Research On Phase Transformation Behaviors And Heat-treatment Processes Of T92Ferritic Steel

T92steel is a novel type of ferritic heat resistant steel developed on the basis ofT91through compositional redesign. In T92steels,1.5%~2.0%amount of W wasadded to realize W-Mo precipitation strengthening, Mo content was reduced to0.3%~0.6%in order to avoid the formation of-ferrite, the content of V and Nb wasaltered to facilitate dispersed distribution of carbonitrides, and0.001%-0.006%B wasadded for the purpose of strengthening grain boundaries. In comparison with T91steels, T92not only has excellent combination of heat conductivity, toughness,formability, and weldability, but also exhibits much higher creep strength attemperatures above600oC, making T92the most promising alternative material forT91steels.There have been sufficient researches focused on the creeping properties,microstructural evolution in creep, and weldability of T92steels, but their phasetransformation behaviors and heat treatment techniques have been rarely studiedfundamentally. Against this background, the present study was mainly targeted atinvestigating martensitic phase behavior of T92steels during continuous coolingprocess and the microstructural evolution in conventional heat treatment. Theprecipitation behavior of M23C6was analyzed by building kinetics models, and Q&Pand Thermo-mechanical Treatment of T92steels were intensively studied. Thefollowing conclusions were achieved.(1) As cooling rate increased, the number of martensite crystal nuclei increasedand martensitic laths became densified as a result of decreasing aspect ratio of lathsand violent collision between laths. With the increasing of cooling rate, the growthactivation energy decreased and supercooling degree expanded, resulting in thestrengthening of parent phase which obstructs the mobility of boundaries, andtherefore leads to the decrease of boundary migration rate, v0. At all cooling rates,martensitic boundary migration rate remained in the range of10-6to10-4, whichindicated the martensite phase transformation occurred by a thermal-activationmechanism.(2) M3C, which occurs by homogeneous nucleation, precipitated duringnormalizing of T92steels, and then was dissolved and replaced by M23C6in the subsequent tempering. The coarsening of martensite laths during tempering wascaused by the migration of martensite lath boundaries, which was driven by theinternal stress occurred in martensite phase transformation. Movement of Y-junctionwas the main mechanism responsible for the coarsening of martensite laths. Thetempering process of T92steel can be separated into three steps. The temperingprocess from10s to120s is defined as the first step, during which the martensitelathes coarsen significantly, and M3C carbides disappear gradually. The temperingprocess from120s to3600s is defined as the second step, during which the growth ofcarbides and the coarsening of lath take place gently. Finally, the process of temperingfrom3600s to3.6×105s is defined as the third step, during which the carbides growrapidly，and equiaxed grains with low dislocation density are formed.(3) On the assumption of site-saturated nucleation, transfusion-controlledgrowth, and soft-collision correction, a kinetics model for the precipitation of M23C6phase in T92steels during tempering was established. The results showed that M23C6grew according to a2-Dimensional mode. Refined martensite laths were beneficial tothe dispersed precipitation of refined M23C6phase. In order to describe moreaccurately the precipitating process of M23C6at various temperatures, atemperature-related growth coefficient was employed. The growth activation energyacquired by fitting was71.60KJ, close to the diffusion activation energy of C in-Fephase.(4) Two martensite phase transformations took place in the Q&P treatment: thefirst martensite phase transformation was incomplete, and part of untransformedaustenite transformed to martensite again in the subsequent cooling process. Theparticipation of C increased the C content in the untransformed austenite. Theenrichment of C increased the stability of austenite, which decreased the start andfinish temperature of the second martensite phase transformation, leading to refinedmartensite laths. Using assumptions of autocatalytic nucleation, interface-controlledgrowth, and collision correction of randomly nucleus distribution, a kinetics model forthe second martensite phase transformation was established, which was a gooddescription of the relationship between volume fraction, f, and temperature, T.(5) Thermo-mechanical Treatment was applied to T92steels, and constitutiveequations were established for hot deformation of T92steels. Using dynamic materialmodel, hot processing map for T92steels was drawn, and the effects of different hotdeformation conditions on the microstructure of T92steels were investigated. Higher m-and η-value was beneficial to dynamic recrystallization and resulted in biggerrecrystallized grains. Hot deformation could effectively refine martensitic structure;the smaller m-value was, the greater the martensitic structure was refined. Refinementof martensitic structure was reflected by shortened martensite lath length, i.e. theaspect ratio of martensite laths increased. When the steel was deformed in theinstability domain E, equiaxed martensite laths, which means aspect ratio of lathsroughly equals1, were realized. Hot deformation effectively facilitated theprecipitation of refined MX carbonitrides, and reduced the precipitation of largeM23C6grains in the subsequent tempering.