Performance Enhancement of Abrasive Waterjet Cutting

(1)

Performance Enhancement

of Abrasive Waterjet Cutting

(2)

(3)

Performance Enhancement

of Abrasive Waterjet Cutting

Proefschrift

ter verkrijging van de graad van doctor aan de Technische Universiteit Delft,

op gezag van de Rector Magnificus prof. dr. ir. J.T. Fokkema, voorzitter van het College voor Promoties,

in het openbaar te verdedigen

op maandag 19 mei 2008 om 10.00 uur

door

Vu Ngoc PI Master of Engineering Hanoi University of Technology, Vietnam

(4)

Dit proefschrift is goedgekeurd door de promotoren: Prof. Dr.-Ing. habil B. Karpuschewski

toegevoegd promotor: dr. ir. A.M. Hoogstrate

Samenstelling promotie commissie:

Rector Magnificus voorzitter

Prof. Dr.-Ing. habil B. Karpuschewski Otto-von-Guericke-Universität Magdeburg, promotor Dr. ir. A.M. Hoogstrate TNO Science and Industry, toegevoegd promotor Prof. dr. ir. J.R. Duflou Katholieke Universiteit Leuven

Prof. Dr.-Ing. H. Louis Leibniz Universität Hannover Prof. dr. ir. A.J. Huis in ‘t Veld Universiteit Twente Prof. dr. U. Staufer Technische Universiteit Delft

Prof. dr. M.A. Guitierrez De La Merced Technische Universiteit Delft, reservelid

ISBN: 978-90-9023096-2

All rights reserved. No part of this publication may be reproduced, utilized or stored in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without permission from the copyright holder.

(5)

(6)

(7)

Acknowledgement

First of all, I would like to express my deep and sincere gratitude to Prof. Dr.-Ing. habil. Bernhard Karpuschewski, my promoter, for his guidance, consideration, and critical review of the present thesis.

I would like to give sincere appreciation to Dr. ir. André Hoogstrate, my adjunct supervisor, for his useful discussion, for his detailed and constructive comments, and for his important support throughout this work.

Then I would like to express my special thanks to Dr. Mohamed Hashish (Flow International Cooperation), Dr. Jey Zeng (OMAX Corporation), Dr. Eric Chalmers (AccuStream Inc.), Dr. Greg Mort (KMT Waterjet Systems Inc.), Dr. Andreas Höfner (GMA Garnet (Europe) GmbH), and Prof. Deng Jianxin (Shandong University), for their discussions, documents, and encouragement. Furthermore I would like to thank Paolo Golfiotti, my Italian MSc student, for his helping in the abrasive recycling experiments.

Not to forget I would like to record my gratitude to the staff members of the department of Precision and Microsystems Engineering, especially Associate Prof. Marcel Tichem, Dr. Sebastiaan Berendse, Harry Jansen and Marli Guffens, for their supports of my works.

My work is a cooperation between the Delft University of Technology and the Vietnamese Government. The work is supported by Training Scientific and Technical Cadres in Institutions Overseas with the State Budget (Project 322) and the Management Centre for International Cooperation (CICAT). I would like to give my appreciation to all members of CICAT and 322, especially Dr. Paul Althuis, Veronique van der Varst, Willemijn van der Toorn, and Ngoc Lien, for their helps and encouragement.

I would like to thank all my colleagues and friends in and outside TU Delft for their encouraging and helping. Very special thanks to Tolga Susuzlu for his help in my experimental work and to Dr. Thieu Quang Tuan and Jeroen Derkx for their review of my thesis. Also, thanks to Nguyen Thanh Hoan for his help by taking pictures of my experimental setup.

Also I would like to give my appreciation to Prof. Nguyen Dang Binh, Associate Prof. Phan Quang The, Associate Prof. Nguyen Dang Hoe and Associate Prof. Vu Quy Dac from the Thai Nguyen

(8)

Last but not least, I would like to thank my mother, my mother-in-law, my sister and brothers, for their love and their encouragement. Thanks to my nephew Vu Quang Dien for his help by designing the cover of my book. I would like to thank my wife, Hong Tham, for her love, patience, enormous support, review of my thesis, and finally for taking care of our children. I also would like to thank my daughters, Thu Trang and My Hanh, for their love and back up.

Delft, May 2008,

(9)

Summary

Abrasive Waterjet (AWJ) Machining is a recent non-traditional machining process. Major part of this technology is a very high-pressure beam of water and abrasives, which is used for machining. The typical water pressure ranges from 300 to 380 MPa and the typical beam diameter varies between 0.6 and 1.2 mm. This technology is widely used in industry for cutting difficult-to-machine-materials, milling slots, polishing hard difficult-to-machine-materials, cleaning contaminated surfaces, etc. AWJ machining has many advantages, e.g. it can cut net-shape parts, no heat is generated during the cutting process, it is particularly environmentally friendly as it is clean and it does not create dust. Although AWJ machining has many advantages, a big disadvantage of this technology is its relatively high cutting cost. Consequently, the reduction of the machining cost and the increase of the profit rate are big challenges in AWJ technology.

To reduce the total cutting cost as well as to increase the profit rate, this research focuses on performance enhancement of AWJ cutting with two possible solutions including optimization in the cutting process and abrasive recycling.

The first solution to enhance the AWJ cutting performance is the optimization of the AWJ cutting process. As a precondition, it is necessary to have a cutting process model for optimization. In order to use that model for this purpose, several important requirements are given. The most important requirement for such a model is that it can describe the “optimum relation” between the optimum abrasive mass flow rate and the maximum depth of cut.

To develop a cutting process model which can be used for the AWJ optimization, many available models have been analyzed. Since the most important requirement for a process model (see above) can be obtained from Hoogstrate’s model, an extension of this model is carried out. The extension model consists of three sub-models including pure waterjet model, abrasive waterjet model and abrasive-work material interaction model. The pure waterjet model enables to determine the energy transfer from pressurized water to the pure waterjet. The abrasive waterjet model is used to calculate the energy transfer from pure waterjet to the abrasive particles. The abrasive– work material interaction model is used to identify the relation between the work material characteristics, the abrasive characteristics and the cutting efficiency in the process of removing work material chips by using the kinetic energy of abrasive particles.

(10)

optimize AWJ systems. The influence of many process parameters such as the water pressure, the abrasive mass flow rate, the nozzle diameter, the abrasive particle diameter etc. have been taken into account. By modeling the work material coefficient, the extension model can be used for various work materials. Also, by giving a model for the abrasive material coefficient, the model can be applied for several most common abrasive types.

Up to now, there has not been a model for the prediction of AWJ nozzle wear. Therefore, modeling the nozzle wear rate has been carried out and a model for the wear rate of nozzles made from composite carbide has been proposed. The model can be used in the optimization problems as well as in the calculation of the AWJ cutting regime.

Based on the extension cutting process model, two types of optimization applications have been carried out. They are related to technical problems and economical problems. The optimization problems have been solved in order to determine the optimum exchange nozzle diameter and the optimum abrasive mass flow rate for getting different objectives including the maximum depth of cut (for technical problems), the minimum total cutting cost and the maximum profit rate (for the economical problems). From the results of these considerations, regression models for determining the optimum nozzle exchange diameter and the optimum abrasive mass flow rate for various objectives have been proposed.

In AWJ machining, there are many cutting process parameters. Therefore, the ways to select other process parameters optimumly have also been investigated. The procedure for the determination of an optimum cutting regime then is given.

