Skip to content

SC 001 - AC losses in HTS tape demo

In this step-by-step tutorial, AC power loss in a high-temperature superconducting (HTS) tape is simulated.

HTS materials are defined as having a relatively high critical temperature of above 77 K. 1

Model definition

The tape model is multi-layered, with the superconducting YBCO (Yttrium barium copper oxide) layer taking up only a small part of the tape cross-section. Hastelloy is used as substrate, which forms the thick middle layer of the tape. Copper is used on the outer layer as a stabilizer.

A reference image of the tape cross-section (not to scale) as well as an image of the actual model in Allsolve model view are depicted below.

Tutorial image

Tutorial image

Model geometry

ElementDimensions
air cylinderradius = 8 cm
tapewidth = 4 mm, height = 95 μm, length = 1 cm
copper layerthickness = 20 μm
silver layerthickness = 2 μm
YBCO layerthickness = 1 μm
hastelloy layerthickness = 50 μm
domainlength = 1 cm
YBCO cross-sectionAYBCOA_{\rm YBCO} = 3.961093.96 \cdot 10^{-9}

Material Data

Magnetic permeability (μ\mu)

  • all domains: μ0\mu_0

Electric resistivity (ρ\rho)

  • Hastelloy: 106 Ω10^{-6}~\Omegam
  • Silver: 108 Ω10^{-8}~\Omegam
  • Copper: 108 Ω10^{-8}~\Omegam
  • YBCO:
    • ρ=EcJc(JJc)n1\rho=\frac{E_c}{J_c}\left(\frac{||J||}{J_c} \right)^{n-1}
      • Ec=100E_c = 100 μV/m
      • n=30.5n=30.5
      • Jc=2.85×1010J_c=2.85\times 10^{10} [A/m2^2]
      • Ic=JcAYBCOI_c=J_c\cdot A_{\rm YBCO}
    • EE0+EJ(JJ0)E\approx E^0+\frac{\partial E}{\partial J}(J-J^0)

Inputs - case 1

  • Frequency f=50f=50 Hz

  • Operation current Iop(t)=0.8 Icsin(2πft)I_{\rm op}(t) = 0.8 ~ I_c \cdot \sin (2 \pi f t)

  • External magnetic flux density Bext(t)=0B_{\rm ext}(t)=0 [T]

Inputs - case 2

  • Frequency f=50f=50 Hz

  • Operation current Iop(t)=0I_{\rm op}(t)=0 A

  • External magnetic flux density Bext(t)=20sin(2πft)B_{\rm ext}(t) = 20 \cdot \sin (2 \pi f t) [mT]

Output results

  • Joule losses in the YBCO region, and in the normalconducting region
Pi(t)=ΩiE(t)J(t) dΩ\begin{equation} P_i(t)=\int_{\Omega_{i}}\boldsymbol{E}(t)\cdot \boldsymbol{J}(t)~\rm{d}\Omega \end{equation}
  • Field visualizations

Step-by-step guide

Here you’ll find a detailed step-by-step tutorial on how to simulate AC Loss in an HTS tape in Quanscient Allsolve.

Step 1 - Create the project and import geometry

  1. Create a new project and name it as HTS tape demo.

  2. Import the model as a mesh file. You can download the file here: htstape.msh

The air cylinder takes up most of the model view, with the small rectangular tape volume visible in the middle.

Tutorial image

Step 2 - Define shared regions

  1. Go to the Properties section.

  2. Define shared regions:

    Region nameRegion typeTarget
    airVolumevolume 5
    copperVolumevolume 6
    silverVolumevolumes 1, 4
    hastelloyVolumevolume 2
    ybcoVolumevolume 3
    tapeVolumevolumes 1 - 4, 6
    normalconductingVolumevolumes 1, 2, 4, 6
  3. Define these Surface shared regions at the end of the tape:

    Region nameRegion typeTarget
    s_copperSurfacesurface 32
    s_silverSurfacesurfaces 13, 25
    s_hastelloySurfacesurface 17
    s_ybcoSurfacesurface 21
    s_tapeSurfacesurface 13, 17, 21, 25, 32

Step 3 - Define the materials

  1. Assign the predefined Air, Silver, Copper and YBCO materials to their corresponding shared regions.

  2. In material properties, change Electric conducticity for Silver and Copper to 1e8.

  3. Change the YBCO material color to purple in order to distinguish it from Hastelloy.

  4. Create the new material Hastelloy and add its properties:

    Material nameColorTarget
    HastelloyDark greyhastelloy region (volume 2)
    Material PropertyValue
    Electric conductivity1e6
    Magnetic permeabilitymu0

Step 4 - Define shared expressions

  1. Edit predefined shared expressions:

    NameUpdated expression
    YBCO_Jc2.85e10
    YBCO_n30.5
  2. Define new shared expressions:

    NameExpression
    freq50
    Bext0
    Aybco3.96e-9
    Iop0.8 * YBCO_Jc * Aybco * sin(2 * pi * freq * t)

Step 5 - Define physics and boundary conditions

Add to the Physics section.

