Precipitation of boride in the Cr-Ni-B ternary phase diagram

[lihaoge]

Dear Bernd

Thank you all the time for your kind reply. Now I want to simulate the effect of cooling rate on boride precipitationin in the Cr-Ni-B ternary phase diagram, For example,Cr2B or CrB. I did the experiment, composition is Ni20.49Cr74B5.51 (at%) and annealed at 1000°C for 1 h. I sent the composition and the SEM picture to you. I want to see the final phase ratio of boride with different cooling rate. Could MICRESS do this job? If it can, could you please tell me how to make the dri. file?
Thank you very much.

lihaoge

[Bernd]

Dear lihaoge,

First thing you need to do is to define the initial microstructure. As I understood, this is the structure heat treated at 1000°C which still does not show borides, and which is subject to cooling afterwards. So, it perhaps is just a single phase multi-grain structure with homogeneous composition.

Then you need to define nucleation conditions, and start the simulation. In any case, the results will strongly depend on the choice of nucleation conditions, and you will need a calibration based on experimental data.

If you furthermore want to simulate the correct anisotropic morphology for the borides, you need to find a suitable anisotropy model for each boride which reproduces the correct morphology. This, however, should be not necessary if you are interested only in the relative phase fractions.

Bernd

[lihaoge1994]

Dear Bernd,

Thank you always for your support.
Now I want to simulate the phase fraction in Cr-Ni-B alloy cooled at different cooling rate with composition designed around eutectic point L (1258 ºC, Ni – 57.9Cr – 4.3B, at%) = (Ni) + (Cr) + Cr2B. I have done some experiments to see the microstructure. Figure shows, in BSE images, the microstructure of the samples cooled at air. Black, dark gray and light gray regions are Cr2B, (Cr) and (Ni), respectively. As the cooling rate slows down, the plate-like Cr2B phase grows longer and thicker but the number decreased.
I defined the initial microstructure and other parameters. I do not know whether it is right. And this error was found.
tWidth value too large.
With the chosen phase-field parameters,
tWidth must be less than -1.690579468357246E-020 s
STOP in routine initTwidth.
Do you have some ideas on my dri. file and this error?
Thank you very much!

lihaoge

[Bernd]

Dear lihaoge,

It seems you messed up the input for phase interaction 3/3. For the interface energy, the input string is completely wrong (I wonder why MICRESS accepts that...), and for the interface mobility the value is huge. This certainly leads to the effect you describe.

Bernd

[lihaoge1994]

Dear Bernd

I have defined the initial microstructure and nucleation conditions. I put the results in the attachment. Now I want to find a suitable anisotropy model for each boride which reproduces the correct morphology but I do not know how to do it. Do you have some ideas about it?
Thank you very much.

lihaoge

[Bernd]

Dear lihaoge,

For borides I guess you should use faceted anisotropy. In MICRESS you can define individual facets by their facet vectors. I personally have no experience with anisotropy of borides, so I cannot provide you with detailed information. You should check out in the Training Examples (T015, T040, T041) how faceted anisotropy works in general.

Bernd

[lihaoge1994]

Dear Bernd

I'm going to simulate the morphology of Cr60Ni37B3 (wt%) at different cooling rate. Figure is a graph showing the molar fraction of all solid phases of Cr-37wt% Ni-3wt% B alloy calculated by Thermo-Calc as a function of temperature. It can be seen that the alloy undergoes single-phase solidification→two phase eutectic solidification→three-phase eutectic solidification stage. It can be seen from the figure that a solidification path that Cr-37wt% Ni-3wt% B alloy may exist is: L→L+Cr2B_ORTH→L+Fcc_A1+Cr2B_ORTH→L+Fcc_A1+Cr2B_ORTH+Bcc_A2→Fcc_A1+Cr2B_ORTH+Bcc_A2
Since the primary phase portion and the eutectic structure portion of the Fcc_A1 and Cr2B_ORTH phase are regarded as two phases during the simulation, this is more advantageous for simulation and research. So I make the dri. fill as:
# 0 -> LIQUID
# 1 -> M2B_ORTH
# 2 -> FCC_A1
# 3 -> M2B_ORTH_2
# 4 -> FCC_A1_2
# 5 -> BCC_A2
But when I check the initial equilibrium which has been calculated in the .log file:
# The linearisation parameters of the phases LIQUID/M2B_ORTH are:
# ---------------------------------------------------------------
1320.7500 ! T0 [K]
3329.6950 ! dG [J/cm**3]
2.6503510 ! dSf+ [J/cm**3K]
2.6503510 ! dSf- [J/cm**3K]
2395.7764 ! dH [J/cm3]
36.999999 ! c0(NI)/LIQUID
91.578783 ! c0(NI)/M2B_ORTH
3.0000002 ! c0(B)/LIQUID
8.4212169 ! c0(B)/M2B_ORTH
13.068809 ! m(NI)/LIQUID
-999.99000 ! m(NI)/M2B_ORTH
99.933178 ! m(B)/LIQUID
-999.99000 ! m(B)/M2B_ORTH
0.0000000 ! dcdT(NI)/LIQUID
0.0000000 ! dcdT(NI)/M2B_ORTH
0.0000000 ! dcdT(B)/LIQUID
0.0000000 ! dcdT(B)/M2B_ORTH

I have decreased the initial temperature from (1835K to 1321K) until dG of LIQUID/FCC_A1 is slightly negative (although this is not in line with the actual).But the dG of LIQUID/M2B_ORTH is very high. If I decrease the initial temperature until dG is slightly negative, the initial temperature will be very small and it is not in line with reality! And the result shows nothing. So maybe somewhere went wrong.Do you have any ideas to deal with this problem? Thanks!

lihaoge

[Bernd]

Dear lihaoge,

It appears that you use different boride phases in the Thermo-Calc and in the MICRESS simulation: In Thermo-Calc it is CR2B_ORTH and in MICRESS M2B_ORTH. What is the reason for that?
Anyway, in the MICRESS simulation you get Ni₂B which probably is not what you want...

