If you download the html generated files 
these scan rules are useless :
+www.website.com/articles/*
+www.website.com/articles/*/*.jpg
+www.website.com/articles/*/*.gif
Normally, you don't have to disable Persistent connection 
or enable Get non-html files(ZIP, images etc).

but there may be a robot.txt file or meta tags excluding 
the download of the images.
Have a look in the hts-log.txt.
Does WinHTTrack build relative or absolute links to these 
images?