I asked this very question in the ChemDraw support fora and here's what I got back.
1 Question:
Which naming convention (IUPAC, etc.) is used to generate names using the Structure to Name tool?
Answer:
The names generated by ChemBioDraw are the ones that we feel are the most appropriate for chemists to use. In most cases, those names are consistent with at least one set of IUPAC recommendations. However, the IUPAC recommendations have changed over time, resulting in conflicts between different sets of recommendations and even within the same set of recommendation. Some IUPAC recommendations have never been accepted by the chemistry community, and are in fact rarely or never seen in real-life usage. Accordingly, there is no simple answer to this question.
2 Question:
I drew a structure, used the Structure to Name command to generate a name, then I used the Name to Structure command to generate the structure, but did not get the starting structure. Why?
Answer:
There is no particular reason to expect this process to work both ways. The name generated by ChemDraw is probably an accurate name for the structure you originally drew. The structure generated from the name is probably an accurate interpretation of the name. The name itself is probably slightly ambiguous, but not incorrect. There are probably other names that would regenerate the original structure. There are also other names that would generate even more dissimilar structures. Currently the IUPAC recommendations offer considerable flexibility when generating chemical names, and that flexibility inevitably leads to ambiguities.
3. Question:
I entered the name [name1], but the structure it gave me was really [name2].
I entered a name that violated IUPAC rules, and it gave a structure anyway!
The structure it gave isn't the one I meant.
Answer:
Name/Structure is designed to give a correct structure based on the name entered. If you enter a name that is ambiguous, you're rolling the dice with regard to the structure you get back. For example, the explosive TNT is really 2,4,6-trinitrotoluene. If you enter "trinitrotoluene", however, you should not expect to get this structure. You might, but you might also receive the 2,3,4-, 2,3,5-, alpha,alpha,alpha-, or any of several other isomers. None of these would be incorrect for the ambiguous name "trinitrotoluene". If you want a specific isomer, you must enter a specific name.
In a similar vein, Name/Structure supports many types of nomenclature that are not condoned by any naming body, including IUPAC, IUBMB, and CAS. There is absolutely nothing ambiguous about a name like 5-chloropentane. The correct name for that compound would be 1-chloropentane, but that does not prevent Name/Structure from interpreting the name it was given and producing a reasonable structure based on that name.
So it would seem that even the mighty ChemDraw makes guesses when it names structures or produces structures from names! These answers come from the ChemDraw FAQ database.
So good luck!