Impact Wear Tester for the Sudy of Abrasive Erosion and Milling Processes

6th International DAAAM Baltic Conference INDUSTRIAL ENGINEERING 24-26 April 2008, Tallinn, Estonia

IMPACT WEAR TESTER FOR THE SUDY OF ABRASIVE EROSION AND MILLING PROCESSES

Tarbe, R.; Kulu, P.

Abstract: Impact wear is common in crushing and mining equipment. For the 1. INTRODUCTION determination of wear resistance of 1 materials in such conditions the centrifugal According to [ ], wear is defined as either type mill based on a disintegrator was mass or volume of material, removed or developed. Results obtained by the tester displaced from a body which is repeatedly enable to analyse impact wear of various stressed in mechanical contact with another materials as well as grindability and body or bodies. abrasivity of used materials Abrasion is sometimes indiscriminately simultaneously. This paper focuses on the associated with erosion, so that the erosive particles are often referred to as development of the testing route and 1 parameters for the tester to obtain results “abrasives” [ ]. with tolerable uncertainty. Three different Erosive wear in which the relative motion groups of materials: steels, cermets and of the solid particles is nearly normal to the solid surface is called impingement erosion coatings were tested. The velocity of the 2 abrasive particles was varied from 40 m/s or impact erosion [ ]. To constraint this to 80 m/s and impact angle was close to term is good to claim additionally that normal. Used abrasive was granite gravel, impact wear is present then abrasive classified according to the standards EN particles are bigger than 2 mm. 13043, EN 13450 and EN 12620. Mechanical contact between solids may Additional tests with different abrasives result from various basic models of relative were made for comparison. The size of motion and impact is one of them. Other abrasive particles was 4/5.6 mm in all models are sliding, and rolling. In each tests. case, large contact stresses may arise, but The tests proofed that cermets have their character, distribution and variation in superior wear resistance in such time are unique. Fig. 1 shows wear charac- conditions, but their uncertainty may be terization based on contact stresses. rather big. Feasible amount of abrasive for Impact wear is characterized by narrow the most of materials to obtain results with and high peak of stress because of very moderate uncertainty was concluded to be short duration of impact. The force F is 15 kg. very big and this is the reason for using The grindability tests demonstrated that the momentums instead of forces in main size reduction of mineral materials as kinematical calculations of impacts: t f brittle materials takes place after impact 3 with 1st circle of pins. I = ∫ Fdt = Δp [ ], where Key words: impact wear, abrasive wear, ti wear resistance, uncertainty, grindability Δp = pf – pi, p – momentum of a particle, I – momentum of the particle, ti – time initial, not for investigating wear phenomena of a) b) working parts. τmax τmax a) b)

t t c)

τmax Fig. 3. Testers using centrifugal acceleration phenomena: a − disintegrator;

Fig. 1. The variationt of the maximum shear b − centrifugal accelerator CAK stress at a point, in three types of contact: a − sliding (the shear stress on the slider is The force of centrifugal acceleration is exploited too in a centrifugal accelerator shown); b − rolling; c − impact [1] tester CAK, developed by I. Kleis [4]. However, it is not impact wear tester tf – time final. Test equipments can be distinguished by because its erosive particles must be the way how the erosive particles are smaller than 2 mm. accelerated in them. The main options are The aim of this work was to develop by gas stream, gravitational acceleration testing machine where is possible to test (Fig. 2) or centrifugal acceleration (Fig. 3). two-body wear process in the conditions of Gas stream acceleration (Fig. 2 b) is very impact wear as well as grindability of common and used by many researchers in abrasive materials to be used. different countries and is the basis for 2. TEST METHODS German standard DIN 50332. 2.1 Impact wear tester a) b) Impact tester DESI is centrifugal type and bases on a disintegrator. The main α components are 2 rotors rotating at opposite directions (Fig. 4).

1st rotor 2nd rotor

α Fig. 2. Erosion testers where particles are accelerated: a − by a gravitation; b − by a 3 gas stream [ ]

Abrasive Centrifugal acceleration principle is used in in disintegrator mills. Erosive particles in 2 disintegrators can be relatively big. Their 3 maximum size depends from the model of a disintegrator and can be more than 10 Fig. 4. The rotors of impact wear tester mm. Their main disadvantage is that they DESI in exploded view. 1 – 1st circle of are built for refining abrasive material and pins; 2 – specimen; 3 – 2nd circle of pins

