page 1
page 2
page 3
page 4
page 5
page 6
page 7
page 8
page 9
page 10
page 11
page 12
page 13
page 14
page 15
page 16
page 17
page 18
page 19
page 20
page 21
page 22
page 23
page 24
page 25
page 26
page 27
page 28
page 29
page 30
page 31
page 32
page 33
page 34
page 35
page 36
page 37
page 38
page 39
page 40
page 41
page 42
page 43
page 44

and 0.083%Al (VM339B). The Mn content wasselected based around commercial highermanganese content steels, summarised fromindustrial data by Subramanyam et al. [12], butits particular value was not thought to be crit-ical, if it was, say, 1% higher or lower. Carbonwas kept low to improve toughness. Ti and Alwere added to tie up residual C, N and O.These alloys cost about half as much as a3.5%Ni steel. The alloys were austenitised at 850oC, waterquenched or air cooled and also tempered for0.1, 1, 10 or 100 h at 450oC. Tempering in therange of 300-500oC increases the ductile-brit-tle transition temperature (DBTT), as found byBolton et al. [13] when measuring this prop-erty in a series of Fe-Mn alloys, albeit withoutdetailed microstructural characterisation workback then. Nasim et al. [14] noted a rapid grainboundary segregation of Mn, P, and N, on age-ing, based on Auger electron spectroscopymeasurements. The control group was in awater-quenched condition. Embrittlementwas the most severe at 450oC.Further, as a continuation of a previouswork, where Nikbakht et al. [7] identified theembrittlement mechanism in Fe-8%Mn as wellas Fe-8%Mn stabilised with Ti and Al, a thirdalloy was studied, to investigate the effect ofNb. A Fe-8.46%Mn-0.24%Nb-0.038%C alloywas vacuum-melted as a 20 kg cast and upsetforged at 1200°C into a 32??76 mm plate. Thisalloy costs about one third of a 3.5%Ni steel.The heat treatment at 600°C on thehomogenised alloy was carried out to intro-duce reverted austenite to improve impacttoughness. Excess C over stoichiometry wasabout 70 ppm, which lowers the DBTT (see Fig-ure 9 of ref. [15]). The alloy was homogenisedas two blocks 32??60??76 mm for 50 h at1100°C, air cooled, and followed by two dif-ferent heat treatment routes for each block.The first consisted of 1.5 h at 850°C, air-cool-ing, and then 4 h at 600°C, air-cooling. The sec-ond consisted of 1.5 h at 850°C, water quench-ing, and then 4 h at 600°C, water quenching.Longitudinal V notch Charpy specimens5??10??55 mm were machined from the heat-treated blocks.Vickers Hardness and Impact TestingVickers hardness was measured using a load of1 kg, at room temperature in the unstrainedpart of the impact-tested specimens. Impacttests were conducted on a Charpy machinewith maximum impact energy of 300 J. Sam-ples of size 5??10??55 mm with a V-notch wereused. The temperatures of impact tests were20°C, -20°C, -100°C, and -196°C. Testing tem-peratures below the room temperature wereobtained by using a chamber cooled by flow ofliquid nitrogen.Scanning Electron MicroscopyThe steel specimens were polished using a 0.05?m colloidal suspension of silica after mechan-ical polishing down to 1 ?m. The microstructure of the steels was exam-ined in a Jeol 6500 FEG scanning electronmicroscope. The SEM was operated at 5-20 kV.The fracture surfaces of the samples wereexamined under the SEM to determinewhether the sample surface had brittle cleav-age type or ductile type failure.RESULTS AND DISCUSSIONTable 1 shows the hardness values of theexperimental alloys VM339A and VM339Bunder various heat treatment conditions. AlloyVM339B is harder than VM339A because itcontains 0.056%C, 0.19%Ti and 0.083%Al. Itwas found that deep quenching the alloysdown to -20, -100 and -196°C does not changethe hardness after the specimens werereturned to room temperature before thehardness measurements were taken.Table 1 includes the hardness values for bothalloys tempered at 450°C with increasing timeand then impact tested. In the case ofVM339A, the hardness increased significantlyafter tempering for 1-100 hours, compared towithout tempering or tempering for 0.1 h. Onthe other hand, little change of hardness hasbeen found when the tempering time variesfrom 1 to 100 h. However, there exists an obvi-ous difference between tempering for 0.1 hand 1 h. There could be two reasons con-tributing to this effect:1. The effect of tempering reaches a satura-tion point after around one hour or some timebefore one hour. Here, a saturation pointmeans no further increase in hardness withincreasing tempering time.2. Experimental limitations meant that heat-ing for 6 minutes did not lead to effective tem-pering, because the specimens were sealed insilica tubes in vacuum, inserted into the pre-heated furnace, and kept for the specifiedlength of time. There might be a heating-upperiod that could take minutes.Impact testing was carried out on both alloysto estimate the ductile-brittle transition tem-perature, which gives the change in behaviourfrom ductile at high temperature to brittle atlower temperature (Table 2). As a generalstatement, ductile fracture initiates at a par-ticular toughness valueMicrographs of fracture surface of alloyVM339A and VM339B (1 h at 850°C, WQ)impact tested at 20°C are shown in Figures 1and 2, respectively. The uneven or rough sur-face of the fracture can be seen at low magni-fication (Figures 1a and 2a). High magnifica-tion shows a ductile type failure for both alloys(Figures 1b and 2b). Tearing and cone shapeddimples were observed in alloy VM339A (Fig-ure 1), which had the higher impact toughnessindicating it is more ductile than alloyVM339B. Micrographs of fracture surfaces forboth alloys at different conditions and tem-pering time were examined which showedductile type fracture with no obvious differ-ence.Micrographs of the surface structure forboth alloys at different conditions and tem-pering time were examined but there were novisible differences in the structures or grain-size.It is possible in Fe-20%Mn to get peculiarstress-strain curves at room temperature, dueto gamma and epsilon phases, like in the ther-mally cycled Fe-8%Mn alloy [16]. Nikbakht etal. [7] summarised the formation of thesephases from relevant literature.The Charpy impact test results of the heattreated Fe-8.46%Mn-0.24%Nb-0.038%C alloyare shown in Figure 3. It should be noted thatthe linear regression lines are used to onlyillustrate the difference between specimensafter the two cooling treatments. The linearregression may or may not represent the realvariation of the impact energy with testingtemperature.These impact test results show that thewater quenched alloy has higher impactenergy. The reason may be that air coolingresulted in the formation of precipitates dur-ing the slower cooling process, which increasesstrength but decreases toughness.In the quenched condition, this third alloyhad a higher impact energy than bothVM339A and VM339B, showing the beneficialeffect of Nb stabilisation. In the air cooled con-dition, however, the Nb-containing alloy hadcomparable impact energy with VM339A andTable 2: Impact energy (in joules J) absorbed by the experimental alloys under various heat treatment and impact test temperatures.AlloysHeat Treatment20oC -20oC -100oC -196oC VM339A850oC 1h, air cooled68588-850oC 1h, water quenched (WQ)7161238WQ + 450oC 0.1 h686011-WQ + 450oC 1 h6261125WQ + 450oC 10 h75389-WQ + 450oC 100 h66457-VM339B850oC 1h, air cooled4356198850oC 1h, WQ5450208WQ + 450oC 0.1 h544915-WQ + 450oC 1 h5242125ANALYSISOFSTEELSMICROSCOPY AND ANALYSISNOVEMBER 201117