Cliff, This is changing the subject a little, but I've noticed using the ABS spreadsheet that there are cases where adding heavier stiffeners to a cylinder actually decreases the depth rating. It's counter intuitive. The optimal stiffener configuration seems to be a happy medium. Do you know why this "over stiffened" effect happens? thanks, Alec ________________________________ From: owner-personal_submersibles@psubs.org on behalf of Cliff Redus Sent: Mon 11/2/2009 3:53 PM To: personal_submersibles@psubs.org Subject: Re: Re: [PSUBS-MAILIST] Re: Stiffiners on Heads Sean I did not know you were working on this. Excellent! Good points on the shape and spacing on inner or light stiffener spacing on cones. Concur. Also it was good to remind anyone that is designing a 1-atm boat with a cone shaped hull section that ABS limits the cone half angle to 60 degrees or less. As to parallel circumferential stiffeners, I did not try this. For sure they would work better than radial but I did not try them in my FEA analysis so I can't comment on their effectiveness or lack thereof. It would not be easy to fabricate this type of inner stiffener. One more point on radial stiffeners, my contention that you should not use radial stiffeners was based on a practical number of stiffeners. If you took an infinite number of radial stiffeners and say they were only the web portion without the outer flange, then you would, in essence, be simulating a head without any stiffeners but with a head wall thickness equal to the web height plus the shell wall, or in other words a very thick shell. This would obviously be stronger. So theoretically there is some number of radial ribs that above this number, the head would reach a deeper max depth. From a practical perspective, this a non realistic number of ribs. Better to just go to a thicker shell wall on the head to get your design depth. Cliff ________________________________ From: Sean T. Stevenson <cast55@telus.net> To: personal_submersibles@psubs.org Sent: Mon, November 2, 2009 1:28:38 PM Subject: Re: Re: [PSUBS-MAILIST] Re: Stiffiners on Heads As some of you are aware, I am in the process of coding the ABS Rules into a comprehensive software package which will perform design optimization, in addition to running "what if" scenarios as per Cliff's spreadsheet. Progress is suspended at the moment due to other commitments, but I plan to return to this by mid-November, so barring any further problems, I will have version 1 of the finished application available by Christmas. Brian: According to the ABS Rules, for a conical shell, as with a cylindrical shell, short-term critical pressures for inter-stiffener strengths are calculated on a bay-by-bay basis. With cylinders, the optimum stiffener configuration is a consistent repeating pattern. With conical shells, this is not the case, as the inter-stiffener spacing (assuming constant stiffener cross-sectional geometry) needs to be smaller towards the larger diameter end. Alternatively, you could hold a consistent spacing but vary the stiffener geometry, or do a combination of both. (My software is intended to determine the least material solution). If you can't wait for the application, modify Cliff's spreadsheet as he suggested, or just plug through the numbers as per the ABS equations. To answer your question regarding the conical transition, this isn't as simple as integrating through multiple cylindrical shells of varying diameters, since there is an axial component to the applied force which is not present in the cylindrical analysis. ABS takes this into account, as you can see from the differing equations, and also sets a limit as to the maximum allowable angle of transition. Taken to the extreme, a conical section would approach the same geometry as an unstayed flat head - not permissible for obvious reasons. Cliff: As for the stiffened head discussion, I would have expected that radial stiffeners would not be advantageous because of the stress concentration and effective "forcing" of a failure mode in buckling (I would, however, employ radial stiffeners in a collision bulkhead which was not a pressure boundary); however, I want to know if the different stiffener designs that you considered included parallel circumferential stiffeners as would typically be used within cylindrical and conical shells? The welding gets difficult, particularly in the case of a foreshortened head (i.e. 2:1 semi-elliptical) where the surface slope increases dramatically, but I can't intuitively see why such stiffeners would be a bad idea. Thoughts? -Sean On Nov 2, 2009, Cliff Redus <cliffordredus@sbcglobal.net> wrote: Brian ABS rules in the chapter on hull stress analysis has a section for stiffened conical frustums. The equations are are harder to implement compared to cylinders or semi-elliptical or hemispherical heads so I did not implement them in the current ABS hull spreadsheet. I believe the solution Vance suggests that of using the large diameter with the same stiffener spacing would be a conservative approach. The actual maximum depth would be something between the results for the small diameter and larger diameter. If you want a more accurate answer, code the ABS equations or get