The other solution to enhance the cutting performance is abrasive recycling. In the present study, GMA garnet, the most popular abrasives for blast cleaning and waterjet cutting, has been chosen for the investigation. The recycling of GMA abrasives has been investigated on both technical side and economical side. On the technical side, the reusability and the cutting performance of the recycled and recharged abrasives have been analysed. The influence of the recycled and recharged abrasives on the cutting quality was studied. Also, the optimum particle size of recycled and recharged abrasives for the maximum cutting performance has been detected. On the economical side, first, the prediction of the cost of recycled and recharged abrasives was done. Then, the economic comparisons for selecting abrasives have been carried out. In addition, the economics of cutting with recycled and recharged abrasives have been studied. Several suggestions for an abrasive recycling process which promises a more effective use of the grains have been proposed. By optimization in the cutting process and by abrasive recycling, the cutting performance can be increased, the total cutting cost can be reduced, and the profit rate can be enlarged considerably. Consequently, the performance of AWJ cutting can be enhanced significantly.

(11)

Samenvatting

Abrasief waterstraal snijden (AWS) is een recent, niet conventioneel verspaningsproces. Het is een technologie waarbij een waterstraal onder hoge druk, gemengd met abrasief, gebruikt wordt voor de verspaning van diverse materialen. De waterdruk ligt tussen de 300 en 380 MPa; de waterstraaldiameter ligt tussen de 0.6 en 1.2 mm. Waterstraaltechnologie wordt veel gebruikt in de industrie voor het snijden van moeilijk bewerkbare materialen, het boren van gaten, het polijsten van harde materialen, het reinigen van vervuilde oppervlakken etc. AWS bewerken heeft vele voordelen waaronder: het maken van “near-net-shape” onderdelen, geen warmte ontwikkeling tijdens het verspaningsproces en het is bijzonder milieuvriendelijk omdat het schoon is en er geen fijnstof of gevaarlijke stoffen vrij komen.

Naast de vele voordelen die AWS snijden biedt zijn de hoge kosten een belangrijk nadeel. Daarom zijn de reductie van de bewerkingskosten en het verhogen van de winstmarge belangrijke uitdagingen in de AWS technologie.

Om zowel de totale bewerkingskosten te reduceren alsook de winstmarge te verhogen, concentreert dit onderzoek zich op de prestatieverbetering van AWS snijden. Twee mogelijke oplossingen worden bekeken: optimalisatie van het bewerkingsproces en hergebruik van abrasief. De eerste oplossing om de prestatie van AWS bewerken te verbeteren is de optimalisatie van het AWS proces. Voorwaarde hiervoor is de beschikbaarheid van een procesmodel van de verspaning voor de optimalisatie. Om een model te kunnen gebruiken voor dit doel moet het aan enkele belangrijke voorwaarden voldoen. De belangrijkst daarvan is dat het model de relatie tussen de abrasief massa stroom en de maximale snedediepte beschrijft.

Voor de ontwikkeling van een procesmodel dat gebruikt kan worden voor de AWS optimalisatie zijn vele beschikbare modellen geanalyseerd. Omdat aan de belangrijkste voorwaarde voor een proces model (zie boven) kan worden voldaan door het model van Hoogstrate, wordt een uitbreiding van dit model uitgevoerd.

Het uitgebreide model bestaat uit 3 deelmodellen: het pure waterstraalmodel, het abrasieve waterstraalmodel en het abrasief-werkstukmateriaal interactie model. Het pure waterstraal model maakt het mogelijk de energie overdracht te bepalen van het samengeperste water naar de pure waterstraal. Het abrasieve waterstraalmodel wordt gebruikt om de energie overdracht te berekenen

(12)

wordt gebruikt om de relatie te bepalen tussen de werkstuk materiaal eigenschappen, het abrasief en de verspaningsefficiëntie tijdens het verwijderen van spanen van het werkstuk door gebruik te maken van de kinetische energie van de abrasieve deeltjes.

Het uitgebreide procesmodel is nauwkeuriger dan het originele model en kan gebruikt worden voor het optimaliseren van AWS systemen. De invloed van diverse procesparameters zoals de waterdruk, de abrasief massastroom, de orifice diameter, de deeltjesgrootte etc. zijn in het model meegenomen. De introductie van een werkstukmateriaal-coëfficiënt maakt het model bruikbaar voor diverse werkstuk materialen. Tevens kan het model gebruikt worden voor verschillende abrasief materialen door de introductie van een abrasiefmateriaal-coëfficiënt.

Tot dusver was er geen model beschikbaar voor de voorspelling van de slijtage van de AWS focusbuis. Daarom is de slijtagesnelheid van de focusbuis gemodelleerd en een model voor de slijtagesnelheid van gesinterde wolfraamcarbide focusbuizen voorgesteld. Dit model kan zowel worden gebruikt voor de optimalisatie van het AWS proces.

Twee types van optimalisaties zijn uitgevoerd, gebaseerd op het uitgebreide procesmodel. Deze zijn gerelateerd aan technische en economische optimalisatie. De optimalisatie functie is zodanig opgesteld, dat de optimale focusbuis wissel diameter en de optimale abrasief massastroom konden worden bepaald. Dit is gedaan voor verschillende doelstellingen waaronder de maximale snedediepte (de technische doelstelling) en de minimale bewerkingskosten en maximale winstmarge (de economische doelstellingen). Gebaseerd op de resultaten van deze overwegingen zijn regressie modellen voorgesteld voor het bepalen van de optimale focusbuis wissel-diameter en de optimale abrasief massastroom voor de verschillende doelstellingen.

Er zijn vele procesparameters in AWS bewerken. Daarom zijn de diverse methodes om de optimale procesparameters te bepalen ook onderzocht. De procedure voor het bepalen van een optimaal verspaningsregiem wordt vervolgens gegeven.

De andere oplossing om de verspaningsprestatie te verbeteren is het hergebruik van het abrasief. In dit onderzoek is gebruik gemaakt van het meest populaire abrasief voor waterstraal snijden en reinigen: GMA garnet. Zowel de technische als de economische kant van het hergebruik van GMA garnet zijn onderzocht. Op het technische vlak zijn de herbuikbaarheid en de verspaningsprestatie van het hergebruikte abrasief geanalyseerd. Hierbij is zowel het batch-gewijze hergebruik van abrasief, alsook het gradueel opmengen van gebruikt en nieuw abrasief geanalyseerd. De invloed van beide recycle-methodes op de verspaningskwaliteit is onderzocht. Tevens is de optimale deeltjesgrootte voor hergebruik bij beide methodes, gerelateerd aan de maximale verspaningsprestatie bepaald. Op het economische vlak is allereerst een voorspelling gedaan van de kosten van beide recycle-methodes; vervolgens is een economische vergelijking voor de selectie van abrasieven gedaan. Daarbij zijn ook de kosten bestudeerd van het bewerken met hergebruikt

(13)

en met toegevoegd abrasief. Er zijn verschillende suggesties gedaan voor een hergebruik proces dat een effectiever gebruik van het abrasief materiaal belooft.

Door optimalisatie van het verspaningsproces en hergebruik van het abrasief kan de verspaningsprestatie worden verhoogd, de totale verspaningskosten worden gereduceerd en de winstmarge aanmerkelijk worden vergroot. Daardoor kan de prestatie van AWS bewerken significant worden verbeterd.

