Anda di halaman 1dari 12

CAESARII

CAESARII

Li
CAESARII
hot load

L2
CAESARII
y
CL = HL + ky



SUSOPE H
. CAESARII CL H





SUS


OPE

EXP
SUS OPE

Thermal Free
OPE

Thermal Free
OPE EXP

CAESARII

Field Installed Load , Theory Installed load()
Field Installed
Load

For CII:
which position is the OPE displacement and SUS displacement leaving
from?

For the OPE displacement, if the spring stiffness is ignored, then the
OPE restraint load and displacement is as same as the hanger report.

If include spring stiffness for the hanger travel design, for the simple
pipe network, then the hanger hot load is still as same as the OPE
restraint load, but the displacement is as same as the EXP displacement.

For the complicated pipe network, then the hanger vertical displacement
can't be located to match any load case. would you explain why?Richard:

Normally, hanger design is done without including the spring stiffness


in the operating case for hanger travel. So in most cases:

CAESARII
1. CAESAR II puts in a rigid support at the hanger location and does a
weight analysis, finding out how much weight wants to fall to the
support. This load is used as the Hot Load.
1 CAESARII
hot load

2. CAESAR II applies a force equal to that Hot Load (but not stiffness)
at the hanger location and then runs an operating analysis. This tells
CAESAR II how much that location wants to move vertically under the
operating loads. So this becomes the "theoretical" hanger movement, and
the "theoretical" cold load is calculated as CL = HL + ky.
2 CAESARII
y
CL = HL + ky

3. CAESAR II adds the CL as a force and the spring stiffness as a


restraint stiffness to all load cases of the model and then runs all of
the normal load cases.
3 CAESARII CL H

So what happens in the SUS case, for example? Well the CL is going to
be different than the HL, so the as-installed weight case (SUS) will not
be balanced. Let's say, for example, that the CL > HL, so it will tend
to lift the pipe causing extension of the spring and thus reducing the
load that the spring reads once it reaches equilibrium. (IF you want to
know what this value is, you can look at the restraint report for the
SUS case, or ask CAESAR II to report the "actual" -- or post-pulling the
travel stops -- Cold Load). So the reported SUS load -- and the
reported SUS displacement -- reflects the impact of having a CL that is
greater or less than the force necessary to exactly balance the pipe
weight at that location.

SUS CL HL SUS CL>HL


(
SUS
CAESARII SUS

Let's say that the HL = 5000 N, Y = 25 mm up, k = 20 N/mm, so the


"theoretical" CL = 5500 N. If the pipe is quite flexible, the excessive
CL might lift the pipe 15 mm, simultaneously reducing the "actual" CL to
5500 - 15 * 20 = 5200 N. When the pipe heats up, it moves up 25 mm.
There are two ways of looking at what happens next:
HL = 5000 N, Y = 25 mm up, k = 20 N/mm, CL = 5500 N
15 mm 5500 - 15 * 20 = 5200 N
25mm

a) The spring load changes to 5200 - 25 * 20 = 4700 N. This is less


than the load necessary to support the pipe, so the pipe sags 15 mm
under load, back to a position 25 mm above the neutral position.
Simultaneously the spring compresses and the load increases back to 4700
+ 15 * 20 = 5000 N.
1 5200 - 25 * 20 = 4700 N
15mm 25mm 4700
+ 15 * 20 = 5000 N.

b) As the pipe begins to move up, the spring decompresses, reducing the
load. As the spring load decreases -- and the load becomes balanced to
the weight, it begins to lose its ability to pull up on the pipe, so the
lift under the weight load goes back to zero (from the previous 15 mm),
and the operating results match the operating for spring travel case
(spring has HL, displacements match).
2

In other words :

Load Case Displacement Due Displacement Total


Spring
to Unbalanced Due to Thermal Displacement
Load
Spring Load

Ope 0 mm +25 mm +25 mm


5000 N
Sus +15 mm 0 mm +15 mm
5200 N
Exp -15 mm +25 mm +10 mm
-200 N

Load Case Spring Displacement Due Displacement Due to Total Displacement


Load Unbalanced Spring Thermal
Load
Ope 5000 N 0 mm +25 mm +25 mm
Sus 5200 N +15 mm 0 mm +15 mm
Exp -200 N -15 mm +25 mm +10 mm

CAESAR II's Hanger Report would show:


CAESARII
Theoretical Cold Load = 5500 N
Actual Cold Load = 5200 N
Hot Load = 5000 N
Vertical Displacement = +25 mm

(Note that in many cases, if the pipe is quite rigid, there will be very
little vertical displacement due to the unbalanced Cold Load, so
Theoretical Cold Load and the load reported in the SUS restraint report
should be very close.)
SUS

