1.1 The abrasive water jet
The first use of high-pressure water jet cutting by N. Franz was in the 1950s [1], for the production of boards from raw trees. In the 1970s, Hashish developed the pure waterjet cutting process for cutting soft materials such as cardboard, fabrics and frozen food [2]. Nowadays this technology is also used to cut plastics or rubber [2], in the medical field [3] and to recycle used tires [4]. In the 1980s, abrasive waterjet technology was developed to cut harder materials such as metals. Abrasive particles were introduced into the jet to obtain an abrasive water jet.
The functioning of this technology is based on several steps (Fig. 1). First, water is pumped at a high pressure and then passes through a small orifice. In this second step, the potential pressure energy is transformed into kinetic energy. In a third step the water passes through the mixing chamber where abrasive particles (garnet, alumina, etc.) are added to the water jet. The last step is the focusing of the jet by passing it through a small diameter tube, called a focusing tube. During that step the energy of the water is progressively transferred to the abrasive particles which are accelerated. A high velocity abrasive water jet is thus obtained [5].
1.2 Material Removal Mechanisms
The work of H. Meng and K. Ludema [7] highlights four modes of material removal: cutting, fatigue, brittle fracture and melting. A. W. Momber and R. Kovacevic [8] point out that these four modes act in combination. The importance of each depends on the angle of impact of the particle, its kinetic energy, its geometry and the properties of the target material.
In his work, T. Sultan [9] introduces material removal according to two main mechanisms:
-
The Impact mechanism is the removal of material by impact which corresponds to brittle fracture (Fig. 2-a). When the energy of the jet is too low, it can cause impacts that do not remove material and only produce micro-cracks.
-
The Erosion mechanism is the removal of material by mechanical action on the surface, which corresponds to cutting (Fig. 2-b). This mechanism is particularly relevant for particles that arrive tangentially to the surface.
The two material removal mechanisms presented by T. Sultan [9] are taken into account in different models that predict the volume of material removed [9, 10, 11]. The studies present in the literature consider material removal mechanisms but there are no studies linking them to the physical and chemical characteristics of the abrasive particles
1.3 Characterization Of Abrasive Particles
The selection of an abrasive in terms of composition and size is made with the intended application accordingly. The different abrasives can also be differentiated by their origin which can be alluvial or quarry, crushed or not, natural or synthetic. The finishing surface of the machined parts may vary depending on the origin of the abrasives [13]. The composition of the grains can vary within an abrasive [14].
According to A. W. Momber and R. Kovacevic [15] the important parameters for characterizing abrasive particles are -without hierarchy- : structure, hardness, shape and grain size. The structure corresponds to the theoretical chemical composition, the crystal system and its parameters [15].
According to T.-M. Oh et al [14], for cutting to be effective, a minimum hardness of the abrasive material is required. Abrasive hardness is the most effective when it’s ratio with the workpiece hardness is close to 1.1. Above this ratio, an increase in the hardness of the abrasive does not appear to significantly improve the material removal performance.
Abrasive grains have different shapes which are difficult to characterize. They depend mainly on the material and the extraction process (alluvial removal, crushing, etc.). A. W. Momber and R. Kovacevic [15] synthesize different indicators of grain shape using optical measurements. H. Heywood [16] proposes the use of an elongation ratio and a flattening ratio, which translate the 3D shapes of particles using their length, width and height [17]. H. Wadell [18] proposes to use sphericity for the edges and roundness for the shape. By assigning a value of 0 for very angular particles and 1 for perfectly spherical particles. J. Safanik [19] provides a visual representation of this (Fig. 6). It is also possible to distinguish abrasive particles by their geometric shapes [6, 20].
The studies partially characterize the abrasive particles. Some studies are concerned with origin or purity. Numerous characterization methods, based on optical measurements or visual criteria, have been proposed without any single method being more effective than the others.
1.4 Cutting Power Of The Abrasive Particles
Oh et al [14] propose to compare many parameters of abrasives in relation to the cutting depth and the material removal rate obtained when cutting granite. In their study, five garnets were studied and characterized according to density, hardness, size distribution, uniformity coefficient (geometry), origin, purity and their type. Their work shows that cutting performance increases first with garnet purity, then with density and finally with hardness.
J. E. Goodwin [21] precises that the abrasive efficiency of particles depends on the angularity of the particles and concludes that the shape of the abrasive is important. Comparing the effect of eroded sand and industrial abrasive, he shows that the material removal rate is 3 to 4 times more effective with the industrial abrasive. Thus, the more edges the abrasive particles have, the more efficient the machining is. This is because the corners of the abrasive grains behave like cutting edges and are therefore able to remove more material than round particles which are more likely to push the material away and deform it [21].
In their work, Agus et al [23, 24] define the parameter 𝑃𝐴𝑏𝑟 (unitless) to evaluate the cutting power of an abrasive (Eq. 1):
𝑃𝐴𝑏𝑟 = (𝐻)𝑎1 × (𝑆)𝑎2 × (𝜌)𝑎3 × (𝑑)𝑎4 × (ṁ)𝑎5 (Eq. 1)
With : H, the hardness of the abrasive (Knoop), S, the shape factor of the abrasive (unitless), ρ the density of the abrasive (g/cm3 ), d the average particle diameter (mm) and ṁ the abrasive flow (g/min). The exponents a1 and a 2 allow weighting according to the material to be cut: a1 = a2 = 0.8 (for steel), while a3= -0.2, a4 = 0.1, a5 = -0.5 are invariable [23, 24].
Agus et al [23, 24] conclude that the hardness of the abrasive material and the shape of the abrasive particles have an important influence on cutting. For hard rocks (granite), the most important abrasive parameter is hardness and to cut softer rocks (marble) it is the shape of the particles.
The studies characterize the cutting power of particles according to their hardness, density and geometry. These studies concern the cutting of rocks and not metals.
Conclusions From The Bibliography
There are some characterization works, but it there are limited in number. These works consider rock cutting and present interesting indicators. In our article, the machining of pockets in titanium will be considered. For this purpose, a characterization of the particles will use existing indicators but will also present new observation techniques (microprobe, etc.). A comparison of the cutting power of different abrasives will then be carried out by considering pocket machining in a Ti Al6V titanium alloy.