Stephan put alot of work into a few posts at least related to your subject right here in the forum.

http://aws.org/cgi-bin/mwf/topic_show.pl?pid=74813;hl=arc%20voltage%20smoke

The topic revolves primarily around aluminum data... But there is still good discussion about the relationship between Arc Voltage and fume production

First, thanks for kicking this to a different thread, I didn't want to overshadow the orighnal thread , but I was curious.

Lawrence, thanks to you too for adding that link.  That didn't capture my full attention the first time around.  Circumstances are different this time.  I'm going to have to print off Stephan's post so I can wrap my brain around it but I did scan through it.

What brings all this up is that our Safety folks have been testing welding fume emissions for manganese levels - a spin off from testing for hexavalent chromium and all the activities that went along with that.  I'm told, that will is the next element to be concerned with.

My immediate reaction was "How can you reduce manganese levels? It's integral to the steel itself."  Richard's response got me thinking there is much more to the story.

I don't think anything will prevent the need for good, positive ventilation when welding.  If it isn't hexavalent chromim, or manganese, it will be something else.  However, the advice on how to prevent cracking - reducing voltage to reduce manganese loss - should also serve to reduce manganese levels in the fumes.

What effect does a different shielding gas have when using the same electrode?  That is, if I'm using, say, Outershield 71M with 100% CO2 and 28 volts, would C25 (75%Argon/25% CO2) and 25 volts (same wire) yield a reduction in manganese loss to fumes?  I assume that is yes and if so, is that primarily due to the reduced voltage?  The reduction in active gas (CO2)? Or a combination of both?  (OS 71M can use either gas per Lincoln's data)