Microstructure Evolution of Multi-Heat Forging and Numerical Simulation for 316 LN Steel

Microstructure evolution has been studied by multi-heat forging experiments and numerical simulation in order to determine the reasonable forging technology of 316 LN steel. The microstructure evolution models were obtained by hot compressive tests and heat treatment tests of 316 LN steels. The one-heat and three-heat upsetting experiments were carried on. Meanwhile, the corresponding numerical simulations were performed. The results show that, the grain uniformity of three-heat upsetting is much better that of one-heat upsetting. The average grain size of three-heat upsetting is smaller than that of one-heat upsetting. So, the forging technology of multi-heat and little deformation should be adopted for 316 LN steel forging. By comparing experimental average grain sizes with simulated average grain sizes for three-heat upsetting, it is found that the simulated values are in agreement with experimental values, which shows that the numerical simulation can be employed to predict the forging microstructure evolution of 316 LN steel.


INTRODUCTION
In recent years, the microstructure evolution of the metals and alloys in hot forming process is a hot problem.Many studies have been carried out to predict the microstructure evolution during forging process (Cho et al., 2005;Ma et al., 2007;Lin and Chen, 2009;He et al., 2010).Ding et al. (2010) and Gu et al. (2011) investigated the microstructure evolution during rolling process.Faraji et al. (2012) and Tang et al. (2012) studied the microstructure evolution during extruding process.
316 LN steel is widely used as a major material for nuclear main pipe.Its plasticity is low and deformation resistance is large.In past, extensive studies have been focused on creep behavior (Kim et al., 2008;Vodárek, 2011;Ganesan et al., 2013), crack (Schwartz et al., 2010;Kang et al., 2011;Zhang et al., 2013), weld characteristic (Kim et al., 2009), cutting performance (Ozcelik et al., 2011) and plasticity at high temperature (Hei et al., 2012).However, the microstructure evolution of multi-heat forging for 316 LN steel has not been reported.Thus, the investigations of microstructure evolution of multi-heat forging are necessary to determine the forging technology of 316 LN steel.
In this study, the hot compressive tests and heat treatment tests of 316 LN stainless steels were performed.The re-crystallization models and grain growth model were derived.The reliability and accuracy of these models were verified by comparing experimental values with simulated results.The influence of one-heat forging and three-heat forging on microstructure of 316 LN steel was studied by upsetting experiments and numerical simulation.Based on these, the reasonable forging technology of 316 LN steel was determined.

MICROSTRUCTURE EVOLUTION MODELS
To simulate the microstructure evolution, the modeling of microstructure evolution process must be completed, namely, the quantitative relationships between material microstructure and deformation technology parameters are established (microstructure evolution models).Then, the models are integrated into the finite element software.The prediction of microstructure evolution is obtained by finite element calculation (Li et al., 2004).
The single-pass and double-pass compression experiments were conducted on Gleeble-1500 D thermal simulator.The heat treatment experiments were performed in heat treatment furnace.The 316 LN steels with the dimensions of Φ 8×12 mm were machined.The chemical composition (wt%) of 316 LN steel used in present study is C≤0.02%,Mn≤2.0%,Si≤0.7%, P≤0.025%, S≤0.005%, Cr = 16-18%, Mo = 2-3%, Cu≤0.1% and N = 0.1-0.16%.For heat treatment, the temperature range was 900-1200°C and the holding time range was 15-60 min and the initial grain size was about 22.5 and 75.5 µm.For single-pass compression, the temperature range was 900-1200°C.The strain rate ranged from 0.005 to 0.5 s -1 .The initial grain size was about 75 and 170 µm.For double-pass compression, the temperature range was 900-1150°C.The strain rate ranged from 0.005 to 0.1 s -1 .The initial grain size was about 75 and 170 µm.In the first pass, the strain was 0.1-0.5.In the second pass, the strain was 0.05.The interpass time was set according to deformation temperature.Based on Sellars model and Avrami equation, the models of re-crystallization and grain growth model were developed by regression of experimental data (Chen, 2010).
The models of dynamic re-crystallization are expressed as follows: The models of meta-dynamic re-crystallization are expressed as follows: The models of static re-crystallization are expressed as follows: Numerical simulation: The processes of one-heat upsetting and three-heat upsetting for 316 LN steels were calculated by coupling finite element method with microstructure evolution models, respectively.The specimen size was Φ 50×75 mm.The initial grain size was same as that of upsetting experiment.According to the actual forging condition, the velocity of upper die was 10 mm/s.The friction coefficient at the die/work piece interface was 0.3.The simulated average grain sizes of one-heat upsetting and three-heat upsetting are shown in Fig. 3.

RESULTS AND DISCUSSION
One fourth of upsetted specimen was used and divided into 8 different regions, as shown in Fig. 4. 1 to 3 represents difficult deformation regions; 6, 7 and 8 represent large deformation regions.The experimental average grain sizes are shown in Fig. 5.The simulated average grain sizes are shown in Fig. 6.The comparisons between experimental average grain sizes and simulated average grain sizes of 8 different regions for three-heat upsetting are shown in Fig. 7.
Figure 5 and 6 show that, in both upsetting experiment and numerical simulation, the grain uniformity of three-heat upsetting is much better that of one-heat upsetting and the average grain size of threeheat upsetting is smaller than that of one-heat upsetting.This is because that, for one-heat upsetting, only dynamic re-crystallization occurs and the grains become fine, but the non-uniformity of billets is not eliminated, the grain uniformity is not good.For threeheat upsetting, besides dynamic re-crystallization during upsetting process, due to solution treatment between heats, the meta-dynamic re-crystallization and static re-crystallization occur.Because the metadynamic re-crystallization and static re-crystallization can eliminate the distortion grains and elongated grains and the grains which have high storage energy, which make grains fine and uniform and the errors which be caused by non-uniformity of the initial grains are eliminated.Thus, the microstructures become more refinement and more uniform by three re-crystallization mechanisms and the mixed grains are also avoided.Therefore, the grain uniformity of three-heat upsetting is much better that of one-heat upsetting.The average grain size of three-heat upsetting is smaller than that of one-heat upsetting.
Figure 7 indicates that the simulated values of average grain size are in agreement with experimental values of average grain size, which shows that, the models of re-crystallization and grain growth model In view of the research above, the forging technology of multi-heat and little deformation should be adopted for 316 LN steel forging.

CONCLUSION
The re-crystallization models and grain growth model of 316LN steels were derived.The reliability and accuracy of these models are verified by comparing simulated results with experimental values.
The forging technology of multi-heat and little deformation should be adopted for 316 LN steel forging in order to improve the grain uniformity and refine the grain.