Input to VIVANA
1. General information
Input to the VIVANA
module comprises two data files.
One of these data files must be an ascii file given by the user, and the other one must be a binary result file from the static solution found by RIFLEX
(<prefix>_ifnsys.sam
). The file provided by the user must be named using the same prefix as was used for the RIFLEX
analysis.
This chapter describes the contents of the user defined input file to
VIVANA
(<prefix>_vivana.inp
). A typical input file for VIVANA
is
listed in Appendix A.
The input to VIVANA
is divided into seven sections:
Theoretical descriptions of the VIV evaluation scheme and the response analysis method are given in the VIVANA Theory Description.
2. Control information
2.1. Data group identifier, one input line
VIVAna CONTrol INFOrmation chvers

CHVERS: character(8)
: VIVANA input file version, e.g. 4.0
Identification of the analysis must be given in three lines of text.
This text will be printed to the vivana.res
file.
2.2. Heading, three input lines
HEADn

HEADn: character, default: CH
: Line n of heading text
Always 3 card images which may all be blank.
2.3. Identification of system and environment, one input line
IDRIS IDSTAT IDENV TEMP IRSTYP

IDRIS: character(6)
: Riser identifier, corresponding to the model established byRIFLEX
moduleINPMOD
and used inSTAMOD
. See Input to INPMOD, Data Group B, Riser type specification in RIFLEX User Guide. 
IDSTAT: character(6)
: Static condition identifier, corresponding to the static analysis carried out byRIFLEX
moduleSTAMOD
. See Input to STAMOD, Data Group A, Principal run parameters in RIFLEX User Guide. 
IDENV: character(6)
: Environment identifier, corresponding to the environment established byRIFLEX
INPMOD
and used inSTAMOD
. See Input to INPMOD, Data Group D: Environmental Data in RIFLEX User Guide. 
TEMP: real, default: 4
: Water temperature \([\mathrm {deg}]\) 
IRSTYP: integer, default: 1
: Option for type of VIV loads to be applied
= 1
: Crossflow VIV loads are applied 
= 2
: Inline VIV loads are applied 
= 3
: Combined crossflow and inline VIV loads are applied

3. VIV frequency analysis
This data group is optional, default values will be inserted if the data group is not given.
3.2. Number of eigenvalues and vectors, one input line
NEIG

NEIG: integer, default: 10
: Number of eigenvalues to be calculated
The number of eigenvalues and eigenvectors to be calculated is the total
number of frequencies. Once these have been found the program selects
the crossflow or inline frequencies as relevant. Approximately 50% of
the calculated frequencies are inline and 50% are crossflow. This
means that in order to study VIV response up to frequency no. 30, the
user must specify NEIG
>= 60.
3.3. Computation parameters, one input line
The parameters below correspond to Lanczos’ method for solution of eigenvalue problems. For a detailed discussion, see NourOmid et al. or Parlett.
TOL MAXLAN

TOL: real >0, default: 1.0e10
: Maximum acceptable relative error in computed eigenvalues 
MAXLAN: integer, default: 8+2*NEIG
: Maximum number of Lanczos steps (vectors) to be used. Note that if specifiedMAXLAN >= 8+2*NEIG
4. Specification of section properties
This data group is mandatory. Property types such as excitation zone
definition, added mass as a function of nondimensional frequency,
excitation coefficient as a function of nondimensional frequency and
response, damping and Strouhal number may be specified by the user.
However, data specified by the user will be disregarded by VIVANA
if
inline VIV is to be calculated. In that particular case, VIVANA
will
use default data irrespective of what is specified in this datagroup.
For the various property types; e.g. added mass, excitation, Strouhal number; an arbitrary number of data sets may be defined in the input file. Within each of these property types the data sets are given unique numbers. The cross section properties for each segment are then defined by referring to these numbers. Appendix A: Typical VIVANA Input File shows an example where several property types are given.
4.1. Data group identifier, one input line
SECTion PROPerty SPECification
NSEGP

NSEGP: integer
: Number of segments for which properties are given. Must be equal to the number of segments in the model specified inINPMOD
4.2. Specification of crosssection property for each segment, NSEGP input lines.
ISEGP IEXCZO IADDMA1 IEXCIT1 IDAMPG ISTROU IADDMA2 IEXCIT2

ISEGP: integer, default: 0
: Segment number (global). 
IEXCZO: integer, default: 0
: Cross section property specification for excitation zone limits. Refers toIPRONO
in Specification of section properties. 
IADDMA1: integer, default: 0
: Cross section property specification for crossflow (IRSTYP = 1
orIRSTYP = 3
) or inline (IRSTYP = 2
) added mass as a function of nondimensional frequency. Refers toIPRONO
in Specification of section properties. 
IEXCIT1: integer, default: 0
: Cross section property specification for crossflow (IRSTYP = 1
orIRSTYP = 3
) or inline (IRSTYP = 2
) excitation coefficient as a function of nondimensional frequency. Refers toIPRONO
in Specification of section properties. 
IDAMPG: integer, default: 0
: Cross section property specification for hydrodynamic damping. Refers toIPRONO
in Specification of section properties.
The program has two possible ways to calculate the damping, see Specification of section properties in this manual :

The damping terms \(\mathrm {R_{sw}}\), \(\mathrm {R_{lv}}\) and \(\mathrm {R_{hv}}\) will be used directly, ref. VIVANA Theory Description, Section The Damping Model, Hydrodynamic damping outside the excitation zone.

The damping is calculated using the excitation curves defined for that section. A still water damping coefficient must be given as input.