This is how 95% of the world does their hanger design. The Jiangsiu
guys do something a little bit different:
95

1. CAESAR II puts in a rigid support at the hanger location and does a


weight analysis, finding out how much weight wants to fall to the
support. This load is used as the Hot Load. (Same as before.)
1 CAESARII

2. CAESAR II applies a force equal to that Hot Load (but not stiffness)
at the hanger location and then runs an operating analysis. This gives
CAESAR II a first pass estimate of how much that location wants to move
vertically under the operating loads.
2 CAESARII
3. CAESAR II selects a spring based upon that HL and movement, and
calculates a CL = HL + ky. CAESAR II then inserts that Cold Load and
the stiffness of the spring into the model and does the Hanger Operating
Travel case again. In this case, two things happen. First the
unbalanced load pulls the pipe up, reducing the load, and secondly the
stiffness in the model reduces the thermal displacements a bit. So the
actual hanger travel is lower than in the other method.
CAESARII CAESARII CL
H


4. Now CAESAR II uses the HL and the newly determined displacement to
select a new spring, which it inserts in the model, does the analysis,
and repeats step 3 until a converged solution is reached. Then CAESAR
II calculates the theoretical CL (not the actual CL, which the hanger
report is reporting) and inserts it a force and the spring stiffness as
a restraint stiffness to all load cases of the model and then runs all
of the normal load cases.

CAESARII HL 3
CAESARII CL

So basically CAESAR II iterates until it figures out the displacement


between the "actual" cold case and the operating case.
CAESARII

So assuming that the same situation as above was done in this way,
CAESAR II would report a CL of 5200, a HL of 5000, and a Y = 10 mm up,
and a k = 20 N/mm. Jiangsiu would order a spring pre-set to a Cold Load
of 5200 N and then install it. So how do they get a way with using a CL
of 5200 N rather than 5500 as in the other example? When they install
the spring, it will initially lift up, reducing the spring load. So
they have to adjust the load in the field (that is why we print out that
warning note in the hanger report), lifting up the pipe and further
compressing the spring until it is reading 5200 N again!! What is the
effect of this? It is identical to as if they had pre-set the spring to
5500 N and just let it drift, after installation, to 5200 N!!
CAESARII CL 5200 HL 5000, and a Y = 10 mm
k = 20 N/mm 5200
5500N
5200 N
5500 N 5200 N
So both cases are the same -- except that in the first, the overload is
added to the spring in the factory pre-set, and in the second it is
added via field adjustment.

So the results will look the same in the CAESAR II reports:


CAESARII
Load Case Displacement Due Displacement Total
Spring
to Unbalanced Due to Thermal Displacement
Load
Spring Load

Ope 0 mm +25 mm +25 mm


5000 N
Sus +15 mm 0 mm +15 mm
5200 N
Exp -15 mm +25 mm +10 mm
-200 N
Load Case Spring Displacement Due Displacement Due to Total Displacement
Load Unbalanced Spring Thermal
Load
Ope 5000 N 0 mm +25 mm +25 mm
Sus 5200 N +15 mm 0 mm +15 mm
Exp -200 N -15 mm +25 mm +10 mm

Except now, since we selected the spring using the iterative method,
CAESAR II's Hanger Report would show:
CAESARII
Field Verified Cold Load = 5200 N
Hot Load = 5000 N
Vertical Displacement = +10 mm

So what are the disadvantages to the Jiangsiu method? They have to


inspect and probably re-adjust every spring in the field to make sure
that it is reading its Cold Load after installation.
CL
What are the advantages? In many cases they can buy a spring with a
smaller travel range. For example, in the first case, we needed a
spring that can handle loads of both 5000 N and 5500 N. In the second
case, we could buy a spring that only needed to handle loads between
5000 N and 5200 N -- so they can save some money.

Regards,

Tom Van Laan, PE


President
(e-mail) tvanlaan@coade.com

COADE, Inc.
12777 Jones Rd., Suite 480
Houston, Texas 77070
(ph) 281-890-4566
(fax) 281-890-3301
(web) http://www.coade.com

would you explain on the bellow questions to us?


1.for the thermol free (HGR) case, is the hanger removed, then caculate
the displacement, or has the preload (use the restraint weight(HGR) load
over there), then get the displacement?

Yes the preload is applied to the hanger location in the "free thermal"
load case for hanger design.

2.For the hanger vendor, the hanger is compressed to a load(Preset


load), and delivered to the site for installation, is the preset load as
same as the theoretical load or not?

Yes, we assume this is the same.