(14)

(15)

Nomenclature

Symbols

Symbol Unit Definition

A m2 _{cross section area}

C € cost

c - coefficient

d m diameter

E MPa elasticity

ec J/m3 specific cutting energy

F N force

Grecy kg/h recycling capacity

g h m depth k - coefficient l m length Nm - machinability number n - number

m kg/s mass flow rate

P w power

Pr € profit

p Pa pressure

Q - quality number

r % reusability

R - abrasive load ratio

Re - Reynolds number

T s time

(20)

η - momentum transfer efficiency coefficient

κ - power transfer efficiency

δ m/s wear

ξ - cutting efficiency coefficient

ρ Kg/m3 _density

Subscripts

Subscript Definition

a related to average

abr related to abrasive

actual related to actual

awj related to abrasive waterjet

c related to cutting com compressible d related to discharge de related to depreciation e related to electrical en related to energy

f focusing tube / nozzle

h related to hour

inc related to incompressible

int related to interest

l related to length la related to labor m related to mass ma related to maintenance max maximum min minimum

msh related to manned shifts

mt related to machine tool

(21)

op related to optimum

ori orifice

ov related to overhead

p related to particle

q related to quality

rech related to abrasive recharging recy related to abrasive recycling

ro related to occupied room

rpl related to replacement

sal related to sale

sqm related to squared meter

sh related to shift

th theoretical

use related to time of use

ut related to utilization

w related to water

wa related to wages

wj related to waterjet

wor related to working

(22)

(23)

1 Introduction

Abrasive Waterjet (AWJ) Machining is a recent non-conventional machining process. In this technology, a very high-pressure beam of water and abrasives is used for machining. This technology is widely used in industry as it has many advantages.

In this chapter an introduction to Abrasive Waterjet (AWJ) Technology is provided. A review of the AWJ history is first carried out to draw a picture of the progress in this technology. Brief descriptions of the schema and the main components of an AWJ system are also given. Advantages and drawbacks of the AWJ technology are then evaluated. Challenges of the technology are discussed in the end.

1.1 Historical review

AWJ machining has been developed from Waterjet machining. The earliest use of the water beam in coal mining was in the former Soviet Union and New Zealand [Summ95]. This mining technique was also used for removing blasted rocks from working areas into collection tunnels.

From 1853 to 1886, pressurized water was used for excavating soft gold rocks. The pressurized water for coal mining was also used in Prussia in the early 1900s and then in Russia in the 1930s [Summ95].

In 1936, Peter Tupitsyn, who was working for the Donetsk Coal Basin in Ukraine, proposed the idea of using a waterjet beam to cut boreholes in the coal bed [Chri03].

In the 1950s, Dr. Norman Franz, a forestry engineer, was the first who studied the use of a waterjet beam as a cutting tool for wood processing [Flow08]. However, the first patent of a waterjet cutting system was granted for the staff of McCartney Manufacturing Company, a division of the Ingersoll-Rand Corp. [Tikh92]. In 1971, the first commercial waterjet machine was introduced into the market by this company [Tikh92].

In 1979, Dr. Mohamed Hashish, who has worked for Flow International Cooperation, invented the abrasive waterjet cutting method by adding abrasives into the pure waterjet [Flow08]. Soon after this, in 1980, abrasive waterjet was first used to cut glass, steel, and concrete [Flow08]. The invention of AWJ led to a huge expansion of applications of cutting with high-pressure water. Since

(24)

materials, cleaning of contaminated surfaces, polishing of hard-to-machine materials, etc.

1.2 Introduction to AWJ Technology

1.2.1 Introduction to an AWJ cutting system

There are two types of waterjets: pure (or plain) waterjet and abrasive waterjet. In pure waterjet cutting, only a pressurized stream of water is used to cut through materials. This type of cutting is used to cut soft materials such as cardboard, leather, textiles, fibre plastics, food or thin plates of aluminium. In AWJ cutting, an abrasive waterjet entrainment system mixes abrasives with the waterjet in a mixing chamber following an orifice (Figure 1.1). The abrasive particles are accelerated by the water stream and then leave the focusing tube (or the nozzle) with the stream. AWJ cutting is used for cutting harder materials such as stainless steel, glass, ceramics, titanium alloys, composite materials, and so forth.

Focusing tube Mixing chamber Abrasive particles Orifice Directional control valve Hydraulic pump Electric motor

Intensifier Attenuator High-pressure water Inlet water Jet former Presure generation system Water preparation system

Figure 1.1: AWJ entrainment system schema

A typical AWJ entrainment system (as shown in Figure 1.1) consists of four main parts: the water preparation system, the pressure generation system, the jet former, and the abrasive supply system. A brief description of these parts is given below:

(25)

The water preparation system is used for supplying purified water for the pressure generation system. Generally, particles larger than 1 μm have to be filtered out to prevent unacceptable wear of the critical parts of the pressure generation system [Hoog00].

• The pressure generation system:

This system is equipped with a pump to ensure a continuous and stable flow of high pressure. Three types of pumps, namely intensifier, crankshaft and direct pumps can be distinguished.

Figure 1.2: Direct pump (Courtesy of Flow International Cooperation)

Direct pumps are used for applications with low pressure such as cleaning, or washing a desk or a work place etc. In a direct pump, the movement of three plungers is transmitted directly from the electric motor (see Figure 1.2).

Figure 1.3: Double-acting intensifier

Intensifier pumps (Figure 1.3) are used for applications with water pressure up to 600 MPa. In an intensifier pump, a double-acting cylinder in which the movement of the piston is driven by a hydraulic system is used. Two small diameter cylinders at each end of the hydraulic cylinder help to

(26)

pressurize the water alternately as the hydraulic piston moves back and forth. By connecting two intensifier pumps in series, the output water pressure can be up to 800 MPa [Susu04].

The third type is the crankshaft pump, which can provide the pressure up to 345 MPa [Chri03]. An example of this pump is shown in Figure 1.4. It is known that the efficiency of crankshaft pumps is higher than that of intensifier pumps because crankshaft pumps do not require a power-robbing hydraulic system.

Figure 1.4: Crankshaft pump (Courtesy of OMAX Corp. Kent, WA) • The jet former:

The jet former is used to transfer part of the hydraulic water energy into kinetic energy of water, and then into kinetic energy of abrasive particles. Figure 1.5 shows a typical jet former for AWJ cutting [Hoog04]. To form the abrasive waterjet, first, the high pressure water is forced through an orifice to create a high speed waterjet. Then the high speed waterjet passes through a mixing chamber, which is installed downstream of the orifice. Because of the Venturi effect, a vacuum is created in the mixing chamber. As a result, the abrasive particles and some air are sucked into the mixing chamber through a feed line. After entering the mixing chamber, the particles are accelerated by the high-speed waterjet (velocity about 600 to 900 m/s) and then passing through a focusing tube (or nozzle).

As mentioned above, the orifice, the mixing chamber and the focusing tube are the main parts of a jet former. Orifices can be made of sapphire, ruby or diamond with a diameter ranging from 0.08 to 0.8 mm [Hoog00]. The lifetime of a diamond orifice is about 1000 to 2000 hours while it is only 40 to 70 hours for sapphire [Koel02]. However, sapphire orifices are most commonly used because they are much cheaper than diamond orifices (the price of a diamond orifice can be $435 while it is

(27)

only $14.5 for a sapphire one [Bart08]).

