Regards,
To clarify a few things, I'm not an engineer, I don't run the model's
through AutoPIPE. I design the systems, one of our engineers runs the
model through AutoPIPE and uses the ASME B31.3 code. The gist of my
last message was that I've never been told to remove a clamp in one of
my models due to leakage concerns. After the model is run and the report
is generated our engineer will peruse it and let me know of any hotspots
that might need tweaking. He will then look over the model himself and
let me know if he thinks there are any problems.
The gasket I was speaking of was specifically a TEF-Steel gasket which
is the only type we've used with the MHP style clamp for high pressure.
They can only be used so long as each time one bolts that clamp together
it stresses the gasket a little bit so the next time it must be
connected tighter until it is bottomed out and will then leak. That
won't happen with normal PTFE or EPDM gaskets. Sorry for the confusion.
Aaron Wolfe
Piping Designer
Paul Mueller Company
P.O. Box 828
Springfield, Mo 65801
(417) 575-9780
E-mail: awolfe@muel.com
-----Original Message-----
From: Al [mailto:alwynk@shaw.ca]
Sent: Sunday, May 23, 2004 2:12 PM
To: PipingDesign@yahoogroups.com
Subject: RE: [PipingDesign] Modeling Hygienic Unions
Interesting discussion.
i am surprised the fittings have to be clamped and unclamped so often to
cause fatigue (the term being used loosely no doubt), but when aaron
says
Autopipe handles the stress just fine, I assume he enters the components
as
standard fittings and the program indicates if the "system" as a whole
is
over or under stress . I can't see Autopipe outcomes as "just fine"
being
more than the system analysis and not much to say about the specifics
around
the fitting performance. (no doubt the unions are modelled as
"flanges").
The tools are very limited there. I would guess the most likely scenario
would be to provide layouts that results in lower bending stresses
around
the fittings and hence
low reaction forces and moments on the fitting to cause leakage.
I understood Caesar defaults to 7x the pipe stiffness for the local
fitting
stiffness. this may or may not be the "magic" number for such widely
varying
kinds of fittings.
This would require experience and trial and error of different layouts
to
find the "sweet spots" like good welding design for low fatigue and
stress.
I agree the objective (and the code intent) is low stress design without leakage.
Have you ever run any of the leakage or rigidity analyses on these? What
are
the results?
Interesting to hear other thoughts on design approaches.
Altecheng
001(780)465-9762
Fax/Msg(780)465-9762
-----Original Message-----
From: Ken Nisly-Nagele [mailto:knislynagele@applied-e-s.com]
Sent: Sunday, May 23, 2004 8:48 AM
To: PipingDesign@yahoogroups.com
Subject: RE: [PipingDesign] Modeling Hygienic Unions
Aaron, Michael, Chris -
My apologies, out of town and slow to answer, so I'd like to reply to
all
here.
I sent two files to Paul to upload. File Clamps&PipingSketch shows a
typical high pressure clamp type union and a typical use of the union in
a
piping system. File ZST shows a typical zero static tee which is
designed
to eliminate dead flow spaces in the piping system.
Aaron addresses many of the questions asked. Welded joints are
preferred,
but can not be completely avoided without sacrificing cleaning,
maintenance,
and inspection, in this high purity water system.
The problem with using some of the products suggested such as braided
hose,
Bradford clamps, or flexible yoke style couplings, is that these are not
ASME BPE compliant in terms of surface finish and elimination of static
flow
areas in their design. I am not sure of the hoses though, and will do a
follow-up on available options with hoses.
We are negotiating with one of the manufacturers for testing the clamps.
It
may offer an advantage to them if they can publish what their clamps can
withstand while others do not have the data. It seems an industry
standard
may be an option here because the design of the clamps are so similar
across
manufacturers and there are only three manufacturers, I believe.
The code I am most concerned with is B31.3 which states the pipe
flexibility
requirements, the BPE code is mostly addressing dimensions and issues
relating to sanitization. B31.3, para. 319.1.1 requires a leak free
system.
