Some questions about slot line (about wavelength, characteristic impedance)

I use HFSS
I have slot line,

1. How to know characteristic impedance of slot line?

I need to know width of slot for which slots characteristic impedance = 100 Ohm
For this I make this model (fig):

50 Ohm -Lump port is in the middle of slot, slot is ended by two 100-Ohm impedances
For some length and width of slot VSWR obtained:

Can I say that characteristic impedance of slot line for f=2.5GHz is about 100 Ohm?

2. How to know its wavelength?
I know this way: measure dependence VSWR of shorted slot line"s length (fig) for one frequency

lambda/2 = distance between minimums = 50mm
Right? Or there is another way, particular maybe in HFSS is some feature?

3. Please, take a look at model:

Lumped port is sheet, wide of lumped port = 1/3 of slot"s wide
How it should be right?

Thanks for help.

Slotline is not modeled with a lumped port in the middle of the line. You should have Lumped ports or waveports at both the ends. And you need to see the field distribution of the ports. The field lines should go from one conductor to the other.
In HFSS, there is any option to plot lamda and you can determine the wavelength.

What is dielectric constant and height of the substrate. There are few IEEE papers on Slotline and there are design equations for calculating the characteristic impedance and slotline wavelength. They are accurate upto 2%.

You might want to look into those references, they are really helpful.

-sv

I've physical model: loop-form slot antenna; feeding: connector is connect to inner and outer sides of loop.
dielectric constant is unknown, height of the substrate = 1.5 mm, width of slot w = 1.5 mm, lambda/2 = 50 mm (f = 2.4GHz) (3-rd fig., obtained for straight slot in the same material), impedance of connector = 50 Om.

I'm trying realize this model in HFSS. To determine dielectric constant I made straight-line model (4-th fig.) and for w=1.5mm, changing dielectric constant I achieve lambda/2 = 50 mm (using fig.3 method).

To feed loop-slot I've to made coaxial and excite it? Why I can't model feeding with a lumped port in the middle of the slot line directing the field lines from one conductor to the other?

sv1437, In those really helpful references results given in grafical form for certain dielectric constants.

Which IEEE papers or references did you go through ?

Because the reference papers by Cohn and Janaswamy.R, have design equations to calculate Zo and effective dielectric constant and slotline wavelength if the dielectric constant is more than 10.
The book on "Transmission line design handbook" has design equations for er>2.2.

You don't know the dielectric constant of the substrate that you are using ? What parameters did you set for the substrate to model in HFSS.

Did you look at the field distribution of the ports ? Are field lines directed from one conductor to the other ?

-sv

Substrate - "glass fiber", simple material for PCB, don't know exactly how this is in english, experimentally I can't measure dielectric constant, using HFSS I can see that for dielectric constant=3.8 lambda/2 is about 50mm (for w=1.5 mm, height of the substrate =1.5mm).

Defining port line, I pointed two points: at opposite sides of sheet from one side of slot to another, field can be seen :

This how I feed ring-like slot-line structure, , I wonder if I right select width of sheet on which port line lies (1/3 of width of slot).

From the figure, most of the field is directed towards one conductor (which is on the top). Did you see the port field display for your lumped port ? (Not the magnitude of the E field, but the field lines).

-sv