ISTROU: integer, default: 0
: Cross section property specification for Strouhal number. Refers toIPRONO
in Specification of section properties. 
IADDMA2: integer, default: 0
: Cross section property specification for inline added mass as a function of nondimensional frequency for the case of combined crossflow and inline (IRSTYP = 3
). Refers toIPRONO
in Specification of section properties. Note that the response frequency iterations will adjust the inline added mass so that the inline response frequency is twice that of the crossflow response frequency.IADDMA2
must be zero for pure crossflow (IRSTYP = 1
) or pure inline (IRSTYP = 2
). 
IEXCIT2: integer, default: 0
: Cross section property specification for inline excitation coefficient as a function of nondimensional frequency for the case of combined crossflow and inline (IRSTYP = 3
). Refers toIPRONO
in Specification of section properties.IEXCIT2
must be zero for pure crossflow (IRSTYP = 1
) or pure inline (IRSTYP = 2
).
The crosssection properties referenced in this data group must be
defined using any of the following PROPERTY
 data groups.
The default value of zero for IEXCZO
, IADDMA1/2
, IEXCIT1/2
,
IDAMPG
and ISTROU
tells the program to use the following:
Crossflow VIV (IRSTYP = 1)  Inline VIV (IRSTYP = 2)  

Excitation zone limits 
\(\hat{f}_{\mathrm {min}}\):0.125 \(\hat{f}_{\mathrm {max}}\):0.3 
\(\hat{f}_{\mathrm {min}}\):0.2 \(\hat{f}_{\mathrm {max}}\):0.9 
Added mass as function of \(\mathrm {\hat{f}}\) 
See the figure with the VIVANA CF added mass model in the Theory Manual, Section Calculation of Response Frequencies, Added mass, CF response. 
See the figure with the VIVANA IL added mass model in the Theory Manual, Section Calculation of Response Frequencies, Added mass, pure IL response. 
Excitation coefficients as function of \(\mathrm {\hat{f}}\) 
A modified version of Gopalkrishnan’s coefficients will be used, ref. the Theory Manual. Section Excitation Force Model, Default excitation coefficient model, CF response. 
A modified version of Aronsen’s coefficients will be used, ref. the Theory Manual. Section Excitation Force Model, Default excitation coefficient model, pure IL response. 
Hydrodynamic damping 
The damping terms \(\mathrm {R_{sw}}\), \(\mathrm {R_{lv}}\) and \(\mathrm {R_{hv}}_{}\) will be used directly, ref. the Theory Manual, Section The Damping Model, Hydrodynamic damping outside the excitation zone. 
Same as for Crossflow VIV. 
Strouhal number definition 
The program calculates the Strouhal number as a function of the Reynolds number for each element in the model, see the Theory Manual, Section Dimensionless Parameters, Strouhal number. 
Same as for Crossflow VIV. 
A constant added mass of 1.0 is recommended for analyses with response
frequencies corresponding to high modes; e.g. above mode 15. The
parameters IMOD_CURVE and IMOD_CONST in
Specification of section properties may be used to specify a smooth
transition between frequencydependent added mass and the constant
stillwater added mass.

A specified transition will also be applied to the default added mass model, see Illustration of transition between frequency dependent and constant stillwater added mass. 
4.3. Excitation zone properties
4.3.2. Number of excitation zones specified
NEXZON

NEXZON: integer
: Number of excitation zones specified in this datagroup.
4.3.3. Specification of excitation zone limits for various sections of the structure, NEXZON input lines.
IPRONO CPRPID FHMIN FHMAX

IPRONO: integer
: Crosssection property specification number for property type Excitation zone properties. 
CPRPID: character(8)
: Text to identify the crosssection property 
FHMIN: real, default: 0.125
: Minimum value of \(\mathrm {\hat{f}}\) in the excitation zone. 
FHMAX: real, default: 0.2
: Maximum value of \(\mathrm {\hat{f}}\) in the excitation zone.
It is not allowed to specify a broader excitation range for a crosssection than the range used for specification of the associated excitation coefficient curves.
4.4. Added mass properties
4.4.2. Number of added mass curves and added mass transition
NADCUR ITRANS_LOW ITRANS_UP

NADCUR: integer >= 0
: Number of added mass curves to be defined by the user. 
ITRANS_LOW: integer >= 0, default: 1000000
: Last mode number in the active VIV direction for which the full frequencydependent added mass curves will be used. 
ITRANS_UP: integer >= 0, default: 1000000
: First mode number in the active VIV direction for which the constant stillwater added mass will be used.ITRANS_UP >= ITRANS_LOW
A smooth transition between the frequency dependent added mass and the
constant stillwater added mass will be applied between ITRANS_LOW
and
ITRANS_UP
, see Figure 1. The specified transition
will also be applied to the default added mass model.
Note that constant added mass is recommended for high modes.
4.4.3. Specification of added mass as a function of nondimensional frequency, repeated NADCUR times
One input line
IPRONO CPRPID NAMPT

IPRONO: integer
: Crosssection property specification number for property type Added mass properties. 
CPRPID: character(8)
: Text to identify the crosssection property 
NAMPT: integer >= 1
: Number of points in \(\mathrm {C_a}\,\hat{f}\) curve.
Constant added mass can be given either by using NAMPT=1
or by using
an constant added mass curve NAMPT>=2
.
Note that if IRSTYP=3
(Combined crossflow and inline VIV) and
constant added mass curve with NAMPT=1
is used for all sections, no IL
frequency iteration is performed. The IL frequency is twice the CF
frequency.
Constant added mass coefficient, NAMPT=1
input lines
ADDMCO

ADDMCO: real
: Added mass coefficient
Added mass coefficient as a function of nondimensional frequency,
NAMPT>=2
input lines
FHAT ADDMCO

