Analysis of Nb-rich and Ti-rich particles in Fe-Nb and Fe-Ti melt additions and J55 steel

Analysis of Nb-rich and Ti-rich particles in Fe-Nb and Fe-Ti melt additions and J55 steel

Info iconThis preview shows pages 1–71. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 10
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 12
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 14
Background image of page 15

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 16
Background image of page 17

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 18
Background image of page 19

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 20
Background image of page 21

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 22
Background image of page 23

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 24
Background image of page 25

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 26
Background image of page 27

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 28
Background image of page 29

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 30
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: I; Matenals nology Laboratory flaw- ' 9 Analysis of Nb—Rich and Ti-Rich Particles in F e-Nb and Fe—Ti Melt Additions and J55 Steel PROTECTED BUSINESS INFORMATION MTL 2005-4(CF) O. Dremajlova, V.Y. Gertsman, J. Li and J .TLBlowk-Ier February 2005 I zithis project was funded by IPSCO. 'bfthis report is restricted to IPSCO. nuutributioin is at their discretion. CANMET-MTL i PROTECTED BUSINESS INFORMATION OANMEfri-MATERIALS TECHNOLOGY LABORATORY ' -' REPORT MTL 2005-4(CF) ANALISIS 0F Nb-RlCH AND Ti-RICH PARTICLES IN Fe—Nb AND Fe—Ti ' MELT ADDITIONS AND J55 STEEL. by O. Dremailova, V.Y. Gertsman, J. Li and J.T. Bowker '5 EXECUTIVE SUMMARY At the request Of IPSCO, Ti-rich and Nb-rich particles present in three different supplied materials including a low carbon J55 steel, and two melt additions, F e-Nb and FeuTi, were characterized. The characterization techniques employed included scanning electron microscopy, electron probe microanalysis and transmission electron microscopy (TEM). FocuSed Ion Beam (FIB) microscopy was used for TEM specimen preparation. In the case of the J55 steel three types of large particles were identified. Type I was shown to be NbC with an f.c.c. structure and lattice parameter of 0.447 11111. Type II particles were confirmed to be (Ti,Nb)N. Complex inclusions containing Fe, Ca, S, Mg, 0 and C surrounded by Ti-l‘iCh particles characterized as TiN or (Ti,Nb)N were labelled as Type III inclusions. Two finer types of particles extracted using a two-stage replica technique were identified as Ti(C,N) and NbC. TEM conducted on samples taken from the Fe—Nb alloy melt addition revealed two phases, the first being a solid solution of Fe in b.c.c. Nb and the second an intennetallic Fer phase. TEM of the Fe—Ti alloy confirmed the presence of four phases: 1“ — solid solutiOn based on oc-Ti; 2nd w solid solution based on [El—Ti; 3rd 7 Ti(C,N) and 4‘h m phase X based on a b.c.c. structure with an approximately 1.3 nm lattice parameter. CANMET-MTL PROTECTED BUSINESS INFORMATION ii CONTENTS EXECUTIVE SUMMARY 1. INTRODUCTION 2. EXPERIMENTAL PROCEDURE 2.1 SCANNING ELECTRON MICROSCOPY 2.2 ELECTRON PROBE MICROANALYSIS 2.3 FOCUSED ION BEAM ANALYSIS 2.4 TRANSMISSION ELECTRON MICROSCOPY 3. RESULTS 3.1 SCANNING ELECTRON MICROSCOPY 3.2 ELECTRON PROBE MICROANALYSIS 3.3 FOCUSED ION BEAM ANALYSIS 3.3.1 References 3.4 TEM RESULTS OF TYPE-l PARTICLE (NB-RICH) 3.4.1 Experimental and Results ' 3.4.2 Conclusion 3.5 TEM RESULTS OF TYPE—Ii PARTICLE (Tl-RICH) 3.5.1 Experimental and Results 3.5.2 Conclusion 3.5.3 Smaller Type-ll Particle 3.6 TEM RESULTS OF TYPE—III PARTICLE 3.7 TEM RESULTS OF FINE PARTICLES 3.7.1 Experimental 3.7.2 Results 3.7.3 Discussion 3.8 TEM INVESTIGATION OF FE-NB ALLOY 3.9 TEM INVESTIGATION OF FE-TI ALLOY 3.9.1 Experimental and Results 3.9.2 Conclusions 3.9.3 References 3.9.4 Appendix 4. GENERAL CONCLUSIONS 5. CONCLUDING REMARKS 6. RECOMMENDATIONS 7. ACKNOWLEDGEMENTS CANNIET-MTL. PROTECTED BUSINESS INFORMATION 1'. INTRODUCTION Twelve fractured samples of carbon steel J55 were received from IPSCO Inc. for quantitative microanalysis cf large Ti-rich and Nb—rich particles. A sample identified as E4098 containing the largest particles was selectedfor study. An additional two alloys, one containing Fe—Ti and the other Fe—Nb, were received for the same investigation. To determine the chemical composition and crystallographic-name of these particles, scanning electron microscoPy (SEM), electron probe microanalysis (EPMA) and transmission electron microscopy (TEM) were used. A Micrion-2500 focused beam (FIB) was used to prepare some of the specimens for TEM analysis. Several other techniques of sample preparation were also employed depending on the material and size of particles. The detailed description of quantitative microanalysis of fine and large Ti-rich and Nb-rich particles that were studied is given below. 2. EXPERIMENTAL PROCEDURE 2.1 SCANNING ELECTRON MICROSCOPY : The Philips XL—3O scanning electron microsoopy (SEM) was used at an accelerating voltage of 20 kV to identify particles present on the fracture surface as well as in the polished sample. 77?