You can’t create directly HLS video transport segments (.ts) from audio raw source. You need to encode it with some encoder and then mux it before sending to hlssink plugin.
One of the problems that you’ll encounter is that the hlssink plugin won’t split the segments with only audio stream so you are going to need something like keyunitsscheduler to split correctly the streams and create the files.
An example pipeline using voaacenc to encode audio and mpegtmux to mux would be as follows:
gst-launch-1.0 audiotestsrc is-live=true ! audioconvert ! voaacenc bitrate=128000 ! aacparse ! audio/mpeg ! queue ! mpegtsmux ! keyunitsscheduler interval=5000000000 ! hlssink playlist-length=5 max-files=10 target-duration=5 playlist-root="http://localhost/hls/" playlist-location="/var/www/html/hls/stream0.m3u8" location="/var/www/html/hls/fragment%05d.ts"