FHAT: real
: Nondimensional frequency 
ADDMCO: real
: Added mass coefficient
4.5. Excitation coefficient properties
Two different ways of specifying excitation coefficients are available
in VIVANA
. The excitation coefficient is used in the response analysis
for calculating the excitation force on the cylinder. Calculation of
excitation force is irrespective of how the excitation coefficients are
specified by the user.
Excitation force is calculated for elements within the excitation zone. The excitation zone is defined by the excitation frequency bandwidth. The excitation frequency bandwidth is by default [0.125, 0.3] for crossflow VIV, but it is possible for the user to define his/her own excitation frequency range for any part of the model, see Specification of section properties. We would normally recommend the user to set the frequency bandwidth equal to the frequency range for which excitation coefficients are provided. Elements outside of the excitation zone will add damping to the system, see the Theory Manual, Section Calculation of Response Frequencies, Added mass, CF response for details.
4.5.1. Specification of excitation coefficient as a function of nondimensional frequency and response amplitude.
PROPerty EXCITATION COEFFICIENT
4.5.2. Number of excitation coefficient data sets to be specified
NLCT01 NLCT02

NLCT01: integer, default: 0
: Number of excitation coefficient data sets to be specified on the same format as the default curves (Type 1). 
NLCT02: integer, default: 0
: Number of excitation coefficient data sets to be specified on the table format (Type 2)
4.5.3. Specification of default format excitation coefficient data (Type 1), 2xNLCT01 input lines
One input line
IPRONO CPRPID NFT01

IPRONO: integer
: Crosssection property specification number for property type Excitation coefficient property. 
CPRPID: character(8)
: Text to identify the crosssection property 
NFT01: integer >= 2
: Number of points in \(\mathrm {C_e(}\hat{f}\mathrm {,\frac{A}{D})}\) curve specification.
Excitation coefficient as a function of nondimensional frequency and
response amplitude, NFT01
input lines
FHAT ACL0 ACLMAX CLMAX CLA0

FHAT: real
: Nondimensional frequency 
ACL0: real
: \(\mathrm {\frac{A}{D}}\) ratio for when \(\mathrm {C_e}\)= 0

ACLMAX: real
: \(\mathrm {\frac{A}{D}}\) ratio for when \(\mathrm {C_e}\)=
\(\mathrm {C_{emax}}\) 
CLMAX: real
: Maximum excitation coefficient 
CLA0: real
: Excitation coefficient for \(\mathrm {\frac{A}{D}}\)= 0
The definition of these four parameters is shown in a figure in the Theory Manual, Section Excitation Force Model, Default excitation coefficient model, CF.
4.5.4. Specification of table format excitation coefficient data set (Type 2). The input data described in the following is to be specified for NLCT02
excitation coefficient data sets
IPRONO CPRPID NFT02

IPRONO: integer
: Crosssection property specification number for property type Excitation coefficient property. 
CPRPID: character(8)
: Text to identify the crosssection property 
NFT02: integer >= 2
: Number of nondimensional frequencies for which an excitation coefficient curve is specified for this data set.
Specification of table format excitation coefficient curves (Type 2).
The input data described in the following is to be specified for NFT02
excitation coefficient curves.
FHTYP2 NPOINT

FHTYP2: real
: Nondimensional frequency. 
NPOINT: integer >= 2
: Number of data points forFHTYP2
Excitation coefficient as a function of
\(\mathrm {\frac{A}{D}}\) for FHTYP02
, NPOINT
inputlines
AD CL

AD: real
: Amplitude to diameter ratio. 
CL: real
: Excitation coefficient
AD
and FHTYP2
must be given in either increasing or decreasing
order.
It is crucial for the analysis that the data covers to complete \(\mathrm {\frac{A}{D}}\) ratio range. That means that the minimum \(\mathrm {\frac{A}{D}}\) ratio should be zero and the maximum value approximately 2.
It is recommended that all extrapolation be done outside of VIVANA
.
If the excitation curves are used to model sections covered with VIV suppression devices, one should note that:

The theoretical model can cover cases with up to approximately
75% coverage of VIV suppression devices. For these cases the bare riser controls the VIV behaviour (frequency, mode etc.)

For larger coverages the straked riser takes control of the VIV
behaviour. Model tests indicate a different physical behaviour than for the bare controlled riser. The behaviour seems to be dependent on pitch and height of the strakes. The responding frequencies and modes are generally lower.
4.6. Damping factor properties
4.6.2. Number of damping factors specified
NDPFAC01 NDPFAC02

NDPFAC01: integer
: Number of damping factor sets for various sections of the structure relative to Venugopal. 
NDPFAC02: integer, default: 0
: Number of damping factor sets for various sections of the structure using user defined excitation curves in table format.
4.6.3. Specification of damping factor sets for various sections of the structure, NDPFAC01 input lines.
IPRONO CPRPID FSTILL FLOWV FHIGHV

IPRONO: integer
: Crosssection property specification number for property type Damping factor properties. 
CPRPID: character(8)
: Text to identify the crosssection property 
FSTILL: real, default: 1
: Factor Venugopal still water damping contribution, \(\mathrm {R_{s\alpha }}\). 
FLOWV: real, default: 1
: Factor Venugopal low velocity region. 
FHIGHV: real, default: 1
: Factor Venugopal high velocity region.
4.6.4. Specification of damping factor for various sections of the structure, NDPFAC02 input lines.
IPRONO CPRPID FSTILL

