Simulation ofAbrasive Water Jet Cutting Based on Unit Event Features

1
Simulation ofAbrasive Water Jet Cutting Based on
Unit Event Features
Andrej Lebar, Miha Junkar University of
Ljubljana, Slovenia

• AWJ cutting, possibilities and limitations
• existing AWJ simulations
• the proposed model
• experimental work
• model implementation and results
• conclusions

2
CUTTING WITH ABRASIVE WATER JET
PRESSURE 276 MPa PUMP POWER 15 kw
CUTTING SPEED (up to) 4600 mm/min WATER
CONSUMPTION 3.8 l/min
3
AWJ cutting, possibilities and limitations
material stacking, 2000
cutting bread, 1999
composite tubes Anhovo, 1995
composite mat., 1999
tool renewal, 1999
drilling hard mat., 2000
4
AWJ - history

• Before 50s several experiments in gold ore
  mining
• 50s Russian water cannon in Syberia mining
• 60s pulsed water jet at 700 MPa
• (Exotech Inc., Virginia, Terraspace Inc),
  industrial experiments
• (O. Imanaka Tokyo) wood cutting (R.Franz,
  Univ. of Michigan)
• 1971 Commercial high pressure water pump (Flow
  Research)
• 1971 Industrial app. cutting of laminated paper
  tubes (Alto Boxboard)
• 1981 Invented and patented abrasive water jet
  (Waterjet Tech. Inc)
• 1990 Invented and patented abrasive waterjet
  milling (WTI., Inc)
• 1991 Suspension jet
• 1995 Cutting with liquid nitrogen (WTI., Inc and
  Praxair)
• 1999 700 Mpa AWJ

5
Existing simulations I.
Fukunishi et al. 1995
6
Existing simulations II.
Yong, Kovacevic, 1997
7
Existing simulations III.
Friedrich et al. 2000
8
PROPOSED MODEL
9
Experiments
10
Unit volume removal g
abrasive grain impact angle
11
UNIT EVENT MASS REMOVAL vs. IMPACT ANGLE AT
DIFFERENT WATER PRESSURE
THEORY

• MEASURED

12
CRATER SIZE vs. IMPACT ANGLE
LONGER ELLIPSOID AXIS IS IMPACT ANGLE DEPENDENT
SHORTER ELLIPSOID AXIS IS INDEPENDENT OF IMPACT
ANGLE
CRATER DEPTH CAN BE CALCULATED FROM ANGULAR
DEPENDENCE OF UNIT EVENT MASS REMOVAL AND CRATER
SIZE MEASUREMENTS
13
MULTIPLE REBOUNDS
14
abrasive grain rebound model (i)
15
abrasive grain rebound model (ii)
16
abrasive grain rebound algorithm
17
ABRASIVE GRAINS DISTRIBUTION

• abrasive grains, recognised objects after image
  processing and fitted beta function

18
POSITIONS OF PRIMARY IMPACTS

• it is assumed that
• traversal velocity of the mixing tube is
  constant and
• abrasive grains are uniformly distributed across
  the mixing tube

19
SIMULATION (i)

• simulation after first 500 unit events

20
SIMULATION (ii)
SAE Al6061-T6

• AWJ machined surface and simulation result.
  Strong striation texture can be observed on both
  surfaces.

21
SIMULATION (iii)
experiment, Momber 98
TOP
roughness
BOTTOM

• another phenomenon is observed in simulation
  step formation

22
MODEL EVOLUTION