On 08/31/2011 12:15 PM, David Mathog wrote: > That never crossed my mind. > > You sure about the flammability? I believe it for the ignition due to > temperature (Fahrenheit 451 and all that). However, I have a gut > feeling (but no data) that sparks are fairly likely to ignite cardboard, > and less likely to ignite a solid plastic sheet (polyethylene or > polypropylene, for instance). Not that I'm expecting sparks, but that > is a real possibility when a power supply fails. Maybe even a brief > flame. Of course paper won't hold up well compared to plastic if it > gets wet. Moisture resistance is not important here though - if the > insides of the computer are dripping, air shroud failure is the least of > my worries.
I'm aware of a machine room fire that was attributed to cardboard dust and the storage of flammable material (paper and cardboard). I wouldn't recommend cardboard or anything else that might generate flammable dust in a high 50-90C airflow environment with low humidity. Supermicro does seem to play pretty fast and loose with a shroud and cooling in general. We had nodes bouncing off the thermal max (and throttling) despite air intake temperatures 30F below the specifications while having very low power load in the node (read that as no expansion cards, one low rpm disk, and the lowest clocked CPU). We did however get them to ship us free shrouds once we complained. Is it really worth wasting even an hour to not get the real shroud? Not sure if this is the one, but they aren't particularly expensive ($13): http://www.provantage.com/supermicro-mcp-310-18003-0n~7SUP91KW.htm _______________________________________________ Beowulf mailing list, Beowulf@beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
