Design of Microwave reactor

[Mirko Scanferla]

Good evening,

I would like to clarify that this is not exactly within my field of expertise. I am a chemical engineer, and my doctoral project involves the development of a catalytic reactor heated using microwaves.

According to my initial idea, the reactor's shape could be as shown in the attached file. The operating frequency is 2.45 GHz, and currently, the microwaves are supplied by a solid-state generator. The connection between the generator and resonator is established through a 7/16 DIN coaxial cable.

The first transition zone, as well as the diameter of the reactor (vertical cylinder), has a diameter of 0.7*lambda (8.8 cm). The connection between the vertical and horizontal cylinders occurs at a height of lambda/3 from the upper end.

The catalytic bed will have dimensions of approximately h = 10 cm. The bed's position will be determined once a simulation (using COMSOL) is conducted, ensuring that the electromagnetic field inside is uniform. The simulation will then be replicated with the catalytic material.

So far, I have obtained a simulation where S11 is nearly zero, indicating that, if I'm not mistaken, almost all the power is reflected. My goal is to achieve the most uniform temperature possible within the catalytic bed.

I would like to ask if the shape and dimensions seem suitable and if you have any suggestions on how to proceed with the work. Additionally, I am interested in knowing if you are aware of any books or guides on the design of such equipment.

Thank you in advance.

 

[FvM]

Most relevant information missing
- coaxial to waveguide coupler, how?
- dielectric properties of reactor content

even the reported s11 is unclear, do you mean
0 dB (full reflection) or
0 (perfect absorption)

 

[Mirko Scanferla]

Good afternoon,

I hope to be as clear as possible.

Coaxial to WG Coupling
I would like to establish an initial connection between the coaxial cable and the waveguide. From my readings, I understand that using a rectangular WR-340 waveguide provides better control over power. Subsequently, a transition can be made to a cylindrical waveguide for improved field propagation. The end of the cylindrical waveguide should be connected perpendicular to the reactor. Theoretically, this perpendicular junction helps protect the generator from potential reflected waves. The junctions are of a mechanical type.
Is this the information you asked for?

Dielectric Properties
The catalytic bed consists of a mixture of Silicon Carbide (80% w) and a catalyst (20% w) based on Nickel and Alumina. Silicon Carbide is essential for wave absorption, as Alumina cannot absorb many microwaves. I am conducting a dedicated experimental campaign because dielectric properties vary with the particle size of the two materials and temperature (operating at approximately 900°C).

S11
S11 refers to reflected power. In another simulation with a different geometry, I managed to reduce S11 to around -8dB. Regarding this, what could be a reasonable limit for this parameter?

Thank you.

 

[FvM]

You stated in post #1 "S11 is nearly zero". Apparently you mean |S11| near 0 dB?

 

[Mirko Scanferla]

you exactly that

 

[FvM]

In the first place I'm asking about the exact design of coax to horizontal waveguide transition. I also wonder if the cutoff frequency of horizontal waveguide is below 2.4 GHz?

 

[D.A.(Tony)Stewart]

I see that you will have many challenges;

1. absorption impedance vs thickness relative to the quartz surface of the cylinder which affects s11.
2. dispersion vs geometry for even power absorption or distribution. (Antenna pattern)
3. Thermal resistance of your lossy dielectric( or conductance) vs thickness and same for thermal conduction to the cylinder, then same from cylinder to external air to determine the total thermal losses and thus power required to elevate the temperature.

I suggest you cannot disperse the energy due to the properties change with fractional wavelengths and varying radiation strength as a function of geometry. This is solved in microwave ovens by using a glass rotating disk. (Pyrex?)
- if you put marshmellows or similar you will be able to observe the hotspots every half wave Peak. Grapes will explode.

- you may have to choose a Lambertian (2D) or more spherical shape container or with a centre feed in a cylinder or a horn antenna to some other shape the disperse energy evenly in this short range.

The chamber must be equal "transmission line length" from the antenna to target to hope to achieve energy transfer evenly then be at N half-lengths from the emitter antenna. Then the antenna must survive 900'C.

The dielectric properties can be measured between parallel plates to a smaller scale with a small air gap. Again the displacement must be close to a half wave or multiples to minimize the return loss < -15 dB which is considered adequate for most applications.

 

[FvM]

I basically agree that the waveguide and reactor shape will hardly achieve uniform power dissipation. Field strength in a cylindrical resonator varies in axial, radial and depending on the mode also angular direction. I would go one or two steps back and consider 1. suitable reactor volume 2. waveguide shape, depending on intended uniformity and material loss factor. Coupler design comes next.