Crushing pins with the specimens are rotors turn with different angular attached onto them. The attachment of the velocities, is possible to have 20 additional specimens is unique and protected by the speed combinations. Estonian utility model no U200600001. Abrasive enters into the center of 1st rotor. After that it slides on impellers (Fig. 5) and 3 gains velocity because of centrifugal force. 2 1 2 0 700 61

v ~900

3 1st circle Fig. 6. Impact angle between particles 2nd cicrle coming from impellers and 1st circle of Fig. 5 Simplified scheme of particles pins: 1 – pin, 2 – specimen, 3 – impeller, v moving in DESI: 1 – pin (specimen – resultative velocity holder), 2 – specimen, 3 – impeller 2.2 Erosion wear resistance calculation In next step it hits the pins/specimens of The specimens are weighed before and the first circle. The 1st ring of specimens after each test. Mass losses of the contains 14 pins or in other words, specimens are determined. The mass loss specimen holders. Two places are occupied of each specimen is divided by the working by the reference specimens, i.e. 12 pins are zone area of the specimen: mass loss to free for specimens of materials to be tested. unit surface area is obtained. The outcome Impact angle between 1st circle of is divided by the density and volume loss is specimens and abrasive particles obtained, which is compared with volume determined partly by the calculations and loss of reference material. This is the route graphical method (Fig. 6), is close to for calculating relative wear resistance (eq. normal. Impact angle between 2nd circle of 1). specimens and abrasive particles ΔV ε = r , where (1) originating from 1st circle of specimens ΔV cannot be determined so easily. The ε - relative wear resistance of material problem is that the impingement from the under investigation; 1st circle of pins adds additional uncertainty factor, because the shape and ΔV - average volume loss per unit surface the coefficient of restitution varies from area of material under investigation; particle to particle. That is why 2nd circle ΔVr - average volume loss per unit surface of pins is optional, i.e. these pins can be area of reference material. Δm Δm removed for the test. ΔV = = , where (2) Angular velocity of rotors is changeable by Aρ bhρ the belt drive mechanism. Possible is to Δm – average mass loss of material under have 5 different angular velocities on both investigation rotor, which gives impacting velocities A – average active surface area of the between abrasive particles and 1st circle of specimen (A = b*h, where b –width of the specimens from 40 m/s to 120 m/s. If both specimen, h –height of the specimen). The formula for calculation of relative Hard rock contains 100 vol % of quartzite. wear resistance is following Cumulative and distribution polygons of Δm bhρ bhΔm ρ used abrasives are given in Fig. 8. They are ε = r = r (3) b h ρ Δm b h Δmρ based on sieve analysis. r r r r r r a) We started by searching of reasonable amount of abrasive from 3 kg per batch. It 100 proved to be too less because the 90 uncertainty of the results was too big. % 80 Finally we determined reasonable amount 70 of abrasive to be 15 kg. In most cases such 60 1 amount of abrasive gives good results with 50 2 small uncertainty. Exceptions are 40 especially materials of very high wear 30 20 resistance, like hardmetals. Then gravel size), F(rock uncertainty can still be very big (even up to 10 50% of their wear resistance value), but it 0 is explainable by brittle materials wear 0.01 0.10 1.00 10.00 nature. They are prone to spalling. Rock gravel size From each material is recommended to test 4 specimens in each batch. b) Abrasive feed rate depends slightly from the abrasive and is usually kept 3 kg/min. 100 90

2.3 Grindability % 80 The size of abrasives usable in DESI can 70 60 be up to 7 mm. 2 1 Used gravels (Fig. 7) were 4/5.6 mm in 50 particle size. Granite gravel is classified 40 according to the standards EN 13043, EN 30

13450 and EN 12620. gravel size), f(rock 20 Hard rock was used in additional tests. 10 0

0.01 0.10 1.00 10.00 Rock gravel size

Fig. 8. Cumulative (a) and distribution (b) polygons of tested abrasives: 1 – granite gravel, 2 – rock gravel

3. RESULTS AND DISCUSSION

3.1 Wear resistance study All tests were made at the velocity of 60 Fig. 7. Overall picture of abrasives: left m/s between 1st circle of specimens and side – granite gravel; right side – hard rock abrasive. 2nd circle of specimens was not in use. The chemical composition of granite gravel As indicated by the studies of hardmetals is 70 vol % of quartz, 10 vol % of feldspar and cermets of various type and and 20 vol % of other ingredients. The composition, their behaviour differ under impact value SZ is 18. studied conditions of wear. The results of testing – relative wear resistance of the Their main problem is too weak bonding WC-Co hardmetals, TiC-NiMo and Cr3C2- strength. The coatings will spall from the Ni cermets are given in Fig. 9. It compares substrate. wear in DESI with wear in CAK. Hardmetal coating WC-17Co (1343V), In the case of abrasive erosion more wear which worked very well in CAK, was resistant is WC-8Co hardmetal, in case of outperformed by fused NiCrSiB coating impact wear – WC-15Co hardmetal. The (1275H). Fused coatings having better high binder content (30 wt%) TiC based bonding are more wear resistant. cermets demonstrated also good resistance Well-known wear resistant steels were in impact wear conditions. The TiC-based tested in DESI (Fig. 11). Their relative cermets are preferable, especially wear resistance was practically the same considering the fact that their density is compared to reference steel 45 and did not lower compared to hardmetals and weight depend considerably from their hardness: wear rate is on the level of hardmetals. ** − the specimens were not heat treated; *** − average hardness after heat treating a) Abrasive erosion b) Impact wear 60 was 42 HRC; **** − average hardness N 50 after heat treating was 46 HRC