someone to do a comprehensive FEA of hull. Coding the equations would be better as they account for all the major failure modes. A simple FEA analysis were you look for stress hot spots using a linear model will not cover items such as fabrication out of roundness de-rating of the strength or stiffener tripping or a sensitivity to the number of lobes in overall buckling. I don't have time to do this update to the sheet but it would be nice if we could get someone to volunteer to add a tab for these stiffended conical sections to the ABS hull spreadsheet. Cliff Cliff Redus Redus Engineering USA Office: 830-663-6445 USA mobile: 830-931-1280 cliffordredus@sbcglobal.com ________________________________ From: Brian Cox <ojaivalleybeefarm@dslextreme.com> To: personal_submersibles@psubs.org Sent: Mon, November 2, 2009 9:12:14 AM Subject: RE: [PSUBS-MAILIST] Re: Stiffiners on Heads Cliff, In the areas where your sub has cones that reduce in size from one diameter to another, are those transitions farely straight forward as far as stressses? In other words is it simply a matter of calculating smaller diameter cylinders to arrive at the ultimate stresses? The reason I ask is because I'm going to have a transition from a 31" diameter to a 51" diameter which will most likely be a 45 degree sloped cone. Brian -----Original Message----- From: owner-personal_submersibles@psubs.org [mailto:owner-personal_submersibles@psubs.org]On Behalf Of Cliff Redus Sent: Monday, November 02, 2009 5:42 AM To: personal_submersibles@psubs.org Subject: Re: [PSUBS-MAILIST] Re: Stiffiners on Heads I concur with Daniel on the rib issue on heads. Radial stiffiners on heads do not increase design depth. Just for fun, a few years back when I did the hull design for the R300, I did some FEA experiments adding radial stiffiners of different designs to try and increase depth. On this boat, the semi elliptical head on the forward end of the pressure hull has the same thickness as the stiffened cylindrical section (0.25 inches). What I found was that every different stiffener design I tried on head actually reduced the heads ability to equally distributed the load. Also it showed that for an elliptical head the highest stress area is in the transition between the short cylindrical skirt and the elliptical portion also known as the knuckle. The only way I could increase the design depth for a give head diameter was to go to a thicker head or a hemispherical head. Cliff ________________________________ From: Daniel Lance <lanceind@earthlink.net> To: personal_submersibles@psubs.org Sent: Sun, November 1, 2009 10:42:36 PM Subject: [PSUBS-MAILIST] Re: Greg Cottrell's Sub Jon, Greg will be first one to admit that if he had it to do over again he would not put stiffening ribs on the heads of his sub. He readily explains that when he began constructing his sub in the late 1980s his enthusiasm for building out paced his knowledge of the engineering principles involved. When we did the refit and added tee bar stiffeners internally in the cylinder section theoretically we increased his dive capability. In reality he still had the problem of the external ribs on the hemi heads.They are undesirable for two reasons. #1 The heads should be allowed to compress in a uniform manner. The ribs prevent that. #2 The ribs were not welded on using full penetration welds. If there is a weld failure any number of unpleasant things could occur. In the end we decided it was not cost effective to remove them . Greg is currently building a new sub incorporating all the lessons he has learned a! long the way. Stiffeners on cylinders are fine. On elliptical heads ,hemi spheres and spheres they are not. Did anyone see stiffeners on the DeepWorkers or on the hemi-heads of the Aquarius? This issue illustrates the potential hazard of designing a sub just by looking thru the Psubs photo gallery or the archive. Because some one else did it and took a picture of it does not mean it is an acceptable method. Unfortunately as a group we lack a peer review mechanism to sort photo gallery or archive entries into categories that could prevent this type of misunderstanding . Dan Lance ************************************************************************ ************************************************************************ ************************************************************************ The personal submersibles mailing list complies with the US Federal CAN-SPAM Act of 2003. Your email address appears in our database because either you, or someone you know, requested you receive messages from our organization. If you want to be removed from this mailing list simply click on the link below or send a blank email message to: removeme-personal_submersibles@psubs.org Removal of your email address from this mailing list occurs by an automated process and should be complete within five minutes of our server receiving your request. PSUBS.ORG <http://psubs.org/> PO Box 53 Weare, NH 03281 603-529-1100 ************************************************************************ ************************************************************************ ************************************************************************
<<winmail.dat>>