| Abstract | U disertaciji su prikazani rezultati istraživanja trošenja u masi abraziva toplinski naštrcanih prevlaka na bazi nikla, krom karbida i volfram karbida nanesenih na podlogu od prethodno toplinski obrađenog čelika 42CrMo4+QT. Prevlake na bazi nikla nanesene su postupkom naštrcavanja pomoću plinskog plamena s istovremenim utaljivanjem (NiCrBSi i NiCrBSi-WC) i postupkom naštrcavanja plazmom prenesenim lukom (NiBSi-WC), dok su postupkom plinskog naštrcavanja velikim brzinama čestica nanesene prevlake na bazi krom karbida (CrCNiCr) i prevlake na bazi volfram karbida (WC-CoCr). Navedene prevlake primjenjuju se u uvjetima gdje se traži visoka otpornost na trošenje i koroziju. Prije glavnih istraživanja, provedena su prethodna ispitivanja koja su se odnosila na istraživanje mikrostrukture i trošenja u masi abraziva s ciljem testiranja ispitnog uređaja i mjerne opreme te određivanja razina ulaznih varijabli za glavna istraživanja. Analiza mikrostrukture prevlaka provedena je primjenom svjetlosnog optičkog mikroskopa, elektronskog pretražnog (SEM) mikroskopa uz energijsko-disperzivnu spektrometriju (EDS) i primjenom rendgenske difrakcijske (XRD) analize. Glavna istraživanja trošenja u masi abraziva provedena su primjenom potpunog faktorskog plana pokusa uz uvažavanje načela randomizacije, blokova i ponavljanja, s ciljem dobivanja statističkih modela za predviđanje ovisnosti izlazne varijable, odziva (gubitak mase nakon trošenja) o ulaznim varijablama i njihovoj međusobnoj interakciji. U postupku modeliranja primijenjene su četiri ulazne varijable, brzina uzorka (mijenjana na tri razine: 1, 1,75 i 2,5 m/s), kut sudara abraziva i trošene površine (mijenjana na dvije razine: 30° i 60°), granulacija abraziva (mijenjana na dvije razine: manja 0,1 - 0,8 mm i veća 0,5 - 1,5 mm) i vrsta prevlake (primijenjeno je pet različitih vrsta prevlake). Na temelju razvijenih statističkih modela utvrđeno je da najveći utjecaj na gubitak mase imaju brzina uzorka i kut sudara abraziva i trošene površine, kao i njihova međusobna interakcija. Sve istraživane prevlake imale su za oba abraziva najveći gubitak mase pri najvećoj brzini uzorka (2,5 m/s) i manjem kutu sudara (30°), dok su najmanji gubitak mase sve istraživane prevlake imale pri najmanjoj brzini uzorka (1 m/s) i većem kutu sudara (60°). Ovakav rezultat posljedica je mehanizma trošenja, promatranog na razini jediničnog događaja, kod kojega pri manjem kutu sudara, čestice abraziva nakon sudara s trošenom površinom nastavljaju kliziti po površini te je kontakt čestica abraziva i trošene površine duži, dok je kod većeg kuta sudara to klizanje kraće jer se čestice odbijaju od trošene površine. Povećanjem brzine uzorka kod oba kuta sudara dolazi do porasta gubitka mase što je posljedica veće udarne energije pri većoj brzini uzorka. Razvijeni statistički modeli predstavljali su i podlogu za provedeni postupak optimiranja primjenom funkcije poželjnosti, s ciljem pronalaska razina ulaznih varijabli (brzina uzorka, kut sudara i vrsta prevlake) s kojima će se postići minimalni gubitak mase s istovremenom minimalnom varijabilnošću gubitka mase tj. minimalnim širenjem pogreške, uz zadane devijacije ulaznih varijabli. Postupkom optimiranja pronađene su odgovarajuće kombinacije razina ulaznih varijabli u pokusu trošenja u masi abraziva (za obje granulacije abraziva), kojima će u dobroj ili izvrsnoj mjeri biti zadovoljena oba postavljena odziva (minimalni gubitak mase i minimalna varijabilnost gubitka mase), što dokazuju dobivene visoke vrijednosti funkcije poželjnosti. Optimiranjem je utvrđeno da bi najmanji gubitak mase uz istovremenu najmanju varijabilnost odziva imala prevlaka NiCrBSi pri brzini uzorka od 1 m/s i kutu sudara od 60° trošenjem u abrazivu manje granulacije, dok bi trošenjem u abrazivu veće granulacije optimalna kombinacija bila - prevlaka NiCrBSi-WC, brzina uzorka 1 m/s i kut sudara od 49,61°. Najmanji gubitak mase i najmanja varijabilnost gubitka mase ostvareni bi bili pri brzini uzorka od 1 m/s i kutu sudara od 30° za CrC-NiCr prevlaku u abrazivu manje granulacije i WC-CoCr prevlaku u abrazivu veće granulacije. Usporedbom gubitka mase istraživanih prevlaka u abrazivu manje i veće granulacije pri kutu sudara od 30°, utvrđeno je da su sve istraživane prevlake pri manjim brzinama uzorka imale veći gubitak mase trošenjem u abrazivu veće granulacije, dok su pri većim brzinama uzorka imale veći gubitak mase trošenjem u abrazivu manje granulacije. Ovo ukazuje