Roland,

A time-varying magnetic field will induce electrical currents in a _conductor_.

Those currents, through Joule (ie, resistive) heating generate heat.

If your dielectric has some conductance, it might warm.

Any specifics? Field flux rate, dielectric type, geometry, etc.

I am heating grains using RF energy. I have read that only electric field helps not the magnetic field. The grains are placed between the parallel plate capacitors (electrode).

Roland,

Well, for an RF wave to be absorbed well by an object, that object needs to be at least as large as the wavelength of the radiated energy.

(which is why an ant is quite happy in a microwave oven)

If the grains are cohered together by sintering, or simply are in contact with each other, they may present a large enough object such that the RF can 'see' the object and dissipate energy into the grains. That will depend on the conductivity of the grains - en masse.

In this case, as we're dealing with RF (not just a time-varying magnetic field) then the electric field dominates the energy density. You can improve the warming rate by reflecting the RF signal back through the material.

May I ask what sort (power, frequency) of signal you're anticipating using?

Time-varying magnetic fields may heat in 2 different ways:

Regarding your setup (dielectric between capacitor's plates), I'd guess that the dielectric is shielded such that it won't be heated directly. But the plates might be heated (currents induced), resulting in indirect heating of the dielectric by conducted heat.