|
Hello Jon,
My basic first run FEA test on the acrylic cylinder
I'm working on configuring into my subs refit, was a very basic pressure
test to see what I could learn from it. My posting of it has resulted in
some very good input, including yours.
My holding means are not reflected as of
yet. Items like the hatch flexing in if it was to weak, and/or
the subs steel CT flexing unevenly have not been evaluated yet. I do
expect to have a retaining flange on the inside. I won't know how that will
effect the model's FEA results. I'll post the results once I've done some more
runs.
I'm currently looking at two basic configurations for attaching and sealing
my acrylic cylinder. The first being the externally mounted SS bars squeezing
two hot dip galvanized then painted, turned steel rings with a large groove
cut into them to receive 1" of the acrylic cylinder, and .125" for a gasket
or Sekaflex. I'm leaning towards using the gasket, since I can more easily
remove the cylinder later for maintenance or replacement. Then use the Sekaflex
in the vertical void inside of that called the radial bearing seal. Since
I can more easily cut that seal, to free the window, and it makes a flush
seal on the top, for easy cleaning, with no junk getting down in a crack or
groove on the bottom seal.
I'm looking at using the A515 GR70 rings instead of 316 SS, since the cost
of SS it prohibitive, and I believe I've put together a good solution for making
the rod attachment areas rust free, with 316 SS sleeves TIG welded into over
size holes that are drilled into the steel rings. Then use Delrin bushings and
washers to finish it out when attaching the prison bars. I can also clean up the
design even further by having the bottom ring have threaded 316 SS sleeves TIG
welded into the oversized holes. This removes the need for nuts and washers
sticking out the bottom.
I don't currently see that I would need to have the axial contraction
compensators shown in Fig. 13.30. Since I'll have at least one end of the
bars that can freely slide in as the sub goes deeper, and also since the S101
and S102 are not using them.
I'm looking at having a couple of elements of Fig. 13.32 on page 769. One being the radial bearing 90 durameter hardness gasket, and the axial bearing gasket also being 90 durameter hardness as the secondary seal, (if I don't use Sekaflex there), if I use the lip seal outside of it. I want to have a primary and secondary seal in my assembly. But I don't know if I can use the radial bearing gasket or Sekaflex seal as the secondary sealing means. I wish to go with the Sekaflex so I can have a perfect fit. Perhaps even cut a curved gland at the base of the flange to lock in the Sekaflex even better in the steel rings. I would like to incorporate the elastomeric 70 durameter hardness lip seal
around the outside also shown in Fig. 13.32 and Fig. 13.30 design C, as the
primary sealing gasket.
I don't know that I really need the flange at all, since we see so
many domes with no raised flange on the inside of the domes, like K-250 dome
assemblies, and many others. I'm aware domes are different, but they also have
pressure squeezing them in and down, much the same as we see on the acrylic
cylinders. I like the raised flange, but if I don't use it I will make the steel
rings lighter, and for the second design concept using UHMW in place of the
upper steel ring, it would make that part less expensive.
For the second design being considered, there are no prison bars, and the
acrylic cylinder has bonded on retaining flanges like you see on the bottom of
the K-250 domes, and in Fig. 13.30 design E and F.
I like design E minus the O-ring and it's groove, for a larger axial sealing surface and less stress on the bonded on acrylic ring, from the retaining V clamp sucking down the said acrylic ring, being that it is resting on the axial sealing surface. 316 SS V clamps like we see on some large front viewports on some Perry subs, there battery pods, and the Triton 1000 battery pods, are used to retain the top and bottom of the acrylic cylinder to the CT and hatch land. You can also see a number of the elements I'm looking at incorporating in
Fig. 21.7 on page 1047 design B.
I'm working on the 1" thickness model currently. I'll keep you
posted. I've obtained a massive amount of FEA data since my last run on
the cylinder, from many sources, that I'm working thru first. Some of this
data will effect how I want to run the FEA on the 1.5" thick model. Once I do
that I'll do the same thing on the 1" model, so I can have a good
comparison.
Regards
Szybowski ------------------------------------------------------------------------------------------------------------------------------------------------------
It looks to me like the bands are appearing where support is called for on the ends of the cylinder. Brent has modeled the top figure as displayed in Figure 13.11 on Page 749 of Stachiw's book. That is the simplist and weakest rendition possible which should also produce the poorest set of performance numbers. However, it's not a realistic or practical approach for a conning tower given the methods Stachiw provides for mounting the cylinder. Brent, I think a more interesting model would be the middle figure on Page 749 which probably represents a more practical mount on the sub, or even the bottom figure which would only require you adding a strengthening rib on each end made out of the same acrylic material. It would be interesting to see if those two blue bands disappear or at least the effects are lessened. The numbers produced by the test seem to match pretty closely what I can extrapolate from Figure 13.15 on Page 753. However, I think it would be more interesting to run your numbers with a t/D of .045 (see bottom page 753) and compare your FEA numbers with the results Stachiw published on Page 754 which were determined by actual testing to failure. You can see that Stachiw's tests matched pretty closely with standards, although the standards (by calculation) were slightly more conservative. Also, when I talked to Reynolds Plastics they told me that anything over 1 inch thick material required special tooling for them to produce. I'd like to see the FEA numbers for all the same dimensions except using 1 inch thick material if you ever find time to redo the test. Jon > > Brent > > You have an enigma around the top and bottom of your cylinder, note > the two thin blue bands. Something like this usually points out that > you might have an issue with your model. Looking at the numbers along > the right side of the image, is this deflection? What are the units? > > R/Jay > |