Font Size: a A A

Measurements And Finite Element Simulation Of Quenching Stress Distribution In Medium Carbon Steels

Posted on:2018-03-14Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y LiuFull Text:PDF
GTID:1361330590455324Subject:Materials Science and Engineering
Abstract/Summary:PDF Full Text Request
The mechanical properties of the medium carbon steels can be improved markedly by quenching and subsequently tempering process in engineering.However,the high level of the internal stress during quenching process is often the cause of distortion and cracking of the treated workpiece.Systematic investigation of quenching stress is very helpful for the design of quenching process.The measurement of the internal stress distribution in a quenched part is the basic method for investigation of quenching.However,experimental measurement has two main limitations:1)for a part with large scale and complex shape,the measurement of the internal stress along depth is not only difficult but also time-consuming;2)In most cases,the cracking of a quenched part is caused by transient stress during quenching,while experiment can only measure the final internal stress?residual stress?,rather than transient stress.Based on these reasons,the computer simulation of quenching was developed to predict the temperature,phase and stress evolution so as to optimize the quenching process,and various theoretical models were established to compare with the experimental results.In the process of computer simulation,transformation-induced plasticity?TP?has significant effect on the simulated results.Taleb measured the variation of transformation with the volume fraction of bainite and was compared with prediction based on functions describing the TP kinetics proposed by Abrassart,Desalos and Leblond,respectively.The results indicated that Abrassart's and Desalos's functions lead to almost the same kinetic which largely underestimates experimental result.But the further comparison of the residual stress predicted by using these TP kinetic functions has not been yet made.According to the hardenability and size of the medium carbon steel parts,the main microstructure of the quenched part can be classified as pearlite+ferrite matrix,bainite matrix,and martensite matrix.Systematical investigation of the quenching stress based on the microstructure matrix is significant importance.Whether the TP should be included during simulation is still not clear now.Oil and polymer are often used as quenchants for medium carbon steel because of the lower susceptible to cracking of the quenched part,and the mechanical properties of the treated part can usually be satisfied.However,oil and polymer are the source of pollutants and harmful for the users;furthermore,in some cases,the mechanical properties of the quenched part cannot be satisfied.Replacement of the oil and polymer quenchant with clean and cheap water has been added in the 21st century Heat Treatment Plan internationally.The technology and equipment of water-air alternately timed quenching?ATQ?have been investigated and popularized by our research group for 15 years in China accompanying with their successfully applied in dozens of special steel plants or companies.However,the fundamental theory concerning the avoiding of cracking by using ATQ is still unclear yet.Thus,by the combination of experimental measurement of temperature,microstructure,and residual stresses with computer simulation,in which modification of the user subroutines UMAT and UMATHT of commercial software Abaqus was established in this thesis,the following3 issues were investigated:1)Whether there exists a better equation which was verified by experiment describing TP;2)Clarify that in what case,the TP should be included based on the experiment and theory;3)According to the cracking phenomenon in engineering,verify why ATQ can avoid cracking.The main results of the present thesis are described as follows:1.Improving the accuracy of FES parameters by experiment and theory:a)by using trial and error method,the heat transfer coefficient?HTC?of water was obtained by fitting with the cooling curves of 50 mm and 60 mm diameter cylinders;b)The TTT curves was modified based on the previous works of Kirkaldy et al.in the calculation of diffusive phase transformation,taking account of the difference in alloy elements between the investigated steels and that on the handbook;c)the residual stress distributions of the quenched cylinders was measured by XRD,and the stress was corrected by the formulas;d)Unitive methods of experiment and theory mentioned above ensured the better correctness and comparability of measurement and prediction on the distribution of residual stress in three kinds of mediun carbon steels with different hardenabilities.2.The comparison between the residual stress distribution predicted by using the TP functions proposed by Abrassart,Desalos,Leblond and us with the experimental results was made.It indicates that the calculated residual stress distribution by using our Ex-Modified normalized TP function was better consistent with the experimental results,which verify that Ex-Modified normalized function better describing TP.The measured residual stress distributions by using XRD of 30 mm and 50 mm diameter42CrMo cylinders demonstrated that the diameter of the cylinder has a significant effect on the residual stress distribution.As for the 30 mm diameter cylinder?martensite matrix?,the maximum tensile stress was located at the 10 mm radius,while as for the 50 mm diameter cylinder?bainite matrix?,the maximum tensile stress located at the core of the cylinder.3.In order to highlight the effect of the transformation stress on the residual stress,three cylinders with the same diameter of 60 mm and length of 240 mm,but different hardenability steels 45,42CrMo,and 40CrNiMo were quenched by the same water quench process in order to ensure that they have the almost same thermal stress.In 45steel,the main microstructure is proeutectoid ferrite and pearlite except the martensite and bainite at the surface,which indicates that 45 steel has lowest hardenability;in42CrMo,there are bainite and martensite,and the amount of martensite gradually increases from the core to the surface,but the amount of bainite still larger than that of martensite,indicating that 42CrMo has higher hardenability;in 40CrNiMo,there are92%martensite and 8%bainite at the core of the cylinder,indicating that 40CrNiMo has the highest hardenability.The maximum tensile and compressive stresses of 45 and42CrMo are located at the core and surface,respectively,and these two stresses of 45steel cylinder are lareger than that of 42CrMo cylinder.As for 40CrNiMo,the maximum tensile stress is located at the subsurface,while the residual stresses at the core and surface are all compressive.The separate calculations of thermal stress and transformation stress indicated that,the three cylinders have almost the same thermal stress,which proved the correctness of our design idea,and also demonstrated that the three different residual stress distributions were determined by transformation stress.Further investigation indicates that the higher of the hardened depth,the more obvious influence of the TP on simulated residual stress.That is,as for 45 steel,there was no significant difference between the calculated and measured stress distribution feature except for the surface even though TP was not included to take into account for 45steel.As for 42CrMo,although the distribution of residual stress is same when include/not include TP in FES,the values of the residual stress differs;while as for40CrNiMo,there is a great deviation between calculated and measures stress distribution when TP is not included in FES.Thus,TP can be ignored in FES for low hardenable parts when the main microstructure is ferrite/pearlite,but TP must be included in FES for the high hardenable parts which consist mainly of bainite/martensite,especially of martensite after quenched.4.In order to explain the cracking of the 42CrMo driver shaft in engineering,the residual stress distributions of 60 mm diameter 42CrMo cylinders after ATQ and DQ were measured and compared,and the quenching stresses of 42CrMo driver shaft were simulated when performing ATQ1,ATQ2,and DQ.The origins of the avoiding of cracking were revealed:a)as for the ATQ treated workpiece,because of the arising of temperature at the surface,the surface of the part undergoes self-tempered,which lead to the lower hardness compared with DQ;b)compared with DQ,the magnitude of the residual stress is lowered by using ATQ,that is,both of the maximum tensile and compressive stresses are lower accompanying with more uniform distribution of the residual stress;c)the dominant stress for cracking of driver shaft is the radical rather than axial or tangential stress.The optimized ATQ significantly lowers the tensile radical stress,which is caused by transformation stress,and thus avoids the cracking of the driver shaft.
Keywords/Search Tags:Medium carbon steel, thermal stress, phase transformation stress, residual stress, Transformation-induced Plasticity(TP), Finite Element Simulation(FES)
PDF Full Text Request
Related items