An Apparatus for Processing Hard Material The invention relates to an apparatus for processing hard material.
There axe numerous industrial uses of super hard materials such as polycrystalline diamond (PCD), natural diamond and Tungsten Carbide (WC). These include tooling for mechanical processing as well as robust substrates for microelectronics in extreme environment applications. Processing these materials is however problematic.
PCD cutting tools are typically used to process non-ferrous metals, wood and rubber. The PCD blanks are cut to shape and brazed into individual holders which are assembled into a cutting tool, often with multiple PCD cutting teeth per tool. Polycrystalline diamond cutting tool blanks can be regarded as a composite material that combines the hardness, abrasion, resistance and thermal conductivity of diamond with the toughness of WC. PCD is a synthetic, extremely tough, intergrown mass of randomly orientated diamond particles in a metal matrix. It is produced by sintering together selected diamond particles at high pressure and temperature. The sintering process is rigidly controlled within the diamond stable region and an extremely hard and abrasion resistant structure is produced. These properties are best utilised in cutting tools for machining a wide variety of materials as well as in wear part applications, where they contribute to improved tool lifetime and offer additional technological advantages such as process reliability and more accurate machining tolerances.
However during the life of the tool the cutting edges wear down (although at a slower rate than traditional cutting tools), which necessitates sharpening of the cutting edges of the PCD teeth. If this is not done then dimensional tolerances are increased and cutting quality is lowered.
One known solution for extending the life of PCD cutting tools comprises electro discharge machining (EDM) in which a wire electrode cuts a diamond by means of spark erosion. A problem with this known solution is that it is low speed and able only to cut simple shapes. Other known conventional techniques suffer from problems such as the requirement for lubricant and excessive tool wear.
Yet a further proposed approach is the use of lasers to cut diamonds. A conventional technique comprises the use of a flash pumped solid state laser
(FPSSL). Flash lamp pumped solid state lasers (FPSSLs) have been widely used in industry for over twenty years operating at a few hundred hertz with millisecond pulse durations providing high power output but low power density
(irradiance). However this arrangement suffers from the problem of low poΛver and efficiency.
The invention is set out in the claims.
Embodiments of the invention will now be described, by way of example, with reference to the figures, of which:
Fig. 1 is a schematic block diagram showing a laser milling apparatus; Fig. 2 is a cut-away perspective view of a PCD composite; Fig. 3 is a graph of removal rate against laser beam pulse irradiance for PCD at a range of puls e repetition rates ;
Fig. 4 is a graph of removal rate against pulse irradiance for tungsten carbide for a range of repetition rates;
Fig. 5 is a schematic block diagram showing a laser cutting apparatus;
Fig. 6 is a diagram of PCD cut depth against test duration for a range of pulse repetition rates;
Fig. 7 is a perspective view of a laser milling application; and Fig. 8 is a side view of a PCDAVC composite and cutting nozzle.
For example the process can be applied to material having a hardness greater than 250 kg.mm" (Vickers Hardness scale), more preferably greater than 500 kg.mm2, more preferably still greater than 1000 kg.mm2. In particular WC has a hardness 1730 kg.mm2 and PCD a hardness 5098 kg.mm2.
In overview the invention relates to a laser configured to process hard material such as natural diamond, PCD and WC using a pulsed laser at very high repetition rates and irradiance. In particular ranges of parameter values have been identified with surprisingly effective operation. As discussed in more detail below it is found that in one embodiment a diode pumped solid state laser (DPSSL) can provide the desired performance.
Two specific processing applications for the apparatus are described below, cutting the initial tool shape and maintaining the tool sharpness.
An apparatus for shaping or milling a hard material is shown in Fig. 1. A laser 10 generates a pulsed beam which is directed by mirrors 12, 14 through an optics array 16, 18 comprising a telescope (beam expander) to a scanner 20 for traversing the beam across a target for example a PCD material to be milled. The beam passes through shaping or clipping apertures or irises 24, 26, 28, 30. In addition an alignment laser 32 for example a HeNe alignment laser generates an alignment beam which is co-axial with the pulsed beam via mirrors 34, 36. The mirror 36 lies in the pulsed laser beam path but is removable once
alignment has been carried out. Iris 23 is used in conjunction with the HeNe alignment beam to aid alignment using back reflections.