“ Back scattered electron (BSE) imaging was carried out to assist in identifying regions richer in high atomic number elements such as niobium. To identify the elements present in particles, energy dispersive X—ray microanalysis was conducted. I' ' 2.2 ELECTRON PROBE MICROANALYSIS Based on the detailed observations made using SEM, as-polished specimens were taken for - further chemical composition analysis using a Cameca SX-51 electron probe micro-analyzer . (EPMA) equipped with four wave-length X—ray spectrometers. It was necessary to use samples in the polished condition. Due to the small particle size, an accelerating voltage of 10 kV was used to minimize the beam—specimen interaction volume. The analysis was performed using a relatively large beam current of 20 HA and counting time of 30 s. 2.3 FOCUSED ION BEAM ANALYSIS For the specific kinds ofinclusion particles that were identified using SEM-EDS analyses, TEM J Specimens were prepared using a Micrion-2500 focused ion beam (FIB) microsc0pe. I The FIB. is perhaps the most powerful tool currently available, which can be used to extract site- - specific TEM specimens. The various FIB-TEM Specimen preparation techniques and their ~~J advantages and shortcomings have already been well documented. With the capability of high— resolution imaging coupled. With st_r_e_ss~free fine ion—beam micro-machining, FIB is by far the most powerful tool in site-specific TEM sample preparation. _ __cANME'r-MT:_L;¢ I PROTECTED BUSINESS INFORMATION 2 2.4 TRANSMISSION ELECTRON MICROSCOPY Samples for TEM investigation were prepared by different techniques. some sections of large inclusions were prepared by FIB. Smaller particles were analyzed by means of two-stage extraction replicas. Thin foils from Fer and Fe-Ti alloys were prepared by combination of _. mechanical polishing and ion milling with high-energy Ar ions. A Philips CM20 FEG TEM equipped with a Schottky field emission gun was operated at a _ voltage of 200 kV. Conventional bright-field TEM image modes were used to visualize the particles. The crystal structure of the phases was determined by means of selected area electron diffraction (SAED) and, for finer particles, convergent beam electron diffraction (CBED). Chemical microanalysis was performed using an Oxford Instruments thin—window energy- dispersive spectrometry (EDS) detector with an INCA System analyzer. Condenser aperture C2 of 50 um and spot size 5, giving the probe size of 10 nm, were employed. Electron energy loss spectroscopy (EELS) was employed for identifying nitrogen in the presence of titanium. Gatan Image Filter model GIF 678 was used for that. TEM was used to analyze large Ti—rich and Nb-rich particles present in the as-polished specimen that were previously observed using SEM and EPMA. In addition fine particles that were present but could not be seen in the SEM or EPMA because of a spot size limitation were analysed after extraction using a two-stage replica technique. Further TEM investigation was performed for Fe-Nb and Fe—Ti alloys. 3. RESULTS The chemical composition of J55 steel supplied by IPSCO is listed in Table 3.1 below. Table 3.1. Chemical composition of 155 steel received from IPSCO. . Ta 013 0.04 0.02 0.04 0.002 S J 0.002%, P — 0.01%, N — 0.008%, B - 0.0002% ' 3.1 SCANNING ELECTRON' MICROSCOPY Individual Nb—rich and Ti—rich particles in the range 2 ,um to 145 ,um were qualitatively analyzed by SEM/EDX. A low magnification SEM image of the fractured sample E4098 is shown in Fig. 3.1(a), revealing two different-fracture morphologies. Two kinds of large particles,'Nb~rich and Ti-rich, were obseIVed on the fracture surface. Nb-rich particles were identified as Type 1 particles and the Ti—rich as Type-II} The maximum size of the Ti—rich particles present on the fractured sample [Figs’g 3j.;l'_(b)_and 3.1(d)] was 14.5 X 14 H.111 and the size of Nb—rich particles was slightly smaller. one "f-‘Nb-rich particles is presented in CANMET-MTL 3 PROTECTED BUSINESS INFORMATION lire surface identified in Fig. 3.1(a). Two EDX spectra Fig. 3.1(c).withi-I_ _ rticles are shown in Figs. 3.1(e) and 3.1(t), reSpectively. - obtained from3N: ri'c A transverse section oh __ _. ample was prepared for further qualitative microanalysis. The as-polished specimen again revealed the presence of Type I and Type II particles as shown in Figs. 3.2 and 3.3, respectively; The Type I Nb—rich particles were in the form of a segregation line of bright rectangular shaped particles. EDX spectra from these particles confirmed the presence of niobium [Fig 32(0)]. A small peak of iron was also present. The Type II large cubical shape particles enriched with titanium were observed in the matrix of the as—polished specimen. The BSE image and EDX spectrum of the Ti-rich particle are presented in Figs. 3.3(a) and 3.303), respectively. Elements such as Nb, N and Mg were also present in this particle. It appears that the Type II particle could be Ti, Nb nitride. A third type of large particle was observed in the as—polished specimen, as revealed in Fig. 3.4(a), containing of a round—shape dark area surrounded by a grey cuboid reminiscent of a ‘ Type II particle. The EDX spectrum of the dark particle identified by the arrows, revealed the presence of elements such as Fe, Ti, Ca, S, Mg, 0 and C, as shown in Fig. 3.400). Similar observations were made for several others dark areas of Type III particles. EDX spectra from the grey cuboids COnfirmed them as Type II particles containing Ti, Nb. and N. 3.2 ELECTRON PROBE MICROANALYSIS The same as-polished Specimen was used to determine the detailed chemical composition of a j, group of elements (Ti, Nb, N, .C, Fe, 0, Ca, Si, Al and B) present in the'large particles (size range from 2 to 5 tam) using EPMA. Prior to the measurements, various standards were scanned. in order to calibrate the microprobe to ensure precise measurements; The standards used in the f current calibration are listed in Table 3.2. The summarized results of four large Ti-rich and three. "" Nb-rich particles are listed in Table 3.3. The following sunnnarizestthegfindings: _ 1. Type I (Nb—rich) particles contained more carbon and iron, and less nitrogen than Type II (Ti—rich) particles. However; the Nb-rich particles are in general quite small which ,3 approaches the interaction volume under the current beam condition. Hence, it is possible that the higher iron and-carbon concentration could be attributed to the surrounding matrix. Further TEM analysis is needed to clarify this matter. No oxygen was observed in Type I - particles. 2. Type II (Ti—rich) particles are mainly-composed of Ti and N. A significant amount of Nb ‘ was also detected: Unlike the Type Iparticles, a small amount of oxygen existed in these particles. ' ' 3. The particle identifiedin the table as Ti-rich *—3 was a Type III particle and contained even higher amounts of oxygen,,carbon and silicon than Type II. ' a CANMETTMTF i easi'mmmmmmam Whamafiaimu a...