traverse speed (mm/s)	threshold pressure (MPa)	pressure constant, k
50	68.8	0.114
150	82.3	0.140
250	86.7	0.171

Equation 7 is similar to the dimensionless cutting Equation derived by Veenhuizen et al (1978). Values for threshold pressure and k are given in Table 4.

Figures 10a, 10b and 10c. Effect of traverse speed on slot depth at different nozzle diameters and water pressures.

5.5 Effect of traverse speed

Graphs of the variations in slot depth with nozzle traverse speed are shown in Figs 10a, 10b and 10c. The graphs indicate an inverse relation between traverse speed and slot depth such that slot depth decreases with traverse speed.

Harris and Mellor (1974) found similar trends whereby at traverse speeds greater than 300 mm/s, slot depth was insensitive to changes in traverse speed but as speed decreased below 200 mm/s, slot depth increased dramatically. On the basis of the apparent trends in Figure 10 it can be shown that:

h µ

or, h = + c 8)

where k is some function of water pressure and c tends to zero.

If the relation were a simple hyperbolic expression as shown in Equation 8, there would be a tendency with increasing speed for the slot depth to approach zero. Intuitively this is to be expected since for any given pressure, a speed must eventually be reached at which the jet will be ineffective in damaging the rock surface. Each of the curves in Figs 10a, 10b and 10c is consistent with this hypothesis but there are strong indications that useful penetrations could still be achieved at quite high traverse speeds.

Figure 11. A schematic diagram of a water jet moving normally across a rock surface.

To define the relation between traverse speed and slot depth consider an idealised case of a water jet with a jet velocity, u_j, moving normally across a flat surface at a nozzle traverse speed, u_t, as is illustrated in Figure 11. Based on this arrangement it can be shown that

where:

u_j = water jet velocity, m/s

and as jet velocity is a function of water pressure and nozzle diameter:

10)

Equation 9 is similar to the energy balance Equation developed by Mohaupt and Burns (1974) and to the General Cutting Equation where a linear relation was assumed between normalised slot depth and the inverse of traverse velocity. But the Equation does not take into account the material properties of rock. Figure 12 is a graph of these terms for the four pressure levels and three nozzle diameters.

Figure 12. Effect of inverse traverse speed
on normalised slot depth.

Although the linear functions shown in Figure 12 appear to be lines of best fit, they do not correlate with the expectations at either extremity of the function. First, it could be expected that as traverse speed approaches zero, the slot depth should approach some limiting depth. With a simple linear relation no convergence to some limiting depth takes place as the inverse of traverse speed approaches infinity (that is as traverse speed approaches zero). Secondly, it could be expected that as traverse speed increases then depth would tend to zero. As the inverse of traverse speed approaches zero (that is as speed approaches infinity) then depth approaches some finite positive depth. Hence, a simple linear model of these terms is insufficient to describe the relation between traverse speed and depth.

A better model to describe the variation in traverse speed with slot depth is that proposed by Veenhuizen et al (1978) where normalised slot depth varies as the inverse of the ratio of jet and traverse velocities. This model has been incorporated into the modified dimensionless cutting Equation of Enever and Tooley (1985). In their Equation, normalised slot depth is assumed to vary as the square root of the ratio of jet velocity to traverse speed that is:

11)

5.6 Effect of multiple jet passes

The final set of tests involved successive multiple passes of a water jet over the same groove. This has several advantages where with a particular jet configuration, a desired slot depth may not be achieved in one pass. Multiple passes of the jet allow time for the water and rock debris to drain leaving the surface exposed for re-application and shock impact by the jet. This is beneficial since some of the energy would otherwise be absorbed by water in the slot.

Figure 13 shows the effect of multiple passes by a water jet on the cumulative increase in slot depth. The graph shows that slot depth tends to some upper limit with the number of passes. Though the initial increases in slot depth are large and of the order of 70%, this rate of increase quickly diminishes after five passes to an insignificant level.

Figure 13. Effect of multiple pass slotting
by a water jet on cumulative slot depth.

Figure 14 shows how the coefficient of variation in slot depth decreases and tends to stabilise with the number of passes. The variability in slot depth after eight passes is probably a reflection of the variation in strength of the test rock. Differences in strength dominate the slot profile at depth because of the energy losses within the slot through friction with the sidewalls, etc.

Figure 14. Effect of multiple pass slotting
on the variation in slot depth.

A graph of the effect of water pressure on multiple pass cutting is shown in Figure 15. The results are consistent with the trend in Figure 13 where for a given water pressure, the cumulative slot depth increases with the number of passes. The graph indicates that the benefits of multiple pass slotting improve with water pressure. As the water pressure increases, so the rate of increase in cumulative slot depth with the number of passes also increases. In addition, the coefficient of variation in slot depth is found to decrease with an increase in water pressure.

Figure 15. Effect of water pressure on slot depth
in multiple pass slotting.

6. CONCLUSION

Based on the results of the test program, the following conclusions can be made.

1. There is little discernible change in slot width with any of the variables tested other than nozzle diameter.

2. Slot depth increases marginally with nozzle diameter at low water pressures but tends to become more significant as the pressure increases. There is evidence of an optimum nozzle diameter above which no useful increase in slot depth is found. This optimum diameter tends to increase with water pressure.

3. The results infer the existence of a critical nozzle diameter. This diameter is defined as the minimum nozzle diameter necessary to affect any significant damage to a rock surface. In Woodlawn Shale, the critical diameter is equal to 0.08 mm.

4. Slot depth increases linearly with water pressure. The slope of this curve increases directly with nozzle diameter and inversely as the traverse speed.

