Dear sirs,
The suggestion of trying a different routing for the
suction line is definitely good. at the outset, the
terminal node i.e. the one at the pump nozzle shall
not be kept as a rigid anchor. Some amount of
displacement does occur at the nozzle by virtue of the
offset from the pump support. Another point: during
maintenance, the anchor should be such that it is not
being disturbed, which will not be possible, in this
case. Third point: API permits very low values as
permitted forces and moments at pump nozzles. This
calls for a flexible routing to render the nozzle node
fairly free of induced reactions. Provision of spring
hangers will surely reduce the stresses and reactions;
but at the same time, the natural frequency of the
system will turn out to be low. The machine frequency
also needs to be compared. This will need to be
verified if restrictions are there on minimum
specified value of frquency.
Summary: Stress anlysis is not an "instant solution".
Try affresh.
Good luck!!!
C. V. Gangadharan
--- Geoff Stone DD&D Australia
<blenrayaust@yahoo.co.uk> wrote:
> Hi Shivakumar,
>
> I would suggest that you may need to re arrange the
> piping in your model.
>
> Another common problem in modelling such nozzles is
> if you make them absolutely
> rigid then enomoius loads and moments will arise. If
> you put the smallest of
> "gaps" in the restraint the nozzle will behave as it
> would in practice and
> accomodate a small amount of strain say .010". You
> may find that the loading is
> more realistic.
>
> Try running the model without this restraint and
> check the strain. WEokr out
> what leg of the piping is causing the problem and
> modify the route or support
> it differently by the use of spring hangars etc.
>
> Geoff Stone
> --- shivakumar angadi <shivku_2000@yahoo.com> wrote:
>
> <HR>
> <html><body>
>
>
> <tt>
> Dear friends,<BR>
> <BR>
> I'm doing analysis for piping where pump is
> getting<BR>
> connected at one end. I have fixed the pump nozzle
> as<BR>
> fixed point. Since there was a huge value for
> moment<BR>
> 'Mx', I have introduced one Ry, rotational
> restraint.<BR>
> Now all the moment has been transferred to this<BR>
> restraint. It is not possible to design the
> support<BR>
> for such a huge moment. Please advise any other
> method<BR>
> to reduce these moments.<BR>
> <BR>
> Thanks<BR>
> <BR>
> shivakumar<BR>
> <BR>
>
__________________________________________________<BR>
=====================================================<BR>
=====================================================<BR>
This archive was generated by hypermail 2.1.8 : Mon Oct 27 2008 - 20:23:58 EDT