da pri manjim brzinama uzorka na gubitak mase više utječe udarna energija (koja je veća za abraziv veće granulacije), dok pri većim brzinama uzorka na gubitak mase više utječe trajanje kontakta čestica abraziva i trošene površine (koji je duži za abraziv manje granulacije). Pri kutu sudara od 60°, prevlake NiCrBSi-WC, CrC-NiCr i WC-CoCr, u pogledu vrste abraziva, imale su jednako ponašanje kao i pri kutu sudara od 30°, dok su prevlake NiCrBSi i NiBSi-WC imale veći gubitak mase u abrazivu veće granulacije pri svim brzinama uzorka. Ovakav rezultat ukazuje da kod NiCrBSi i NiBSi-WC prevlaka, trošenjem pri većem kutu sudara, snažniji utjecaj na gubitak mase ima udarna energija u odnosu na duljinu trajanja kontakta čestica abraziva i trošene površine. Usporedbom gubitka mase prevlaka i podloge trošenjem u masi abraziva, utvrđeno je da su sve istraživane prevlake imale manji gubitak mase u odnosu na podlogu pri kutu sudara od 30° u abrazivu manje granulacije, pri svim brzinama uzorka. Trošenjem u abrazivu veće granulacije, pri kutu sudara od 30°, sve istraživane prevlake imale su manji gubitak mase u odnosu na podlogu pri svim brzinama uzorka, dok je jedino pri brzinama od 1 i 1,75 m/s prevlaka NiCrBSi-WC imala veći gubitak mase u odnosu na podlogu. Pri kutu sudara od 60°, trošenjem u abrazivu manje granulacije pri najmanjoj brzini uzorka (1 m/s) uočeno je da su sve istraživane prevlake (osim NiBSi-WC) imale veći gubitak mase u odnosu na podlogu. Pri najvećoj brzini uzorka (2,5 m/s) i kutu sudara od 60°, sve istraživane prevlake (osim NiCrBSi-WC) imale su manji gubitak mase u odnosu na podlogu trošenjem u abrazivu manje granulacije. Usporedbom gubitka mase prevlaka i podloge pri kutu sudara od 60° trošenjem u abrazivu veće granulacije, sve istraživane prevlake imale su veći gubitak mase u odnosu na podlogu pri brzini uzorka od 1 m/s, dok su pri brzini uzorka od 2,5 m/s sve istraživane prevlake (osim NiCrBSi prevlake) imale manji gubitak mase u odnosu na podlogu. Manji gubitak mase podloge u odnosu na prevlake, pri određenim kombinacijama razina ulaznih varijabli, posljedica je manje tvrdoće podloge koja je bolje apsorbirala udarnu energiju |
| Abstract (english) | This dissertation presents the results of investigating the wear by abrasive particles of nickel-, chromium carbide-, and tungsten carbide-based thermal spray coatings deposited to a previously heat-treated steel 42CrMo4+QT substrate. Nickel-based coatings were deposited by flame spraying with simultaneous fusing (NiCrBSi and NiCrBSi-WC) and plasma transferred arc (NiBSi-WC), while high velocity oxygen fuel was applied on chromium carbide-based coatings (CrC-NiCr) and tungsten carbide-based coatings (WC-CoCr). These coatings are used in conditions requiring high wear and corrosion resistance. Prior to the main investigation, preliminary studies of microstructure and wear by abrasive particles were conducted to assess the testing device and measuring equipment, as well as to determine the input variables for the main investigation. Coating microstructure analysis was conducted using the light optical microscope, a scanning electron microscope (SEM) with electron dispersive spectroscopy (EDS) and X-ray diffraction (XRD) analysis. The main investigation of wear by abrasive particles was conducted using a full factorial design, taking into account the principles of randomization, blocks and replications in order to obtain the statistical models that can predict the dependence of the output variable i. e. response (mass loss after wear) on input variables and their interaction. There were four input variables applied in the modeling process, sample velocity (three levels: 1, 1.75 and 2.5 m/s), impact angle of the abrasive and worn surface (two levels: 30° and 60°), granulation of the abrasive (two levels: lower 0.1 - 0.8 mm and higher 0.5 - 1.5 mm) and type of coating (five different types of coating). The derived statistical models indicate that the greatest influence on mass loss is exerted by the sample velocity and the impact angle of the abrasive and worn surface, as well as their interaction. All investigated coatings underwent the highest mass loss at the highest sample velocity (2.5 m/s) and lower impact angle (30°). The lowest mass loss was observed in all investigated coatings at the lowest sample velocity (1 m/s) and higher impact angle (60°). This result is due to the wear mechanism, observed at the unit event level, in which at the lower