Whether IGBT's or FET's are best suited is dependant on a bit more thought
FET's have the benefit of their drive-cct not being so demanding and that they switch fast and lower on-state losses. BUT their SOA is actually pretty shite
IGBT's have higher on-state losses, switch slower BUT have a more rectangle SOA (this one fact makes them GREAT!!!)
General rule of thumb is anything below 50V (maybe 100V dependant on current level) then FET's, IGBT's come into their own with higher voltage blocking, saying that look at what yr current is demanding and yr SOA, you may find you may need to goto IGBT's
The first problem I see is the field winding in series with the bridge.
1) a H-bridge IF operating as a voltage-source (from what you are saying yrs is) NEEDS to be a voltage-source, that means big capacitance close to the silicon of the H-bridge.
Inductors == Current-source
Capacitors == Voltage-source
Inductance in the DC-link is VERY BAD!!!! so bad infact...
as it stands as you start to PWM: you turn an diaganal pair of switches on and current builds up in the motor AND the windings. WHAT happens when the PWM turns the switch's off, yes the diodes of the H-bridge will allow the motor free-wheel current to flow BUT what path is there for the winding-inductor..
The path of It will jack the voltage up and break over the silicon, path completely destroying yr bridge within 1-PWM cycle
IF you can make it a parallel excited arrangement then you are onto a winner. Get a DC-link capacitor as well (is needed) and then re-gen can start being a reality.
say you have the DC-link at 12V (well 13V from a lead-acid) and you stop PWM'ing AND the machine's rotor is still turning (say just inertia or because down a hill) what will happen is when the backEMF of the machine (which is proportional to speed) exceeds the DC-link the rectifier-aspect of the H-bridge will kick in and it will start to "pump" the link up.
With a 12V battery there as well that will start to take the charge, thus charging the battery up.
What I would do is put a brake-cct across the DC-link.
The car-battery will only be able to obsorb (via charging) so much re-gen energy, anymore and it will just keep pumping the link up, so put a comparator cct across the link to trip at say.. 15V and reset at.. 14V, this then fires another IGBT/FET that is across the DC-link (with a resistor in series with it) to dump the extra energy