Kuemmlee H. Siegl G. Woywode P. - Influence of Elastic Foundation Structures on the Rotor Dynamics.pdf
(
1245 KB
)
Pobierz
77133302 UNPDF
INFLUENCE OF ELASTIC FOUNDATION STRUCTURES ON THE ROTOR
DYNAMICS OF DRIVE TRAINS
Dr. Horst Kuemmlee
Siemens AG
13624 Berlin, P.O. Box 13 03 00
Germany
Dr. Günter Siegl
Siemens AG
13624 Berlin, P.O. Box 13 03 00
Germany
Dr. Peter Woywode
Siemens AG
47015 Duisburg, P.O. Box 10 15 07
Germany
Abstract
-
For large motor-compressor drive trains the
influence of the foundation and the supporting structure
is essential for the rotor dynamics and vibration behav-
ior of the complete system and must be taken into
account during the design process. This paper is a
contribution to the understanding of how rotating ma-
chines that were originally designed for operation on
rigid concrete foundations can be operated safely on
flexible foundation structures. It will be shown, that
despite the physically necessary reversal of some of
the originally demanded criteria, a safe plant operation
can be achieved. Examples of large motor-compressor
systems show the principal difference between a rigid
and a soft installation. The interaction of all system
components is modeled in detail and compared to
common Standards, like DIN 4024. The paper will give
a guideline for performing such a complete system
evaluation and discuss the acceptance criteria in a
case study of a typical large motor-compressor applica-
tion.
These units are normally the heart of the complete oil
& gas plant. Therefore very high requirements must be
demanded for this type of drive train to get a maximum
of availability and reliability.
II. DESIGN BASIS
In general the vibration design of rotor trains is based
on international Standards like IEC or API. Normally
additional specific customer requirements have to be
considered for the mechanical design of rotors for elec-
trical machines, compressors and other driven ma-
chines. The demands on vibration levels have been
tightened over the years. For electrical motors IEC
60034, API 541 and API 546 are the most common
regulations, for compressors API 617 is normally used.
The requirements regarding the location of bending
critical speeds are defined as well as their verification
proof in these standards. The standards for the oil &
gas industry are recognized worldwide. In general
these standards have to be complied with for contracts
coming from this economic sector. Experience shows
that these requirements can be achieved regarding the
rotor dynamic design. Sometimes problems can arise if
the mounting of the machine is different between cus-
tomer’s site and manufacturer’s test field regarding
dynamic behavior. This paper discusses some exam-
ples of different mounting conditions for the rotor dy-
namic behavior of drive trains. It will be shown that
Eigenvalues were shifted in the originally defined speed
range including separation margins and how, by appro-
priate system evaluation, possible problems can be
avoided.
I. INTRODUCTION
For large electric motor compressor trains the indi-
vidual machine cannot be considered as an independ-
ent unit. The interactions between the components and
the surroundings should be taken into account for vi-
bration purposes. The electric motor must also be able
to run uncoupled and coupled to a loading machine in
the manufacturer’s test field, as well as coupled to a
complete drive unit with the compressor and skid in the
manufacturer’s test field and finally on site at the end
destination. Fig. 1 shows a motor-compressor unit for a
natural gas application.
A. Rotor
The finite element model of the rotor is always simpli-
fied to rotating beam elements with all rotational
masses, inertias and other restraints like magnetic pull
of the electric motor included. For high-speed applica-
tions the gyroscopic effects of shaft and components
must be taken into account as well as the flexibility of
“stiff-coupled” components.
B. Bearing system
Fig. 1: Motor-compressor unit for natural gas application
For the rotor dynamic design of a machine shaft it is
normally assumed that the machine is mounted on a
very stiff concrete foundation. This means that the
dynamic characteristic of the foundation is decoupled
from the rotor train, and no interaction between the
natural frequencies of rotor and foundation exists. The
usual standard bearing conditions are the speed-
dependent oil film stiffness and damping combined with
bearing pedestal stiffness and modal masses of the
bearing shell and pedestal, as well as the modal mass
of the base frame (Fig. 2).