“ ' ' _CANMET-MTL PROTECTED BUSINESS INFORMATION 4 3.3 FOCUSED ION BEAM ANALYSIS The FIB microscope was used to prepare TEM specimens containing large Ti-rich and Nb-rich particles. Figures 3.5(a) and 3.5(b) show FIB secondary electron images of the Type I and Type 11 particles taken from a mechanically polished sample surface. Once these particles were identified in the FIB, precise ion beam milling was used to create trenches around it and leave a thin membrane of approximately 20 x 10 x 5 pm in size. At this point, the feature of interest was isolated in this thin slab. This thin slab was subsequently cut free in the FIB using the same ion beam. A so-called “lift-out” process was used to transfer this small slab to a carrier, which was subsequently mounted on a TEM copper grid. The mounted slab containing the feature of interest was then further thinned to become electron transparent using the FIB. Details of the technique can be found in open literature [1—3]. TEM specimens from all three types of inclusions were prepared using this technique. Figures 35(0) and (d) show examples of the small sample approximately 20 x 10 x 5 pm in size ' containing the feature of interest that had been lifted-out and mounted onto a carrier. This sample will be further milled (thinned) in a FIB microscope so that the feature of interest (particles) will be electron transparent in TEM. 3.3.1 References l. M.W. Phaneuf, and J. Li, FIB Techniques for Analysis of Metallurgical Specimens, Microscopy and Moroanalysis 6 (Suppl 2: Proceedings) (2000), 524-525. 2. J. Li, G.S. McMahon and M.W. Phaneuf, “FIB Techniques for Microscopy Applications”, Proceedings of the Microscopy Society of Canada, vol. XXVIII, (2001), 26-27. 3. Jian Li, V.Y. Gertsman, and J. Lo, “Preparation of Transmission Electron Microscope Specimens from Ultra—Fine Fibers by a FIB Technique”, Microscopy and Moroanalysis 9 (2003), 888-889. ‘ 5 PROTECTED BUSINESS INFORMATiON I = "MM-a Vi 2 A \ Table 3.2. standards and corresponding reflecting crystals. Standards Table 3.3. Chemical compositions of large particles obtained by EMPA. Particle Mm Ti Nb 5 '5; - :- --: .' _: ._ ' .i"':-f'i :.' i Ti—rich-l 0 - Ti-rich -4 3 -~.' ; Nb-rich -2 2.5 * Type III particle CANMET-MTL PROTECTED BUSINESS INFORMATION 6 (a) Low magnification SEM image of as-received sample. :mm ¥%&”m{fffi?€:¥fi (d) BSE image shown Ti-rich particle. m E‘OSEHHGIB] [55.4453 kcoum, EIDBB-flinclzl anusu ' ' Rev D 2 4 5 a 10 keV n 2 4 . s e 10 (e) EDX spectrum of Nb-rich particle. (f) EDX spectrum of Ti—rich particle. Fig. 3.1. SEM images and EDX spectra of fractured ass-received E4098 sample. CANMET—MTL_ _; Luau! 7 PROTECTED BUSINESS INFORMATION (b) High magnification BSE image of Type 1 particles. Fig. 3.2. Type Iparticles (Nb—rich). CANME‘IT” . PROTECTED BUSINESS INFORMATION 8 kCounts E40993:l Blinc3 _ _ FS=BS4B keV (c). EDX spectrum of Type I particles. Fig. 3.2 (continued) CANMET-MTL PROTECTED BUSINESS INFORMATION - (a) BSE image of Type II particle. kCaunts . E 4UBBIFI B Iinc1 3 F5 $533 Ti ke'V' u 2 , 4 s '- a w (b) EDX of Type [I particle. - . Fig. 3.3. . Type II particle. CANMET L 10 PROTECTED BUSINESS INFORMATION f Type 111 particle. 1mage 0 (21) ESE F5 =37SD EdflflfllFlfllincIEb kCaLmts ke‘v' Isle. ed by the arrow above of Type 1]] part indicat (b) X- ray spectrum of dark area 165. Type IEI part' 4 1g. 3 F 16 L. T M. T E. M N. A. C 11 PROTECTED BUSINESS INFORMATION (a) FIB secondary electron image showing the (b) FIB secondary electron image showing the Type-l inclusions to be examined in TEM. " Type-II inclusion to be examined in TEM. (0) Small sample containing Type—l inclusion. (d) Small sample containing Type-HI Inclusion is mounted on a carrier. inclusion. Inclusion is mounted on a carrier. Fig. 3.5. FIB secondary electron image showing lifted—out specimen containing the feature of interest to be FIB thinned. CANMET-MTL' PROTECTED BUSINESS INFORMATION 12 3.4 TEM RESULTS OF TYPE-l PARTICLE (Nb-RICH) 3.4.1 Experimental and Results The sample was prepared by FIB—milling. The left photo of Fig. 3 .4.1 represents almost the entire sample, while the right photo is a higher magnification image of the particle in question. Several locations on the inclusion were examined by EDS (Table 3.4.1). Two other regions on the sample exhibiting somewhat similar contrast were also analyzed (spectra 4 and 5, see Table 3.4.1), but those appeared to be simply ferrite matrix. A series of diffraction patterns was taken from the particle at different goniometer tilts (Fig. 3.4.2). The EDS data (see Table 3.4.1) suggest that the particle is Nb with a little Ti and Fe (the latter signal could have come from the surrounding matrix and/ or smearing of the material during F [B- milling). However, diffiaction patterns could not be indexed as metallic Nb, which has a b.c.c. structure with a = 0.33 nm, but are perfectly matched by the NbC f.c.c. structure with a = 0.447 nm. Careful re-examination of the EDS spectra indeed revealed a small carbon peak (an example is shown in Fig. 3.4.3). Apparently, the C signal is very weak because the sample is thick and heavy element (Nb) absorbs most of characteristic C X—rays. 3.4.2 Conclusion The inclusion is niobium carbide with a little bit Ti (and possibly Fe). menu-rt. .,,.-...-.. PROTECTED BUSINESS INFORMATION stilts as given by the ]NCA software. CANMET ° ° 9 9. n {I g n G 8354 S355 Fig. 3.4.2. Selected area electron diffraction patterns indexed as an f.c.c. structure with a = 0.447 11111, Le. NbC. 'MlTL . .2 t “m . 7 5 3 S . em. . 