Most of AWJ nozzles are made from composite carbide materials. They are available on the market under specific product names such as ROCTEC 100 and ROCTEC 500 from Kennametal Inc. ROCTEC composite carbide is a very dense, sintered, tungsten carbide based hardmetal. The common inner diameter of the focusing tube is from 0.5 to 1.5 mm, and the common length is from 70 to 100 mm.

• The abrasive supply system:

The abrasive supply system is used for accurate supply of abrasives with a pre-required mass flow rate. In practice, there are many types of abrasives which are used in AWJ machining. They can be garnet (for example Barton garnet (a trade mark of Barton Mines Company) and GMA garnet (a trade mark of GMA garnet Pty Ltd) – two most common garnets), olivine, aluminum oxide, silica-sand etc. Generally, in AWJ machining, the abrasive mass flow rate is about 0.08 to 0.5 kg/min (15 to 30 kg/h [Trum97]) and the abrasive size varies between 0.1 and 0.3 mm.

Figure 1.5: A typical jet former for AWJ cutting [Hoog04]

1.2.2 Parameters of an AWJ machining process

There are many parameters involved in an AWJ machining process. In general, these parameters can be divided into two groups: process parameters and target parameters [Momb98]:

• Process parameters:

The process parameters include parameters relating to the forming of the AWJ beam. These parameters can be sorted into four following sub-groups [Momb98]:

high pressure water orifice

abrasive supply

mixing chamber focusing tube

(28)

-Mixing parameters including focusing tube (or nozzle) diameter and focusing tube length.

-Abrasive parameters including abrasive material, abrasive particle size, abrasive shape, and abrasive mass flow rate.

-Cutting parameters including standoff distance, impact angle, traverse rate and number of passes. • Target parameters:

The target parameters consist of parameters related to the target of the machining. These parameters are the work material, the depth of cut and the cutting quality.

1.2.3 Advantages and disadvantages of AWJ Technology

AWJ cutting has various advantages over other non-conventional techniques such as laser and Electrical Discharge Machining (EDM). The advantages can be presented as follows:

-AWJ can machine a wide range of materials including titanium, stainless steel, aerospace alloys, glass, plastics, ceramics, and so on.

-AWJ can cut net-shape parts and near net-shape parts.

-No heat is generated in the cutting process. Therefore, there is no heat-affected area and thus no structural changes in work materials occur.

-AWJ cutting is particularly environmentally friendly as it does not generate any cutting dust or chemical air pollutants.

-The abrasives after cutting can be reused which allows for possible reduction of the AWJ cutting cost.

-Only one nozzle can be used to machine various types of work materials and workpiece shapes. -AWJ machining can be easily automated and therefore can be run with unmanned shifts.

Although AWJ cutting is a truly useful machining process and can be used for various applications, the technology still has two major disadvantages:

-The total cutting cost is relatively high;

-The cutting quality is not always satisfying and unstable.

1.3 Challenges in AWJ Technology

(29)

prominent disadvantage. In AWJ cutting, the total cutting cost depends on many cost components such as machine tool cost, abrasive cost, nozzle wear cost, wages including overhead cost and so on. High AWJ cutting cost, for example, in Europe, the cutting cost per hour is about 150…200 (€/h), makes the AWJ business less competitive. As a result, the reduction of the total cutting cost and cutting time as well as the increase of the profit rate (or profit per hour) in AWJ machining are big challenges for this technology.

(30)

(31)

2 State of the art in optimization of AWJ machining

As addressed in Chapter 1, one of the biggest disadvantages of AWJ cutting is its high cost. The AWJ cutting cost per hour, for example, can be 150…200 (€/h) in Europe. Therefore, finding solutions to reduce the total cutting cost to increase the profitability for AWJ users is an important task of AWJ technology.

In the AWJ cutting cost, the abrasive cost (including disposal cost) is usually the largest component (Figure 2.1). This can amount to 20% up to 70% of the total cutting cost, depending on parameters such as the abrasive mass flow rate, the number of cutting heads, the abrasive price, the AWJ system’s cost and so on. However, the abrasives after cutting can be reused, which can reduce the abrasive cost and the disposal cost.

Machine tool cost (23.94%)

Abrasive cost (53.98%)

Water cost (1.31%) Nozzle cost (3.05%) Orifice cost (0.92%)

Wages including overhead cost (16.89%)

Figure 2.1: A typical AWJ cost breakdown [Hoog06]

In practice, the AWJ optimization and abrasive recycling are two main ways to increase the profitability for AWJ users. Especially, optimization can reduce both the cutting time (or increase the cutting performance) and the cutting cost and can increase the profit rate. Therefore, AWJ optimization and abrasive recycling have been the objectives of many studies.

The optimization problems in AWJ machining can be divided into two categories including AWJ technical optimization and AWJ economical optimization. The technical optimization, based on the

(32)

process parameters in order to fulfill the maximum cutting performance or the minimum cutting time. The economical optimization, based on the economical relations as well as the physical relations between the process parameters, aims at the optimum values of the process parameters for getting the minimum total cutting cost per product (or per unit length of cutting) or the maximum profit rate.

Up to now, there have been many studies on both AWJ optimization and abrasive recycling. To have a clear picture on this, a literature review is carried out. The review is split into three parts: AWJ technical optimization in Section 2.1, AWJ cost calculation and cost optimization in Section 2.2, and abrasive recycling in Section 2.3.

2.1 State of the art in AWJ technical optimization

In the AWJ cutting process, there are various factors affecting the material removal process or the cutting performance. These factors include the jet-parameters (the water pressure, the orifice diameter, the focusing tube diameter, the focusing tube length, the abrasive mass flow rate, the abrasive size, the abrasive shape and type) and the cutting parameters (e.g. the standoff distance, the workmaterial, the feed speed).

2.1.1 Optimum combination of focusing tube and orifice diameter

0.5 1 1.5 2 2.5 3 20 30 40 50 60 70 80 90

Focusing tube diameter d f (mm)

Material removal rate Q (mm

3/s) p w=240 MPa; dori=0.25 mm v f=1.67 mm/s; lf=50 mm R=0.3; AlMgSi0.5 0 20 40 60 80 100 20 25 30 35

Focusing tube length l f (mm)

Maximum depth of cut h

max (mm) p w=240 MPa; v=1.67 mm/s d ori=0.25; df=1.2 mm R=0.3; AlMiSi0.5

Figure 2.2: Focusing tube diameter versus material removal rate [Blic90]

Figure 2.3: Focusing tube length versus maximum depth of cut [Blic90]

H. Blickwedel [Blic90] investigated the relationship between the focusing tube diameter and the volume removal rate. The author notes that the final abrasive particle velocity depends on the density of the abrasive-water-air mixture: a denser mixture creates a higher particle velocity. Also,

(33)

as the focusing tube diameter increases, the density of the mixture decreases and therefore the particle velocity decreases. However, a small focusing tube diameter leads to more interactions between particles and nozzle wall, and particles with each other and thus reduces the particle velocity. Therefore, an optimum value of the focusing tube diameter exists for the material removal rate (see Figure 2.2).

H. Blickwedel [Blic90] proposed an optimum ratio between the focusing tube diameter and the orifice diameter as follows:

= …3 4 f ori d

d (2.1) U. Himmelreich and W. Riess [Himm91] confirmed that the above ratio is a good value for AWJ formation. E.J. Chalmers [Chal91] observed that the maximum depth of cut will occur for the ratio of nozzle to orifice diameter of 3. Zeng and Munoz [Zeng94] also reported that the highest cutting performance is achieved when using the following optimum combination of focusing tube/orifice: 3.3 (0.023”/0.007”), 3.2 (0.032”/0.01”), and 3.14 (0.044”/0.014”).