The laser milling approach uses a Starlase AO2 Nd: YAG Q-switched DPSSL at the fundamental wavelength of 1064 nm available from Powerlase Limited, Crawley, West Sussex. This pulsed laser offers average powers up to 220 W at a range of repetition rates and pulse durations between 3 to 50 kHz and 20 to 200 ns respectively. The output beam power is varied using any appropriate attenuator unit (not shown) and then collimated with a Galilean telescope 16, 18 and directed into a galvanometric scanner 20 (ScanLab HurryScan25 available from Scanlab GmbH, Munich, Germany). The scanner is fitted with an 80 mm focal length f-theta telecentric objective lens with a working target area of 25x25 mm. AU of the processing work is performed in air at standard atmospheric conditions and no gas assist is used.
In operation the system is first aligned using the alignment laser 32 and the removable mirror 36 is removed. The pulsed laser 10 is then activated and carries out an operational cycle as described in more detail below. The pulsed beam is scanned across the target via scanner 20.
The scanner 20 can be understood in more detail from Fig. 7 in which the scanner includes steering mirrors 50, 52 and a flat-field lens 54. As can be seen from the inset 56, as a result of scanning the pulsed beam, overlapped laser pulses are provided allowing any desired complex milled shape. The scanning speeds are determined as a function of the laser repetition rate and the amount of laser pulse spatial overlap. The amount of pulse overlap is used to control the amount of "heat" that any particular section of the target receives; when a pulse hits the target the material will heat up and some material will be vaporised, and after the pulse the target begins to cool down. Whilst this
cooling process is happening, the next pulse arrives and where there is an overlap the material reaches the point of vaporisation more quickly than for the initial pulse (because the material temperature after the first pulse was higher than ambient) so there is more of the laser pulse available for vaporisation. However, there is a limit to this process - if there is too much residual heat energy left (stored) in the target, it will become a liquid and a melt pool is formed which is much less controllable and therefore undesirable in the milling process. Furthermore, carefully controlling the overlap allows control of the quality of the material surface after the milling process has finished - higher overlaps tend to give better smoother finishes. Accordingly an appropriate scan rate can be determined by experimentation for a given set up to achieve the desired level of milling.
The specific operational parameters selected are described herein in relation to two specific materials, PCD and WC. For example the laser may be used to mill (and as discussed in more detail below, cut) a material such as that shown in Fig. 2 which includes a layer of PCD 200 upon a substrate of WC 210 to form a composite 220 with an edge 230 to be cut/milled and a taper angle Θ240. The laser cutting and milling process must produce and maintain a cut that enables the PCD to be used as a cutting tool. Fig 2 shows the required sharp edge 230 on the diamond side of the PCD material. This edge must be very straight and have as small a radius as possible. The sharp cutting edge is only required on the cutting face 250 and not on the other sides of the PCD part.
The exact operational parameters are material dependent and rely on factors such as thermal conductivity, density, heat capacity, vaporisation temperature, specific heat of vaporisation and reflectivity of target surface, and preferred ranges of specific materials as discussed below.
In the case of milling PCD the preferred irradiance range is 107 to 109 Wcm"2, more preferably 100 MWcm"2 to 200 MWcm"2, the preferred pulse duration is 47 ns to 160 ns, more preferably 120 ns to 160 ns, the preferred repetition rate is in the range 10 kHz (47 ns) to 50 kHz (160 ns), more preferably 40 kHz
(120 ns) to 50 kHz (160 ns), and a removal rate of up to 9 mm3/min is achieved using the specific laser apparatus described above although higher removal rates are possible with higher power lasers. Fig. 3 shows in more detail the removal rate achieved against irradiance for a range of repetition rates from 10 to 50 kHz.
Surprisingly the ranges of parameters described above provide especially good performance results. This is because of the physical mechanism that takes place during laser milling in which the laser pulse provides a first melting stage where the surface of the work piece is raised to the vaporisation temperature followed by a material removal stage where vaporisation occurs in a controlled manner. In particular the laser pulse must be sufficiently powerful (i.e. have enough irradiance) and have sufficient duration to raise the material temperature above the melting point and up to the vaporisation point (near the boiling point) of the material. From this point in the duration of the pulse there is vaporisation and material is removed from the target in a controlled manner.
