Cyclic Symmetry Analysis

Cyclic Symmetry Analysis
Mechanics & Simulation Support Development Group
Overview
model
modelcourtesy
courtesyofofPADT
PADT
Why Cyclic Symmetry?

Theory
Implementation
Commands
Supported Analysis Types
Analysis Guide
Example Models
Future Efforts
F.A.Q.s
Why Cyclic Symmetry?
Reduced model size

Reduced solution time
Widely used in rotating
machinery industry:
Blades
Vanes
Cases
Applications:
Turbines
Gas
Steam
Hydro-Electric
F119 engine image courtesy

of Pratt & Whitney
http://www.pw.utc.com/
Pumps
Fans
3
Theory
F (t )
Full Structure Dynamic Equation:
M &x& + C x& + K x = F
Finite Fourier Series Expansion in
Complex Exponential Form:
M
F j (t ) = f p (t )e i ( j 1) pa
p =0
M
U j (t ) = u p (t )e i ( j 1) pa
a=
i = 1
j = 1,L , N
p = 0,L , M
p =0
N : Number of substructures
: Angle of substructures
(Substructure => Sector)
f p : Coefficients corresponding to the j-th substructures

4
2
N
Theory (continued)
Transformation of the displacements and forces
into finite Fourier series:
U = T up
T=I
F =T fp
ipa
I
e
T =
i ( N 1 ) pa
I e
i ( j 1) pa
=>
j = 1,L , N
Pre-multiplying the dynamic equation by the conjugate transpose of T => T
&& p + T CT u& p + T KT u p = T T f p
T MT u
p is the nodal-diameter
M = N/2
p = 0,L , M
for even
M = (N - 1)/2
5
for odd N
Theory (continued)
The cyclic symmetry transformation relating the
displacement {U} and Fourier components {u} leads to a set
(p = 0, , M) of uncoupled cyclic substructure problems.
M = N/2 for even N
M = (N-1)/2 for odd N
The cyclically symmetric problem is solved on a single

substructure enforcing the compatibility boundary
conditions between the cyclic substructures.
Two most commonly employed solution methods are Duplicate

Sector and Complex Hermitian
For faster performance, ANSYS employs the Duplicate Sector
method
Constraint equations used to enforce cyclic symmetry boundary
conditions on the sector edges
Combine the solutions (real & imaginary) from the two sectors to
obtain the full structure solution
6
Implementation
Basic Sector
Low Edge / High Edge
Duplicate Sector
Constraint Equations
Basic Sector
BASIC
SECTOR
One part of a pattern that, if repeated N

times in cylindrical coordinate space, yields
the complete model
The angle (in degrees) spanned by the
basic sector should be such that N = 360,
where N is an integer.
Meshed or unmeshed geometry
Nodes and elements (no geometry)
User-defined coupling and constraint
equations permitted on internal nodes
Cyclic coordinate system can be any
defined cylindrical system
8
Low Edge / High Edge
HIGH EDGE
LOW EDGE
Edge having the algebraically lower in the

R--Z (cylindrical) coordinate system is
called the low edge
Edge having the higher is called the high
edge
Edge entities stored in components
Lines, areas, nodes
Matched or unmatched node patterns

Internal nodes are those not included on
the low and high edges
Duplicate Sector
Created automatically during

solution
Nodes and elements at same
geometric location as the basic
sector
Loading, boundary conditions, and
coupling and constraint equations
automatically copied from the basic
sector
Option to remove the duplicate
sector, however
The duplicate sector is needed for
displaying cyclic symmetry results
in /POST1 !!
10
Constraint Equations
Used to enforce cyclic symmetry boundary
conditions on the sector edges
Automatically generated during solution
Automatically deleted after solution
NOTE: Nodal coordinate systems of all
nodes on the low and high sector edges
are automatically rotated to be parallel with
the cyclic coordinate system
A
U High
cos k
B =
U High sin k
A
sin k U Low
B
cos k U Low
where,
k = nodal diameter
= sector angle (2 / N)
11
Commands
CYCLIC
Configures the database for cyclic symmetry
CYCOPT
Specifies cyclic symmetry solution options
/CYCEXPAND
Graphically expands elements and results
CYCPHASE
CYCGEN
CYCSOL
/EXPAND
Determines max/min results quantities from

modal analysis frequency couplets
/PREP7
/SOLUTION
/POST1
CYCLIC
Yes
No
No
CYCOPT
Yes
Yes
No
/CYCEXPAND
Yes
No
Yes
CYCPHASE
No
No
Yes
12
CYCLIC
Automatic mode
Automatically detects the number of
sectors, the sector angle, and the cyclic
coordinate system based upon the
existing solid or finite-element model
Manual mode
User must create and verify low and
high edge component pairs
Component names
name_mxxl, name_mxxh
name_uxxl, name_uxxh
m used for potentially matched node
patterns
u used for potentially unmatched node
patterns
name can contain up to 11 characters
xx is the component pair number
starting at 01
13
CYCOPT
CYCOPT,NODDIA
CYCOPT,NODDIA
CYCOPT,DOF
CYCOPT,DOF
CYCOPT,TOLER
CYCOPT,TOLER
CYCOPT,MOVE
CYCOPT,MOVE
Nodal diameter solution ranges