2.1.2 Optimum focusing tube length

Figure 2.3 shows the relation between the focusing tube length and the maximum depth of cut [Blic90]. The depth of cut, at first, increases linearly with the increase of the nozzle length. This is because a certain acceleration distance is necessary to accelerate the injected abrasive particles [Momb98]. Beyond this critical acceleration distance, the friction due to the spreading water jet increases. This leads to a reduction of the particle velocity and therefore a decrease of the depth of cut [Momb98]. The optimum acceleration distance, as noted by M. Heβling [Heβl88], depends strongly on the abrasive material density. Figure 2.4 shows the relation between the focusing tube length and the maximum depth of cut for different abrasive materials [Heβl88]. It is observed that round steel cast abrasive material is most influenced by the nozzle length while broken abrasive material and quartz sand are only lightly affected (Figure 2.4).

H. Blickwedel [Blic90] suggested the optimum focusing tube length lf,op based on his experimental results: , 25 50 f op f l d = … (2.2) M. Hashish [Hash91] indicated that the depth of cut and the kerf width both depend on the length of the focusing tube. The depth of cut and the kerf width reduce as the focusing length increases up to a length of about 50 to 70 times of the focusing tube diameter. Also, it is noted that no change in the depth of cut and the kerf width occurs when the focusing tube length increases

(34)

Generally, the wear of the focusing tube is affected by the tube length. M. Hashish [Hash94] addressed that the nozzle exit bore wear rate reduces as the nozzle length increases. This conclusion was also confirmed later by K.A. Schwetz et al [Schw95] and M. Nanduri et al. [Nand00]. Figure 2.5 illustrates the relation between the tube length and the exit bore diameter wear rate [Nand00]. It follows that when the tube length is smaller than a certain value (in this case around 75 mm), a decrease of the tube length will lead to a significant increase of the exit bore wear rate. Beyond this value, the exit bore wear rate is almost unaffected by the tube length.

In practice, the length of the focusing tubes is determined for both a high cutting performance and a long nozzle lifetime. The nozzle lengths are standardized in some common sizes of 76 mm (3”), 89 mm (3.5”) and 101.6 mm (4”). It is known that the most commonly used nozzle length is 76 mm, offering the best cost-to-wear-life ratio [Chal06].

0 50 100 150 0 10 20 30 40 50 60 70 80

Focusing tube length l

f (mm)

Maximum depth of cut h

max

(mm)

Steel cast, angular (7400 kg/m3) Steel cast, round (7400 kg/m3) Quartz sand, round (2650 kg/m3)

p w=200 MPa; vf=20 mm/s d ori=0.6; dp=600 µm m a=30 g/s 20 40 60 80 100 120 0 5 10 15 20 25 30

f (mm)

Exit diameter increase rate (%)

p w=310 MPa; ma=3.8 g/s d ori=0.38; df0=1.14 mm Nozzle material: WC/C o

Abrasive: aluminum oxide #80

Figure 2.4: Focusing tube length versus maximum cutting depth [Heβl88]

Figure 2.5: Nozzle length versus nozzle exit bore increase rate [Nand00]

2.1.3 Optimum abrasive mass flow rate

Typical relations between the abrasive mass flow rate and the maximum depth of cut are shown in Figure 2.6. It follows that the depth of cut, at first, increases as the abrasive mass flow rate increases. However, when the abrasive mass flow rate exceeds a certain value, the depth of cut will drop (Figure 2.6). This relation can be explained by the following equation [Hash89]:

η = ⋅ + 1 / wj awj a w v v m m (2.3) In which, vwj is the velocity of water leaving the orifice, vawj is the velocity of abrasive particles

(35)

leaving the nozzle, η is momentum transfer efficiency, m_a is the abrasive mass flow rate, and mw is the water mass flow rate.

Previous studies ([Mill91], [Clau98] and [Susu06]) indicate that the momentum transfer efficiency η decreases as the abrasive mass flow rate increases. Hence, it can be deduced from Equation (2.3) that an increase of the abrasive mass flow rate can lead to a decrease of the abrasive particle velocityv_awj. Moreover, an increase of the abrasive mass flow rate also results in an increase of the impact frequency of particles. Therefore, a critical value of the abrasive mass flow rate exists at which the benefit of the impact frequency balances the loss in particle velocity [Zeng94]. This critical value is the optimum abrasive mass flow rate for the maximum depth of cut.

0 5 10 15 20 25 14 16 18 20 22 24 26 28 30 32 34

Abrasive mass flow rate (g/s)

Maximal depth of cut h

max (mm) d_f=0.8 mm d f=1.2 mm d_f=1.6 mm p w=240 MPa; vf=1.67 mm/s d ori=0.25 mm; lf=50 mm AlMgSi0.5

Figure 2.6: Abrasive mass flow rate versus maximum depth of cut [Owei89]

The optimum abrasive mass flow rate for the maximum cutting performance (or for the maximum depth of cut) depends on many parameters. These are the water pressure [Chal91], [Guo94a], [Guo94b], orifice diameter [Chal91], [Guo94b], the focusing tube diameter [Chal91], [Guo94a], [Hoog05] and the focusing tube length [Guo94a].

Table 2.1: Optimum abrasive to water mass flow rate [Chal91]

Nozzle/orifice combination ma/mw for hmax ma/mw for 0.85 h⋅ max

0.76 mm / 0.25 mm 0.3 0.17

1.14 mm / 0.38 mm 0.19 0.12

1.65 mm / 0.53 mm 0.19 0.1

Figure 2.7 shows the effects of jet-parameters on the optimum abrasive mass flow rate according to experimental data of Guo [Guo94a]. It follows that the optimum abrasive mass flow rate increases with the increase of the water pressure (Figure 2.7a), of the water mass flow rate (Figure

(36)

mass flow rate and the focusing tube length is shown in Figure 2.7d.

To determine the optimum abrasive mass flow rate, E.J. Chalmers [Chal91] found that cutting with the ratio of nozzle to orifice of 3:1 results in the maximum depth of cut at a specific value of

/

a w

m m for a given size of the nozzle. In addition, to avoid excessive use of abrasives, Chalmers [Chal91] assumed the optimum depth of cut is defined as occurring at 0.85hmax. The optimum abrasive to water flow rate is shown in Table 1 [Chal91].

100 150 200 250 300 350 5 5.5 6 6.5 7 7.5 8 8.5 9 Water pressure p w (MPa)

Optimal abrasive mass flow rate (g/s)

d ori=0.25; df=0.95 mm v f=1.67 mm/s; dp=355 µm minersiv 15 20 25 30 5 5.5 6 6.5 7 7.5 8 8.5 9

Water mass flow rate (g/s)

p w=240 MPa; vf=1.67 mm/s d f=0.95 mm; dp=355 µm minersiv a) b) 0.5 0.7 0.9 1.1 1.3 1.5 5 6 7 8 9 10 11 12

Focusing tube diameter d f (mm)

p w=240 MPa; v=1.67 mm/s d ori=0.25 mm; df=0.95 mm d p=355 µm; minersiv 20 30 40 50 60 70 80 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10

f (mm)

p w=240 MPa; v=1.67 mm/s d ori=0.25 mm; df=0.95 mm d p=355 µm; minersiv c) d) Figure 2.7: Effect of factors on the optimum abrasive mass flow rate [Guo94a]

(37)

In spite of recent efforts, the optimum abrasive mass flow rates are predicted for specific combinations of the focusing tube and the orifice diameter only. No model has been developed for determination of the optimum abrasive mass flow rate for more general combinations.