0.. w. . \rl. mu. Q, .me 8 . M 5 /l\ e 3 2 .G .r n S 4 : 3 i. '1 F o . a , .m.w 6 K .m 5 a 3. ...e ‘ S a a . D U PROTECTED BUSINESS INFORMATION- CANMET-MTL. w» PROTECTED BUSINESS INFORMATION D 1 uil Scale 42‘] ds Cursor: 0.803 keV (12356 (:13) Fig. 3.4.3. LOW—energy part of EDS spectrum. CANMETeMTL PROTECTED BUSINESS INFORMATION 16 3.5 TEM RESULTS OF TYPE-II PARTICLE (Tl-RICH) 3.5.1 Experimental and Results Sample was prepared by F IB milling. The left photo of Fig. 3.5.1 represents almost the entire sample, while the right photo is a higher magnification image of the particle in question. Several locations on the inclusion were examined by EDS (Table 3.5.1) and two selected area diffraction pattern Were taken at different gonimeter tilts. The inclusion is a single crystal of Ti(+Nb)N. The particle shape (see Fig. 3.5.1) is consistent with such identification and selected area diffraction patterns (see Fig. 3.5.2) confirm that. EDS results are presented in Table 3.5.1. Spectra were acquired in two different regimes, 0—20 keV and 0—10 keV. The latter range is a little more sensitive for light elements, but may miss Nb, which is definitely present in small amount. Nitrogen concentration is less than would be expected from the equiatomic stoichiometry, but this is a common effect in EDS analysis due to absorption of light element X—rays by heavier atoms in the sample. Because of that, the measured concentration of N is higher at the foil edge (see Table 3.5.1) where the sample is thinner. Carbon signal was only occasionally detected and was probably due to surface contamination. Iron signal most likely came from spurious X-rays from the surrounding steel matrix and/or frOm some smearing of the material during FIB milling. 3.5.2 Conclusion This inclusion is titanium nitride with some Ti substituted by Nb and small amount of Zr. CANMET-MTL-j. ' Sectrum18* PROTECTED BUSINESS INFORMATION I results as given by INCA system. ~ 0"” I Results in at. % Nb Spectrum” 1 Spectrum 2* I Sectrum 3* Spectrum 4 Spectrum 5 S ectrum 6* Spectrum 7* Spectrum 8 Sectrum 9 Spectrum 10* Spectrum 1 1"= Sectrum 12 Spectrum 13 Spectrum 14* S Spectrum 16 ectrum15* Spectrum 17 * - spectra acquired in the 0—10 keV range only CANMET-MTL PROTECTED BUSINESS INFORMATION Fig. 3.5.1. Bright-field TEM images of the inclusion. Locations of EDS measurements are #2: _ S384 Fig. 3.5.2. Selected area electron diffiaction pattei‘ns indexed as an f.c.c. structure CANMET-MTL 18 shown. - B.'TiN. S385 (iii: 3 PROTECTED BUSINESS INFORMATiON Fig. 3.5.4. Enlarged TEM image of the particle. Fig. Sale 6' area electron diffraction pattern — corresponds to [100] f.c.c. ' ' ' with a = 0.424 11111, i.e. TiN: ' CAN MET-MT L. _ . v «- PROTECTED BUSlNESS INFORMATION 20 Table 3.5.2. EDS measurements on two areas of the particle. Ti 74.6 65.1 Nb 13.7 6.0 N 9.9 28.9 No carbon was found in the particle. at. % 66.1 7.8 26.1 wt% 3.6 TEM RESULTS OF TYPE-Ill PARTICLE The TEM sample was prepared by FIB milling. The sample contains an inclusion of Ti (carbo—)njtride along with some other particle. The presence of C in the inclusion could not be reliably confirmed, since a high level of carbon was also measured in the steel matrix (see Tables 3.6.1, 3.6.2), indicating contamination of the sample (probably by the epoxy used to mount the FIB sample during preparation). The shape (Fig. 3.6.1) and chemistry (Table 3.6.1) of the main inclusion suggest that it is titanium nitride or carbo-nitride. As mentioned above, carbon measured on the inclusion could have come from epoxy contamination sincc its level is almost as high on the steel matrix (see locations 4, 5 and 15 in Table 3.6.1). Selected area diffraction pattern, SAEDP (Fig. 3.6.2) confirms the characteristic f.c.c. structure with the lattice parameter about 0.425 nm, which is closer to TiN (a = 0.424 nm), than TiC (a = 0.433 nm). This inclusion also contains other metallic elements, which apparently substitute for Ti — a large amount of Nb, a little of Zr and a trace of Fe (this could be due to smearing on the surface during the FIB milling). Since the measurements on the neighbouring particle (locationstl, 2 and 3 in Fig. 3.6.1) produced widely different results (see Table 3.6.1), more detailed EDS investigation was carried out on this area (Fig. 3.6.3 and Table 3.6.2). This confirmed that composition of the particle is variable. Compositional surface plots corresponding to the data of Table 3.6.2 are plotted in Fig. 36.4 for easier visualization. ' Pay attention that some points at the edges of the measured grid are on the TiN and steel matrix. Apparently, there is also superposition of different phases under the beam at some locations. The following are the main observations. The particle core contains high concentrations of Mg (up to 32 wt %), 0 (up to 6.7 wt .%), and no N. The particle also contains Ca (up to 36 wt %) and S (up to 16 wt %), but these elements are concentrated mostly off the particle centre. Probably, there are substitutiOns Mg <-> Ca and O <—> S in this compound. There is a spike of Fe (up to 36 wt %) in the particle centre. Concentrations of Si (up to 4.4 wt %) and C (up to 28 wt %, though this figure may notbe very reliable quantitatively due to contamination — see above) are also higher than in; the surrounding phases. TEM imaging (Fig. 3.6.5) shows that this particle is actually comprised " "alicrystallites, with a relatively large crystallite in the middle. Diffraction patterns (Fi __ .Ould not be indexed as Mg-based compounds, such as MgO, MgS, FegMgO4, MgS,-O 'eyen CagFeMgOS. However, among Ca—based compounds there are many, " erplanar spacings in the range observed on the diffraction patterns (for example .eoxides). Very complex diffraction patterns may also be due to superp -c_rystallites or even different phases. ' ' -' GANMET-MTL PROTECTED BUSINESS INFORMATION & I: w - 9 ,6 l. a. i. w. - ! .tjsii": 'I t. ' .dr. 4: . S457 Fig. 3.6.2. SAEDPI-fibrrrzthemain particle in Fig. 3.6.1. Indexing corresponds to " ' crystal structure with a a 0.425 nm ' " CANMELf i? - PROTECTED BUSlNESS INFORMATION ‘ 22 ' dBUnm ' Fig. 3.6.3. STEM image with locations of EDS grid measurements (see Table 3.6.2). This area corresponds to region with labels 1, 2 and 3 on Fig. 3.6.1. PROTECTED BUSINESS INFORMATION Fig. 3.6.4. Concentration surfaces corresponding to the EDS grid measurements ._ _.= (see Fig. 3.6.3 and Table 3.6.2). . eANME'rl-MTL PROTECTED BUSINESS INFORMATION _. O n I' "s: I N (wt.%) odwwbmmwmw 5 {wt.%) CANMET-MTL 3"in ‘ V 24 0 (wt.