1.2 1.1 40 1.0 1.1

ε 1.0 1.0 0.9 30 0.8 0.8 20

0.6 10

Relative wear resistance, resistance, wear Relative 0.4 0

0.2

i Relative weighted wear resistance, o Ni o C iMo 0N 0 iMo NiMo 2 -8C 0N 0 - C 0N 2 0.0 WC-8Co 3 -5 W C-15Co -5 WC-15iC- iC r3C W iC T T C Cr3C2-3 TiC-30NiMoT Cr3C2-20Ni Hardox 400 Hardox 600 Vanadis 6** Vanadis 6*** Weartec** Weartec**** Fig. 9. Relative wear resistance of Fig. 11. Relative wear resistance of steels hardmetals and cermets at abrasive erosive in DESI. (a) and impact wear (b) 3.2 . Grindability study The relative wear resistance of coatings Depending on the particle composition and made by Tafa’s HVOF system JP-5000 and initial defects, a significant size reduction compared with CAK tests is given in Fig. takes place (Fig. 12). The initial particle 10. As it follows from Fig. 10, the coatings do not suit into high-energy impact a) conditions presented in DESI. 100.0 a) Abrasive erosion b) Impact wear 90.0 12 % N 80.0 10 2 1 70.0 8 60.0 6 50.0 3 4 40.0 2 30.0 Relative wear resistance, resistance, wear Relative 0 20.0

5 V H gravel F(granite size), 9 3 5 -Co 4 7 343V 275H 2 2 10.0 12495 1 1 1 134 1 0.0

2495+25WC 1 0.01 0.10 1.00 10.00 Fig. 10. Relative wear resistance of powder Granite gravel size coatings at abrasive erosion and impact wear b) proofed that the increase of particles velocity 2 times decreases their size d50 4 100 times (compare curves 1 and 3 in Fig. 13). 90 . It correlates with kinetic Work-energy 80 theorem [3]. 70 60 4. CONCLUSIONS 50 40 3 2 30 1 • The main advantage of impact wear tester 20 DESI compared to erosive testers is f(rock gravel size), % size), gravel f(rock 10 simultaneous wear study of various 0 materials as well as of grindability of 0.01 0.10 1.00 10.00 abrasive materials. Rock gravel size • Reasonable amount of abrasive is 15 kg. Such amount of abrasive gives results with good uncertainty. Fig. 12. Cumulative (a) and distribution (b) • DESI suits perfectly for testing polygons of grinded abrasives: 1 – granite hardmetals and cermets. In the case of gravel after impact with 1st circle of pins; steels and coatings, their wear resistance 2 – granite gravel after impact with 2nd is practically same and is similar to circle of pins; 3 – hard rock after impact reference steel (steel 45). Smaller angular with 1st circle of pins velocities of rotors and/or lighter size decreases considerably after impact abrasives should be used. with pins of the first road, the mean • Impact wear tester DESI is suitable device for grindability study of brittle particle size d50 of granite gravel decreases from 5 mm to 0.4 mm. 2nd circle of pins abrasive materials. Size reduction of does not cause considerable size reduction abrasive material is considerable. (compare curves 1 and 2 in Fig. 12). Similar size reduction behaviour was 5. REFERENCES noticed for hard rock abrasive. In case of 1. Engel, P. A. Impact wear of materials. hard rock, additional velocities of 40 m/s Elsevier Science Publishers, New and 80 m/s were studied (Fig. 13.). It was York, 1986. 100 2. Blau, P.J. Glossary of Terms. In ASM 90 Handbook, Friction, Lubrication and

% 80 Wear Technology ,Vol. 18, 942. 70 3. Serway, R. A. Physics for Science & 60 3 2 1 Engineers with Modern Physics. 50 Saunders college publishing, United 40 States of America, 1996. 30 4. Kleis, I., Kulu, P. Solid Particle 20 F(rock gravel F(rock size), Erosion Occurrence, Prediction and 10 0 Control. Springer-Verlag, London, 0.01 0.10 1.00 10.00 2005. Rock gravel size

Fig. 13. Cumulative distribution polygon of hard rock grinded in various velocities between abrasive and 1st circle pins: 1 – 40 m/s; 2 – 60 m/s; 3 – 80 m/s