The next refinement step for shaft trains mounted on
sub-critical foundations is to add a simplified frame
structure and modal masses for the foundation. For stiff
coupled machines it is necessary to model both rotors
for motor and compressor in coupled condition (Fig. 4)
Fig. 4: Motor and compressor rotor and simplified frame
structure on sub-critical foundation
Fig. 2: model of sleeve bearing and bearing support
For super-critical foundations or machines running
near to a significant foundation Eigenvalue a complete
system analysis including motor frame and foundation
structure is required (Fig. 5). Soft mounted parts - like
motor hubs or auxiliaries – or parts where detailed
structural information is not available – like the gear
box in the example figure – are modeled as rigid mass
and inertia connected by spring elements.
C. Foundation
For foundation design the stiffness of the machines
normally will not be considered. The mass and mass
moment of inertia, e.g. of the motor, will be modeled
using single mass points. The connection of the mass
points can be realized using rigid beams to the foot
point of the machines.
In an ideal case the stiffness of turbo machinery
foundations should be higher than the stiffness of the
machinery oil film and bearing housing, respectively
casing stiffness. For large turbo machinery trains this
cannot be realized in every case due to arrangement
and cost/space restrictions.
In accordance with DIN 4024 there are two kinds of
foundations. First there are so called rigid foundations,
meaning the first Eigenvalues are higher than the op-
erational speed. In most cases so called flexible foun-
dations can only be realized, that means the first or
more Eigenvalues are within the operational speed
range. In that case especially the stiffness of the foun-
dation may influence the rotor dynamic calculation.
To avoid foundation caused vibrations the first Ei-
genmodes of the foundation should be at the minimum
20 % lower or 25% higher than the operational speed.
The higher mode should have a separation margin of
more than 10 %.
The second design criteria also used for foundation
design are effective vibration amplitudes at the ma-
chinery bearing housing and/or casing in accordance
with DIN ISO 10816. For motor design vibration, veloc-
ity in accordance with DIN ISO 10816-3, should not
exceed 4.5 mm/s for rigid foundations and 7.1 mm/s for
flexible foundations. The values are for Zone B/C which
means during operation. For the response analysis an
imbalance quality of Q=6.3 and a structural damping of
2 % in accordance with DIN 4024 should be assumed.
D. Evaluation steps for system analysis
Fig. 5: Complete system analysis including foundation
structure
III.
EXAMPLES
FOR
SYSTEM
VERIFICATION
The system units discussed normally have life times
of some decades and a high grade of investment.
Therefore in many cases only parts of the plant were
renewed after reaching their life time or due to per-
formance improvements. The following chapter deals
with a revamp of an old compressor station and the
second part with planning for a new train.
In the first design stage, motor and compressor ro-
tors are handled as individual units connected by linear
spring elements, which represent the surroundings like
base frame and foundation structure (fig. 3).
A. Revamp of a motor-compressor set for an air
separation plant
Fig. 6 shows an example of a revamp of a stiff cou-
pled motor-compressor train for an air separation plant
mounted on a concrete table foundation. An old syn-
chronous motor is substituted by a new type, the com-
Fig. 3: Motor rotor on stiff foundation
pressor remains. The foundation must be modified to fit
the new motor frame.
structural deformation of the housing is negligible
compared to the piles.
6000
8
8
8
9
7
7
7
[1/MIN]
4140.0
4000
3600.0
3060.0
6
6
6
6
5
1
4
4
5
4
2000
4
3
3
3
1
2
2
1
2
1
1
1
0
Fig. 6: Revamp of an air separation compressor train
0
1000
2000
3000
4000
5000
ROTATIONAL SPEED
[1/MIN]
The existing compressor is rigid coupled to the new
motor rotor and this cannot be changed in the new
design due to a natural frequency location on the com-
pressor side, because of this stiff coupling condition
between motor and compressor the interaction of both
rotors cannot be neglected for the evaluation of dy-
namic behavior. The complete shaft string has to be
treated as a statically undefined support system.