Add the Magnetism φ and Magnetism H physics before moving on to set up their interactions.

Physics 1 - Magnetism φ

  1. Set the Magnetism φ target:

    PhysicsTarget
    Magnetism φair region (volume 5)
  2. Add a constraint interaction to Magnetism φ:

    Interaction nameInteraction typeTargetValue
    ConstraintConstraintpoint at the outer edge of the air cylinder (point 26)0
  3. Add an External field interaction to Magnetism φ:

    Interaction nameInteraction typeTargetValue
    External fieldExternal fieldouter surface of the air cylinder (surface 26)[0; Bext; 0]

    Tutorial image

  4. Add a Lump I/V cut interaction to Magnetism φ:

    Interaction nameInteraction typeTargetActuation modeCurrent
    Lump I/V cutLump I/V cuta loop around the tape cross-section (curves 46, 48, 50, 51)CurrentIop

    Tutorial image

Physics 2 - Magnetism H

  1. Set the Magnetism H target:

    PhysicsTarget
    Magnetism Htape region (volumes 1 - 4, 6)
  2. Add H-φ coupling to Magnetism H.

Step 6 - Select simulation options

  1. Go to the Simulations section.

  2. Add a new simulation.

  3. Set Analysis Type to Transient.

  4. Set Timestep algorithm to Implicit Euler.

  5. Set Start time to 0 [s].

  6. Set End time to 0.02 [s].

  7. Set Timestep size to 0.001 [s].

  8. Set Solver mode to Direct solver.

  9. Select the imported mesh as the mesh for your simulation.

Step 7 - Add simulation outputs

  1. Add custom value outputs for joule loss integrals:

    NameOutput expression
    Pybcointegrate(reg.ybco, transpose(E)*j, 2)
    Pncintegrate(reg.normalconducting, transpose(E)*j, 2)
  2. Add custom value outputs for net currents:

    NameOutput expression
    Itotlump.I
    Iybcointegrate(reg.s_ybco, on(reg.ybco, compz(j)), 2)
  3. Add the j field output with ybco region as target (volume 3).

  4. Toggle Skin only on the j field output.

Step 8 - Modify the simulation script & run

  1. Open the simulation Script.

  2. Enable Scripting mode.

  3. Change the φ interpolation order to 1. Example image

  4. Replace the first line of the autogenerated magnetism H formulation with the following Newton-Raphson linearization:

    rho = 1 / par.sigma(df.j)
    dedj = rho * qs.eye(3) + (expr.YBCO_n - 1.0) * rho / qs.max(df.j * df.j, 1e-40) * df.j * qs.transpose(df.j)
    dofe = rho * df.j + dedj * (qs.curl(qs.dof(fld.H)) + var.curl_dof_Hs - qs.curl(fld.H) - var.curl_Hs)
    form += qs.integral(reg.ybco, dofe * (qs.curl(qs.tf(fld.H)) - var.curl_tf_Hs))
    form += qs.integral(reg.normalconducting, qs.inverse(par.sigma(df.j)) * (qs.curl(qs.dof(fld.H)) + var.curl_dof_Hs) * (qs.curl(qs.tf(fld.H)) - var.curl_tf_Hs))

    Example image

    The Newton-Raphson linearization helps in speeding up the convergence of nonlinear systems, leading to faster runtimes. In some cases, the nonlinear system does not converge at all with fixed point iteration unless linearized.

  5. Change timestepper tolerance to 1e-4 and j field output region to reg.s_ybco.

    Example image

  6. Save the script.

  7. Run the simulation.

Step 9 - Visualize results

  1. Add a visualization for the j field.

  2. Add Glyph to the j field visualization with options as below.

Example image

References

Footnotes

  1. https://en.wikipedia.org/wiki/High-temperature_superconductivity