Shunt-to-tank clearance changes magnetic collection, the oil passage and mounting heat transfer. Compare loss and available branch flow before selecting a gap. [1]
Hold winding excitation fixed across clearance candidates
Use the same winding geometry, current condition and material basis when comparing clearances. Let the model recompute flux redistribution as the shunt moves. Imposing an unchanged incident flux on every candidate can be a useful isolated submodel exercise, but it may miss the change in coupling caused by the clearance itself.
Measure clearance from defined surfaces and include brackets, tolerances and local protrusions. The nominal drawing distance may not describe the smallest oil throat or the closest magnetic region.
Build a magnetic–thermal comparison
| Clearance candidate | Magnetic record | Thermal record |
|---|---|---|
| Smaller gap | Tank field, shunt peaks and redistributed losses | Available oil path and mounting heat transfer |
| Intermediate gap | Same quantities under the same excitation | Same thermal model with updated geometry |
| Larger gap | Field bypass and neighbouring-wall response | Flow access and local temperature distribution |
Compare candidates rather than presuming that the smallest gap is best. The magnetic field method does not supply the oil-flow boundary; obtain flow from the connected hydraulic system. [1] [2]
Update the thermal model honestly
If oil flow is resolved, include the actual passage and surrounding route rather than a short isolated gap with arbitrary inlet conditions. If a simplified convection boundary is used, state how its coefficient was obtained and whether changing the clearance invalidates it. Reusing one coefficient for all gaps can suppress the very cooling effect being investigated.
Include heat transfer through supports and into the tank where relevant. A mounting change made to achieve a new gap can alter conduction independently of oil access. That is part of the practical design comparison, but it should not be mislabelled as a pure clearance effect.
Inspect the worst local result
Track tank, shunt and support temperatures and the location of their maxima. A lower tank-average loss can coexist with a hotter shunt end or a restricted oil channel. Keep the location map across the clearance sweep so a migrating hot spot remains visible.
For an independent screen, assume a 10 mm-gap branch carries 0.0050 kg/s oil and removes 100 W; a candidate 8 mm gap has 90 W but only 0.0035 kg/s after its assumed network solution. Use assumed cₚ = 2,000 J/(kg·K) and identical inlets. From ΔT = P/(ṁcₚ) [2], outlet rises are 10.00 and 12.86 K. Against an assumed 12 K outlet-rise cap the baseline passes and the narrower candidate fails despite 10 W lower heat. Retain the baseline pending local surface-temperature checks. These fictional network outputs illustrate the coupled decision; actual flow, bypass and wall convection must be measured or solved.
Select clearance against the required load cases and construction variation. Preserve source, oil-flow and mounting conditions with the chosen magnetic and thermal limits.
References
- Z. Jurković, B. Jurišić and T. Župan, Multi-Step Approach for Fast Calculation of Magnetic Field in Transformer Tank Shields, Energies 17, 1378, 2024. DOI: 10.3390/en17061378.
- U.S. Department of Energy, DOE-HDBK-1012/3-92, Thermodynamics, Heat Transfer, and Fluid Flow, Volume 3, 1992.