simulating two materials with different lattices

Dear Dr.Steve/Dr.Axel and all,

I am simulating two materials with different lattice (bcc and fcc).
First case is a system where the bcc lattice is in the center of fcc lattice.
Second case is a system where bcc lattice material is in interface with fcc lattice material.

I were success to simulate those two cases (at temperature 300K), however I felt the results was not right.
When I checked the RDF function, I saw the two material have gone to the liquid phase.

From experimental data, it should be in a crystalline state at this 300K room temperature.

I checked to see the log file, and found the temperature have become very high, more then 2000K.
Actually I want to simulate with different elevated temperatures.
The simulation used the script:

................
read_data twomaterials.dat
................
dt = 0.0005
thermo 1000
.................
.................
.................

velocity all 300.0 dist gaussian
................
#fix 1 mat1 npt term 300. 300. 10. iso 0.01 0.01 0.01
#fix 2 mat1 npt term 300. 300. 10. iso 0.01 0.01 0.01
fix mat all npt term 300. 300. 10. iso 0.01 0.01 0.01
..............
.............
run 100000
#end

My questions to all of you:

1) May you explain to me what and happened?
2) Is my script input wrong?
3) How to make sure the temperature is fixed during npt simulation?

For your information:
However, if I simulate the two materials separately ( not mixed), everything is good.
Only if i mixed the two material then the results were bad.

Thanks for your reply.

best regards
Arkundato
Physics student

First, you should verify you have set up the correct geometry
with no strong overlaps. Second, you should run with NVE
and see that it is stable, and you get good thermo output.
Then try NVT or NPT. Also with 2 different lattices you
need to worry, for a periodic system, how you satisfy
periodic boundaries for both materials.

Steve