As a result the governing parameters for removing material by ablation are the irradiance or power density of the pulse and the pulse duration. In some cases pulse duration is directly linked to the laser repetition rate; at high repetition rates the pulse length is long and at low repetition rates the pulse length is shorter and so in the embodiment described the range of repetition rates described above forms a governing parameter by virtue of the direct
relationship to pulse duration. However, in other embodiments the pulse duration itself can be appropriately controlled separately from the laser repetition rate to achieve the desired working conditions.
The importance of irradiance rather than merely pulse energy arises because the pulse energy must be concentrated in a short duration pulse - on the order of nanoseconds - in order to reduce conductance losses into the bulk of the target material which would leave less energy for actually processing and can indeed cause liquid melt pools to form. In the case of PCD, it is found that irradiance levels below the lower limit of the above specified range have little effect on the material other than to melt it such that there is limited removal of material. Conversely the upper limit of pulse irradiance is governed by plasma absorption effects which block the delivery beam to the target (laser induced absorption waves, LAW).
Similarly, as the laser repetition rate is increased the pulse duration increases whilst the irradiance and pulse energy decrease such that the maximum repetition rate limit is governed by the lower threshold of irradiance. Similarly again the lower limit on the pulse repetition rate is also governed by the onset of LAW as the laser pulse irradiance increases at lower repetition rates until an absorbing plasma is produced as discussed above. However a further factor needs to be taken into account namely that as the repetition rate of the laser is decreased, the pulse duration decreases and the irradiance increases. The material removal per pulse will increase as the irradiance increases (because the material will reach the vaporisation temperature more quickly) but the pulse duration is reduced so this increased removal rate works for less time. In addition, as the repetition rate is decreased, although the amount removed per pulse is higher there will be less pulses so the overall removal may not be higher.
Accordingly higher repetition rates are found to be the dominant factor producing the fastest rate of removal of PCD. It is also found that the base of the milled area becomes smoother as the laser repetition rate is increased. This result determines the best laser milling conditions and shows that a high power nanosecond-kilohertz operating regime is effective for milling PCD.
Turning to WC, lower removal rates are achieved than for PCD. Lower frequency repetition rates (which have high pulse irradiance) produce higher removal rates - whereas the highest repetition rate of 50 kHz cannot be used in some cases for ablation at the available pulse irradiances since there is no material removal at all. Referring to Fig. 4 the relationship between pulse irradiance and removal rate is shown for a range of different repetition rates. Surprisingly the preferred range for irradiance is 108 Wcm"2 to 109 Wcm"2, more preferably 500 to 700 MWcm"2 for pulse duration 47 ns to 160 ns more preferably 120 ns to 160 ns and repetition rate 10 to 50 kHz (47 to 160 ns), more preferably 10-3O kHz and the highest removal rates occur in the 10- 20 kHz (47 to 63 ns) range.
Again, this results from the underlying physical mechanisms involved as described above. For the 50 kHz case, although the pulses are relatively long the pulse irradiance is low, and in many cases the vaporisation temperature for WC is not reached within the pulse duration. Only the maximum pulse irradiance reaches the material removal stage. In fact the removal rate generally drops to zero at 50 kHz repetition rate - the pulse irradiance is insufficient for the material to reach its vaporisation temperature, so there is no vaporisation and therefore no material removal. This means that the minimum pulse irradiance boundary condition is not reached. Only at lower repetition rates is material removed where the pulse irradiance level is high enough to
reach the vaporisation temperature within the duration of the pulse. For example, for the 30 kHz case, whilst the pulse durations are shorter, the pulse irradiance is much higher and the material removal stage is reached more quickly within the duration of the pulse.
It is interesting to note that for both of these cases the laser output power may be the same. This means that by maintaining the laser output power and by changing other laser parameters a great deal of improvement to the material removal rate can be made.
Turning to a further application of the tool, cutting is described below with reference to Fig. 5 which shows a laser cutting apparatus of similar nature to that shown in Fig. 1, like reference numerals relating to like parts. The principal difference is the provision of a gas assisting cutting head 40.
