Ned Simmons wrote:

On Thu, 24 Jun 2010 11:01:48 -0700, Winston
wrote:

On 6/24/2010 7:01 AM, Ned Simmons wrote:
On Thu, 24 Jun 2010 00:29:27 -0500, Don Foreman
wrote:

On Wed, 23 Jun 2010 22:07:50 -0700,
wrote:



Hokay but the drum still has 2.8 x the seam length of the jerrycan.

That's a good thing. The total force is divided by the length of the
seam.

More is really better?

In this case, yes.


I thought the seams represented the weakest link.

I don't know that for a fact, but let's assume they are. Two inches of
weak seam will resist more force than one inch.


I stumbled across a claim that the flat top and bottom
of the drum go noticeably convex at only 5 PSIG.

5 PSIG is about 11.5 feet of water head. Head necessary to get the
stuff out of the drum is only 35" or 1.26 PSIG. "Noticibly convex" is
markedly different from "BOOM" so about 0.252 NC (Noticably Convex)
seems safe enough to me.

And the 35" of head represents the hydrostatic pressure that the
bottom head sees anyway when the drum is full. Add a few more inches
to get over the rim of the drum, and a bit more to get a reasonable
flow thru the spout, and I don't see a problem, conceptually anyway.
The real challenge is in making damn sure the drum is never
over-presssurized.

A failed gas regulator shouldn't cause anything worse than
minor disappointment. I am lazy by nature and would be
likely to leave the air connected. This approach would
make me nervous.

I think the proposal is to meter the air in, not the liquid out. The
spout is unrestricted and the liquid runs out as long as there's
enough pressure in the drum to overcome the head that results from the
highest point in the plumbing.

Just get the Harbor Freight air operated piston pump, they are designed
to be output throttled i.e. oil dispenser guns. Close off the outlet and
the air piston just stalls, the fluid line is pressurized ready to
dispense again. The drum is never pressurized.