2.1.4 Optimum abrasive particle size

The effect of abrasive particle sizes on the depth of cut was investigated by J. Ohlsen (Figure 2.8a) [Ohls97]. This effect can be classified into two cases: brittle behaving materials (e.g. glass in Figure 2.8a) and ductile behaving materials (e.g. AlMgSi0.5 in Figure 2.8a). In the former case, the workmaterial seems less sensitive to the impact frequency [Momb98]. Therefore, the maximum depth of cut increases with the increase of the particle diameter. In the latter case, for small particles (smaller than 100 μm), a larger particle diameter causes a higher depth of cut. This is because a larger particle means a higher kinetic energy, i.e. ∝ 3

p p

E d [Momb98]. In contrast, for larger particles (larger than 100 μm), an increase in size of abrasive particles can lead to a reduction of the maximum depth of cut (Figure 2.8a). Momber et al. [Momb98] noted that this phenomenon was also observed by Nakamura et al. [Naka89], Guo et al. [Guo92] and Momber et al. [Momb96].

H. Oweinah [Owei89] investigated the effect of the abrasive particle diameter on the depth of cut for various abrasive mass flow rates (Figure 2.8b). It is concluded that large particles have a significant influence on the depth of cut when the abrasive mass flow rate varies, while smaller particles are not sensitive against the changes of the abrasive mass flow rate (Figure 2.8b).

0 100 200 300 400 500 600 15 20 25 30 35 40 45

Abrasive particle diameter d

p (g/s)

max (mm) AlMgSi0.5 Glass p w=300 MPa (AlMgSi0.5) p w=100 MPa (glass) d ori=0.25; df=0.9 mm v=1.67 mm/s; m a=5 g/s garnet 0 5 10 15 20 0 10 20 30 40 50 60 70 80

max (mm) d_p=0.25 mm d p=0.029 mm p w=400 MPa; vf=0.83 mm/s d ori=0.25; df=1.08 mm AlMgSi1; corundum a) [Ohls97] b) [Owei89]

(38)

2.1.5 Optimum standoff distance

The effect of the standoff distance on the maximum depth of cut was first investigated by R.E. Barton [Bart82]. It was found that the depth of cut decreases almost linearly with the increase of the standoff distance. Figure 2.9 shows the relationship between the standoff distance and the maximum depth of cut [Blic90]. R.A Tikhomirov et al. [Tikh92] reported the same result for the relation between the standoff distance and the maximum feed speed. The authors noted that at a small increase of the standoff distance, the maximum feed speed first remained constant and then decreased according to an almost linear relation [Tikh92]. The effect of the standoff distance on the depth of cut was also confirmed by Blickwedel [Blic90], Kovacevic [Kova92] and Guo et al. [Guo94b]. In addition, Guo et al. suggested that the optimum standoff distance is about 2 mm [Guo94b]. 0 10 20 30 40 50 5 10 15 20 25 30 35 40 45 Standoff distance (mm)

max (mm) p_w=300 MPa p_w=200 MPa d ori=0.25; df=1.2 mm l f=50 mm; vf=1.67 mm/s m a=8 g/s

Figure 2.9: Standoff distance versus maximum depth of cut [Blic90]

2.2 State of the art in AWJ cost calculation and cost optimization

2.2.1 State of the art in AWJ cost calculation

• Study of J. Zeng and T. J. Kim:

To calculate the cutting cost per length, J. Zeng and T. J. Kim [Zeng93] first introduced a model for prediction of the feed speed:

⎛ ⋅ ⋅ ⋅ ⎞ = ⎜_⎜ _{⋅ ⋅ ⋅} ⎟_⎟ ⎝ ⎠ 1.15 1.25 0.687 0.343 0.618 m w w a f f N p m m v C q h d (2.4)

(39)

The cutting cost per length was then determined by the following equation: = h l f C C v (2.5) Where, Ch is the total hourly cost ($/h), which is calculated as follows:

= + + + +

h mh lh th ph dh

C C C C C C (2.6) In which Cmth is the machine hourly cost, Clh is the labor hourly cost, Cth is the material hourly cost which considers the abrasive cost, water cost, focusing tube cost and orifice cost, Cph is the power hourly cost, and Cdh is the cost of maintenance and disposal.

In this study, many cost components were taken into account. Additionally, the effects of many jet-parameters on the cutting cost were also investigated through a model for prediction of the feed speed. After all, the effect of the number of jet formers as well as the effect of the nozzle wear on the cutting cost was still not well-understood.

• Study of D.A. summers et al.:

To compare the AWJ cutting cost per part in both cases with and without abrasive recycling, D.A. Summers et al. [Summ01] carried out a study in which the influence of many cost parameters on the total cutting cost, e.g. the abrasive cost, the disposal cost, the power cost, the water cost and the nozzle wear cost were investigated. Also, the optimum cutting performance was predicted by a tabulated method. The authors concluded that by cutting with the recycled abrasives using particles larger than 100 µm, the cutting cost can be reduced significantly. Nevertheless, the effects of several cost components such as the machine cost, the labor cost and the maintenance cost were not considered. Besides, although the nozzle wear cost was taken into account empirically, there is still no model for calculation of the nozzle wear.

• Study of M. Hashish:

M. Hashish [Hash04] compared the cutting cost in two cases: with water pressure of 400 MPa and 600 MPa. The effects of the water pressure on different cost elements such as abrasives, pump and machine maintenance, water, power and the nozzle wear were studied.

Hashish noted that the feed speeds when cutting at 600 MPa should be at least equal to those at 400 MPa while using 33% less abrasives and water. The author also found that cutting at a pressure of 600 MPa can save the total cutting cost 10 % to 25 % over that at 400 MPa [Hash04]. Although the study compared the cutting cost when cutting with high and low pressure, the effect of water pressure on the AWJ system’s utilization and the nozzle wear were not investigated. In practice, cutting with high pressure can increase not only the pump maintenance cost but also the downtime due to the pump’s maintenance. As a result, the total available cutting time when cutting

(40)

due to the high pressure, the nozzle wear also increases (in the study the nozzle wear was constant) and leads to an increase of both the nozzle wear cost and the downtime because of replacement of the nozzle. These effects of high pressure on the utilization and on the nozzle wear cost should therefore be taken into account.

2.2.2 State of the art in AWJ cost optimization

• Study of P. J. Singh and J. Munoz:

P.J. Singh and J. Munoz [Sing93] noted that the AWJ cost optimization problem is very complicated to solve because there are a lot of parameters affecting the total cutting cost. However, local sub-optimization can be used as a solution for the problem. For cost analysis, the authors divided the cost elements into three main components: the operating costs, the labor costs Cl, and the capital investment costs Ce [Sing93]. The operating costs consist of the abrasive cost Ca, the power cost Cp, the water cost Cw, the focusing tube cost Cf, orifice cost Cori, and maintenance cost Cmai. The total cutting cost per centimeter is then determined as follows [Sing93]:

⋅ = ⋅ ⋅ , 10 60 h l c f cf C C v k (2.7) Where, vf is the feed speed (mm/s) calculated by the model by Zeng and Kim (Equation 2.4); kcf is

a contour factor which considers the necessary slow down of the system during turns; Ch is the

total cutting cost per hour which is calculated by the following equation [Sing93]:

= + + + + + + +

h a p w f ori mai l e

C C C C C C C C C (2.8) A sub-optimization problem is performed by considering the orifice diameter as an independent variable. Other parameters are then chosen based on this variable [Sing93]. From the results of the optimization problem, the authors concluded that use of smaller orifices is more cutting efficient, i.e. the cutting length per unit of power is higher. However, larger orifices are more cost efficient as the feed speed can be increased so that the labor cost and the capital cost are reduced [Sing93]. In addition, the authors found that cutting with multiple-heads reduces the total cutting cost since the combination of the higher efficiency of smaller orifices with higher throughput of multiple-heads [Sing93]. It is noted that careful cost analysis and cost optimization can save 10 to 30% of the total AWJ cutting cost [Sing93].

