As you said, CHARSET=UTF8 is recommended. Please use it.
The only reason that CHARSET=UTF8 might not work for your application is for retrieving fixed-width CHAR columns, and your application has difficulty with the trailing spaces that are appended by the Teradata Database when using CHARSET=UTF8. For this situation, CHARSET=UTF16 is the recommended alternative.
The CLIENT_CHARSET connection parameter is only intended for compatibility with legacy applications that erroneously stored non-Latin character data into Latin columns. Please do not use it for any other purpose. Please avoid it.
As you said, CHARSET=UTF8 is recommended. Please use it.
The only reason that CHARSET=UTF8 might not work for your application is for retrieving fixed-width CHAR columns, and your application has difficulty with the trailing spaces that are appended by the Teradata Database when using CHARSET=UTF8. For this situation, CHARSET=UTF16 is the recommended alternative.
The CLIENT_CHARSET connection parameter is only intended for compatibility with legacy applications that erroneously stored non-Latin character data into Latin columns. Please do not use it for any other purpose. Please avoid it.