%) Carbon Fig. 3.6.4 (continued) 2 5 PROTECTED BUSINESS INFORMATION {Mm Lam; '3 i Fig. 3.6.5. Bright—field TEM images at different goniometer tilts of the particle containing Mg, Ca, S, 0, etc. CANMET' PROTECTED BUSINESSINFORMATFON _ '_ 26 S459 (SAEDP) S460 (SAEDP) S462 (SAEDP) S467 (CBEDP) Fig. 3.6.6. SAEDPS and convergent beam electrqn‘ diffraction pattern (CBEDP) from the crystallites comprising the particle in. Fig. 3.6.5. CANMET-MTL CANMET 9m Wm md 06 5; od 06 ed HE ad ad 5o ad PROTECTED BUSINESS INFORMATION , _ .. Oymymopmg _ $.23 A”; :3 E168 mmm $3.9m 033; 28 PROTECTED BUSiNESS INFORMATION 5wa 38m 62% SEE 62¢ GEE GEE @sz Q75 6V2: AUVZMH fiwa 38m NEH—ha mvoww 'MTL. E, .5 £32 mom 3:3 03$ CANMET . .3%;1§33S\§ «its. .3 3 ‘ a: w .5 . :3". “.33... PROTECTED BUSINESS INFORMATION 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 .90.. 0.0. an 30 N m T A M R 0 F W. s S E W S U B D E T c F. T 0 R D.- GSEEOV dom 03$: CANMET-MTL PROTECTED BUSINESS INFORMATION NA od H E ed 06 H; 0.0 0.0 58 803383 dom OBS CAMMEii-J” 32 PROTECTED BUSINESS INFORMATION . CANMET-MTL-r Eta/i“ .Inl—‘i . PROTECTED BUSINESS INFORMATION carefully peeled off andalong'wnh'the extracted particles coated with an evaporated carbon film, approximately 20 nm in thickness. The carbon coated acetyl cellulose film was placed on Cu grids and then dissolved using a series of acetone—based washes leaving only the carbon ‘ coated replica behind. Table 3.7.1. TEM samples. No inclusions are visible on replica No inclusions are visible on replica No inclusions are visible on replica No inclusions are visible on replica Very dark etch, tranSparency of the replica is insufficient — too thick 2nd iece ofBioden was used on same area to et a thinner re lica 3.7.2 Results Several Ti—rich and Nb—rich particles were found on replica F during TEM examination. Figure 3.7.1. shows the particles examined. EDS results, as given by the INCA software, are presented in Table 3.7.2 and Fig. 3.7.2. It should be noted that the results in Table 3.7.2 should be considered semi—quantitative, especially 'with regards to the light elements. There are several reasons, why more accurate measurements are impossible on the given sample. First, the thickness of different particles under the electron beam varies greatly; therefore, X—ray absorption'could be different for different particles. This effect generally leads to ' underestimation of the amount of light elements, such as C, N and 0. On the other hand, the carbon support film, as thin as it is, contributes to the carbon signal, thus increasing the measured concentration of the element. The relative proportions of metallic elements is much more reliable though. Examples of diffraction patterns used for crystallographic analysis are shown in Fig. 3.7.3. ' - ' ' CANMET—MTij; i: PROTECTED BUSINESS INFORMATION 3.7.3. Discussion Although the chemical composition varies from particle to particle (see Table 3.7.2), the overall results obtained suggest that there are two different, though closely related types of secondary ~rich compound, containing carbon and nitrogen. It also and a little bit of iron. Hereafter, we shall call this phase titanium carbo—nitride” for brevity. Pa ' ' Fig. 3.7.1) 4 cubes or parallelepipeds, Which usually suggests a cubic crystal structure. Particles of the other phase are shapeless, and frequently, they are connected to the crystals of Ti carbo- nitride [see e.g. Fig. 3.7. l (c)-(f)], which suggests that they nucleate heterogeneously on the surface of the Garbo-nitrides. These are niobium-rich carbides with no (or little and thus non- measurable by EDS) nitrogen. The titanium content in this phase is much lower than in the carbo-nitrides, but it contains more iron and, sometimes, a little of aluminum and a trace of manganese. Oxygen is also present in (or on) thesc particles, but it could come from surface oxide. We shall call this phase “niobium carbide”. Fig. 3.7.3) indicate that both phases have the face-c group ij‘m), with the lattice parameter about 0.43 0.45 nm for the Nb carbide. These results are consi databases. Thus, binary compounds of the element with the following parameters _ TiC a = 0.433 nm, TiN a = 0.424 mm, NbC a =~ 0.447 nm. The lattice parameters of titanium carbo—nitrides depends on the relative proportion of C and N — the higher the nitrogen content the smaller is the lattice parameter; for example, TiC0jN03 a = the addition of Nb increases the lattice parameter. meters in this particular case would be emistry imply variability in the crystal lattice stent with the data known from the reference 5 in question have the f.c.c. crystal lattices 0.430 nm, TiCOJNflj a = 0.426 nm. Evidently, More precise measurements of the lattice para meaningless since variations in the particle ch parameter. ' ‘ 35 PROTECTED BUSINESS INFORMATION (b) Fig. 3.7.1. Examples of the inclusions (precipitates) on the extraction replica. Numbers indicate 7 locations of EDS measurements. Figures 3.7.l(a) and (b) show the same region, but Viewed at different goniorneter positions — 1b is tilted so that the Ti(C,N) particle is in the “cube~on” - orientation. ' - I cAumer MTL PROTECTED BUSINESS INFORMATION 36 Fig. 3.7.1 (contii'riued CANMET-MTL PROTECTED BUSINESS INFORMATION 1) ( 1 (cont mued) 7 3. Flg CANMET-M-TL 035.80 32 mm Subcomm owfltméfiwo EL Hm 838mm 2353 $380 em Suboon . 