What kind of stem seal did they use? Was it the type where you have a cone
that squeezes into a sleeve?
I hadn't seen any short stem cryo models. It might be worth me getting one to
take it apart. Were these custom or off the shelf?
-Bob
On May 30, 2016, at 10:24, Ben Brockert <wikkit@xxxxxxxxx> wrote:
To be clear, basically every vehicle I've worked on had short stem cryo ball
valves, and most of them did not leak. Torque requirements are another issue,
but it is not a universal truth that all short stem ball valves leak out the
stem.
On Monday, May 30, 2016, Robert Watzlavick <rocket@xxxxxxxxxxxxxx> wrote:
Agree that drilling out the stem would make sense. Another option would be
to use a 1/4 shaft and 3/8 bore instead of 3/8 shaft and 1/2 inch bore as I
noticed there are standard seals and bearings made for those sizes. I could
also go with an aluminum shaft (I did in my prototype) but you have to be
super careful to maintain the surface finish and not let it get scratched.
There are also wear issues but anodization would likely help. The original
Swagelok design has an OD near the ball of 3/8 OD but the rest of the stem
is 1/4 inch. However, the original stem is also designed to be inserted
from within the ball cavity which limits its overall length.
I sketched up a bearing-less design that still uses the PTFE spring seals
uses close-fit PTFE bushings instead of bearings. I concluded that it might
be difficult to hold tolerances on the PTFE parts and that you would really
have to load up the bushings radially to keep the shaft from wobbling
around. That's essentially the original stem packing arrangement along with
it's high torque requirements (which leaks at cryo temps with a short stem
configuration).
I would be interested to see the the seal design on this XCOR valve:
http://aerospace.xcor.com/media/5766/fuel-valve.jpg
I was originally going to duplicate that design and I even purchased a small
air cylinder but I thought I could save some weight and complexity by using
an electric servo motor instead of the air cylinder. I also would have had
to have a separate regulator and double-acting solenoid.
-Bob
On 05/27/2016 11:09 PM, Paul Breed wrote:
The stem limit may be rigidity against the horn turning it not actual
torque...
Also how much spring can you handle in the drive....
Rotating shafts that run aricraft controls are almost always hollow tubes...
On Fri, May 27, 2016 at 5:41 PM, Lars Osborne <lars.osborne@xxxxxxxxx>
wrote:
RE: drilling out the stem.
If you have margin on the stem yielding from torque, wouldn't it be better
in this case to reduce the OD as much as possible? A smaller OD will
reduce the friction of the stem seal.
Thanks,
Lars Osborne
On Wed, May 25, 2016 at 6:26 PM, Robert Watzlavick <rocket@xxxxxxxxxxxxxx>
wrote:
Here are the weights of the new design (lb):
Body (6061) 0.077
Stem nut (SS) 0.005
Bearings/Sleeves (SS) 0.012
Stem (SS) 0.023
Body bolts/nuts (6061) 0.015
Servo 0.120
Ball/seats (SS/PTFE) 0.027
End caps (-6 male, 6061) 0.066
Servo hub (6061) 0.009
Total: 0.354 lb
Weighing all the items together comes out to 0.324 lb (some scale
resolution issues above). I'm not sure where I got the 0.27 lb
number from before - that may have been without the
servo. And I guess the stem didn't weigh as much as I remembered.
Brackets will add another 0.02 or so. Switching from SS to aluminum body
bolts saved a lot of weight.
For the main LOX and fuel valves, a pin connects them together since they
are effectively on a common shaft. So I can't drill the pin hole any
deeper or the pin might slip into one or the other stem. I could use a
longer pin but that wouldn't help the weight situation. But for the vent
valves, I could drill out the stem somewhat.
-Bob
On 05/25/2016 11:54 AM, Paul Breed wrote:
Any reason you can't drill out the center of the SS Stem?
The stuff in the middle is adding no real strength...
IE Turn it into a tube....
Leaving 0.075 wall would reduce the weight ~35%..
0.050 wall would save > 50%
On Wed, May 25, 2016 at 9:21 AM, Robert Watzlavick
<rocket@xxxxxxxxxxxxxx> wrote:
I'll weigh the components tonight. The heaviest component by far is the
SS stem.
-Bob
On May 25, 2016, at 10:59, Paul Breed <paul@xxxxxxxxxx> wrote:
I'd be interested in the mass break down of your final valve, ie
weight of all the components....
I was looking at building a plug valve with bearings to support it and
a hollow 3D printed plug....
Paul
On Wed, May 25, 2016 at 5:57 AM, Robert Watzlavick
<rocket@xxxxxxxxxxxxxx> wrote:
I never though to ask McMaster but they responded quickly:
The vendor for the PTFE seal is Bal Seal Engineering Co Inc and their
part number is 100MB-012-T or X124209.
I can't find that part number in their catalogs but the MB describes
the spring. The 012 is probably the series and T corresponds to
Virgin PTFE. Attached is the datasheet I received from them.
I recall our conversation about leakage a couple of years ago and my
quick fix was to use two of them and a pressurized cavity with both
U-cups facing toward the pressurization port. That did wonders for
the leakage and almost completely eliminated it. Now that I've
improved the bearing situation, maybe it will be better. I could add
a 2nd seal in this design but it would shrink the distance between
the bearings down. I could always make the valve body longer if
needed but that means redesigning a few other things on the vehicle.
I'm using these shafts:
https://www.servocity.com/html/3_8__precision_shafting.html
and it says they are 303 SS with a 10 RMS micron finish. There are
better shafts out there but the 303 makes it easy to cut for threads
for the nut. They tend to have some very small nicks in them from
rattling around in the bag so I used graduated sanding pads to polish
them up:
http://micro-surface.com/index.php/products-by-type/soft-touch-pads/micro-mesh-soft-touch-pad-variety-packs.html#
These go all the way up to 12000 grit so I hopefully I am improving
the surface finish - it looks almost chrome plated when I'm done with
them.
I suspect though that due to shrinkage, the seal locks itself around
the shaft and it is the OD of the seal turning against the bore that
becomes the new sliding surface. The bore looks pretty smooth,
mainly because I ran the last pass by hand instead of using the mill
power feed but I would like to polish it up some more if anybody has
a good suggestion. I thought of making my own 0.500 sanding pad but
I don't want to oval the bore or take too much off since it is
6061-T6. I've seen rotary grinding stones - maybe those would work
without taking too much off. I did find these although 140 grit
seems pretty low:
http://www.artcotools.com/precision-diamond-pin-gx-141c.html
-Bob
On 05/24/2016 10:20 PM, David Gregory wrote:
Another note: some amount of leakage for spring energized seals in
dynamic aoolications is quite common in my experience. Often the
leakage will subside after the motion stops. Also, I think I've
mentioned before I've seen higher leakage with lower pressures due
to insufficient preload (at higher pressures the seal is assisted by
the fluid pressure). What surface finish do you have on the shaft?
16 or better is typically called for.
David