3.If the stiffness is included, then the field installed load is get
instead of the theoretical load. what's the difference of the load,
especiall the text note as :.....VERIFY THAT THE ACTUAL READING DURING
THE PIPING COLD CONDITION CORRESPONDS TO THE "FIELD INSTALLED LOAD"
LISTED IN THIS TABLE.

The difference between the theoretical and actual installed loads is


detailed in Chapter 6 of the Technical Reference. Typically the
difference is very small, but may be large if the piping is relatively
flexible and/or the "Theoretical Installed Load" is significantly
unbalanced from the hot load. Yes the actual reading during the piping
cold condition corresponds to the "field installed load".

If no on site verification, that means that the hangers design fails


too.

>Hi Richard AY
>I was questioned on the vertical movement of hangers. the local
>software's vertical movement of the hangers is the movement from the Sus
>to OPE. CAESARII's movement is from the installation position to the
>Ope, it sounds that the Ope is comes from SUS.
>
>Do you agree with us? and how do you explain on the vertical movement of
>Hanger?
>The vertical movement shown in the hanger report is the movement at the
>hanger node from the "free thermal case" used in designing the hanger.
>This could (and probably will) differ from the displacements shown in
>your real load cases. This is because the real load cases include both
>the hanger stiffness and the preload.

HI Richard Ay
If so, how do you order your hanger from the manufacturer. how many data you need to offer to
the vendor? do you think this vertical movement is ok?
Typically you only need the data from the "hanger report". The load and and travel in the
individual "real" load cases should fall within the working range of the hanger (unless the user
fouled things up by say adding a T2 vector that wasn't considered in the hanger design).

If you wanted to be complete, also provide the restraint summary for all load cases at the hanger
location.

Hi Richard Ay
Right now, one China Power CII user is facing a question like this:
if he don't include the spring stiffness for the analysis, a few hangers
is designed as 2 instead of 1 hangers.

If they include the spring stiffness, then the hangers is designed as


they expected, but for the vertical movement has some difference to the
previous hangers.

Would you explain why? and how to get an accurate and correct spring
hanger design for the power industries?

If he don't include the spring stiffness for the analysis, a few hangers
is designed as 2 instead of 1 hangers.
I presume you are talking about the load case "operatng for hanger
travel"? Yes I would expect different results based on whether or not
you included the stiffness (as designed) or ignored the stiffness (the
default behavior). As to whether the design yields 1 or 2 springs is
simply a function of the load and the spring table. Unless the system
is particularly sensitive to the hangers, I would expect the 2 hanger
situation to yield smaller hanges, about equivalent to the single larger
hanger.

Again, I would expect differences in the movements between the "include


stiffness" and "ignore stiffness" conditions.

On CII's application over China Power industries. as Tom and Bill had been to China for this
issure, and COADE added include hanger stiffness too. as we are having more and more sales
over China power industries.

Still I had been asked a lot of the reliability of CAESARII. so my questions are:
If the user start to running CII as the other process industries engineer did to analysis main
steam/reheat and others major pipeline system, do you think it's dangerous to get the wrong
hanger/too much constant hangers and others compared the China Power in house software.
1) "do you think it's dangerous to get the wrong hanger/too much constant hangers and others
compared the China Power in house software" No. CAESAR II has been verified over and
over and over again. I can't vouch for what China Power's in house software does, so yes there is
a possibility that the two programs may yield different results.

If they are having cold spring with vertical pipeline + Hangers, do they need to run in 2 steps as
COADE's guideline, if they run in one step, how is the result?
2) "If they are having cold spring with vertical pipeline + Hangers, do they need to run in 2 steps
as COADE's guideline, if they run in one step, how is the result?" Not any more, since we have
separated cold spring (CS) and hanger preloads (H) from the concentrated force vector (F1). You
still need to be careful with your load case setup, since you don't want cold spring to affect the
hanger design.

How did Simens/Alstom and other big Power engineering corps. over the world'a applications of
CAESARII?

I did trust CAESARII, would you offer us some application sample input data for a reference?

3) "How did Simens/Alstom and other big Power engineering corps. over the world'a applications
of CAESARII?" I can't answer this one. Customers never come back to us and tell us how they
are using the software.

per our client requirement, would you help them make the decision on the
attached job, if this is your job, are going to include the spring
stiffness or not?

As you know, the option to "include spring stiffness" in the hanger


design (which includes the stiffness in the free thermal case and causes
the hanger design to iterate), was added for Jiang Su, in Version 4.40.
The idea behind this option is that by including the spring stiffness in
the hanger design algorithm (which necessitates the iterations), you
fine tune the design and end up with less variability in the final
hanger loads.
This particular job illustrates this theory. Just review the attached
output files, which prove the point. For this job, yes I would include
the spring stiffness in the hanger design.

Anda mungkin juga menyukai