Mailing List Archive Mailing List Archive
Me two cents worth: A pressure housing such as a dive light battery cannister subjected to external pressure loading can fail structurally by a couple of distinct mechanisms. Either the limits on the material of construction will be exceeded and the material will yield or break. (Yielding is like bending a fork, if you apply a little force to the fork it will return to it's original position when the force is released without deformation; however, if you apply enough force, you will have a bent fork after the force is released: it has yielded). Failure in material yielding will ultimately lead to failure of the housing. The second mechanism of failure is buckling. Buckling failure is primarly a function of the geometry of the housing (e.g. diameter, wall thickness, length) and the stiffness (modulus) of the material. In other words a large diameter housing will fail before a smaller diameter housing (with equal wall thickness). Theoretically, a very thin wall can support a substantial pressure load, however due to imperfections in manufacture or out-of-roundness of a circular tube, the failure will typically occur at lower pressures than theoretically predicted for perfect cylinders. The stiffness of the material describes how much it will deform for a given load. For example aluminum is about three times more stiff than acrylic, a plastic typically used in light cannisters. Steel is about ten times more stiff than acrylic. Plastics also "creep" more under constant loading. Reinforced plastics (fiberglass) gain a little in stiffness and strength due to the fiber reinforcements embedded in the plastic matrix. So material yielding is a function of the properties of the material and buckling is a function of the geometry of the housing and the stiffness properties of the material. In addition, there are several modes of buckling. In design, both mechanisms, yielding and buckling, must be evaluated with additional safety margin. Onward! A rectangular pressure housing has by definition flat sides. If you take a flat plate and rigidly clamp all around the edges and apply a distributed force (pressure: #/sq.inch) to one side of the plate, the plate will deform inward with the maximum deflection at the center of the plate. The highest stress in the plate will occur at the edges of the clamped plate. The edges of the plate will be subjected to bending forces (bending moments) in addition to membrane forces (stretching). (Clamp a yardstick over the edge of a table and press down on the end; you are applying a bending moment equal to the applied force times the distance of the applied force to the table edge) The total stress in the material is the sum of the stress from bending and the membrane stress from stretching. The center of the flat plate has very little stiffness from geometry and it will deform substantially. If the plate was not clamped around the edges but was supported on "knife-edges" the highest stress would be in the center of the plate. On a rectangular pressure housing, such as the Neutralite, you do not have symmetry in the loading around the housing. You have high membrane stresses in the center of the low stiffness flat sides, with high stresses at the connections of the sides (primarily due to bending). The lack of symmetry induces buckling because the forces are not evenly distributed around the housing. You are likely to be at the limits in yielding at the center of the flat sides if the deformation is not restricted (i.e. battery supports the side walls). The key is symmetry. You do not get it where you have distinct zones of stress (i.e. bending dominated v.s. membrane dominated). Although rectangular housings have been used at shallow depth because they can offer a high packing efficiency, they are unsuitable for deep housings. A circular housing has symmetry around the housing. This is the distinct difference compared to rectangular housings. The bending forces are diminished; any yielding is in compression. The circular shape does not induce buckling. The forces on the housing are evenly distributed around the body of the tube. A flat endcap is definitely not as good as a hemispherical endcap. A flat endcap if bonded or screwed to the circular housing will induce bending moments in the end of the body (see clamped plate above). A hemispherical endcap does not induce bending moments. Its better, but not necessarily required. A relatively thick endcap "floating" on o-ring seals can be designed such that very little bending force is transmitted into the body. This is one reason why I prefer to see the bottom endcaps on light cannisters also "floating" on o-ring seals. A circular flat endcap is symmetrically loaded; a rectangular flat endcap is not. Penetrations through an endcap disturb symmetry somewhat. The stresses in a circular endcap which is not bonded to the body of the housing are the greatest at the center of the endcap. Guess where most people put the cord or switch penetration? If the circular body is somewhat out-of-round, it will lose symmetry (induces lobes) and an applied pressure will tend to further distort the walls inward at the major diameter (largest diameter) and relatively "outward" at the minor diameter (smallest diameter) (e.g. bourdon tube in your pressure gauge). Endcaps bonded on may pop off due to these unsymmetric forces. Another reason to put the endcap on the bottom "floating" on o-ring(s). Spot weld your spring clamps (stainless steel) to a thin strip of stainless screwed into the bottom cap just like the lugs on the top. Now the body is supported by a step in the endcap and not the glue bond. Got to run; sorry don't have time to spell check. Doug Chapman
Navigate by Author:
[Previous]
[Next]
[Author Search Index]
Navigate by Subject:
[Previous]
[Next]
[Subject Search Index]
[Send Reply] [Send Message with New Topic]
[Search Selection] [Mailing List Home] [Home]