IPRONO: integer
: Crosssection property specification number for property type Damping factor properties. 
CPRPID: character(8)
: Text to identify the crosssection property 
FSTILL: real, default: 1
: Still water damping factor
If a section has damping defined as NDPFAC02
, the section must have
excitation coefficients given on table format.
The still water damping coefficient is calculated using the empirical formulae
\(\mathrm {c_{sw}=\frac{\omega \pi \rho D^2}{2}(1+(\frac{A}{D})^2)F_{still}}\)
The value of \(\mathrm {F_{still}}\) must be found using curve fitting of the data points \(\mathrm {c_{sw}}\) as a function of the \(\mathrm {\frac{A_{}}{D}}\)ratio. These data points can be found from still water decay tests.
4.7. Strouhal number properties
PROPerty STROuhal SPECification
The user may allow the program to calculate local Strouhal number for an element as a function of the Reynolds number. The program then interpolates in a user defined \(\mathrm {S_t(Re)}\) curve if given, or in a default \(\mathrm {S_t(Re)}\) curve.
Alternatively, the user may specify a constant Strouhal number for all
the elements in a segment. This is done by specifying STROU
. STROU
will then be used as the Strouhal number for all the elements in the
segment.
NSTRSP

NSTRSP: integer
: Number of Strouhal number sets specified in this data group.
4.7.1. Specification of Strouhal number sets, NSTRSP data sets.
IPRONO CPRPID NPUDSC STROU

IPRONO: integer
: Crosssection property specification number for property type Strouhal number properties. 
CPRPID: character(8)
: Text to identify the crosssection property 
NPUDSC: integer, default: 0
: Number of points in user defined StrouhalReynolds number curve.
NPUDSC < 0
: Default curve will be used 
NPUDSC = 0
: Constant Strouhal number 
NPUDSC > 0
: User defined curve will be used


STROU: real, default: 0.19
: Fixed Strouhal number
Dummy for
NPUDSC
\(\mathrm {\neq }\)0

5. Structural damping specification
This data group allows additional material and slip damping to be
specified for some or all segments in the system. This structural
damping is read from separate files and is applied in addition to the
VIV response analysis RELDAM
. The structural damping is given as a
function of the response curvature and is therefore updated during the
response iterations.
At present, Newton  Raphson response iterations must be used, i.e.
IFRIT = 2
: in VIV response analysis.
5.2. Number of specifications, one input line
NSPEC

NSPEC: integer, default: 0
: Number of segments with detailed structural damping specification
5.3. Detailed structural damping specification, NSPEC input lines
IGSEG CHFILE

IGSEG: intger > 0, default: 0
: Global segment number 
CHFILE: character(256)
: File with detailed structural damping specification, See the file description. The same file may be specified for multiple segments.
Detailed debug information about the applied structural damping may be
found in the file <prefix>__eledam.asc
.
6. VIV response analysis
6.2. Response parameters, one input line
RELDAM IOPFRC IPRINT

RELDAM: real, default: 0
: Relative structural damping
RELDAM = 0.1
gives 10% relative damping.


IOPFRC: integer, default: 1
: Force switch
IOPFRC = 0
: Forces calculated using stiffness matrix 
IOPFRC = 1
: Forces calculated using curvature and axial strain. This option requires that the axial and bending stiffness is linear for all elements. If nonlinear stiffness is found, forces, stress and fatigue will not be calculated. Curvature time series can still be printed, see VIV fatigue analysis.


IPRINT: integer, default: 1
: Print switch
IPRINT = 1
: Final results are printed to Matrix Plot file 
IPRINT = 2
: Final results, results from the final iteration and axial force and bending moments are printed to Matrix Plot file 
IPRINT = 5
: Final results, detailed results from all iterations and axial force and bending moments are printed to Matrix Plot file.

If the forces are calculated using curvature and axial strain (IOPFRC = 1 ), shear stiffness cannot be modelled and must be set equal to zero for all cross sections.

6.3. Response parameters, one input line
IFRIT MAX_ITER CHILIM CONLIM SCAINI IOPTRY

IFRIT: integer, default: 2
: Response iteration method
IFRIT = 1
: Fixed point 
IFRIT = 2
: Newton  Raphson


MAX_ITER: integer, default: 30
: Maximum number of iterations 
CHILIM: character(6), default: AMPNOR
: Convergence criterion
CHILIM = AMPNOR
: Norm of amplitude change for translations. The norm is equal to the squared sum of the amplitude changes normlized with the product of the number of nodes and squared the average diameter. 
CHILIM = DIFMAX
: Maximum difference. This is the maximum absolute change in amplitude; i.e. phase change ignored. 
CHILIM = NONE
: No convergency test.MAX_ITER
iterations will be performed. 
CHILIM = 5
: AsAMPNOR
, Norm of amplitude change for translations 
CHILIM = 3
: AsDIFMAX
, Maximum difference 
CHILIM = 0
: AsNONE
, No convergency test 
CHILIM = 1
: Quadratic norm. Not recommended. 
CHILIM = 2
: Relative maximum difference. Not recommended. 
CHILIM = 4
: Unbalanced force. Not recommended.


CONLIM: real, default: 0.0001
: Convergence limit for the iteration. Recommended values 0.001  0.01 forCHILIM = AMPNOR
.
CHILIM = NONE
or0
: Dummy 
CHILIM = DIFMAX
or3
: Dimension \(\mathrm {[L]}\) 
All other values of
CHILIM
: Nondinemsional


SCAINI: real, default: 0.5
: Scaling factor for the initial response estimate
SCAINI = 0
: Scale the corresponding mode shape by the average \(\mathrm {\frac{A}{D}*D}\) for zero excitation in the excitation zone, weighted by the mode shape. The IL part of combined CF and Il loading,IRSTYP = 3
, is scaled with half the value found for CF. 
SCAINI > 0
: Scale initial response estimate so that the maximum amplitude is \(\mathrm {SCAINI*D_{avg}}\) for CF loading,IRSTYP = 1
, IL loading,IRSTYP = 2
and the CF part of combined CF and Il loading,IRSTYP = 3
. The IL part of combined CF and Il loading,IRSTYP = 3
, is scaled with half this value.


