any limits on mechanical seals?

Suppose you have a large space station (OK, let's say a colony) with a
rotating portion and a stationary portion, both pressurized, and with
constant traffic back and forth between them. Obviously you need a
large mechanical seal between them, and I have some questions about that
which I hope someone can answer:

- In general how likely is this to be a problem? Suppose we want a
corridor 20 m wide between the two portions -- how much bigger is that
than any seal we've built before?

- How can I estimate the leak rate through the seal?

- Can I characterize the problem only by the pressure difference between
the inside and outside? In other words, is a seal that holds 1 ATM
against a vacuum the same as one that holds 2 ATM against 1 ATM?

- What type of seal would you expect this to be? Radial shaft?
Labyrinth? Rotating face? Something else? (I only vaguely understand
these different types of seals, so even small insights will be
appreciated.)

Many thanks,
- Joe

On Wed, 05 Jul 2006 09:01:50 -0600, in a place far, far away, Joe
Strout made the phosphor on my monitor glow in such a
way as to indicate that:

Suppose you have a large space station (OK, let's say a colony) with a
rotating portion and a stationary portion, both pressurized, and with
constant traffic back and forth between them. Obviously you need a
large mechanical seal between them, and I have some questions about that
which I hope someone can answer:

- In general how likely is this to be a problem? Suppose we want a
corridor 20 m wide between the two portions -- how much bigger is that
than any seal we've built before?

Unknown, but I don't see any theoretical problems.

- How can I estimate the leak rate through the seal?

Depends on design. I see no reason it couldn't be designed to zero
leak.

- Can I characterize the problem only by the pressure difference between
the inside and outside? In other words, is a seal that holds 1 ATM
against a vacuum the same as one that holds 2 ATM against 1 ATM?

In terms of the basic mechanics and strength requirements of the seal,
I'd think so. The issue is that there may be problems on the vacuum
side from material sublimation or welding. Not to mention radiation
(and perhaps monatomic oxygen, if it's in a suitable
environment--e.g., LEO). Just off the top of my head.

- What type of seal would you expect this to be? Radial shaft?
Labyrinth? Rotating face?

Rotating face would be my first choice, but again, just off the top of
my head. Think of a giant (hold breath) O-ring. Just don't do it in
January. You'd probably heat it, particularly on the space side,
though rotisserie effect might ease the issues..

In article ,
h (Rand Simberg) wrote:

- How can I estimate the leak rate through the seal?

Depends on design. I see no reason it couldn't be designed to zero
leak.

That's a bold thought -- my understanding is that pressure seals are
assumed to always leak at some rate, and that it was just a matter of
how expensive you want to make it to get the leak rate lower. But
that's a stance I'll question more carefully now.

- Can I characterize the problem only by the pressure difference between
the inside and outside? In other words, is a seal that holds 1 ATM
against a vacuum the same as one that holds 2 ATM against 1 ATM?

In terms of the basic mechanics and strength requirements of the seal,
I'd think so. The issue is that there may be problems on the vacuum
side from material sublimation or welding. Not to mention radiation
(and perhaps monatomic oxygen, if it's in a suitable
environment--e.g., LEO). Just off the top of my head.

Thanks, that's what I was thinking too. So if we wanted to, say, build
working models of such seals on Earth, we could do it, and tests thereof
would be as accurate as tests of any other machinery intended for space
on Earth -- i.e., you just have to watch out for the standard effects of
the space environment.

- What type of seal would you expect this to be? Radial shaft?
Labyrinth? Rotating face?

Rotating face would be my first choice, but again, just off the top of
my head. Think of a giant (hold breath) O-ring. Just don't do it in
January. You'd probably heat it, particularly on the space side,
though rotisserie effect might ease the issues..

Thanks again. I confess that the space-environment thing makes me a
little nervous when elastic materials (like an O-ring) are in such a
critical path, but it's still worth looking into. Failure of an O-ring
in this case wouldn't be catastrophic; you'd just start venting
atmosphere at a prodigious rate, but it would still take many years to
depressurize an entire space habitat. No serious damage would result.

Of course, *replacing* the O-ring, or otherwise doing maintenance on
this giant rotating seal, could be a real PITA. Certainly something to
plan for up front.

I don't suppose anyone knows any mechanical seal engineers (if there is
such a thing) sufficiently interested in space to kick around some of
the details with me?

Thanks,
- Joe

P.S. It's probably a test of maturity to see how long you can use the
phrase "giant mechanical seal" before a silly image pops into your head.
(I didn't fare very well on this test.)

On Wed, 05 Jul 2006 10:20:38 -0600, in a place far, far away, Joe
Strout made the phosphor on my monitor glow in such a
way as to indicate that:

In article ,
h (Rand Simberg) wrote:

- How can I estimate the leak rate through the seal?

Depends on design. I see no reason it couldn't be designed to zero
leak.

That's a bold thought -- my understanding is that pressure seals are
assumed to always leak at some rate, and that it was just a matter of
how expensive you want to make it to get the leak rate lower.

Well, zero is a small number. Let's say negligible. How often do you
you have to fill your car tires (which contain double atmospheric
pressure)?

It also depends on what you're trying to keep from leaking and what
temperature it's at (e.g., O-rings and hot gases, or hydrogen or
helium, which will seemingly tunnel through anything). But we're just
talking about an atmosphere of air here.

