Shown below are the first two tanks I made using two different fastening
methods. The tubing is 2.250" X .049" wall 6061-T6 purchased from Wicks
Aircraft Supply. The ends are also 6061-T6 purchased from McMaster-Carr.
You can't tell from this photo but there's a Viton o-ring in there too
which actually seals the end piece.
The end piece in the tank on the left was fastened using .125" diameter
aluminum flathead aircraft rivets (two are shown.) Each rivet was
pressed into a blind hole which was drilled through the tubing and into
the end piece. Each hole was finished with a reamer to insure the
right interference fit, and then countersunk. As complicated as that
sounds, it only took a couple of hours to do both ends. The result
was a very cool looking tank (I don't think that comes across in the photo).
However, the design had a few shortcomings. To have material to drill
into requires a thick (and heavy) end piece. In addition, using the
flathead rivets with .049" wall tubing was a marginally acceptable application
(according to the engineering data in the rivet catalog.) Still,
using longer rivets, it might be a suitable technique for creating thicker-wall
tanks.
No, that isn't a very small tank on the right. It's a 6061-T6
.500"OD X .032" wall tube with a "groove" made by replacing the scoring
wheel in a small tubing cutter with wheel on the left. Actually,
it's not a groove. It's an annular indentation or crimp (not sure
what name to give it.) Rather than removing material, the tool deforms
the wall of the tube inward as shown below in cross section. The
wall of the tube retains almost its full thickness. It might not
be clear in this photo but the metal in the indentation becomes elongated
when formed so that it is thinner than the wall thickness and is thinnest
at the deepest part. Even at that point it's not really thin, about
75% - 80% of the tube wall thickness. Also, the wheel was unhardened
carbon steel and flattened after a few test grooves.
Since the hope is to make tanks, I made a similar indenting
wheel for a larger tubing cutter. I bought one at Sears for $19.95
which can cut tubing up to 2.5" in diameter. I made the wheel on
a lathe from tool steel, heat treated it, and then had to make spacers
to hold it precisely in position in the tube cutter.
Before machining the end closures, I wanted to understand and experiment
with the new tool. First, I wanted to see if it worked AT ALL on
larger tubing so I made some test runs. Using tubing left over from
the rivet tank experiment, I made my first groove. The tool worked
perfectly! It was easy to produce just the result I wanted. I learned
a couple of other things as well. The first was I got some
idea of the profile of the formed metal on the inside of the tube.
I needed to know this so that I could machine a complementary groove in
the end closure. The second thing was that forming the groove is
much easier if a lubricant is used. I used an aluminum tapping oil.
Below are photos of the side view and ID of the test tube.
I thought that looked pretty good. The left photo above shows
that, in addition to the indent made directly by the wheel, the tube wall
adjacent to the crimp was drawn in slightly as well. Using a straightedge
and feeler gauge this measured about .009". The indent into
tube ID therefore measured .014" (.072" from OD surface to indent peak
- .009" for the "draw in" -.049" wall thickness). I believed that
if an end closure was present to support the tube the "draw in" would be
approximately zero. Therefore, the indent would penetrate beyond
the ID surface .023" (.072" - .000" -.049"). I also believed that
while being formed it would actually push in elastically a little farther,
so the complementary groove in the end closure needed to be a little deeper
to end up with a deformation of .023".
Here's the result of the first test with an end closure. First,
notice the "draw in" adjacent to the groove really IS zero! The OD
measured exactly the same right next to the groove as elsewhere on the
OD (the measurements were made before the tube was cut for the cross section).
The clearance between the end piece OD and the tube ID was about .002".
The groove in the end closure is .050" deep by .090" wide at the surface
with 60 degree "V" profile with a generous radius at the bottom.
The profile of the wheel was a 90 degree peak .025" high with a small radius.
Here's a cross section of the crimp. The peak of the indent protrusion
measured about .018" from the inside surface. A sharp eye will also
notice a problem. Note that the end piece had a shoulder. My
positioning of the tool on the outside of the tube was a little too far
to the right, not centered exactly above the groove in the end piece.
Because of the shoulder, the end piece could not reposition itself as the
indent formed so the profile is asymmetric.
It was my intention to cut out a quarter of the end piece and tube to view the indent formed into the groove. However, as soon as the tube was cut through it sprung open by the amount you see above and the end piece was free to fall out. My first thought was that forming the crimp caused this to happen. To determine how bad a problem this might be I cut through a tube with no crimp which is shown on the right below.
As you can see there is no difference. Forming the crimp did
not strain the tube causing it spring open.