IOPTRY: integer, default: 0
: Option to retry response iterations if the response iteration does not converge. Under development.
IOPTRY = 0
: the analysis will continue and the nonconverged results used 
IOPTRY = 1
: a second attempt will be made with the other response iteration method. May not be combined with data group Structural damping specification.

6.4. Response parameters, one input line
IOPTSH NUDDF ADLIM REXPAL REXPAH

IOPTSH: integer, default: 0
: Option for combining response frequencies
IOPTSH = 0
: Response frequencies act concurrently, i.e. space sharing 
IOPTSH = 1
: Response frequencies act consecutively, i.e. time sharing


NUDDF: integer, default: 0
: Number of dominating frequencies given in user defined frequency ranking. Not used forIOPTSH = 1
. 
ADLIM: real, default: 0.01
: Amplitude limit for including frequencies in the calculated response, normalized by the minimum diameter \(\mathrm {[1]}\). 
REXPAL: real, default: 0
: Cutoff excitation parameter ration for frequencies below the identified dominating frequency \(\mathrm {[1]}\). 
REXPAH: real, default: 0
: Cutoff excitation parameter ration for frequencies above the identified dominating frequency \(\mathrm {[1]}\).
Analysis of frequency content in time series from VIV experiments with high response frequencies shows that the response process is often somewhere between concurrent and consecutive. The response frequency at a location shifts between different consecutive values. However, other frequencies often dominate at other locations. Further work is needed before firm recommendations may be given.
For cylinders with constant diameter in linearly sheared flow the values
REXPAL = 0.2
and REXPAH = 1.0
agree well with experiments.
6.5. User defined frequency ranking, NUDDF input lines
IDOMFRQ

IDOMFRQ: integer
: Frequency number
The most dominating frequency is to be specified first, then the second most dominating frequency is specified etc.
If NUDDF = 0
the program will give the possible response frequencies a
ranking according to an excitation parameter for each frequency.
\(\mathrm {E_i=\sum\limits_{j=1}^{N_{Ex}}l_j\,D_j^2\,u_j^2(\frac{A}{D})_{C_e=0,\,j}}\)
where

\(\mathrm {E_i}\) is the excitation parameter for frequency i.

\(\mathrm {l_j}\) is the length of element j within the excitation zone.

\(\mathrm {D_j}\) is the diameter of element j.

\(\mathrm {u_j}\) is the flow velocity at the midpoint of element j.

\(\mathrm {N_{Ex}}\) is the number of elements in the excitation zone.

\(\mathrm {(\frac{A}{D})_{C_e=0,\,j}}\) is the normalised response amplitude that give \(\mathrm {C_e=0}\), given as a function of the local value of \(\mathrm {\hat{f}}\) for frequency i.
7. VIV fatigue analysis
7.2. Control data, one input line
IOPFAT NSECT NPCS IOPPR TSIOPPR CHTSPRN

IOPFAT: integer, default: 0
: Option for fatigue calculation
IOPFAT = 0
: Determined by parameterIOPTSH
in VIV response analysis.
For
IOPTSH = 0
, concurrent response frequencies, the default is 1, rainflow counting. 
For
IOPTSH = 1
, consecutive response frequencies, the default is 2, Rayleigh distributed stress cycles


IOPFAT = 1
: Fatigue calculated in time domain using rainflow counting.
Currently not available for
IOPTSH = 1
.


IOPFAT = 2
: Fatigue calculated in frequency domain using Rayleigh distributed stress cycles.
Currently not available for
IOPTSH = 0
. 
May only be used for SN curves with a single straight line and bilinear SN curves.


IOPFAT = 3
: Fatigue calculated directly from constant stress amplitude.
Not available for
IRSTYP = 3
orIOPTSH = 0
and multiple response frequencies.



NSECT: integer, default: 0
: Number of crosssections to be specified below
NSECT = 0
: All elements 
NSECT > 0
: Additional data is given in the end of this subsection.


NPCS: integer, default: 8
: Number of points around crosssection where fatigue is calculated,1 ≤ NPCS ≤ 16
, see Illustration of points around the crosssection. For definition of local axis, see RIFLEX User Guide. 
IOPPR: integer, default: 0
: Print option for fatigue results
IOPPR = 0
: Print fatigue results only for most critical point in the crosssection. IfTSIOPPR >0
, time series is printed for one of two concurrent element ends. 
IOPPR > 0
: Print fatigue results for allNPCS
points. IfTSIOPPR >0
, time series is printed for all elements and both element ends.


TSIOPPR: integer, default: 0
: Time series print switch. Dummy ifIOPTSH = 1
.
TSIOPPR = 0
: No print of tension, y and z curvature time series to file 
TSIOPPR = 1
: Print of tension, y and z curvature time series to file 
TSIOPPR = 1
: Print of tension, y and z curvature time series to file, skip fatigue calculation.


CHTSPRN: character(256)
: Time series file name
For NSECT
> 0 and IOPFAT = 1
, Rainflow counting method, fatigue
damage will be calculated at the specified cross sections only.
Otherwise, fatigue damage will be calculated for alle elements. The cross section specification may e used to specify stress concentration factors (SCFs) that differ from the default values (DSCFA, DSCFY and DSCFZ) given below
Tension and curvature time series are printed for the NSECT
specified
elements or for all elements if NSECT = 0
. The format of the time
series file is given in Appendix B: Format of Time Series File.
Note that the two input lines of subsection SN curve data must be given
below even if NSECT > 0
.
7.3. Stress time series data, one input line. Given if IOPFAT = 0
or IOPFAT = 1
.
TSLEN DT IRSNO

TSLEN: real, default: 0
: Length of stress time series to be generated for fatigue calculation \(\mathrm {[T]}\).
Minimum length is 60.


