How does Alt-Ctl-F1 really work, down deep?

I have been working on a 'kiosk' type application. I want no user logins (for security reasons), but need the alternate TTYs.
1) replace the bash shells on tty2-5 with 'stty onlret' for proper \n handling.
2) in rcx.d start 'appX > /dev/tty2', 'appY >/dev/tty3' ...
3) remove shell on tty1

Now the Ctl-Alt-Fx stops working. I assume that there must be someone requesting keyboard input to keep the keyboard active.
Where does the C-A-Fx processing take place in the stack between the bios kbd reader/kernel module/application?

Can I just start a process that sucks in keypresses to keep the kbd stack alive? "cat /dev/tty1 | /dev/null" (or something like that)?

Can I write an application that reads the keyboard directly, without concern of which tty is active? I didn't find a generic /dev/kbd.

BillS