025E330 E mm 538% H. $33.0 £2 a. uwumnénrmo EL 23me om Saboomm uuwfimo DZ a. oflbfiéfimo EL 83wa 3 8260mm $55228 E 2 Saboogm $330 32 02%-830 E 03$“ng 265an ofidfizfposzfiv nzdvszé ofinomdéizfiuvszfiv offiomdzizdvszé 38 Aznoxpzfiv Ochfiomfizv N H 5,38on AZdvmn—ZHC Z. Subcomm .5nt 3588338 QO .msgm 03mqu PROTECTED BUSINESS iNFORMATlON 39 PROTECTED BUSINESS INFORMATION Spectrum 11 Fig. 3.7.2. Examples of EDS spectra from titanium-rich carbo-nitride (11), niobium—rich carbide (32) and mixture of both (20). Cu signal is coming fiom the TEM support grid. ;, CANMETe-IMTL [2:22) I! a , M mammam & KM. (80 (b) Fig. 3.7.3. Examples of diffraction patterns: (a)_SAED_p_attem from the crystal shown in Fig. 3.7.103), (b) CBED pattern corresponding to EDS-31" tipn 32 from the particle shown in Fig. 3.7.1(f}. Both diffraction patterns are ind e 1 With lattice parameters 0.43 and 0.45 nm for the patterns in '3" respectively. cANMET-MTL, 41 PROTECTED BUSINESS INFORMATION 3.3 TEM INVESTlGA IONO F" NB:.:_ALLOY The TEM sample was: .th material designated as 'Fe-Nb alloy. Material is extremely brittle and i o prepare TEM foils. Eventually, electron—transparent foils were prepared byme _, ' carp :lishi'ng followed by ion-milling. To keep the samples from falling apart, copper gltied to them with epoxy at intermediate stages of preparation. The samples were examined TEM CMZOFEG. Figure 3.8.1 shows one of the area examined. EDS results (Table 3.8.1) and diffraction analysis (Figs. 3.8.2 and 3.8.3) revealed two phases in the sample: WV“ .JséWW-a‘z-Wu 1' ». new 1. solid solution of Fe in b.c.c. Nb (lattice period about 0.33 nm) and 2. intermetallic Fer — the so—called u—phase having a trigonal (rhombohedral) structure with periods with a = 0.493, c i 2.683 nm. These are exactly the two phases expected for the given composition according to the equilibrium phase diagram [T.B.Massalski, Binary alloys phase diagrams. Materials Park, Ohio: ASM International, 1990, p.1732—l733]. Small amounts of Al were detected in both phases, While Fer also contains a little of Si and Ti. Zuni Fig. 3.8.1. Bright-field TEM image with locations of EDS measurements. CANMET-MTI— PROTECTED BUSINESS INFORMATION I 42 Table 3.8.1. EDS results. Results in wt % Results in'at. . 96.0 1.4 0.0 65.6 . 3.7 64.6 96.6 65.7 65.3 Spectrum 1' Spectrum 2 Spectrum 3 Spectrum 4 Spectrum 5 S I ectrum 6 S447 Fig. 3.8.2. Example of selected area electron diffraction pattern from niobium-rich phase (corresponds to EDS spectra 1 and 4). :.c._c. Nb with a = 0.33 nrn. . CANMET-MTL. PROTECTED BUSINESS INFORMATION S455 Fig. 3.8.3. Examples of selected area electron diffraction patterfis from iron-niobium phase (corresponds to EDS spectra 2, 3, 5, 6). Indexing — Fer (p—phase), trigonal (rhombohedral) with a = 0.493, c = 2.683 11111. CANMETrMITL PROTECTED BUSINESS INFORMATION '. I 2 44' 3.9 TEM INVESTIGATION OF Fe-Ti ALLOY 3.9.1 Experimental and Results The TEM sample was prepared from the material designated as F e—Ti alloy. First, a thin slice was cut with a diamond disk saw. Then, 3—m1n—diameter disks were cut out of the slice and mechanically polished on both sides, followed by dimple grinding on one side. Finally, one disk was thinned to perforation using Ar ions with energy of 5 keV. The sample was examined in TEM CMZOFEG. Typical bright~field TEM images are shown in Fig. 3.9.1. To save time and resources, some analytic measurements m selected area electron diffraction (SAED), energy—dispersive X-ray spectroscopy (EDS) and electron energy~loss spectroscopy (EELS) — were performed without recording TEM images. Therefore, not all SAED patterns and spectra labels are present on the micrographs. The results of EDS measurements (Table 3.9. 1) suggested that different phases were present in the sample. It should be noted that the presence of nitrogen could be suspectedin some spectra, but it was not included in quantitative analyses due to the following reason. It is known that the N K peak almost coincides with Ti L peak. Moreover, in the given case the alloy also contains vanadium, which also has L peaks in the same area of the EDS spectrum. The deconvolution procedure does not work reliably in this case and produces widely different results depending on the input parameters. Therefore, nitrogen was excluded when the spectra were quantified, and Table 3.9.1 simply gives the prOportions of metallic elements in the phases studied. The issue of nitrogen was addressed by using EELS (see below). Crystal structure of the phases was determined by means of SAED. One group of diffraction patterns matches the h.c.p. structure of cc~Ti (Fig. 3.9.2). This can clearly be seen e.g. fiom the hexagonal pattern S398, which is an indication of the crystal lattice being oriented in such a way that the basal plane of the hexagonal lattice is perpendicular to the incident electron beam. Such a pattern could also be produced by an ‘f.c.c. or a h.c.c. crystals oriented along‘the <111> direction, but that would require the lattice parameters of 0.72 nm (f.c.c.) and 0.36 nm (h.c.c.), which do not correspond to any of the known phases with a chemical composition anywhere close to the given one. This phase consists primarilyof Ti with additions of Al and V, and a little bit of Fe (see Tables 3.9.1 and 3.9.2). Some SAED patterns (Fig. 3.9.4) are cons carbide or nitride; EDS spectrum wasii contains both C and N, quantification'resu that this phase is cubic titanium carboéri CANMET-MTL 45 PROTECTED BUSINESS INFORMATION One set of S reflections on 3.4 However, there- patterns S3 77 and'S", reflections that d_ nitride (see Fig. 3. 5'); ho ittle confusing. The most prominent square pattern of atch TiN or Ti(C,N) in the cube-on orientation. tes that do not match the main pattern. Moreover, he ame location at different goniometer tilts have strong y of the interplanar spacings of the titanium (carbo-) .'.would match ot—Ti. The EDS measurement at this place (spectrum 11 in Tab] e'sponds to 0t~Ti. Apparently, there was a (small) Ti(C)N crystal within the Ct-Tl g _ _ hlch was not noticed during TEM examination, but was accidentally brought under- the beam when tilting the goniometer. Finally, there is a group of diffraction patterns (Fig. 3.9.6), which is hereafter referred to as phase X. The reason for this is the following. From the appearance of the patterns, which have very closely spaced reflections, it is obvious that the crystal structure is long—periodic with very large interplanar spacing. All the phases of binary [l] and ternary [2] equilibrium and metastable compounds containing'Ti and Fe, A], or N with even remotely resembling chemistry known in the literature were tried, but none matched the given SAED patterns. Specifically, the following phases were checked: or-Ti, B—Ti, TiFe, TiFeg, TiN, TlgN, 