Willard Strap Analysis Update 9 30 09 by 7UDEc2g

VIEWS: 6 PAGES: 27

									Figure 1 – NSTX Upper Umbrella Assembly Upgrade Design: Version 4
Figure 2 – Single Segment 3-Strap Assembly Solid Model: Version 4
                              By = .3T
                                                                    ux = .018”
    I = 130 kA                Bz = 1 T
                                                                    uy = .30”




                 JS
                                                  FLorentz

                                                                 Non-Linear
                                                                 Large Deflection


     Heat Gen         Transient Thermal (ANSYS)


                                                        Tnodes




                      Transient Thermal (ANSYS)




Figure 3 – ANSYS Multiphysics Analysis Block Diagram
Figure 4 – Single Segment 3-Strap Assembly FEA Model: Mesh
Figure 5 – Single Segment 3-Strap Assembly Electric Model Results: Voltage
Fig. 6 – Single Segment 3-Strap Assembly Electric Model Results: Current Density
Figure 7 – Single Segment 3-Strap Assembly Electric Model Results: Joule Heat
Fig. 8 – Single Segment 3-Strap Assembly Thermal Model Results: Temperature
    Study: Determine Current Best-Practice to Perform
    Magnetostatic Analysis in ANSYS 12.0 WorkBench
•   New SOLID236/237 magnetic analysis elements
     – Have both Magnetic Vector Potential (MVP) and Line Edge method capability.
       Replaces SOLID97 and SOLID117.
     – Compatible with WB generated Electric, Thermal, and Static Structural analyses
       meshes.
•   No 3D MVP or Line Edge contact elements
     – Requires conformal mesh with shared nodes across the joints, which makes
       modeling assemblies including frictional and pressure-dependent electric and
       thermal contact impossible, or
     – Non-conformal/ dissimilar mesh, with duplicate nodes across the joint. Magnetic
       coupling using CPINTF command requires nearly-matched meshing, which is
       difficult to achieve in a large assembly.
•   Above problems are greatly reduced if modeling the air enclosure, and
    modeling the magnetic coupling across the joints, are not necessary
     – May be valid for materials with a relative magnetic permeablity = 1.
     – Goal: Prove with a comparison study.
 Merged Volumes




Outer-most Lamination Arch Segment with Air Enclosure: Solid Model
                                            Conformal Mesh:
                                            Nodes shared at Interface
                                            (perfect magnetic coupling)




Outer-most Lamination Arch Segment with Air Enclosure: Mesh
Arch Segment w/ Air Magnetostatic Model Results: Current Density (A/m^2)
Arch Segment w/ Air Magnetostatic Model Results: Joule Heat
         SOLID236:
         LINE EDGE METHOD




Arch Segment w/ Air Magnetostatic Model Results: Magnetic Flux (Metal +Air)
Arch Segment w/ Air Magnetostatic Model Results: Magnetic Flux (Metal Only)
Arch Segment w/ Air Magnetostatic Model Results: Current Density
Arch Segment w/ Air Magnetostatic Model Results: Lorentz Forces (N)
Arch Segment w/ Air Magnetostatic Model Results: Magnetic Flux (Metal Only)
Arch Segment w/ Air Magnetostatic Model Results: Lorentz Forces (N)
      SOLID186
      Stress and reaction force results closely
      agree with hand-calculated values.




Arch Segment w/ Air Static Structural Model Results: von Mises Stress (Pa)
       SOLID236
       LINE EDGE METHOD




Arch Segment _No Air - Magnetostatic Model Results: Magnetic Flux (Tesla)
Arch Segment _No Air - Magnetostatic Model Results: Current Density (A/m^2)
Arch Segment _No Air - Magnetostatic Model Results: Lorentz forces (N)
Arch Segment _No Air - Magnetostatic Model Results: Magnetic Flux (Tesla)
Arch Segment _No Air - Magnetostatic Model Results: Lorentz Forces (N)
        SOLID186
        Stress and reaction force results closely
        agree with hand-calculated values.




Arch Segment _ No Air - Static Structural Model Results: von Mises Stress (Pa)
               Conclusion
• SOLID117 results are not valid, but
  SOLID236 results using line edge method
  agree with hand-calculated classic solution
  values.
• No difference between results with air
  enclosure modeled and without. Note:
  Modeling without air enclosure is valid only
  for cases where all materials have a
  relative magnetic permeability = 1.

								
To top