thx,
PeterH, i hope so. As sensor i want to use mpx7002dp, 2% accuracy with zero-auto. Pressurerange -20mbar - +20mbars. Usually i think, "stuff" has its best accuracy at the end of its range, so thats why i wanted 20mbars.
I also have an mpx5001dp, which has its range up to 100mbars, but accuracy is around 5%, maybe that would be an option too. if i have a fitting setup id experiment to see if using weight/density/temperature->volume gives same results as calculation via pressures.
cycle, i think you linked something i already have in my first post 
http://www.ebay.de/itm/290756617726?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2649
zoomkat, thx for the idea. it looks nice but also like a lot of work and hard to do, as you need the equipment, metall, tools to create something like that.