impact angle, abrasive particles continue to slide on the surface after the impact with the worn surface and the contact of abrasive particles and worn surface is longer, while at the higher impact angle this sliding is shorter because the particles bounce off the worn surface. Increasing the sample velocity at both impact angles leads to an increase in mass loss due to higher impact energy at higher sample velocities. The developed statistical models were the basis for the numerical optimization procedure as well, using the overall desirability function approach, with the aim of finding the levels of input variables (sample velocity, impact angle and type of coating) to achieve minimum mass loss with minimum propagation of error, using estimated standard deviations for the input variables. Optimal combinations of levels of input variables (for both abrasive granulations) were found by the optimization procedure, to meet both goals (minimum mass loss and minimum propagation of error), i. e. to maximize the overall desirability function. The optimization determined that the minimum variation in the mass loss would have NiCrBSi coating at the sample velocity of 1 m/s and the impact angle of 60° by wear in abrasive of lower granulation. By wear in abrasive of higher granulation, the optimal combination of factor levels would be - NiCrBSi-WC coating, sample velocity 1 m/s and impact angle of 49,61°. The minimum variation in the mass loss would be achieved at the sample velocity of 1 m/s and the impact angle of 30° for CrC-NiCr coating in abrasive of lower granulation and for WC-CoCr coating in abrasive of higher granulation. Comparing the mass loss of the investigated coatings in the abrasive of lower and higher granulation at the impact angle of 30°, it was found that all investigated coatings underwent higher mass loss by wear in the abrasive of higher granulation at lower sample velocity, while at higher sample velocity they underwent higher mass loss in the abrasive of lower granulation. This indicates that at lower velocity, the mass loss is more affected by the impact energy (which is higher for an abrasive of higher granulation), while at higher velocity the mass loss is more affected by the contact duration of the abrasive particles and the worn surface (which is longer for an abrasive of lower granulation). The NiCrBSi-WC, CrC-NiCr and WC-CoCr coatings, in terms of the type of abrasive, behaved the same at both impact angles (30° and 60°), while the NiCrBSi and NiBSi-WC coatings underwent a higher mass loss in the abrasive of higher granulation at all sample velocities. This result indicates that NiCrBSi and NiBSi-WC coatings, in a higher impact angle wear, the impact energy of the abrasive particles affects the mass loss more in relation to the duration of contact between the abrasive particles and the worn surface. By comparing the mass loss of the coating and the substrate by wear in the abrasive mass, it was found that all investigated coatings underwent lower mass loss compared to the substrate at the impact angle of 30° in the abrasive of lower granulation, at all sample velocities. By wearing in the abrasive of higher granulation, at the impact angle of 30°, all investigated coatings underwent less mass loss compared to the substrate at all sample velocities, while only at 1 and 1.75 m/s the NiCrBSi-WC coating underwent a higher mass loss compared to the substrate. At the impact angle of 60°, wear in the abrasive of lower granulation at the lowest sample velocity (1 m/s) indicated that all investigated coatings (except NiBSi-WC) underwent a higher mass loss compared to the substrate. At the highest sample velocity (2.5 m/s) and the impact angle of 60°, all investigated coatings (except NiCrBSi-WC) underwent lower mass loss compared to the substrate by wear in the abrasive of lower granulation. By comparing the mass loss of the coatings and the substrate at the impact angle of 60° by wear in the abrasive of higher granulation, all investigated coatings underwent a higher mass loss compared to the substrate at the sample velocity of 1 m/s, while at the sample velocity of 2.5 m/s all investigated coatings (except NiCrBSi) underwent lower mass loss in comparison to the substrate. The lower mass loss of the substrate relative to the coatings, with certain combinations of the levels of input variables, is due to the lower hardness of the substrate which better absorbs the impact energy |