DT: real, default: 0
: Time step to be used in the stress time series \(\mathrm {[T]}\). 
IRSNO: integer, default: 31415
: Seed for generating random phase angles.
The time unit is usually seconds.
A warning will be written if the user specifies a time step DT which is larger than\(\mathrm {\frac{2\pi }{\omega _{max}\times 20}}\), wheer \(\mathrm {\omega _{max}}\) is the largest of the discrete response frequencies.
If default TSLEN
and DT
are used, TSLEN
and DT
are calculated by
VIVANA
. DT
will be set to 1/30 of the shortest crossflow response
period with significant response. TSLEN
will be set to 200 times the
period of the highest ranked crossflow response frequency, multiplied
by the shortest crossflow response period with significant response
divided by the longest crossflow response period with significant
response.
7.4. Crosssectional data, one input line
DSCFA DSCFY DSCFZ ASI WSTI THI RFACT

DSCFA: real, default: 1
: Default stress concentration factor for axial force contribution. 
DSCFY: real, default: 1
: Default stress concentration factor for bending about Yaxis. 
DSCFZ: real, default: 1
: Default stress concentration factor for bending about Zaxis. 
ASI: real, default: 0
: Optional crosssection area.
= 0.0
: Use the values specified in or derived from the cross section properties given inINPMOD


WSTI: real, default: 0
: Optional section modulus.
= 0.0
: Use the values specified in or derived from the cross section properties given inINPMOD


THI: real, default: 0
: Optional wall thickness.
= 0.0
: Use the values specified in or derived from the cross section properties given inINPMOD


RFACT: real, default: 0.001
: Factor between the stress unit \(\mathrm {[S]}\) used to define the SN curve and the force and length units \(\mathrm {[F]}\) and \(\mathrm {[L]}\) chosen inINPMOD
.
\(\mathrm {S\times RFACT=\frac{F}{L^2}}\)

If \(\mathrm {kN}\) and \(\mathrm {m}\) were chosen
as force and length units while the SN curve is given in
\(\mathrm {MPa}\), RFACT
should be set to 0.001
.
If the SI units \(\mathrm {N}\) and \(\mathrm {m}\)
were chosen for force and length and the SN curve is in
\(\mathrm {MPa}\), RFACT
should be set to 1.0E6
.
Note that the stress concentration factors are included in the reported stress distributions.
7.5. SN curve data, two input lines
NSNCRV RELDUR

NSNCRV: integer > 0
: Number of selected SN curves 
RELDUR
: real, default: 0.0`: Relative duration / probability of the current condition.0.0 ≤ RELDUR ≤ 1.0
. IfRELDUR > 0.0
, the fatigue damage contributions, the calculated fatigue damage scaled byRELDUR
, will be written to the_vivana.mpf
file.
CHIDSNi...CHIDSNnsncrv

CHIDSNi: character(6)
: Identification of selected SN curve i, i = 1, …,NSNCRV
7.6. Crosssection specification, NSECT input lines.
CHILIN ISEG IEL IEND SCFA SCFY SCFZ

CHILIN: character(8)
: Line identifier 
ISEG: integer/character
: Segment number in line.
ISEG > 0
: Local segment in the specified line 
ISEG = 'ALL'
: All segments in the specified line. IEL must then also be `ALL'.


IEL: integer/character
: Local element number in specified segment.
IEL > 0
: Local element in the specified line and segment 
IEL = 'ALL'
: All elements in the specified segment(s). IEL must beAL
if ISEG isALL
.


IEND: integer
:
IEND = 1
: Crosssection at end with smallest node number checked. 
IEND = 2
: Crosssection at end with largest node number checked.


SCFA: real, default: DSCFA
: Stress concentration factor for axial force contribution. 
SCFY: real, default: DSCFY
: Stress concentration factor for bending about Yaxis. 
SCFZ: real, default: DSCFZ
: Stress concentration factor for bending about Zaxis.
This input is used to specify lines / segments / elements and stress concentration factors (SCFs) that differ from the default values (DSCFA, DSCFY and DSCFZ) given above.
If SCFs are given several times for the one element, the last one given is used.
the stress concentration factors are included in the reported stress distributions. 
7.8. SN curve, 2xNOFC input lines
CHIDSN NOSL LIMIND FATLIM TREF KEXP

CHIDSN: character(6)
: Identification of SN curve 
NOSL: integer ≤ 5
: Number of straight lines defining the SN curve. 
LIMIND: integer, default: 0
: Fatigue limit indicator
LIMIND < 0
: Fatigue limit in terms of stress cycles is specified. 
LIMIND = 0
: No fatigue limit. 
LIMIND > 0
: Fatigue limit in terms of stress range is specified.


FATLIM: real, default: 0
: Fatigue limit, interpretation dependent onLIMIND
.
LIMIND < 0
: Logarithm of number of stress cycles for which the SN curve becomes horizontal. 
LIMIND = 0
:FATLIM
is dummy. 
LIMIND > 0
: Stress range level for which the SN curve becomes horizonal.


TREF: real, default: 0
: Reference thickness for thickness correction
TREF = 0
: No thickness correction


KEXP: real, default: 0
: Exponent for thickness correction.
KEXP = 0
: No thickness correction

TREF and KEXP must either both be zero, no thickness
correction, or both positive, thickness correction included.

7.9. Fatigue capacity curve constants
RM1 RC1 RMi RNCi

RM1: real
: Slope of the SN curve.
First curve segment for
NOSL > 1

Total curve for
NOSL = 1
.


