© MatCalc (2008)
 
 
 
 
 
 
 
 
   
 
   

 

A4: Heat treatment of a complex tool steel

(download: application_script_9Cr_steel.mcs)

Abstract

Martensitic 9-12 wt.% Cr steels are favoured grades for high temperature components in thermal power generation industry. Welding is the major joining and repair technology of such components. The creep strength of 9-12 wt.% Cr steels is mainly based on the fine distribution and long-term stability of different populations of precipitates. After the quality heat treatment, i.e. in normalised and tempered condition, the steels show a tempered martensitic microstructure with finely dispersed precipitates. Due to thermal influence by the welding processes the initial optimized microstructure is significantly altered. Most of the precipitates are dissolved during heating and virgin martensite is formed on cooling. The heat-affected zone (HAZ) of martensitic steels is often afflicted with inferior mechanical properties. A post-weld heat treatment (PWHT) is mandatory and can recover the base material microstructure and properties of the ‘as-received’ condition to a certain extent. Although especially during long-term service, the HAZ of martensitic steels is prone to premature failures.


The following application example shows how the production and fabrication process consisting of cooling after casting, heat treatment (HT), welding process (WP) and post weld heat treatment (PWHT) can be simulated with the software MatCalc. Furthermore in the example is shown how simulation results can be compared with experimental data to validate the obtained simulation results. Therefore, the same heat treatment which is simulated with the software MatCalc was also performed on real specimens.

Experimental

The simulated material is a modified 9Cr-1Mo cast steel. The chemical composition of the investigated material is shown in Table 1.

 

Table 1: Chemical composition

 

 

Heat treatment

The simulated time temperature sequence consisting of cooling after casting, austenitisation, tempering, welding cycle and post weld heat treatment resembles the real industrial heat treatment extended by the welding process with the additional post weld heat treatment necessary for these types of steel. The whole simulated time temperature sequence is shown schematically in the following figure.

 

 

Precipitate evolution

To compare the simulated results with the real precipitate evolution in this steel, additionally to the simulation work microstructural investigations were performed at defined points, as shown in Fig. …. The results of the microstructural investigations are shown in the following Figure.

 

 

Simulation setup

The simulation setup doesn’t differ compared to the simulation shown and explained in (e.g.) Tutorial 15 and Tutorial 16. The main items of a successful system setup remain the same, namely:

  1. Create a workspace
  2. Load the database(s) and select the phases which have to be considered
  3. Define the chemical composition
  4. Define the precipitation domain(s)
  5. Define the precipitate phases(s)
  6. Define the time-temperature schedule (if necessary!)
  7. Create the output window(s)
  8. Finalise the system setup defining simulation end time/temperature and start the simulation

Of course, in the present case the setup becomes more complex due to the complex alloy which makes it necessary to consider a couple of precipitate phases which can appear during the heat treatment (i.e. MX, M3C, M7C3, M23C6). Furthermore in the present case the material undergoes several austenite/ferrite(martensite) phase transformations, which are also taken into account.
How a new workspace can be created, databases are loaded and the chemical composition of an alloy is defined will not explained here. (The databases used in the present example are the mc_steel.tdb and the mc_sample_fe.ddb.) Also the definition of precipitation domains, precipitate phases as well as the definition of the output windows is explained in the Tutorials. Information about nucleation sites are given in the following Table.

 

 

 

The following figures are only shown for reasons of traceability to summarize the most important simulation settings regarding the precipitation domains and precipitate phases.

 

 

 

 

However, at the beginning of this page you can find the script (*.mcs) for this example. Therefore, you can go trough the script step by step if there are procedures still not clear.

Much more interesting is the definition of the time-temperature sequence for the simulation, due to the new heat treatment functionality of version 5.22 (and newer releases). The new functionality allows the definition of heat treatment segments. Therefore, heat treatment definition by defining a table containing time-temperature couples belongs to the past! The Edit heat treatment window can be opened pressing Alt+F8.

 

 

 

Heat treatment

As already mentioned, the functionality allows to define heat treatment (HT) segments. In MatCalc three ‘variants’ of heat treatment segments can be selected:

  • Segment 1: to define the HT segment, the end temperature and the heating or cooling rate have to be defined.
  • Segment 2: to define the HT segment, the heating or cooling rate and the duration of the segment have to be defined.
  • Segment 3: to define the HT segment, the end temperature and the duration of the segment have to be defined.

 

Additionally, for each HT segment the prevailing precipitation domain and the storage interval can be defined respectively. If no changes are made MatCalc takes the settings of the previous segment.

 

 

Furthermore, using the pre- or post-script option additional settings can be implemented into the heat treatment. In the present case the change of nucleation sites for the precipitate phases due to the phase transformation from austenite to ferrite/martensite after the cooling process (cooling after casting) is shown in the following figure.

 

 

Attention: For the first heat treatment segment start temperature and precipitation domain must be defined!

 

Results

The obtained results of the numerical simulation are shown in the following. Agreement of the simulated radius evolution of the precipitate phases compared to the obtained experimental data can be designated as very good. The measured particle size after PWHT for M23C6 and MX precipitates are 42nm and 11nm respectively, whereas, numerical simulation results in 40.2nm for M23C6 and 13.1nm for MX precipitates.

 

 

 

 

 

 

 

‹‹ to the Top ››  ‹‹ to the Application files ››