100
[%]
1
80
1
60
1
1
40
4
4
1
20
6
4
4
1
6
6
9
9
7
4
2
5
5
1
7
6
2
2
3
1
3
3
0
8
8
8
1
0
1000
2000
3000
4000
5000
ROTATIONAL SPEED
[1/MIN]
Fig. 7: Standard model for the rotor-compressor shaft
string
Fig. 8: “Critical speed chart“ of drive train system on a
very stiff foundation
(
(
p
(
p
(
In the first design stage the rotor train is modeled on
a rigid foundation represented by spring elements be-
low the bearings (Fig. 7). Fig. 8 shows the critical
speed chart with the Eigenfrequencies and the related
modal damping of the system as a function of the rota-
tional speed. The operational speed range including the
separation margin is free of natural bending frequen-
cies, according to API standard 546.
Imbalance response calculations according to API 546
show no inadmissible resonance amplifications in the
operational speed range. Fig. 9 shows the in-phase
imbalance at the motor active part and Fig. 10 shows the
out-of-phase imbalance reaction. The resonance
amplification of the system is different for the individual
imbalance settings; also the interaction between motor
and compressor is affected.
The next step in the refinement of the vibration model
is to take into account the modal mass of the machine
housing and the foundation table (Fig. 11). In this case
the foundation stiffness is also modeled as linear
springs. The machine housing is modeled as a 3
dimensional rigid body, the mass and moments of inertia
were concentrated in the center of gravity. The rotor
bearing nodes were coupled by rigid links. Some
compressor manufacturers use only vertical steel piles
for height adjustment of the motor and compressor. This
leads to weak supporting stiffness values which can
affect the rotor-dynamic behavior of the system
significantly. This model assumes that the dynamic
Fig. 9: Shaft vibrations caused by in-phase imbalance at
motor active part
Fig. 10: Shaft vibrations caused by out-of-phase imbal-
ance at motor active part
5
1
5
7
9
Fig. 11: Model extended by housing and foundation plate
masses
Due to these additional degrees of freedom additional
Eigenfrequencies can be created and these may fall into
the operational speed range. This can cause the risk that
the originally planned separation margin of the API
cannot be met any more. In the present application no
problems occurred (compare Fig. 12 to 14).
Fig. 14: Shaft vibrations caused by out-of-phase imbal-
ance at motor side
6000
8
8
8
The next step in refinement of the vibration model is to
take into account that the motor and compressor are
mounted on a table shaped concrete or steel foundation
(Fig. 15). The foundation table of this frame itself seems
to be very stiff but it must be taken into account that this
big mass rests on soft and high lateral supports. In
connection with the large additional masses of the
machines and the additional aggregates/auxiliaries this
always produces a complete system with structural
Eigenvalues and related mode shapes that have to be
considered during planning of the installation. In the
process alternating forces between the rotor and the
remaining structure can lead to unforeseen resonance
behavior.
The critical speed chart of this system shows a large
number of Eigenvalues which are also located within the
operational speed range (Fig. 16). The Eigenvalues of
the system generated by the interaction with the
foundation must be taken into consideration because of
possible interference by the rotor modes. The example
shows that normally the separation margins for
calculated foundation modes defined in several
standards cannot be met. The severity of each individual
foundation mode on the rotor-dynamic behavior of the
drive train must be examined and checked for
interference with the rotor. In simplified rotor-dynamic
models only the damping of the oil film is included.
Therefore pure foundation modes lead to zero damping
and high local amplifications. To build a sufficiently
accurate model of the complete system to verify
amplification factors, material damping for foundation
and frame structures must be taken into account.