Bernd

[lihaoge1994]

Dear Bernd

Thank you all the time for your kindly support. I changed the Cr2B phase in Micress and the simulation start to run. Bt the dG of LIQUID/CR2B_ORTH_2 is still positive and the eutectic Cr2B did not appear. The result is only the Cr2B of the primary phase. Should I continue to lower the temperature until it negative? Is there any other influence factor besides temperature?

# The linearisation parameters of the phases LIQUID/CR2B_ORTH_2 are:
# ------------------------------------------------------------------
1430.0000 ! T0 [K]
18.241137 ! dG [J/cm**3]
2.0031783 ! dSf+ [J/cm**3K]
2.0031783 ! dSf- [J/cm**3K]
5656.7176 ! dH [J/cm3]
47.212073 ! c0(NI)/LIQUID
2.7967145 ! c0(NI)/CR2B_ORTH_2
1.1043081 ! c0(B)/LIQUID
9.3742126 ! c0(B)/CR2B_ORTH_2
-4.5952488 ! m(NI)/LIQUID
-999.99000 ! m(NI)/CR2B_ORTH_2
220.84495 ! m(B)/LIQUID
-999.99000 ! m(B)/CR2B_ORTH_2
0.0000000 ! dcdT(NI)/LIQUID
0.0000000 ! dcdT(NI)/CR2B_ORTH_2
0.0000000 ! dcdT(B)/LIQUID
0.0000000 ! dcdT(B)/CR2B_ORTH_2

# Minimum undercooling for stable growth, seed type 3: 0.9984134 K [r=1.000000 mic.]
The composition of NI cannot be used for CR2B_ORTH_2 as a thermodynamic condition!
The composition of B cannot be used for CR2B_ORTH_2 as a thermodynamic condition!

# The linearisation parameters of the phases LIQUID/FCC_A1_2 are:
# ---------------------------------------------------------------
1430.0000 ! T0 [K]
-94.531650 ! dG [J/cm**3]
0.97229610 ! dSf+ [J/cm**3K]
0.48318420 ! dSf- [J/cm**3K]
985.21850 ! dH [J/cm3]
47.213458 ! c0(NI)/LIQUID
55.475541 ! c0(NI)/FCC_A1_2
1.1043453 ! c0(B)/LIQUID
1.40919201E-03 ! c0(B)/FCC_A1_2
3.3585866 ! m(NI)/LIQUID
10.393362 ! m(NI)/FCC_A1_2
-79.676809 ! m(B)/LIQUID
-97048.914 ! m(B)/FCC_A1_2
3.53283127E-02 ! dcdT(NI)/LIQUID
-1.05500541E-02 ! dcdT(NI)/FCC_A1_2
-4.75600037E-03 ! dcdT(B)/LIQUID
8.54304038E-06 ! dcdT(B)/FCC_A1_2

Another question is how to set up time input data? I found that if I set up the time input data as this time:

# Finish input of output times (in seconds) with 'end_of_simulation'
# 'regularly-spaced' outputs can be set with 'linear_step'
# or 'logarithmic_step' and then specifying the increment
# and end value
0.0005
linear_step 0.001 1
end_of_simulation
# Time-step?
# Options: fix ...[s] automatic automatic_limited
automatic_limited
# Options: constant from_file
constant
# Limits: (real) min./s, [max./s], [phase-field factor], [segregation factor]
1.E-7 1.E-5
# Coefficient for phase-field criterion 1.00
# Coefficient for segregation criterion 0.900
# Number of steps to adjust profiles of initially sharp interfaces [exclude_inactive]?
20


The simulation start to run and I can get the results. But I think the time interval is too short, and I increased it as:


linear_step 1 400
end_of_simulation
# Time-step?
# Options: (real) automatic [0<factor_1<=1] [0<=factor_2] [max.] [min.]
# (Fix time steps: just input the value)
automatic 0.9 0.9 1.E-1 1.E-5
# Number of iterations for initialisation?
5

and decreased cooling rate from -2000K/s to -200K/s. And there is no results this time. Do you know why this happen?

Thank you very much!

lihaoge

[Bernd]

Dear lihaoge,

First of all I have seen that you use a a 1D-extension to the concentration field (1d_far_field). This makes absolutely no sense and should be removed. Then, the warnings in the .log output will not appear anymore!

Another important point is that you should not use such high static anisotropy for interface 0/2. In the case of metallic anisotropy, the value 1.0 is extremely high anisotropy, while in case of faceted anisotropy 1.0 means no anisotropy...

I believe that your second boride phase (CR2B_ORTH_2) does not appear because you have a lot of primary boride (CR2B_ORTH). Thus, secondary boride should not form at least until the primary phase is not completely overgrown. You should try to start with less primary seeds. Then the secondary phase perhaps grows easier.

With respect to the change of time stepping, you use wrong syntax which probably leads to extremely small time steps (check in .TabT). Therefore simulation probably gets extremely slow (is this what you mean by "there is no results this time"?). Please use the syntax with "automatic_limited" like in your attached input file!

Bernd