Database logistics help

Hello,
I have a database in which I track production units at an agricultural farm.
I am having trouble to create a set of related tables that have random relationships.
Here’s the setup
I have a table called transfergroup
PK = TGID
this table gives rise to several “types” of transfer which live in a separate table and get a unique id by TGID+”type”
My problem comes where i used TGID+”type” to graft onto a new TGID+”type”2 so in the cyclical relationship I have trouble referring to the differences between the starting TGID and the resulting TGID any help to be offered today?