7
7
7
[1/MIN]
4140.0
4000
3600.0
3060.0
6
6
6
6
5
5
4
5
4
4
9
2000
4
3
3
3
2
2
9
2
1
9
0
0
1000
2000
3000
4000
5000
ROTATIONAL SPEED
[1/MIN]
100
[%]
80
9
60
9
40
4
4
6
20
4
4
1
6
6
7
2
5
5
9
1
7
6
7
2
2
3
1
3
0
8
8
0
1000
2000
3000
4000
5000
[1/MIN]
ROTATIONAL SPEED
Fig. 12: “Critical speed chart“ of a drive train system
including housing stiffness on a very stiff
foundation
(
(
p
(
p
(
Fig. 13: Shaft vibrations caused by in-phase imbalance
at motor side
Fig. 15: Concrete foundation with motor and compressor
as in previous model
1
5
3
8
5000
S
S
T
R
R
R
Q
Q
Q
P
P
P
[RPM]
O
O
O
N
N
N
M
M
M
4000
L
L
L
K
K
K
3600.0
J
J
J
I
I
I
H
H
H
G
G
G
F
F
F
3000
E
E
E
U
D
D
D
C
C
C
9
B
B
B
9
9
9
A
A
A
2000
8
8
8
7
7
7
6
6
6
5
5
4
4
5
1000
3
3
3
4
2
2
2
1
1
1
0
Fig. 18: Shaft vibration at E-motor bearings for out-of-
phase imbalance
0
1000
2000
3000
4000
5000
ROTATIONAL SPEED
[RPM]
60
4
[%]
40
4
9
20
9
4
9
9
4
U
4
5
5
6
6
6
3
5
5
B
C
A
6
3
C
C
8
3
8
0
D
D
D
3
0
1000
2000
3000
4000
5000
[RPM]
ROTATIONAL SPEED
Fig.16: ”Critical speed chart” of the machine set up on a
concrete table foundation
Despite the big number of natural modes near the
operational speed indicated in the critical speed chart the
motor shows no significant resonance amplifications
(Fig. 17 and 18).
The vibration level at the foundation is verified by
vibration velocity at six selected points (Fig. 19). Fig. 20
and 21 show the diagrams for in-phase and out-of-phase
imbalance at the motor rotor. The results show that
resonance peaks at the foundation (Fig. 22 and 23) can
be excited without creating high shaft vibrations.
Dependent on the imbalance distribution, local
foundation resonances can occur.
Fig. 19: Overview of the complete model with num-
bered sensor points.
Fig. 17: Shaft vibration at E-motor bearings for in-phase
imbalance
Fig. 20: Foundation velocities at sensor points 1 to 6 for
in-phase imbalance at motor rotor
T
T
S
B
S
B
T
A
S
A
8
1
E
F
G
J
K
L
Q
E
F
G
J
K
L
O
Q
1
E
F
G
J
K
L
O
Q
2
7
H
I
M
N
O
P
R
1
2
7
H
I
M
N
P
R
2
7
H
I
M
N
P
R
Plik z chomika:
ziolek6661
Inne pliki z tego folderu:
Wang S. - Handbook of Air Conditioning and Refrigeration.pdf
(11053 KB)
Maluf N. Williams K. - Introduction to Microelectromechanical Systems Engineering.pdf
(4324 KB)
Barron Randall F. - Industrial Noise Control and Acoustics.pdf
(3373 KB)
Crowe D. Feinberg A. - Design for Reliability.pdf
(31452 KB)
Fraden Jacob - Handbook of Modern Sensors Physics, Designs, and Applications.pdf
(14658 KB)
Inne foldery tego chomika:
Access
Adam Bodnar - Wytrzymałość Materiałów
ceramika
Chemistry
Comprehensive Organic Synthesis [9 volumes] (1991)
Zgłoś jeśli
naruszono regulamin