The laser piercing and cutting approach uses the higher power Starlase AO4 Nd: YAG Q-switched DPSSL at the fundamental wavelength of 1064 nm available from Powerlase Limited, Crawley, West Sussex.. This pulsed laser offers average powers up to 420 W at a range of repetition rates and pulse durations between 3 to 50 kHz and 20 to 200 ns respectively. The output beam power is varied using an attenuator unit, collimated with a Galilean telescope 16, 18 and then directed by mirror 42 into an Anorad XYZ motion stage 11, available from Anorad UK, Rockwell Automation, Basingstoke, UK. This stage moves the target in the XY directions and the focussing head in the Z direction. The Anorad system is granite mounted and has linear drives capable of a top speed of 2 m/s with an accuracy of +/-1 μm over an XY travel of 450 x 450 mm. The laser beam is focus sed with a variety of lenses 46 of focal length 100 mm to 2O3 mm for example a lens of focal length 149 mm producing at best focus a 200 μm diameter spot. The cutting head 40 allows a co-axial gas
jet 48 to be used to assist the cutting process which can be compressed air, oxygen or nitrogen and can be supplied to the work-piece at pressures up to 10 Bar.
The cutting technique adopted comprises reactive fusion cutting, known as "melt burn and blow", in which a laser beam creates a melt pool and a co-axial gas jet blows the liquid out of the bottom of the cut, the gas jet reacting exothermically with the molten material, adding another heat source to the process and accelerating the creation of the melt pool and hence the cutting speed. This approach provides excellent perpendicular edge quality using either oxygen or air as the gas jet.
In practice it is found that the cutting operation has two distinct phases. The pierce-through phase occurs at the start of a cut-line where a percussion drilled hole is made. For most of this operation the hole is blind and the debris from the drilling operation is thrown up out of the hole entrance, resulting in an area of dross on the material surface around the hole. The Cutting phase follows the pierce-through. The laser cutting head is moved over the material at a constant speed and the material is cut in a single pass. An angled cut front is established and this is where the laser beam is absorbed. The laser beam is waveguided through the thickness of the material.
Operation of the system is principally as discussed above in relation to laser milling, however with specific operation parameters as discussed in more detail below with specific reference to PCD.
In particular the preferred pulse repetition rate is the range 10 to 50 kHz or more preferably 40 to 50 kHz and the pulse duration is preferably 30 to 200 ns, more preferably 100 to 200 ns. The average laser power 300 W to 1 kW, more
preferably 350 to 400 W and the irradiance is preferably in the range 106 to 10s Wcm"2, more preferably of the order of magnitude 100 MW cm"2 and most preferably 110 MWcm"2 (piercing) 118 MWcm2 (cutting). The preferred assist gas pressure is in the region 1 to 10 Bar, more preferably 8 Bar. Referring to Fig. 6 the relationship between PCD depth and test duration during piercing is shown for a range of repetition rates.
In terms of the scanning speed of the beam across the material to be cut, it is preferable to operate at higher speeds because not only does this increase production rates but is also allows less time for the heat to diffuse sideways and creates narrower Heat Affected Zones (HAZ). The optimum speed range is found to be 21 to 27 mm/minute with the best edge quality obtained around 24 mm/minute. The limit on the cutting speed is the point at which the gas jet can no longer eject molten material, at which point the react diffusion cutting operation fails.
In case of composite PCDAVC materials of time type shown in Fig. 2, it is found that at 100 MWcm"2 the PCD removal rate is 7.6 mm3/min whilst WC is just at the threshold of removal. For cutting purposes this means, with the WC side upwards, close to the cutting nozzle and best focus, the preferred range for irradiance is 10 to 10 Wcm"" most preferably 120 MWcm" , more preferably 100 to 200 MWcm"2, repetition rate 40 to 50 kHz, most preferably 45 Hz, pulse duration 41 ns to 200 ns, more preferably 155 ns to 200 ns, pulse energy 6.7 mJ and oxygen gas assist at 8 Bar. These values are preferred for direct WC cutting as well.
However it is found that when cutting with the WC side upwards, striations are caused by the reactive fusion gas cutting; process. This problem can be overcome by turning over the material so that the PCD side is closest to the
cutting nozzle. The best focus position is maintained in this orientation, namely half way through the WC. As a result the assist gas goes through the PCD layer in order to reach the WC, so that there is much less deposited on the PCD, leaving it mostly clean and dross free. This approach provides a significant improvement, reducing the PCD striations and providing a straight, sharp edge on the cutting edge. The PCD layer is dross free and no discontinuity is created between the PCD and the WC layers. The removal of the striations reduces degradation of the straightness of the cutting edge. Non formation of dross ensures that there is reduced interference Λvith the brazing process by which the PCD cutting tools are mounted in to a holder prior to use which can otherwise result in a weaker joint. Removal of a step or discontinuity at the PCD-WC interface removes a weakness at the interface between the materials which can be a source of premature failure of the cutting tool. It is further noted that although some striation is still present but the other layer is dross free, the striations assist the brazing process by providing a larger surface area for the brazing process to adhere to.