Joe Strout wrote:
Of course, *replacing* the O-ring, or otherwise doing maintenance on
this giant rotating seal, could be a real PITA. Certainly something to
plan for up front.
For maintenance you could start up the stationary part or stop the
moving part. Then you just put in a temporary inner seal and work
on the O-ring to your heart's desire.

Or the connector consists of a tube with an O-ring on both ends, then
you can couple the tube either to the still or to the moving part and
always have one O-ring at rest relative to the tube and the connected
part. Then maintenance ought to be easy too.

You can minimize leaks further by pressurizing the connection
only when needed.

Also, for long term survival I think that the air locks pose the bigger
problem. And accidents of course.

OTOH, earth loses a bit of air too (http://tinyurl.com/nxqqu) but during
the last few billion years it wasn't much of a problem, so, depending on
the size of your station, a few spoonful per day may be tolerable.

Lots of Greetings!
Volker
--
For email replies, please substitute the obvious.

In article ,
Volker Hetzer wrote:

Joe Strout wrote:
Of course, *replacing* the O-ring, or otherwise doing maintenance on
this giant rotating seal, could be a real PITA. Certainly something to
plan for up front.
For maintenance you could start up the stationary part or stop the
moving part. Then you just put in a temporary inner seal and work
on the O-ring to your heart's desire.

Well, yeah, except that we're talking REALLY large parts here -- it
would take months to spin down the habitat, and the same amount of time
to spin it up again, during which your 10,000 residents are getting
mighty grumpy. Spinning up the stationary portion would be easier, but
may not be possible, as it wouldn't be designed to take the loads.

Or the connector consists of a tube with an O-ring on both ends, then
you can couple the tube either to the still or to the moving part and
always have one O-ring at rest relative to the tube and the connected
part. Then maintenance ought to be easy too.

That's an interesting idea. Seems to me it would double the leak rate,
and the failure rate, but it may be worth it for easy maintenance.

You can minimize leaks further by pressurizing the connection
only when needed.

I don't think so. Both modules are always pressurized, and there isn't
a lot of point to making the connection double as an airlock. Assume
the connection is needed pretty much constantly anyway.

Also, for long term survival I think that the air locks pose the bigger
problem. And accidents of course.

Could be, but one issue at a time please.

OTOH, earth loses a bit of air too (http://tinyurl.com/nxqqu) but during
the last few billion years it wasn't much of a problem, so, depending on
the size of your station, a few spoonful per day may be tolerable.

Quite so; I don't expect the leak to be a problem, but it should be
estimated and figured into the resupply needs.

Best,
- Joe

Joe Strout wrote:
In article ,
Volker Hetzer wrote:
Or the connector consists of a tube with an O-ring on both ends, then
you can couple the tube either to the still or to the moving part and
always have one O-ring at rest relative to the tube and the connected
part. Then maintenance ought to be easy too.

That's an interesting idea. Seems to me it would double the leak rate,
and the failure rate, but it may be worth it for easy maintenance.
Not necessarily double. Since only one of them has to be in operation,
they would, when alternated regularly, only accumulate half the duty-hours
each.

Planning an SF story? :-)

Lots of Greetings!
Volker
--
For email replies, please substitute the obvious.

Joe Strout wrote:

In article ,
Volker Hetzer wrote:

Joe Strout wrote:
Of course, *replacing* the O-ring, or otherwise doing maintenance on
this giant rotating seal, could be a real PITA. Certainly something to
plan for up front.
For maintenance you could start up the stationary part or stop the
moving part. Then you just put in a temporary inner seal and work
on the O-ring to your heart's desire.

Well, yeah, except that we're talking REALLY large parts here -- it
would take months to spin down the habitat, and the same amount of time
to spin it up again, during which your 10,000 residents are getting
mighty grumpy. Spinning up the stationary portion would be easier, but
may not be possible, as it wouldn't be designed to take the loads.

So why not design it to take the loads? You are making design
decisions (it won't be designed so) in advance of setting your design
requirements. Generally this is a Bad Idea.

D.
--
Touch-twice life. Eat. Drink. Laugh.

-Resolved: To be more temperate in my postings.
Oct 5th, 2004 JDL

In article ,
Volker Hetzer wrote:
Also, for long term survival I think that the air locks pose the bigger
problem.

Needn't be that big a problem. So long as the doors open inward, they
won't open if there's any significant pressure difference. You will have
to be careful about protecting the door seals from damage, and you'll
probably have to replace the seals occasionally.

For bonus points, each airlock has three doors. The outermost one is
normally left open, with protective covers over its seals. When, and only
when, the seals of the middle door need replacing, you carefully close the
outer door, and do the work on the middle door in shirtsleeves. (If you
ever have to replace the outer door's seals, that work will have to be
done in spacesuits... but that should be extremely rare, especially since
you don't care much about slight leaks there.)
--
spsystems.net is temporarily off the air; | Henry Spencer
mail to henry at zoo.utoronto.ca instead. |

Henry Spencer wrote:
In article ,
Volker Hetzer wrote:
Also, for long term survival I think that the air locks pose the bigger
problem.

Needn't be that big a problem. So long as the doors open inward, they
won't open if there's any significant pressure difference.
I got the impression that the OT wanted less leakage than what
can be achieved by running pumps for a few minutes. I cannot imagine
that such an airlock, maybe used twice a day, would leak less than the
seal he wants to be airtight.

Lots of Greetings!
Volker
--
For email replies, please substitute the obvious.