5. There appears to be a minimum water pressure necessary to cut rock. This pressure, termed the threshold water pressure, appears to be only marginally affected by nozzle traverse speed. The threshold pressure is of a similar order as the minimum measured value of compressive strength or about twelve times the lowest measured value of tensile strength. Values for threshold pressure were found to vary from 68.8 to 86.7 MPa for traverse speeds ranging from 50 to 250 mm/s.

6. There appears to be a linear variation in normalised slot depth with water pressure which takes the form

k = - 1

where k was found to equal approximately 0.14 for Woodlawn Shale though the value varied slightly with traverse speed.

7. Slot depth decreases with traverse speed in a hyperbolic fashion. Normalised slot depth varies as the inverse of traverse speed to some power. Even though a simple linear model may appear to correlate well with the data, it is limited in its prediction of slot depth.

8. Slot depth can be increased by multiple passes of a water jet. Despite this it can be an inefficient process especially after several passes. Total slot depth becomes asymptotic to some depth value with the number of passes. Although initial increases in slot depth with each pass can be large and of the order of 70%, the increase in depth quickly reduces after five passes. The efficiency of multiple passes in terms of slot depth tends to increase with water pressure.

ACKNOWLEDGMENT

The assistance of Professor FF Roxborough, Ross Marbles and the technical staff of the School of Mining Engineering, The University of New South Wales, is acknowledged as well as the financial support of CRA.

SYMBOLS

C_d: discharge co-efficient

d_c : critical nozzle diameter, mm

d_n : nozzle diameter, mm

k : constant

k : pressure constant

h : slot depth, mm

p_n : water pressure, MPa

p_t : threshold pressure, MPa

r_f : fluid density, kg/m³

s_d : standoff distance, mm

u_j : water jet velocity, m/s

u_t : nozzle traverse speed, mm/s

W' : specific hydraulic energy, MJ/m

REFERENCES

Brook, N and Summers, D A, 1969. The penetration of rock by high-speed water jets, Int J Rock Mech Min Sci and Geomech Abst, v6.

Crow, S C, 1973. A theory of hydraulic rock cutting, Int J Rock Mech Min Sci and Geomech Abstr, 10:567-84.

Enever, J R and Tooley, S M, 1985. A laboratory assessment of the hydraulic cuttability of an Australian black coal, The AusIMM Bulletin, 290:65-70.

Fairhurst, C E and Deliac, E P, 1986. Water-jet assisted rock cutting - the effect of pick traverse speed, in Proceedings 8^th International Symposium on Jet Cutting Technology (BHRA: London).

Fleming, D H and Calaman, I, 1951. Production jet-piecing of blastholes in magnetic taconite, Min Eng, July.

Fowell, R J, Anderson, I and Waggott, A,1988. Investigation into the dynamic rock cutting characteristics of boom tunnelling machines, in Proceedings 29^th US Symposium Rock Mechanics, University of Minnesota (Balkema: Rotterdam).

Harris, H D and Mellor, M, 1974. Cutting rock with water jets, Int J Rock Mech Min Sci and Geomech Abstr, 11:343-358.

Hood, M, 1975. Water jet assisted drag-bit cutting of hard rock, Research Review (Chamber of Mines of South Africa: Johannesburg).

Hood, M, Li, X, Salditt, P and Knight, G, 1991. An advanced system for rock tunnelling - results from a field experiment, in Proceedings 6^th American Water Jet Conference (Water Jet Tech Assoc).

Lin, B, Hagan, P C, and Roxborough, F F, 1990. Massive breakage of rock by high pressure water jets, in Proceedings 10^th International Symposium on Jet Cutting Technology (BHRA: London).

Mazurkiewicz, M, Sebastian, Z and Galecki, G, 1978. Analysis of the mechanism of interaction between high-pressure water jet and the material being cut, in Proceedings 4^th International Symposium on Jet Cutting Technology (BHRA: London).

Mohaupt, U H and Burns, D J, 1974. Machining un-reinforced polymers with high velocity water jets, Experimental Mechanics.

Morris, C J and MacAndrew, K M, 1986. A laboratory study of high pressure jet assisted cutting, in Proceedings 8^th International Symposium on Jet Cutting Technology (BHRA: London).

Nikonov, G P, 1971. Research into the cutting of coal by small diameter water jets, in Proceedings 12^th US Symposium Rock Mechanics (SME-AIME: Littleton).

Rehbinder, G, 1976. Some aspects of the mechanisms of erosion of rock with a high speed water jet, in Proceedings 3^rd International Symposium on Jet Cutting Technology (BHRA: London).

Sellar, J G, 1991. Resonance: The pulsing laser - an irresistable miner, in Proceedings International Symposium Mine Mechanisation and Automation (Colorado School of Mines).

Taylor, C D and Thimmons, E D, 1989. Water-jet-assist improves shearer performance, Eng and Min J, April, pp 37-53.

Taylor, C D, Thimmons, E D and Kovscek, PD. 1989. Evaluation of Water Jet Assisted Cutting Capability on Longwall Shearers, USBM Report RI-9270.

Veehuizen, S D, Cheung, J B and Hill, J R M, 1978. Waterjet drilling of small diameter holes, in Proceedings 4^th International Symposium on Jet Cutting Technology (BHRA: London).

Walsh, J M, Shrefler, R G and Willig, F L, 1953. Limiting conditions for jet formation in high velocity collisions, J Applied Physics, 24(3):349-359.

Wang, F D and Wolgamont, J, 1978. Application of water jet assisted pick cutter for rock fragmentation, in Proceedings 4^th International Symposium on Jet Cutting Technology (BHRA: London).

Original version of manuscript published in Proceedings of Western Australia Conference on Mining Geomechanics, June, 1992 (WASM: Kalgoorlie).