PyFEA Quasi-Static Simulations for thermal & magnetic characteristics

[wgbowley]

1. Replace old API calls with new validate ones
2. Use correct replacement materials due to deprecation.
3. Use reference design to validate pipeline behavior (cmore)

[wgbowley]

This one really doesn't mean anything currently. It might for prototype gemma but this is really were PyFEA comes in for prototype beta:

Finite element quasi-static simulations (thermal -> Asymptotic temperature, inductance vs current & flux linkage vs current)
Keeping it simple for now as the simple simulations are easy to validate fast and mostly just parameters that are geometric base like inductance, flux density, field strength over the machine

[wgbowley]

I think this is reasonable for this implementation:

Folder PATH listing
Volume serial number is 0000005E 6A01:1985
C:.
│   evaluator.py
│   readme.md
│   
├───00_wound_armature
│       model.py
│       parameters.uiv
│       readme.md
│       
└───01_pcb_armature
        model.py
        parameters.uiv
        readme.md

Important

The parameters.uiv will control those models. Its just a matter of the user changing the import in evaluator.py and changing the path to parameters.uiv for which ever their using and they can get it working.

[wgbowley]

Seems reasonable for a pcb motor:

[version]
format: 0.1.0


[notes]
type: "Electromagnetic Parameter File"
application: "Finite Element Model | Tubular Ironless PCB Armature Linear Motor"
coordinate_system: "axisymmetric (Z-R) (Axial-Radial)"


[model]
environmental_material: "air"
line_voltage:   24 (V)


[armature.core]
number_slots:       6 (∅)
number_sub_slots:   5 (∅)

rad_clearance:      1 m(m)
pcb_edge_clearance: 0.5 m(m)

[armature.slot]
rad_thickness: 5 m(m)

[armature.sub_slot]
material:           "copper"
trace_thickness:    70  u(m)
rad_fill_factor:    0.8

axi_len_substrate:  1.0 m(m)
axi_layers:         2 (∅)


[stator.tube]
length: 300 m(m)
rad_thickness: 1 m(m)

[stator.dipole]
material: "NdFeB"
grade: "N52"

axi_length: 10 m(m)
rad_thickness: 5 m(m)

[wgbowley]

Because I am most interested in the pcb design right now I'll do the model for that first and than come back and do the normal wound topology next.