8’(~Ti2N), TlgAl, TiFezAl, TiFeAlg, TlgFCgAlgz, plus numerous titanium oxides. Then, the search of the database {3] was performed with compounds containing light elements from H to B, which may not be detectable by EDS, also included. For example, hydroxides might give similar EDS spectra to intermetallics: it is possible that 0 could be overlooked because its peak overlaps with V, whereas H of course cannot be detected by EDS at all. No structures matching the obtained SAED patterns were found among hundreds of compounds checked. Figure 3.9.? shows two examples of attempted matching the interplanar spacings of two titanium hydroxides. It is obvious that some experimental spacings, especially those corresponding to the shortest reciprocal vectors, do not fit. interestingly, the experimental diffraction patterns from this phase , . can be indexed as a b.c.c. structure with the period of about 1.3 11111. More detailed determination of the phase structure would require significant additional experimental and computing efforts. Now, let us return to the issue of nitrogen. Limited EELS studies were performed to detect the possible presence of this element.’ Typical EELS spectra are shOwn in Fig. 3.9.8. It was found that Ot—Ti contains N [see Fig. 3.9.8(a)]. Incidentally, the two spectra on this plot (#11 and #14) are almost identical, which confirms the above interpretation of the SAED patterns on Fig. 3.9.5 as coming mostly from Ot-Tl. The edge corresponding to N is clearly visible. Quantification of the spectra gives the atomic ratio of NT 2 0.065 i 0.011. It is known that N has a large solubility in a-Ti, even though the exact solubility limit has not been established yet-[1]. Probably, the observed concentration of N is within the equilibrium solubility in Ot—Ti. On the other hand, N has much lower solubility in [3—Ti [1], and indeed the N edge was not detected in the corresponding spectrum [see Fig. 3.9.8(b)]. The Fe edge is present in the B-Ti EELS spectrum and absent in or-Ti, in agreement with the EDS results(see Tables 3.9.1 and 3.9.2) and with the expected equilibrium solubility [1]. Phase X exhibits a tiny N edge [see Fig. 3.9.8(c) and (d)]; quantification yields the atomic ratio N/Ti = 0027 i 0.012 (average of measurements at five locations). The Fe edge has a higher relative intensity in Fig. 398(0) than in Fig. 3.9.8(b), in agreement with the EDS results that the Fe concentration is higher in phase X mant—Ti. ' '- cANMET-MTL . mmdw ._ ._., ....._... . . PROTECTED BUSINESS INFORMATION 46 3.9.2 Conclusions Analytical TEM, i.e. diffraction studies coupled with EDS and EELS analyses the following crystalline phases in the alloy: 1. Solid solution on the or—Ti base with the hop. crystal lattice (L: = 0.295 mm, c m 0.468 hm) and approximate chemical composition (wt %): N — 3, Al — 3, V — 1, Fe < 0.3, Ti — balance. 2. Solid solution on the B—Ti base with the b.c.c. crystal lattice (L: = 0.325 nm) and approximate chemical composition (wt %): AI — 6, V — 4, Fe — ‘19, Ti — balance. ' 3. Titanium carbo—nitride with the f.c.c. structure (a z 0.43 nm) and the ON ratio of ~2:1. 4. Phase X, probably having a long-periodic structure based on the b.c.c. lattice with a: 1.3 nm. Approximate chemical composition (wt 0/o): N — 1, A] — 4, Si w 0.3, V — 2, Fe — 25, Ni — trace, Ti H balance. ' - 3.9.3 References 1. TB. Massalski, Binary alloys phase diagrams. Materials Park, Ohio: ASM International, 1990. " 2. V. Raghavan, Phase diagrams of ternary iron alloys. Metals Park, Ohio: ASM International; Indian Institute of Metals, 1987. 3. Electron Diffraction Database. NIST/Sandia/ICDD, 1997. 47 PROTECTED BUSINESS INFORMATION _, ' Other . detected Corresponding 'F e elements diffraction attern ‘ S362, S363 S362, S363 trace Ni trace Ni C . trace Ni S366 8369,8370 S368 8372,8387—399 _ S373 8374,8403-405 S376 3377,3400—402 8378 S379 S381 trace Ni Nitrogen was not included in the analysis. Table 3.9.2. Average compositions of phases (wt %). Nitrogen not included. Table 3.9.3. Composition of the titanium carbo-nitride (at. %) estimated --- . taking into account X-ray‘ absorption. CANMET-MTL PROTECTED BUSINESS lNFORMATION Fig. 3.9.1. TEM images showing locations of EDS measurements (eds *), selected area diffraction patterns (S***) and EELS analyses (eels#*).‘ CANMET-MTL $372 E S3 97 Fig. 3.9.2. Diffraction patterns matching Ot-Ti (h.c.p. structure with a = 0.295 nm and c = 0.468 nm). __. CANMEI'I't-IMTI. PROTECTED BUSINESS INFORMATION cANMET-Mry _ 51 I PROTECTED BUSINESS INFORMATION I; I. I I I 1 n I 4. ‘I I. .5 _ co 14F ‘ 7 (12:15] 5! S369 S370 Fig. 39.3. Diffraction patterns matching B-Ti (b.c.c. structure with a = 0.325 nm). | i I I I i E I I T 31 3 : " . cANME-Tf-MTL n PROTECTED BUSINESS INFORMATION (150] S374 S379 Fig. 3.9.3 (continued) - '- MTL... CANMET- 53 PROTECTED BUSINESS INFORMATION - (011)- - W?- 501 [a] S381 S403 Fig. 3.9.3 (continued). 7 1 CANMETS—MTL. 54 PROTECTED BUSINESS INFORMATION S404 S405 Fig; 3.9.3 (contifined) MTL CANMET 55 PROTECTED BUSINESS INFORMATION S364 S365 Fig. 3.9.4. Diffraction patterns matching Ti(C,N) (fee. structure with a z 0.43 11m). 1: QAN ME_T-_-..MT .I... . 56 PROTECTED BUSINESS INFORMATION S406 M 36. 1 . S407 Fig. 3.9.4. (contifiliéd- 2%. “Q m N fix «.5. a “3% MTL. T a: M N A C PROTECTED BUSINESS INFORMATION £00 I _ _ 5 (1:11 a. I * o . _ w '1 ‘ a 6 S408 ~ ‘ (i311 ' ' - '° (lie?) .,_: 9 . [I131 .43. S409 Fig. 3.9.4 (continued). '_ CANM'ET—MTL .: I. .5: giggihl xii aw ~.E§;:. wt»? saixxi g.) . if _ Fir rim, rim. _ '58 S410 Fig. 3.9.4 (continued). MTL PROTECTED BUSINESS INFORMATION CANMET PROTECTED BUSINESS INFORMATION $377 .3: ' -- . " (one: ms. . - _ ' :4" S400 Fig. 3.9.5. Diffraction patterns fi‘om the grain in Fig. 3.9.203). Apparently — superposition of B—T and TiN(C). Spot indexing fl TiN with a : 0.425 nm; rings also drawn corresponding‘to the interplanar spacings of TiN. cANMET-MIL. i§<s ;. nU . . AU . 4 . S ._... . fi 6 . r e . ‘c m . noilunl S .. 0.. B , 1... . “.0 6 . m m ® .~ .fl S 5 . ,9. 0 9. as m 3 . . a . I _ g 2 F . 0 . r . 