A modern approach to coupled neutronic simulations

Jonathan Shimwell

To perform automated parametric multiphysics analysis of breeder blanket designs with an aim of optimising the design.

Objective

To perform automated parametric multiphysics analysis of breeder blanket designs with an aim of optimising the design.

Current procedure for obtaining volumetric heating

• Neutron and photon heating simulated and results passed to engineering codes

• Parametric neutronics models created from surfaces equations and boolean operations

• Structured mesh overlaid on to the geometry        

Images provided by P. Pereslavtsev

 Suggested procedure for obtaining volumetric heating

• Neutron and photon heating simulated and results passed to engineering codes

• Parametric CAD models created using Python scripts and automatically converted to neutronics models

• Unstructured mesh applied to the geometry        

 Benefits of this approach

• Heating results obtained for discrete materials

• CAD models generated can be used as a geometry for simulations in many disciplines

• Unstructured mesh requires less elements and memory

Images provided by B. Collings and J. Naish

Method of reducing memory and element requirements for a structured mesh

Images provided by B. Collings

Selection of design parameters

Parametric CAD construction

Neutronics simulation for TBR

Converstion to unstructured mesh

Neutronics simulation for volumetric heating

Converstion to engineering mesh

Simulations to find stress and temperature

Evaluate design

    Converstion to      CGS

Predict performance and uncertainty

Demonstration parameter study

• CAD geometry for 100 different versions of the HCPB breeder blanket were produced

• Parameters varied were lithium ceramic bed height and neutron multiplier bed height

Neutron multiplier

Cooling plates

Lithium ceramic

Demonstration parameter study

• Different geometric parameters avaiable

Demonstration parameter study

Demonstration parameter study

• Different neutron multipliers (Be12Ti, Be) and different 6Li enrichment of Li4SiO4

60%

70%

80%

90%

100%

Be

Be12Ti

6Li enrichment

Be

Be12Ti

TBR

Energy multiplication

Interpolation of results

Results for lithium ceramic Li4SiO4 with 90% 6Li and Be12Ti as the neutron multiplier

Interpolation of results

Gaussian processes were used to fit TBR and energy multiplication values obtained by MCNP simulation

Tritium production

Energy amplification

Software able to interpolate TBR values and confidence

curl -L jshimwell.com/predicttbr/80-Be-15-60

Visualization

Fully automated parametric model production, simulation and visualization of results.

Meshing

Fully automated hex meshing of the geometry to abaques format for use with MCNP 6 unstructured mesh

Volumetric heating

Li4SiO4

Be12Ti

Heating W/cm3

with 60% 6Li enrichment

Volumetric heating

Heating values obtained on unstructured mesh that conforms to material boundaries. 

y

Additional details - cooling channels

y

Future simulations

y

Parametric DEMO reactor

18 coils

• Magnetic equilibrium solved

• Plasma ripple kept below limit
•  
• Model includes
  • poloidal field coils
  • toroidal field coils
  • gravity supports
  • intercoil connections

pythonOCC , Salome

16 coils

Isotope creation

Isotope creation

Conclusion

• Multi-parameter optimisation will be used to find blanket designs that merit future investigation.

• The approach to coupling neutronics inputs to engineering simulations has been overhauled

• Future work will include automated simulations to obtain stress and temperature of materials

L. Lu - McCad

Y. Qiu - Salome

P. Pereslavtsev - DEMO model

A. Haussler - unstructured mesh

F. Hernandez - HCPB model

C. Zeile - HCPB model

G. A Spagnuolo - Benchmarking

Acknowledgments

S. McIntosh - Gaussian process

T. Eade -unstructured mesh

T. Barrett - supervision

L. Packer - supervision

Volunteers needed

12.00 - 2.00 (lunch time)

Wednesday the 7th December

Current PhD students and posters need for the open day