RC1: real
: Constant defining the SN curve. First segment or total curve. 
RMi: real
: Slope of curve segment i, i = 2, …,NOSL

RNCi: real
: Transition point between curve segment (i1), and i, i =2, …,NOSL
.
The transition point is given as the logaritmic value, see Figure 3.

The SN curves defined by the input parameters are always assumed to relate the stress range, \(\mathrm {\Delta S}\), to the number of cycles before failure, N.
A straightlined SN curve in loglog scale is in general defined as
\(\mathrm {N=C\times (\Delta S(\frac{t}{t_{ref}})^k)^m_{}}\)
or
\(\mathrm {\log N=\log C+m\log (\Delta S(\frac{t}{t_{ref}})^k)_{}}\)
where:

\(\mathrm {N}\): Number of cycles to failure

\(\mathrm {\Delta S}\): Stress range

\(\mathrm {t}\): Cross section thickness

\(\mathrm {t_{ref}}\): Reference thickness

\(\mathrm {k_{}}\): Exponent for thickness correction
The two input parameters used to define SN curves are directly found in the equation above, namely.

\(\mathrm {RC=\log C\quad }\) (always positive)

\(\mathrm {RM=m\quad \quad }\) (always negative)
If the user has a SN curve without having these parameters explicitly defined, they can be calculated as follows; see also Calculation of SN curve parameters.
Using the two points A and B on Calculation of SN curve parameters to define the straight line, we have
\(\mathrm {\log N=\frac{\log N_2\log N_1}{\log\Delta S_2\log\Delta S_1}\times \log\Delta S\frac{\log N_2\log N_1}{\log\Delta S_2\log\Delta S_1}\times \log\Delta S_1+\log N_1}\)
Hence
\(\mathrm {RM=\frac{\log N_2\log N_1}{\log\Delta S_2\log\Delta S_1}\quad }\) (always negative)
\(\mathrm {RC=RM\times \log\Delta S_1+\log N_1\quad }\) (always positive)
The relation between these parameters specified for different unit systems is easily found from the equations above.
8. VIVANA print flags
8.2. Print flags, one input line
IPRELF IPRSTF IPRDRG IPRRSP IPRCNG

IPRELF: integer, default: 1
: Print flag for printing local element forces to the_vivana.res
file.
IPRELF = 0
: No print 
IPRELF = 1
: Print


IPRSTR: integer, default: 1
: Print flag for printing local element stresses to the_vivana.res
file.
Interpreted as
IPRELF
.


IPRDRG: integer, default: 1
: Print flag for printing drag amplification factor to the_vivana.res
file.
Interpreted as
IPRELF
.


IPRRSP: integer, default: 1
: Print flag for printing VIV response (displacements) to the_vivana.res
file.
Intepreted as
IPRELF
.


IPRCNG: integer, default: 1
: Print flag for printing convergence during the response analysis to the_vivana.res
file.
Interpreted as
IRRELF
.

9. VIVANA MPF print flags
9.2. Print flags, one input line
IPRCRV NSNAPP DTSNAP

IPRCRV: integer, default: 0
: Print flag for printing curvature to the_vivana.mpf
file.
IPRCRV = 0
: No print 
IPRCRV = 1
: Print


NSNAPP: integer, default: 1
: Number of snapshpts to be printed on the_vivana.mpf
file.
NSNAPP = 0
: No print of snapshots 
NSNAPP < 0
: Default values will be used, see below.


DTSNAP: real, default: 0
: Time step for print of snapshots
DTSTEP ≤ 0
: Default values will be used, see below. 
DTSNAP > 0
: Snapshots will be printed atDTSNAP
intervals.

The default printing of snapshots depends on how response frequencies
are combined, see input prameter IOPTSH
in VIV response analysis.
For concurrently acting frequencies, IOPTSH = 0
, the default is 20
snapshots and the default time step for the snapshots is 1/20 th of the
period of the highest ranking response frequency with significant
response.
For consecutively acting frequencies, IOPTSH = 1
, the default number
of snapshots is \(\mathrm {MAX(20,8*NFREQ\_SIGNRES)}\), where
NFREQ_SIGNRES
is the number of response frequencies with significant
response. Snapshots are allocated to the different response frequencies
depending on their predicted relative durations, but with at least one
snapshot per response frequency. The default time step will distribute
the snapshots in each response period.
10. Random number generator
In version 4.18 and later, the algorithm for generating pseudorandom numbers may be selected by the user. Both methods are expected to be acceptable for the moderate amount of random numbers needed in VIVAAN. The default, legacy method should be used if backwards compatibility with previous versions is required. Note that the default value may change in a future release. The choice of random number generator will apply to:  generation of stress time series for fatigue calculations using the Rainflow method.  output of curvature time series consistent with the stress time series generated for the Rainflow method.
10.2. Random number generator input, one line
CHRAN

CHRAN: character (7), default: TWISTER
: Choice of random number generator
= 'LEGACY'
: Legacy random number generator used. Results will be consistent with previous VIVANA versions. 
= 'TWISTER'
: Mersenne Twister’ random number generator used. Results will NOT be consistent with previous VIVANA versions.

