80-90% is expected but more than that is difficult to achieve in practice. check your feed and ports. In simulation you may get more than 100% efficiency that is absolutely incorrect. So, if you are getting that high efficiency then check your model very carefully, try to change some dimensions to see the effect that you are still getting the similar results. If for all set of values you get similar results then the results may be true or you could do the simulations in another software to verify the results.

So what's the problem? HFSS can even calculate 110% . Usually that is not an error, but a meshing inaccuracy (both in antenna and on outer air-box boundary) combined with lossless model.

Add all these factors, make an experiment, see what you missed.

Rahul Agrawal What do you specifically mean by "scanning rate enhanced" and "good scanning rate"? I guess you mean rate of change of main lobe angle with frequency, but then what is so "good/enhanced" about it? The frequency scanning rate is dependent on how slow/fast your wave propagation in the transmission line (in your case SIW) is. Basically, you can play with the propagation constant of the SIW to control the scanning rate (imaginary part of the fisrt harmonic of the modulated fast wave) and radiation efficiency (leakage constant in the fast-wave, mainly determined by the perturbations). Now coming to the number 90%, my approach would be:

1. calculate input reflection loss as percentage in terms of input power

2. calculate output dissipated power as percentage in terms of input power

3. calculate dielectric loss (the substrate) in the antenna aperture as percentage of input power

4. similarly simulate your conduction loss (from top and bottom metal layers and the vias) in percentage of input power.

5. now, add up the above 4 quantities and substract it from 100. The number should represent radiation eficiency and can be used to verify your already simulated radiation efficiency. If they match then its okay or else there is problem.

PS: there can be some loss to higher order mode in your second port of the SIW. So inter-modal conversion loss (although can be small) but must be included into the above calculations.

Before claiming that 90% is a problem and blindly pointing to some dimensions or simulation shit or trying out prarametric sweep and even to that matter another software to verify it, be an engineer and believe in antenna maths. I would suggest you to back-calculate that number through an extensive power budget analysis of your structure.

also, I am a bit amazed by your graph of radiation efficiency vs frequency graph because generally for any planar transmission line based LWA (assuming the antenna has a very good braodband matching) the radiation efficiency decreases with frequency. The reasons are:

----the conduction loss and dielectric loss increases with frequency (very obvious)

-----leakage rate decreases because the filling factor of the mode increases with frequency: the modal energy is shifting towards the center of the transmission line, hence the effect of the surface perturbations (whose work is to modulate the slow wave inside the line to a fast wave radiating to the outer medium) decreases on the propagating mode, resulting in decrease of radiation.

hence (law of conservation of energy), theoretically the radiation efficieny must decrease with frequency, while the power dissipation to the other end of the line increases (this can be evident from S21 of your simulations).

However, if you have optimized your perturbation to have an enhanced leakage rate at higher frequencies, then I accept your graph is correct.

Thank you so much for the detailed explanation. It is very helpful @ Utpal Dey sir