[PSUBS-MAILIST] MK 2 Brushless Thruster

[Sean T. Stevenson via Personal_Submersibles]

Alan, I am unclear as to what purpose the inner seal is serving, given that you have the same fluid at the same pressure on both sides?  If you are compensating with positive pressure, an outer seal failure will result in losing some compensating oil to sea, until the pressure is equalized to sea pressure, at which point you may have communication through the outer seal, but no pressure differential to drive fluid transfer. As such, only incidental admission of seawater will occur, but you have a communication channel to your motor compartment bypassing the inner seal, which puts the motor at risk.  An inner seal failure alone in your embodiment is entirely inconsequential, and if it is there as a failsafe in case of accidentally developing negative compensation pressure, you defeat that purpose by allowing the intermediate volume to communicate with the motor chamber.

When I design mission-critical sealing arrangements, I use the following rule-of-thumb: Use a seal to change fluid, or to change pressure, but not both simultaneously. This ensures that any incidental leakage as a result of dynamic seal movement is usually inconsequential. Accordingly, I might look at your design to see if I could somehow arrange to have the motor compartment at a slight positive pressure for compensation (using either central compensation or a compensation tube or bladder with spring bias to elevate the pressure), and then the inner seal sealing between that oil at slight positive pressure and oil at sea pressure in the isolated intermediate volume so that incidental leakage is oil to oil and inconsequential (using an unbiased compensator for this volume), and then the outer seal sealing that oil at sea pressure against seawater at sea pressure, so no differential exists. If you do that, here is the failure analysis:

An inner seal failure will cause oil to move from the motor compartment to the intermediate compartment as a result of the positive pressure. If the displacement of the unbiased compensator on the intermediate chamber is limited, and the displacement of the biased compensator on the motor compartment is large enough to bring the unbiased one on the adjacent chamber to that limit, then you simply end up with the intermediate chamber also being compensated at positive pressure, and the outer seal does a little more work but maintains the arrangement's integrity.

Conversely, if you had an outer seal failure, seawater is permitted to communicate with the oil in the intermediate chamber, but there is no pressure differential to drive mixing of the fluids, and in any case, the intermediate chamber remains isolated from the motor compartment which is at slight positive pressure, sealed by the inner seal.

Anywhere you employ seals with zero pressure differential across them, they need to be of a spring energized type since the pressure can't do it.

Anyway, just a few things to think about.

Sean


On April 1, 2016 3:05:09 AM MDT, Alan James via Personal_Submersibles <personal_submersibles at psubs.org> wrote:
>Updated version with Hanks suggestion of a channel directto the motor
>cavity.I have also added a bearing at the wiring end of the motor
>shaft.I am not sure about this. It would be easier engineering wise, to
>locate the motor ina central back bearing & have the holes for the
>locating pins a bit over sized,rather than a tight fit. That way they
>would only serve to stop the rotation of the back section of the motor.
>According to my propeller calculator, adding anothershaft bearing will
>reduce the power output by 1.5%.Alan
>
>
>------------------------------------------------------------------------
>
>_______________________________________________
>Personal_Submersibles mailing list
>Personal_Submersibles at psubs.org
>http://www.psubs.org/mailman/listinfo.cgi/personal_submersibles
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.whoweb.com/pipermail/personal_submersibles/attachments/20160401/805f8d06/attachment.html>