Adûnaic derived verbs are evidentally constructed by adding a verbal suffix to an Adûnaic root. There are only two unambiguous derived verbs attested: azgarâ- and ugrudâ- (SD/439). Both begin with a vowel, but it is not clear whether this is true of all derived verbs. Only the first of these, azgarâ-, has any attested inflections. These inflections have similar glosses but different forms in each draft of the Lament of Akallabêth:

First draft	azgaranādu	“was waging war‽”	SD/311
Second draft	azagrāra		SD/312
Final draft	azaggara	“was warring”	SD/249

Given the glosses, these appear to be examples the continuative-past tense, but this is the only unambiguous instance of this tense. It isn’t clear whether the different forms are due to changes to the verb stem, changing grammatical rules or changing verb tenses (or some combination of these). This makes the formations almost impossible to analyze.

My best guess is that the first two examples are, in fact, in the continuative past, with changing rules between the draft and later Adûnaic grammars. I think that the final form is instead a simple past tense, and that Tolkien neglected to change the gloss. Based on this (extremely speculative) analysis, I would use the follow rules to conjugate Adûnaic derived verbs:

Tense	Formation	Examples
aorist	identical to the verb stem with a short final vowel	azgarâ- → *azgara
continuative-present	?	?
past	double the middle consonant, add a stem vowel to break up clusters of more than three consonants and shorten the final vowel	azgarâ- → azaggara (SD/247)
continuative-past	displace the second stem vowel before the middle consonant and repeat the derivative suffix with a short vowel	azgarâ- → azagrāra (SD/312)

As with biconsonantal verbs, the aorist formation is a guess based on the aorist form of the triconsonantal-verb. There isn’t enough information on the continuative present to guess how it might be formed. See the entries for the individual verb tenses for further discussion.