Degrees of freedom (DOFs) for constraint
equations from low to high edges
Tolerance for matching node algorithm
Move nodes to be precisely matched
Status and default reset options
Defaults:
All applicable nodal diameters solved
All applicable DOFs for constraint equations
(1e-6) X (characteristic model length) for
matching node tolerance
Do not move edge nodes
14
CYCOPT (continued)
TOLERANCE BOUNDARY
FOR NODE PAIRING
User
User control:
control:
CYCOPT,MOVE
CYCOPT,MOVE
User
User control:
control:
CYCOPT,TOLER
CYCOPT,TOLER
PRECISELY MATCHED
NOTE: IF ONE OR MORE HIGH EDGE
NODES ARE OUTSIDE THE TOLERANCE
SPHERE, THE UNMATCHED LOGIC IS
USED AUTOMATICALLY
15
/CYCEXPAND
/CYCEXPAND,,AMOUNT
/CYCEXPAND,,AMOUNT
/CYCEXPAND,,WHAT
/CYCEXPAND,,WHAT
/CYCEXPAND,,EDGE
/CYCEXPAND,,EDGE
/CYCEXPAND,,PHASE
/CYCEXPAND,,PHASE
Number of sectors to expand

Element components to expand
Controls for averaging across
sector boundaries
Phase angle shift
Status and default reset options
U j = U A cos{( j 1)k + } U B sin{( j 1)k + }

j = Sector number
k = Nodal diameter
= Sector angle (2/N)
= Phase angle
Uj = Response of the full structure for sector number j

UA = Basic sector solution
UB = Duplicate sector solution
16
CYCPHASE
CYCPHASE,ALL
CYCPHASE,ALL
CYCPHASE,LIST
CYCPHASE,LIST
CYCPHASE,PUT
CYCPHASE,PUT
CYCPHASE,STAT
CYCPHASE,STAT
Maximum and minimum results for

displacement, stress, and strain
List max/min at each node
Put max or min results on basic sector for
plotting
Summary status
U i ( ) = U iA cos U iB sin
Summary of Modal Cyclic-Symmetry Phase Angle Sweep

Load Step = 3 Sub Step = 10
SEQV:
Node
2
MINIMUM
Value
4.29534
Phase Ang (DEG)

36.0000
Node
62
17
MAXIMUM
Value
57134.22623
Phase Ang (Deg)

308.0000
Supported Analysis Types

ANTYPE,STATIC
ANTYPE,STATIC
STATIC Cyclic Symmetry

(with cyclic loading only)
Small Deflection
Large Deflection
ANTYPE,MODAL
ANTYPE,MODAL
MODAL Cyclic Symmetry

Small Deflection Stress-Free
Small Deflection Pre-Stressed
Large Deflection Pre-Stressed
18
STATIC Cyclic Symmetry
Cyclic applied loads

Cyclic applied displacements (constraints)
Nodal diameter zero solution
Small Deflection
Large Deflection
NOTE: It is recommended that this method

be used instead of the CPCYC command
for static cyclic symmetry.
19
MODAL Cyclic Symmetry
To obtain the large deflection

pre-stressed static solution:
Include large deflection effects with
NLGEOM,ON
Include pre-stress effects with
PSTRES,ON
To obtain the large deflection

pre-stressed modal solution, two
options are available:
1. Issue SOLVE command to update
the geometry and solve
2. Update the geometry (via the
UPCOORD command) and issue
the PSOLVE,EIGLANB command
20
Analysis Guide
Cyclic Symmetry Analysis chapter in

the Advanced Analysis Techniques
Guide
Provides central location for
documentation of all supported cyclic
symmetry analysis types
Main sections:
Understanding Cyclic Symmetry
Analysis
Cyclic Modeling
Solving a Cyclic Symmetry Analysis
Post-processing a Cyclic Symmetry
Analysis
Includes modeling process flowcharts

Old modal cyclic symmetry section
removed from the Structural Analysis
Guide (previously section 3.14 at
release 6.0)
21
Example Models
models
modelscourtesy
courtesyofofPADT
PADT
model
modelcourtesy
courtesyofof
Mechanical
MechanicalSolutions,
Solutions,Inc.
Inc.
22
Example Models (continued)
/CYCEXPAND,,PHASE
CYCOPT,DOF
23
Future Efforts
Static Cyclic Symmetry
(with non-cyclic loading)
ANTYPE,BUCKLE
ANTYPE,BUCKLE
ANTYPE,HARMIC
ANTYPE,HARMIC
ANTYPE,TRANS
ANTYPE,TRANS
Linear Buckling Cyclic Symmetry

Harmonic Cyclic Symmetry
Transient Cyclic Symmetry
Heat-Transfer Cyclic Symmetry
24
F.A.Q.s
Q1:
A1:
Q2:
A2:
Q3:
A3:
Q4:
A4:
What is the best way to ensure precisely matched node pairs?