As the above sub-optimization study was carried out by considering only one variable, the applications are therefore limited. The effect of process parameters as well as the effect of cost elements on the total cutting cost should be taken into the sub-optimization problem.

(41)

J. Zeng and J. Munoz [Zeng94] presented a study on optimum selection of the abrasive mass flow rate in order to fulfill the minimum cutting cost. In this study, the total cutting cost per length

l

C ($/m) is calculated according to the approach of Zeng and Kim [Zeng93]:

+ ⋅ ⋅ = ⋅ , 60 60 h a a m l f C m C C v (2.9) Where, Ch is the total hourly cost ($/h) excluding abrasive cost; vf is the feed speed (m/min); m is a

the abrasive mass flow rate (kg/min) and Ca is the abrasive cost per kilogram ($/kg).

The total cutting cost per unit length Cl1 when cutting with ma1and vf1 can be compared to Cl2,

when cutting with m_a₂, vf2 by the following equation [Zeng94];

+ ⋅ ⋅ = ⋅ + ⋅ ⋅ 1 1 1 1 1 2 2 2 2 2 60 60 a a l f l a a f C m C C v C C m C v (2.10) The ratio Cl1/Cl2 was determined for three different combinations of the orifice and focusing tube

diameter with Ch1=Ch2=$62 and Ca1=Ca2=$0.59/kg ($0.32/lb) (see Table 2.2) [Zeng94].

Table 2.2: Relative cost using different abrasive mass flow rates [Zeng94]: Orifice/tube combination 0.113 kg/min 0.227 kg/min 0.34 kg/min 0.454 kg/min 0.567 kg/min 0.68 kg/min 0.177/0.584 1.04 1.00 0.254/0.81 1.23 1.00 1.00 0.356/1.12 1.31 1.11 1.00 1.02 1.01

The optimum abrasive mass flow rates found for various combinations of the orifice and focusing tube diameter are 0.227 kg/min for 0.177/0.584, 0.34 kg/min for 0.254/0.81, and 0.454 kg/min for 0.356/1.12 [Zeng94].

A main advantage of this method is that the optimum abrasive mass flow rate can be determined rather easily. However, the results are valid for pre-set combinations of orifice and nozzle diameter only. Also, the effects of the abrasive mass flow rate on the nozzle wear and on the abrasive disposal cost were not investigated. Moreover, the effects of cost elements on the optimum values of abrasive mass flow rate should also be taken into account.

• Study of M. Mono:

For solving the AWJ cost optimization problem, M. Mono [Mono97] first introduced graphical relations between the water pressure, the abrasive mass flow rate, the feed speed and the surface roughness. In his cost model, various cost elements were taken into account such as the abrasive

(42)

surface roughness, graphical relations between the depth of cut, the ratio between the feed speed and the abrasive mass flow ratev_f /m_a, the depth of cut and the minimum cutting cost were constructed.

After all, the use of the graphical relations for selecting AWJ parameters is not straightforward. More important, the approach is only valid for cutting with aluminum [Mono97] and at fixed values of water pressure and surface roughness.

• Study of A. Henning and E. Westkämper:

To find the optimum values of the abrasive load ratio R for getting the maximum cutting performance and the minimum cutting cost per meter, A. Henning and E. Westkämper [Henn04] gave a cost study in which many cost components were taken into account. These cost elements consist of the electricity cost, the water cost, the abrasive cost, the labor cost, the occupancy cost, the nozzle wear cost, the orifice cost, the revenue etc. It was found that the optimum values of the abrasive load ratio for the maximum cutting performance and for minimum cutting cost are different (see Figure 2.10). The authors concluded that, in many cases, an abrasive load that is between these optimum points can be chosen as the optimum abrasive load ratio.

0 5 10 15 20 25 30 35 40 0 2 4 6 8 10 12 14 16 18

Abrasive load ratio R (%)

Cutting performance (m/h) Cutting cost (

/m) 0 4 8 12 performance cost

Figure 2.10: Cutting performance and cutting cost versus abrasive load ratio [Henn04] The effect of the hydraulic power on the cutting performance as well as on the cutting cost per meter was also investigated in the study. It was found that the optimum cost to abrasive load ratio decreased when cutting with higher hydraulic power [Henn04]. Also, it was noted that the optimum abrasive mass flow rate never exceeded 1.3 kg/min (within experiments in the study). Moreover, in the study, the cutting cost calculation when cutting with multiple cutting heads was carried out. The authors found that it is possible to gain more benefits from high power by cutting with multiple

(43)

cutting heads [Henn04].

• Study of U. Andersson and G. Holmqvist:

Recently, U. Andersson and G. Holmqvist [Ande05] have carried out a study on strategies for cost and time effective AWJ cutting. In their cost structure, the cost elements are classified into two groups: fixed costs and running costs. The fixed costs include the AWJ system cost (including software) and the labor cost. The running costs consist of the abrasive cost, the water cost, the electricity cost, the cutting head cost (including nozzle, orifice, valve etc.), and the maintenance cost. In addition, the influence of other factors such as the utilization, the economic life of the AWJ system and the interest are also considered.

Andersson and Holmqvist found that the fixed cost is usually half or up to two thirds of the total cutting cost per unit length. They also noted that cutting with two cutting heads instead of one can reduce the cutting cost significantly.

It is noted that three factors have to be considered to reduce the total cutting cost. These are the optimized abrasive mass flow rate, the optimized lifetime of cutting head consumables (focusing tube, orifice, valve etc.), and optimized water pressure [Ande05]. However, in their cost structure, the effect of these factors on the cutting cost was still neglected.

2.3 State of the art in AWJ abrasive recycling

In the AWJ cutting process, the breaking (or the fragmentation) of abrasive particles occurs in two stages: first, during the mixing process (due to interactions between particles and the walls of the mixing chamber and the focusing tube and between particles with each other), and second, during the cutting process (because of the interactions between particles with the workmaterial and particles and each other). Therefore, understanding of the fragmentation of abrasive particles is highly relevant to a study on abrasive recycling.

The fragmentation of abrasive particles has been studied intensively. G. Galecki and M. Mazurkiewicz [Galec87] were the first who studied the fragmentation during the mixing process. The authors found that a large number, i.e. 70 to 80%, of initial particles are disintegrated during the mixing process [Galec87]. They also noted that this number depends on the initial abrasive size, the water pressure, the abrasive mass flow rate and the focusing tube diameter.

T.J. Labus et al. [Labu91] carried out a fundamental research in which the influence of the process parameters on the particle size distribution after the mixing process and after the cutting process was investigated. It was found that low water pressure levels (from 0 to 205 MPa) can have more affect on the main mass fraction change than those by high pressure levels (from 274 to 342 MPa).

