Microwave Heating Simulation
Theory: heat from dielectric loss
Overview
I'm designing an industrial microwave dryer. How do I link the electromagnetic and thermal analyses?
Microwaves shake molecules in the material (especially water) with the electric field, producing heat through dielectric loss. The basic flow is to compute the field distribution electromagnetically, derive the heat-generation distribution, and pass it to the thermal analysis. But permittivity changes with temperature and moisture, so heating changes the generation pattern too. Whether you include this two-way coupling can change results greatly. First pin down three things: how deep it reaches, what pattern it heats in, and whether absorption rising with temperature causes runaway.
Heat generation and penetration depth
$\varepsilon'$ is relative permittivity, $\varepsilon''$ the loss factor, $E_{rms}$ the RMS field and $\lambda_0$ the free-space wavelength. $D_p$ is the depth at which power falls to 1/e; in thicker material the centre heats poorly.
A simple runaway model
If dielectric loss rises exponentially with temperature, the threshold is where the heat-generation curve touches the heat-loss line. Beyond it there is no steady temperature and the temperature keeps climbing.
A cooker born of a melted chocolate bar
The oft-told origin of the microwave oven is that around 1945, Raytheon engineer Percy Spencer, working near a radar magnetron, noticed the chocolate bar in his pocket had melted. He tried popcorn and eggs and confirmed that microwaves could heat food. The first product was the size of a refrigerator, but household units shrank in the 1960s–70s and spread worldwide. Today microwaves also serve industry: drying timber and pharmaceuticals, vulcanising rubber and sintering ceramics.
Worked examples
Example 1: penetration depth
| Material (ε′, ε″) | 2.45 GHz | 915 MHz |
|---|---|---|
| Water (78, 10) | 17.2 mm | 46.1 mm |
| Beef (52, 17) | 8.4 mm | 22.4 mm |
| Ice (3.2, 0.003) | 11.6 m | 31.1 m |
Permittivities are typical values, assumed the same at 915 MHz (they actually vary with frequency). At 1 kV/m water absorbs 1.36×10⁶ W/m³, heating at 0.33 K/s. The half wavelength is 6.1 cm in air, 0.69 cm in water and 0.85 cm in beef — the spacing of standing-wave highs and lows.
Microwaves reach 11 m into ice, yet it barely heats?
It reaches deep precisely because it's barely absorbed — ice's loss factor is about 1/3,000 of water's. So when heating frozen food, parts that melt first become water, absorption jumps and those spots overheat. The melted regions heat further and melt their surroundings — runaway thawing — sometimes cooking the outside while the inside stays frozen. Defrost modes pulse power on and off to let heat conduct and even out.
Example 2: thermal runaway
A 10 cm³ sample with 26 cm² surface and 20 W/m²K heat transfer, loss factor 0.01 at 25 °C rising e-fold per 100 K (2.45 GHz). The threshold field is 11.8 kV/m:
| Field | Result |
|---|---|
| 10 kV/m | Steady at 64 °C |
| 11.5 kV/m | Steady at 94 °C |
| 12 kV/m | No steady state (142 °C at 3 h, still rising) |
| 13 kV/m | Passes 300 °C in about 66 min |
A 4% increase in field decides whether a steady temperature exists. In materials whose loss rises with temperature, such as ceramics being sintered, slight field non-uniformity can trigger local runaway, causing cracking or melting.
Analysis workflow
- Measure (or gather) permittivity and loss factor as functions of temperature, moisture and frequency.
- Solve the field including cavity and waveguide to obtain heat generation.
- Pass generation to the thermal model, update permittivity with temperature and re-solve the field (two-way coupling).
- Include turntables, mode stirrers or conveyor motion to evaluate averaging.
- Validate with fibre-optic thermometers or IR cameras (metal thermocouples disturb the field).
“Only the prototype's corners scorch”
In industrial microwave heating of food, products in square trays scorched only at the corners. Field analysis showed the electric field concentrated at corners and edges, where waves entering from two or three directions overlapped and generation rose. As moisture left and temperature rose, the loss factor changed and the imbalance grew. Rounding the tray corners and adjusting layout and power fixed it. Corners are common starting points for local overheating in microwave heating.
Common mistakes
Mistakes and fixes
| Mistake | Effect | Fix |
|---|---|---|
| Constant permittivity | Runaway and unevenness missed | Make it a function of temperature and moisture |
| Solving the field once | Generation pattern drifts | Re-solve with updated properties |
| Grid coarse relative to wavelength | Standing waves lost | Below 1/10 of in-material wavelength |
| Measuring with metal thermocouples | Field disturbed | Fibre-optic thermometers |
| Setting temperature by power alone | Runaway past the threshold | Control power by temperature |
I'd like to learn related topics.
Check electromagnetic basics with the skin depth calculator and microwave transmission simulator. For related heating analyses see microwave heating basics, dielectric heating, RF heating and induction heating.
Related Topics
Experience the theory firsthand with the interactive simulator for this field
All Simulators