Issue the CYCLIC command before meshing the basic sector. Subsequent AMESH and
VMESH commands will attempt to automatically create precisely matched node pairs.
I am using the CYCOPT DOF option. Why do my deformed shape plots look strange?
Turn off averaging across the sector boundaries using the EDGE option on the /CYCEXPAND
command.
What is the correct method for removing the duplicate sector?
Use the UNDOUBLE (or OFF) option on the CYCLIC command.
Why is the maximum X-stress from CYCPHASE less than the maximum X-stress from
my expanded results plot using /CYCEXPAND?
CYCPHASE currently has several limitations:
Q5:
A5:
Q6:
A6:
The RSYS command has no effect. Component values reported by CYCPHASE are in the global Cartesian
coordinate system. (However, /CYCEXPAND does support RSYS.)
Response results are valid only for the first cyclic sector.
Midside node values are node considered when evaluating maximum and minimum values.
There appear to be many functional and ease-of-use enhancements to cyclic symmetry

but have there been any solver performance improvements?
Yes, the sparse solver has been enhanced to use a symbolic equation assembly process.
This offers many performance benefits for (modal) cyclic symmetry analyses which typically
contain large numbers of constraint equations and by default use the sparse matrix solver and
Block Lanczos eigenvalue extraction method.
Is there any way to glue two dissimilar meshes together in my basic sector and
perform a large deflection analysis?
Yes, you may use surface-to-surface bonded contact within the basic sector. The contact
information is automatically copied to the duplicate sector.
25
F.A.Q.s (continued)
Q7:
A7:
Are there any basic sector mesh restrictions for using automatic detection on the
CYCLIC command?
Yes, automatic detection is valid for models meshed as follows:
Q8:
A8:
All elements meshed into volumes only

All elements meshed into areas only (no volumes selected)
All 3-D solid elements only (no volumes or areas selected)
All 3-D shell elements only (no volumes or areas selected)
All plane 2-D elements only (no volumes or areas selected).
Examples requiring manual specification include models meshed with a combination of

volumes and areas, or models meshed with lines.
How can I animate a traveling wave?
Use the PHASE angle option on the /CYCEXPAND command combined with the /SEG,MULTI
and ANIM,,1 commands to create graphic frames at different phase angles. Refer to the
following example macro:
/USER,1
/SEG,DELE
/SEG,MULTI
frames=arg1
angle=360/frames
*DO,phase,0,360-angle,angle
/CYCEXPAND,,PHASEANG,phase
/REPL
*ENDDO
/SEG,OFF
ANIM,,1,arg2
26

Cyclic Symmetry Analysis

Uploaded by

Copyright:

Available Formats

Cyclic Symmetry Analysis

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cyclic Symmetry Analysis

Uploaded by

Copyright:

Available Formats

Cyclic Symmetry Analysis

Mechanics & Simulation Support Development Group

Why Cyclic Symmetry?

Why Cyclic Symmetry?

Reduced model size

F119 engine image courtesy

Full Structure Dynamic Equation:

(Substructure => Sector)

f p : Coefficients corresponding to the j-th substructures

Pre-multiplying the dynamic equation by the conjugate transpose of T => T

The cyclically symmetric problem is solved on a single

Two most commonly employed solution methods are Duplicate

One part of a pattern that, if repeated N

Low Edge / High Edge

Edge having the algebraically lower in the

Matched or unmatched node patterns

Created automatically during

Determines max/min results quantities from

Nodal diameter solution ranges

Number of sectors to expand

U j = U A cos{( j 1)k + } U B sin{( j 1)k + }

Uj = Response of the full structure for sector number j

Maximum and minimum results for

Summary of Modal Cyclic-Symmetry Phase Angle Sweep

Phase Ang (DEG)

Phase Ang (Deg)

Supported Analysis Types

STATIC Cyclic Symmetry

MODAL Cyclic Symmetry

STATIC Cyclic Symmetry

Cyclic applied loads

NOTE: It is recommended that this method

MODAL Cyclic Symmetry

To obtain the large deflection

To obtain the large deflection

Cyclic Symmetry Analysis chapter in

Includes modeling process flowcharts

Example Models (continued)

Linear Buckling Cyclic Symmetry

What is the best way to ensure precisely matched node pairs?

There appear to be many functional and ease-of-use enhancements to cyclic symmetry

All elements meshed into volumes only

Examples requiring manual specification include models meshed with a combination of

You might also like