Transformer 3.5kW 100kHz?

I need design and produce transformer
Primary side
DC 315V switched on full bridge frequency about 100kHz
n=N1/N2=1.2
because leakage inductance on primary enters the resonance tank, I need to know leakage inductance.

I can not help but do not know how to grasp it?
What the core using, several toroids , what materials?
Classical ferrite transformer core?
So large core are commonly available?
There is software for the transformer design so that I do not have to count everything manually?

ok i try it differently

entered
U=315V
f=100kHz
P=3500W

I suppose
T transformer charge core time 3us
I will try

Iron Powder toroid T200-52B
AL= 155nH/N2
Ae=2,32cm2
NT Number toroids in transformer 5 pcs
Smax induction of saturation 0,2T

Number of turns for full saturation core
N=U*T/Smax/Ae
N=315*3E-6/0,2/(5x2,32E-3)=4 turn for saturation on 0,2T
Induction
L=Al*N^2*NT= 155E-9*4^2*5=12,53uH

Ripple Current
I=U*T/L
I=315*3E-6/12,53E-6=75A
Energy storage in core on 1pulse
WP=1/2*L*I^2
WP=0,5*12,53E-6*75^2=0,035629j
Energy storage in core on 1S
w=wp*f
W=0,035629*100000=3562W


Result
Over transformer composed of five 5pcs T200-52B
N1= 4 turn
N2= 5 turn
will pass 3,5kW


Is it a complete nonsense to approach this as follows?
If not,
It's material Iron Powder 52 good choice?
4 turns will come to me like very little?

WHAT use wires?
Skin effect for 100khz have Skin Depth arount 200um.
Idelal cable id varnished Copper Litz Wire for example 10x0,4mm, but I do not buy it in a small one
as is the case with unvarnished coper wire litz?

When designing a ferrite or powder core, you have to look primarily for core losses versus T, not only saturation. Powder core manufacturers have many curves and tables about core losses, for a first estimation of transformer designs there may also tables of power handling versus core size and frequency. Did you check if your design is in the ball park?

I presume a 3500W converter will be designed as forward or push-pull type. Respectively you want high magnetizing inductance and low energy storage in the core, or in other words, high µr core. Your core selection doesn't seem reasonable in this regard. For transformer application, ferrite core may be a better choice in general.

As for litz wire source, there are specialized catalog distributors supplying in small quantities. Using stranded magnet wire is another option, particularly for low winding count.

You say Bmax = 0.2T at 100kHz? i.e. +/- 200mT ? this will make the cores pretty hot - have you checked the core loss graphs..?

Normally 100KHz => ferrite 100mT max B (each way )

- - - Updated - - -

200mT = 2000 gauss? from memory => losses = 5000 mW per cm3, way to high to be realistic I'm sorry. from datasheet.

T200-52B is only u = 75, thus you will store a fair bit of energy in the core - if you want to do that - OK - but a real transformer stores very little energy - you usually want to get it all to the output - so if you are seeking a Tx with low magnetising current ( hi Lmag ) then you need ferrite... (can still be toroids though )

- - - Updated - - -

**broken link removed**

Honestly, I knew it would not be good solution.
The problem is, I do not know when the energy is stored in the transformer core and when not.

If I simplify situation on this case

transformer is matching fransformer for impedance matching and Separating Protective Transformers becouse R have water cooling


Why am I in my consideration not using ferrite?
Ies ferrite have high u and easy to overcome.
Transformers are solved by an air gap, but toroids do not have air gap.


Just as I said at the beginning I do not know how to keep it whole, I would welcome some tutorial how to proceed with the design and what is important

really you want ferrite, 3F3, 3C96, N87 at 100kHz, design for 80mT peak, for losses ...

yes, i think ferrite core is good for this
use E or PQ or RM core shape
you can use copper foil winding if currents is high
is it serial device or single sample for testing?