If I rightly remember FB is in multiples of the block sizes, that is just how DASD stores the files on disk, it must fit in that multiple block size, which increases speed and throughput on the Mainframe. If the data file is not within the boundary of multiple block sizes (This has nothing to do with the actual size of the data), the DASD system just access files in blocks of 256 bytes...there will be a host of special fields inserted into the data file to describe the blocking and so on...which will get inserted when transferred to the mainframe and that data gets transferred to magnetic tape backups...
There should be a script available on the Mainframe to convert it using JCL (Job Control Language)..ask the Mainframe administrator to do it for you...
By the way it should be noted that the character set you used in your data file, just be aware that the mainframe uses EBCDIC character set...There are plenty of tools out there that can convert from ASCII data to the format to be readable by the mainframe, just something to bear in mind of...If the data gets converted that could impact the file size...Thought it would be worth mentioning and important!
There is a Unix/Linux utility that can convert the data to a fixed block size using the dd
utility, although I do not think it would be the right way to do it...
Here's a useful link that will help you in understanding this. And also here on SO a similar user was asking about MVS/TSO data...
Hope this helps,
Best regards,
Tom.