In my concept I can see you don't really need to make the welded on rings SS, since there and not any sharp corners to try to keep paint on. I do prefer the SS, but mild or cold rolled round rod stock I think would work just fine. I liked the round stock instead of the square stock, since to start with it gives you a undercut area to put a weld bead in, and there was no corners, which made entry and exit nicer. Not to mention the shape of the upper rings, where they mold into the Sekaflex is a better shape for that purpose. But for a design like Peter has, when adding one or more O-rings, I currently like the square stock. Also consider that these metal round rod rings are under mostly a compressive load.
The upper sealing rings in my first concept model are seating them selves in a relatively soft material, so I was thinking that would relieve some of the localized stresses on the upper rings. The surface area of the hatch that does translate into downward force to the hatch land, as I currently see it, is always going to out strip the pressure that is going into the gap between the hatch and hatch land until it goes metal to metal under extreme pressure. Also being that there are two upper rings, and that they are in a arch, they are a lot stronger then say if the rods went in a straight line. It might be a good idea to lay in larger weld beads there. I was just thinking of putting less heat there, and doing a fast TIG bead. But since a thicker weld bead might be better, and this whole arrangement was designed for a lower cost sub, perhaps a TIG machine wouldn't be an option. So a MIG or stick bead, that could easily be larger, could be used.
For more support, and a larger area to spread out the forces on the cured urethane sealant, you might like this below model better. It's using 1/2" round stock for the upper sealing rings. I also removed a quarter inch of the upper hatch plate thickness. Since in that configuration your adding a lot of metal, and part of the reason for that plate being as thick as it was, was to allow for the gland to be cut into it, and not unduly weaken it. So since the gland is not being cut and you want a lighter hatch, I removed the material. Look at the hatch on Idabel. It doesn't even need a lift spring, it's so lite. He did put in a conical hatch seating surface however. Idabel's hatch land is also very thin.
Regards,
Brent
Jay and Alec both make good points. I also wonder if the 1/4 inch beads
become the weak point of the sub. Given your rendering, I can see why
Peter used square stock. What becomes evident to me is that you've only
got 1/4 inch material at the cross section of the hatch ring. With
square stock you would have a consistent thickness which would be better
resisting overall pressure. Regarding Peter, you have to also keep in
mind that he did not dive his subs at relatively great depth.
Jon
Sent: Friday, September 04, 2009 5:57 AM
Subject: RE: [PSUBS-MAILIST] No Machining Hatch Sealing Concept
The reason an O ring seals so well is that, so long as there is an initial seal, increasing pressure expands the width of the O ring thus increasing it's sealing ability. In other words, the higher the pressure the better it seals. In this case the two SS rings that sit on the Sikaflex won't expand. The Sikaflex seat will deform under pressure, but in the opposite way you want it to. Pressure will tend to expand the Sikaflex to deepen the recess under the SS rings and tunnel under them. I'm not saying it wouldn't work at all. I think it would, but only at low pressures.
At least that's my take. Very nice renderings though!
Alec
This is an idea I've been wanting to make some CAD models and drawings of for some time now. We have this big oval or rectangle steel plate doors for the hyperbaric chambers, and it's expensive to machine the doors and seating flanges flat, plus cut the O-ring groove. Even then the welded in door frame can and usually warps some. Once welded in the door frame is very hard to machine. The concept shown in the below pictures and drawings, is showing a K-350 type hatch and hatch land, that have two rolled 1/4" SS rod rings, that are TIG welded onto the hatch flange, and two on the hatch land. The two on the hatch land are to form a dam for Sekaflex, that you caulk in there and level off. Then to allow the upper SS rings to seat down into the Sekaflex enough to take up any warp in both flanges and not stick to the Sekaflex while it is curing, you apply mold release to the SS rings, and then spray them with PVA mold release agent and let that dry. Then before the Sekaflex is caulked into the lower rings, you lower the hatch and figure out what spacers you need to keep if from going down to far into the urethane. Then add the Sekaflex of the type of hardness you want, and lower the hatch, which is attached at the hinge to keep every thing lined up, and set the hatch down on your taped or clamped on spacers, and let it cure.
One could use just one upper SS sealing ring as you can see in the drawing of a large WWII ballast tank flapper valve. But I like the primary and secondary sealing means, and it evens out the stresses more.
" Stand still. The trees ahead and bush beside you are not lost." ~ Albert Einstein
Regards
Szybowski