As Aaron comments, the unions usually work initially, but often leak.
If
the system is designed to cycle thermally, and will leak after "x"
number of
cycles, does that meet B31.3? It seems to me since the intent of
B31.3's
flexibility requirements is to end up with a system that can cycle
thermally
and still not leak, these joints can not be shown to comply.
Ken A. Nisly-Nagele, P.E.
Project Engineer, Mechanical
Applied Engineering Services, Inc.
7999 Knue Road
Indianapolis, IN 46250
317-585-8920
-----Original Message-----
From: Aaron Wolfe [mailto:aaronwolfeman@hotmail.com]
Sent: Friday, May 21, 2004 8:46 AM
To: PipingDesign@yahoogroups.com
Subject: RE: [PipingDesign] Modeling Hygienic Unions
Ken,
In my experience the clamps hold up well to thermal expansion....If the line is designed with sufficient change of direction to handle the force. The MHP style could be used in high pressure situations. But over time they just leak. I work in mostly bio-pharm systems. And the places where clamps are required are for removable spools for sterilization, calibration of an instrument, or connection to a piece of equipment. It is standard practice to minimize the use of such clamps because like I said, eventually they just leak. This is due to the gasket being repeatedly fatigued by clamping and unclamping. A quick replacement and they are good for another round. Being Autoplant users, we use AutoPIPE for the "stress test" and it seems to handle the clamp fitting just fine.
Regards,
Aaron Wolfe
Piping Designer
Paul Mueller Company
P.O. Box 828
Springfield, Mo 65801
(417) 575-9780
E-mail: awolfe@muel.com
-----Original Message-----
From: Ken Nisly-Nagele [mailto:knislynagele@applied-e-s.com]
Sent: Thursday, May 20, 2004 5:43 PM
To: PipingDesign@yahoogroups.com
Subject: RE: [PipingDesign] Modeling Hygienic Unions
These clamps are made by Tri-Clover, Waukesha, or Swagelok, and are
rated to
200 psig. By meeting code, I mean to demonstrate that the unions will
not
leak during sterilization. I'm not concerned about the clamps working
at
the design pressure, but rather, I'm concerned with the clamps leaking
when
the piping in which they are installed undergoes bending during thermal
expansion and causes the faces of the ferrules to spread apart and
result in
leakage. With standard flanges there are tools to predict leakage, but
I am
not aware of a way to check for leakage with a union of this type. It
may
be that the only solution is to ask the manufacturers to test and
publish
the data as a group or individually.
Ken A. Nisly-Nagele, P.E.
Project Engineer, Mechanical
Applied Engineering Services, Inc.
7999 Knue Road
Indianapolis, IN 46250
317-585-8920
-----Original Message-----
From: Christopher Wright [mailto:chrisw@skypoint.com]
Sent: Thursday, May 20, 2004 5:00 PM
To: â¡
Subject: RE: [PipingDesign] Modeling Hygienic Unions
>My apologies, my initial posting was not sufficiently clear. No problem. Are you talking about clamps like the Grayloc or Tri-Clover variety?
>The question is one of
>applying the unions and demonstrating compliance with the code.
I think the easy for doing this is to show that your clamp meets the
code
rules for a pressure which produces a unit axial load equal to the
(calculated axial pressure stress + the calculated bending stress)/pipe
wall thickness. This is fairly conservative, but it'll meet the U-2(g)
requirements without a lot a extra work.
Christopher Wright P.E. |"They couldn't hit an elephant at chrisw@skypoint.com | this distance" (last words of Gen. ___________________________| John Sedgwick, Spotsylvania 1864)http://www.skypoint.com/~chrisw
Yahoo! Groups Links
Yahoo! Groups Links
Yahoo! Groups Links
Yahoo! Groups Links
Yahoo! Groups Links Received on Mon May 24 09:23:00 2004
This archive was generated by hypermail 2.1.8 : Mon Oct 27 2008 - 20:24:03 EDT