4 S ‘ . t .I -MTL N mu. T A M R O F m S S E m S U B D W. C E T O R P T E M, N A C: 61 PROTECTED BUSINESS INFORMATION S325 ' 3362' S363 - S366 3373 .1ffraction patterns from phase X. cAume'T-MII... Z .. . . , RU . . . 7 u . _ . .5 . 4 . . fls .. cu 3376 S411 MTL PROTECTEDBUSNESSINFGRMAHON T E M. N A. C. 63 PROTECTED BUSINESS INFORMATION S415 - S416 S417 Fig. 3.9.6 (continued). gfiacings on diffi'action pattern S411 with long-periodic figure — T12F€OQ2H23 nm), F 601-123 (P63/mmc, 0:10.799, c=0.813 CANMET-MTL PROTECTED BUSINESS INFORMATION 1 200 1000 0 0 8 coo 0 AU 0 0 D 0 6 4 2 _\ x 2:58 ooo 600 700 800 900 Energy Loss (eV) 500 (a) 120 100 0 0 0 8 6 4 000—. x E550 000 CANMET—MTL PROTECTED BUSINESS ENFORMATION 010) CO CDC) 4:. O O 0.3 O 0 CCD counts x 1000 0 466' 500 600 700 800 ' 900 _ Energy Loss (eV) (0) CCD counts x 1000 “400- 450 500 550 600 650 Energy Loss (eV) (6») i Fig. 3.9.8 (continued). GANM-ETFMTL PROTECTED BUSINESS INFORMATION . ' 66 3.9.4 Appendix One interesting observation was made. At an intermediate step of ion-milling, some material was sputtered and redeposited on the TEM sample. This redeposit crystallized into ultra-fine crystallites with h.c.p. oc-Ti crystal structure (Fig. 3.9.9). Crystal lattice of the redeposit does not necessarily reflect crystal phases existed in the sample prior to sputtering; however, its chemical composition bears a signature of elements present in the sample. Interestingly, in some places there is a significant amount of molybdenum (Table 3.9.4). Apparently, Mo was not a preparation artefact and could not come from anywhere but the sample itself. Therefore, it is believed that Mo contained in the sample (at some locations), though it was not encountered during further examination of the limited area available for TEM analysis. Fig. 3.9.9. Selected area diffraction from ultrafine-grained oc—Ti. Table 3.9.4. EDS analysis results from the sample after intermediate ion—milling step. Spectrum 2 Spectrum 3 Spectrum 4 Spectrum 5 Spectrum 6 7- ectrum 7 CANMET-MTL _....... .... - t. -A-Wmmmm\wwm 67 PROTECTED BUSINESS INFORMATION :QGENERAL CONCLUSIONS 1- Three types of large nannies-were identified in the matrix of 155 steel and characterized as follows: - -. ' a. Type I - Nb-rich, NbC with a small amount of Ti (and possible Fe); f.c.c. crystal structure with a = 0.447 nm. b. Type 11 - Ti-rich, (Ti,Nb)N with a small amount of Zr; f.c.c. crystal structure with a = 0.424 nm. ' c. Type III - Ti—rich (Ti,Nb)N surrounding or bordering a complex inclusion containing Fe, Ca, S, Mg, 0 and C. 2. Finer particles in the steel are also of the two f.c.c. phases based on titanium (carbo-) nitride and niobium carbide. 3. The Fe-Nb alloy examined in TEM revealed two phases in the thin foil sample: 1) solid solution of Fe in b.c.c. Nb (lattice period about 0.33 nm) and 2) intermetallic Fer — the so-called n-phase having a trigonal (rhornbohedral) structure with periods with a = 0.493, c = 2.683 nm. 5‘ ' 4. Analytical TEM, i.e. diffraction studies coupled with EDS and EELS analyses revealed the following crystalline phases in the Fe—Ti alloy: a. Solid solution on the ct-Ti base with the hop crystal lattice (a = 0.295 11m, 0 = 0.468 nm) and approximate chemical composition (wt %): N —— 3, Al — 3, V — 1, Fe < 0.3, Ti — balance. b. Solid solution on the B-Ti base with the b.c.c. crystal lattice (a = 0.325 nm)'.and approximate chemical composition (wt %): Al — 6, V — 4, Fe — 19, Ti H balance. c. Titanium carbo-nitride with the f.c.c. structure (a z 0.43 nm) and the C:N ratio of ~2:1. (1. Phase X, probably having a long-periodic structure based on the b.c.c. lattice with a m 1.3 mn. Approximate chemical composition (wt %): N — 1, Al — 4, Si — 0.3, V —— 2, Fe — 25, Ni — trace, Ti — balance. . 5. CONCLUDING REMARKS The F e-Ti and Fe-Nb melting additions are certainly the source of the carbide— and nitride- forming elements, i.e. metallic elements - Ti and Nb, as well as ,C and N, which are contained in the F e—Ti alloy. However, it is highly unlikely that the coarse inclusions found in the steel were transferred directly from the melting additions. Thus, niobium carbide particles have not been encountered in the Fe-Nb .alloy,_.though.they Were found in the steel samples. Titanium carbo- nitride is present in the Fe-Ti array," the particles have different shape and chemical . CANMETEMTL PROTECTED BUSINESS lNFORMATION . -' i . composition than the cuboid titanium nitride are abundant in the steel. It is almost impossible for such huge changes in composition and morphology to occur by diffusion in the solid state. It is possible that the titanium":icarbo€nitride particles were not completely melted during processing and their remnants served as nuclei for the cuboid particles in the steel. However, thorough compositional measurements on a couple of typical inclusions in the steel did not reveal any'variation from the centre to periphery, which would be expected for growth on existing nuclei of different chemistry. On the other hand, the cuboid Ti nitrides in the steel are frequently associated with inclusions of non-metallic origin containing Mg, Ca, S, 0, etc. (could those be ceramic particles?) Apparently, the non-metallic particles serve as nuclei for the cuboid inclusions. 6. RECOMMENDATIONS 1. It is possible that Type II and Type III particles are the same type and that depending on how the particle is sectioned for specimen preparation it will determine whether the complex - inclusion containing Fe, Ca, S, Mg, 0 and C is observed. Further study may clarify this issue. 2. Further work is required to characterize the nature of Phase X in the Fe-Ti alloy. 7. ACKNOWLEDGEMENTS The authors gratefully acknowledge MTL technologist Catherine Bibby for TEM sample preparation, which involved different complicated replica and thin foil techniques. - CANMET-MTL_ . am ms; » hyahywawn kwka/A‘A?‘ '- aflmw ...
View Full Document

Page1 / 71

Analysis of Nb-rich and Ti-rich particles in Fe-Nb and Fe-Ti melt additions and J55 steel

This preview shows document pages 1 - 71. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online