The range of parameters provide surprisingly good operation "because the high irradiance, high repetition rate laser pulses produce molten and vaporised WC for the oxygen to react with. This burning reaction proceeds outwards from the laser focal spot in all directions, causing striations until the turning reaction moves too far away from its fuel source (the focussed laser beam and co-axial gas jet). By that time the laser cutting head has moved to a. fresh section of PCD where the burning process starts again. As the cutting speed is increased to that of the "burning reaction" speed, there is a decrease in striation formation.
The pulse irradiance used in the successful cutting trials (118 MWcm"2) is just above the threshold for WC removal by vaporisation, as shov^n in Fig. 4. It is
believed that the cutting process is being enhanced by this vaporisation, with the oxygen assist gas reacting directly with the WC vapour rather than the melt. This could lead to a much more exothermic reaction which would result in faster cutting.
In one preferred approach the cutting process is applied in multiple passes, for example first and second passes, where each may use, for example, the settings described above, with the PCD layer uppermost/closest to the cutting nozzle. In that case the first pass is used to cut through the PCD and the second pass is used to increase the quality of the PCD cutting edge.
In one preferred implementation, dynamic focus changes between the different passes is used to provide improved results. For example in the first pass the focus position may be located within the WC layer as described above. However for the second pass the focus can be moved up to the uppermost surface of the PCD. It is found that this approach in particular increases the quality of the PCD cutting edge. It is also possible to move the focus position for the second pass to an alternative location within either the WC layer or the PCD layer.
In a further improvement a negative taper may be formed on the cutting edge of the material so that the leading cutting edge overhangs the lower edge as shown. in Fig. 8. In particular it can be seen that a cutting nozzle designated very generally 70 is provided near a PCD/WC layered composite 76 with the PCD layer 72 uppermost and closest to the nozzle. The leading cutting edge 74 off the PCD layer overhangs the WC layer such that a negative taper designated generally θ is created. It is further found that changing the focus position fronx a point designated generally X, half way through the WC layer to a point designated generally Y, at the upper surface of the PCD layer can help with the
production of a negative taper. The amount of taper will vary with the specifics of the PCD and the material that the PCD will be used to process (wood, copper, aluminium, etc). The taper angle is typically 7 degrees and not normally greater than 15 degrees,
The approaches described above give rise to a number of advantages: it is found that it is possible to process hard materials at superior rates to conventional technologies, achieving comparable quality without the issues of tool wear and lubrication to contend with. This technology can both cut and mill these materials in concurrent processing - offering new flexibility for manufacturing design.
Much higher energy intensities are now possible with DPSSLs and the nanosecond-kilohertz regime of operation greatly improves many challenging laser materials processing applications. The short pulses diminish thermal effects and improve process quality, while higher fluences improve material coupling and process efficiency. DPSSLs also offer a combination of good beam quality, high efficiency, rugged construction and long diode lifetime. This allows manufacturing on both the macro and micro scale. Laser cutting of PCD is possible with DPSS lasers at much higher cutting speeds than alternative technologies.
For example in comparison with EDM, a laser cutting rate of 24 mm/minute is achieved which is four times faster than the corresponding rate using EDM. Yet further the use of the laser and scanner stage allows omnidirectional cutting.
Laminate structures can be cut using the technique such that PCD and WC disks achieve a cut quality of the same quality as EDM and better than for
FPSS. Such laminate structures include Syndite ™ available from deBeers and single layer or composite structures with thicknesses preferably in the range 0.5 mm to 3.2 mm, for example composite structures of thickness 1.6, 2.0 or 3.2 mm with 0.5 mm PCD on a WC substrate.
It will be appreciated that the technique described herein can be extended to any appropriate hard material and using any laser capable of achieving the specific parameters set out herein. Any material processing application can be adopted in addition to cutting and milling such as indelible marking, drilling of PCD material, for example with a fine hole to make wire drawing dies and 3D milling to create 3D shapes.