(44)

mixing chamber geometry does. The authors noted that for a garnet #80, the main particle breakdown process is a shift from 180 micron particles into 63 micron particles or less. Particles which have the size from 75 to 150 microns do not seem to be affected during the cutting process. The workmaterial thickness also influences the particle disintegration. It was concluded that the recycling is more applicable for thin workmaterials than for thick ones, since more of the main abrasive mass fraction remains intact [Labu91].

H. Louis et al. [Loui95] investigated the effect of cutting parameters on the particle size distribution after the cutting process. The average particle size after cutting is found to be a bit smaller than that after the focusing tube. The influence of workmaterial types on the fragmentation of the abrasive particles was also investigated in this study. The authors noted that cutting stainless steel can reduce the average particle size more than when cutting aluminum. Also, the effect of the abrasive material on the fragmentation of particles after the cutting process was investigated with two types of abrasives (garnet and olivine). It is observed that olivine produces a bit smaller average particle size than garnet [Loui95]. Finally, the effect of the cutting quality was discussed. The authors found that high quality cutting causes a bit smaller average particle size than rough cutting [Loui95].

J. Ohlsen [Ohls97] carried out a systematic study on the recycling of Barton garnet. To evaluate the fragmentation of the abrasive particles, Ohlsen introduced a “disintegration number” which is defined as follows: φ = − , , 1 ap out D ap in d d (2.11) In which, dap,in and dap,out are the average diameter of input and output particles, respectively.

There are many process parameters that affect the magnitude of the disintegration number, for example, the water pressure, the abrasive mass flow rate, the abrasive particle diameter, the focusing tube diameter and the focusing tube length. The effects of these parameters are discussed below (see also Figure 2.11a through 2.11f).

It was observed that the disintegration number increases linearly with the water pressure (see Figure 2.11a). The abrasive mass flow rate affects the particle disintegration significantly only when this rate is smaller than a certain value (4 g/s). Above this value the influence on the particle fragmentation is negligibly small (Figure 2.11b). Figure 2.11c shows an almost linear relation between the initial particle diameters and the disintegration number. Figure 2.11d describes a monotonous decrease of the disintegration number with the increase of the focusing tube diameter.

(45)

100 150 200 250 300 350 400 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Water pressure p w (MPa) Disintegration number (−) d p=0.045−0.063 mm d p=018−0.25 mm d p=0.5−0.71 mm d ori=0.25; df=0.9 mm m a=5 g/s; garnet 0 2 4 6 8 10 12 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5

Disintegration number (−) d ori=0.25; df=0.9 mm p w=300 MPa; garnet a) b) 0 0.2 0.4 0.6 0.8 0.1 0.2 0.3 0.4 0.5 0.6

Abrasive particle diameter d p (mm) Disintegration number (−) d ori=0.25; df=0.9 mm p_w=300 MPa; m_a=5 g/s 0.5 1 1.5 2 0.2 0.25 0.3 0.35 0.4 0.45 0.5

Focusing tube diameter d f (mm) Disintegration number (−) d ori=0.25; df=0.9 mm m a=5g/s; pw=300 MPa d_p=0.18−0.25 mm c) d) 100 150 200 250 300 350 400 110 120 130 140 150 160 170 180 Water pressure p w (MPa)

Outlet abrasive particle diameter (µm)

Conventional chamber design Optimized chamber design

d_ori=0.25; d_f=0.9 mm m a=5 g/s; dp=0.18−0.25 mm 20 40 60 80 100 0.1 0.2 0.3 0.4 0.5 0.6

Focusing tube length l f (mm) Disintegration number (−) d ori=0.25; df=0.9 mm m a=5 g/s; pw=300 MPa d p=0.5−0.71 mm e) f) Figure 2.11: Effect of parameters on the abrasive particle disintegration [Ohls97]

(46)

The effects of the focusing tube geometry (Figure 2.11e) and the focusing tube length (Figure 2.11f) on the fragmentation are small. Although the focusing tube length increases 5 times, the disintegration number increases only about 10%.

J. Ohlsen [Ohls97] reported that particles smaller than 60 µm lead to a very small depth of cut, poor cutting quality and can cause abrasive clogging in the mixing head. Moreover, the author found that the cutting performance and the cutting quality of the recharged abrasives are slightly better than those of the new abrasives. The reason is that the particle size distribution of recycled abrasives lies in the range from 125 to 180µm. This range of the particle size can lead to the maximum depth of cut and a lower surface roughness.

M. Kantha Babu and O.V. Krishnaiah Chetty [Babu03] introduced a study on the recycling of a local garnet (origin: Southern India). The authors found that the reusability (or the recycling capability which is determined by the percentage of abrasives that can be reused) with the particles larger than 90 µm is 81, 49, 26 and 15% after the first, second, third and fourth recycling, respectively [Babu03]. The effect of recycled abrasives of three cycles on the depth of cut, on the surface roughness and on the kerf width was investigated. It was observed that the maximum depth of cut of the first and second recycled abrasives is approximately 82 and 79% of the new abrasives. Also, cutting with the first and the second recycled abrasives can reduce both the surface roughness and the kerf taper [Babu03].

In practice, after recycling, the abrasives (recycled abrasives) can be used as a new abrasive or used as addition to new abrasives. The process in which new abrasives are added to recycled abrasives is called abrasive recharging. The recharging aims at maintaining the amount of input abrasives, so as to increase the cutting performance or to maintain the maximum cutting performance at all times.

M. Kantha Babu and O.V. Krishnaiah Chetty [Babu02] carried out a study on abrasive recharging. In their study, the recycled abrasives (with the size more than 90 µm) were recharged with new of abrasives at 20, 40, 60, 80 and 100% of the recycled abrasive mass. The influence of the recharging on the depth of cut, on the surface roughness, and on the kerf width for cutting with aluminum was investigated. It was noted that an increase of the added new abrasives up to 40% led to a significant increase of the depth of cut and a slight increase thereafter [Babu03]. Consequently, for getting maximum depth of cut, the recharging at 40% of the recycled abrasive mass is recommended [Babu03]. It is found that the surface roughness is minimum at 60% recharging of recycled abrasives with the size larger than 90 µm. Also, the top and bottom kerf widths increase marginally when the amount of added new abrasives increases [Babu03].

Performance Enhancement of Abrasive Waterjet Cutting

Performance Enhancement

of Abrasive Waterjet Cutting

Performance Enhancement

of Abrasive Waterjet Cutting

Proefschrift

Acknowledgement

Summary

Samenvatting

Contents

Nomenclature

1

Introduction

1.1

Historical review

1.2

Introduction to AWJ Technology

1.2.1

Introduction to an AWJ cutting system

1.2.2

Parameters of an AWJ machining process

1.2.3

Advantages and disadvantages of AWJ Technology

1.3

Challenges in AWJ Technology

2

State of the art in optimization of AWJ machining

2.1

State of the art in AWJ technical optimization

2.1.1

Optimum combination of focusing tube and orifice diameter

2.1.2

Optimum focusing tube length

2.1.3

Optimum abrasive mass flow rate

2.1.4

Optimum abrasive particle size

2.1.5

Optimum standoff distance

2.2

State of the art in AWJ cost calculation and cost optimization

2.2.1

State of the art in AWJ cost calculation

2.2.2

State of the art in AWJ cost optimization

2.3

State of the art in AWJ abrasive recycling