12. Structural damping file description
The file CHFILE
specified in Structural damping specification contains a
description of the structural material and slip damping of the
crosssection.
12.1. Material danping
12.1.1. Material damping control data, one input line
RJ_MTRL RN_MTRL FREQREF_MTRL TEMPREF_MTRL NTENS_MTRL NCURV_MTRL NFREQ_MTRL NTEMP_MTRL

RJ_MTRL: real, default: 0
: Material damping constant. Dummy in the present version. \([\mathrm {\frac{FL}{(F/L^2)^nL^3}}]\) where n is the input parameter RN_MTRL. The material damping energy per volume per cycle is \(\mathrm {RJ\_MTRL(\Delta \sigma )^{RN\_MTRL}}\) . 
RN_MTRL: real, default: 0
: Material damping exponent. Used in interpolation of the material damping data given in the table below. 
FREQREF_MTRL: real, default: 0
: Reference frequency for the material damping data given in the table below. Dummy ifNFREQ_MTRL = 0
. IfNFREQ_MTRL > 0
the frequency scaling factor corresponding toFREQREF_MTRL
should normally be 1.0. \([\mathrm {Hz}]\) 
TEMPREF_MTRL: real, default: 0
: Reference temperature for the material damping data given in the table below. Dummy ifNTEMP_MTRL = 0
. IfNTEMP_MTRL > 0
the temperature scaling factor corresponding toTEMPREF_MTRL
should normally be 1.0. 
NTENS_MTRL: integer > 0, default: 0
: Number of static effective tension levels in the material damping table 
NCURV_MTRL: integer > 0, default: 0
: Number of curvature amplitudes in the material damping table 
NFREQ_MTRL: integer >= 0, default: 0
: Number of frequencies in frequency scaling curve 
NTEMP_MTRL: integer >= 0, default: 0
: Number of temperatures in the temperature scaling curve
12.1.2. Static effective tension levels, one input line
TEMS_MTRL(1) .... TENS_MTRL(NTENS_MTRL)

TENS_MTRL(itens): real, default: 0
: Static effective tension level nr. items for material damping \([\mathrm {F}]\)
The static effective tension levels must be given in increasing order.
12.1.3. Table of material damping data, NCURV_MTRL input lines
CURV_MTRL(icurv) DAMP_MTRL(icurv,1) .... DAMP_MTRL(icurv,NTENS_MTRL)

CURV_MTRL(icurv): real, default: 0
: Curvature amplitude level nr. icurv for material damping \([\mathrm {1/L}]\) 
DAMP_MTRL(icurv,itens): real, default: 0
: Energy loss per cycle per length for static effective tension TENS_MTRL(itens) and curvature amplitude CURV_MTRL(icurv) \([\mathrm {(FL)/L}]\)
The curvature amplitude levels must be given in increasing order.
Linear interpolation is used on the values \(\mathrm {DAMP\_MTRL(icurv,itens)/(CURV\_MTRL(icurv)^{RN\_MTRL})}\) The first / last values are used for static effective tension or curvature amplitude below / above the range given.
12.1.4. Frequency scaling of material damping, two input lines
FREQ_SCAL(1) ... FREQ_SCAL(NFREQ_MTRL)

FREQ_SCAL(ifreq): real, default: 0
: Frequency value nr. ifreq for the frequency scaling of the material damping \([\mathrm {Hz}]\)
SCALE(1) ... SCAL(NFREQ_MTRL)

SCALE(ifreq): real, default: 0
: Scaling factor corresponding to frequency FREQ_SCAL(ifreq)
Frequency values must be given in increasing order.
Linear interpolation is used for intermediate frequency values. The
first / last values are used for frequency values below / above the
range given. If the scaling factor corresponding to the reference
frequency FREQREF_MTRL
is not 1.0, the resulting scaling will be the
ratio between the scaling factor for the response frequency and the
scaling factor for the reference frequency. Specifying the same scaling
factor for all frequencies will thus result in no frequency scaling.
12.1.5. Temperature scaling of material damping, two input lines
TEMP_SCAL(1) ... TEMP_SCAL(NTEMP_MTRL)

TEMP_SCAL(itemp): real, default: 0
: Temperature value nr. itemp for the temperature scaling of the material damping
SCALE(1) ... SCAL(NTEMP_MTRL)

SCALE(itemp): real, default: 0
: Scaling factor corresponding to the temperatire TEMP_SCAL(itemp)
Temperature values must be given in increasing order.
Linear interpolation is used for intermediate temperature values. The
first / last values are used for temperature values below / above the
range given. If the scaling factor corresponding to the reference
temperature TEMPREF_MTRL
is not 1.0, the resulting scaling will be the
ratio between the scaling factor for the current temperature and the
scaling factor for the reference temperature. Specifying the same
scaling factor for all temperatures will thus result in no temperature
scaling.
12.2. Slip damping
The slip damping is found as the area enclosed by nonlinear momentcurvature cycles. The RambergOsgood model is used to define the momentcurvature curves.
12.2.1. Slip damping control data, one input line
TEMPREF_SLIP NTENS_SLIP SCALE_SLIP

TEMPREF_SLIP: real, default: 0
: reference temperature for slip damping, Not used in present version. 
NTENS_SLIP: integer >= 0, default: 0
: Number of static effective tension levels slip damping is given for 
SCALE_SLIP: real, default: 1.0
: Scaling factor for the resulting slip damping.
12.2.2. RambergOsgood parameters, NTENS_SLIP input lines
TENS_SLIP CY MY ETA GAMMA C_SLIP

TENS_SLIP: real, default: 0.0
: Static effective tension \([\mathrm {F}]\) 
CY: real, default: 0
: RambergOsgood Cy parameter \([\mathrm {1/L}]\) 
MY: real, default: 0
: RambergOsgood My parameter \([\mathrm {FL}]\) 
ETA: real, default: 0
: RambergOsgood eta parameter \([\mathrm {1}]\) 
GAMMA: real, default: 0
: RambergOsgood gamma parameter \([\mathrm {1}]\) 
C_SLIP: real, default: 0
: Curvature for which slip first appears. No energy loss for curvatures below this level. \([\mathrm {1/L}]\)
For a given static tension T
and curvature amplitude C
, linear
interpolation is used between the damping found for curvature C
at
effective tension values bracketing T
. If T
is below / above the
range given, the damping for C
at the first / last tension value is
used.