Work with FinLiv v. N

Consortium FADOFF - Work with FinLiv v.N December 20131

Work with FinLiv v. N

Jan Papuga

The development of FinLiv application was realized thanks to the support of TACR agency


Current state 2 application versions currently

FinLiv v. N released February 2014 – developed mostly by Jan Papuga

S-N curves – recording, viewing, regression analysis (Basquin model, model according to Kohout-Věchet)

integrates the basis of static experiments data manipulation – storing, viewing through interface, saving

FinLivX a modification of FinLivK (released 30. 6. 2013)

prepared by Jakub Hasman as a post-processing of his bachelor thesis

implements the functionality related to inclusion of static experiments – their record, viewing, analyses based on CSN standard

Expected inclusion of e-N curves prepared by a PhD student at the VŠB-TUO university


What is FinLiv

An application intended to save experimental data Fatigue experiments

Both S-N and e-N curves

No crack growth currently

Static experiments – under development

Built in Visual Basic for Applications with a common MS Excel interface

An intended pre-processor for managing data later saved to on-line FinLiv database


FinLiv file content

form sheet – the

basic graphical

interface enabling

the input of data

data sheets – any

other sheets with

description of test

group sets

User can pick the selected

test group


Test group vs. curve

Set of all experiments realized with one batch of material

Treatment relates to the preparation of semi-product related to all individual curves

Test group

Related to a particular test group

Treatment relates to any thermo-mechanical operation encountered during preparing the specimens

Curve

Exception:

ID_REF – relates to every individual curve, because the sources providing the data can differ during more extensive experimental campaigns


Thermomechanical treatment

Two times within one form operations during preparing semi-product operations while preparing the final specimen

Defined type of the operation environment during the treating possible commentaries for whatever not included variables

Hint shown while mouse hovers over the cell describing the meaning of the parameter

Parameters to be filled in potentially are marked by salmon colour


Specimen

Approx. 20 different pre-configured specimen designs SP_V1 – SP_V12

based on the configuration, thenumber of geometric parametersand their meaning differ

Related to every curve, can differwithin one test group

SP_ORIENT – describes the orientationof the specimen axis to the majordimensions of the semi-product


Description of input loading

Up to 4 load channels can act simultaneously

The same colour convention valid for inputs

Load modes tension plane bending rotating bending pressurizing torsion thermal load

Load types harmonic loading constant loading triangular trapezoidal sampled load input general load load blocks Gauss load

Hints included to describe the meaning

well done

to be finalized

This view relates to

fatigue experiments


Description of input loading

Up to 2 load channels can act simultaneously

The same colour convention valid for inputs

Load modes tension compression torsion plane bending

Load types 0-max constant loading

Hints included to describe the meaning

This view relates to

static experiments


Master and Slave load channels

Master parameter is that one taken as a reference for this kind of loading (and used for graphic output therefore)

Other parameters of the loading (on the same channel or on other channels) are either its multiples or constants

N

a aa

ft=10 fb


Master and Slave load channels II

In version M, an additive constant can be added to the end of formula a curve with lower stress being constant = C = 100

note: only the additive constant can be added (+) or subtracted (-). No other

operations are allowed allowed operators: x ~ multiple, + ~ addition, - ~ subtracting multiple should precede any further operation – it is the marker that the

input is a formula and not a constant value only one multiple and one addition or subtracting allowed in one formula

N

a aa

ft=10 fb

ah/2-C/2mh/2+C/2

LD_V1: x0.5-50

LD_V2: x0.5+50


Individual test data input END parameter for S-N curve

describes the condition of the end of the experiment two lifetimes can be recorded

for the final break of the specimen for the END

Unfinished experiments marked by I symbol in the appropriate column


Action buttons - Save

Process the data either in data sheets or in the form

Save transfers the current setup (both

test group information and curve information currently depicted) in the form sheet to the appropriate data sheet

!!!doesn’t saves the file!!!


Action buttons - Show

Starts a form to select the curves of a given test group to be send to a separate file for a regression analysis

Access to another test group than currently selected enabled by clicking the three dots “…”

The backslash “\” shows that the content is a test group and not a curve

The final file created is described in this presentation later on


Action buttons – Back Up

A similar selector tool as by Save, but only test groups listed

Any selected test group is detected and saved as a separate csv file

The range of saved data is based on the number of cells continuously

non-empty on the line CURVE_ID the number of continuously non-

empty cells in the column B of the appropriate data sheet


Action buttons – Import

Pushing this buttonresults in this message box:

A file has to be selected then based on the previous choice: Any *.csv file previously saved or

prepared to work as the data sheet

A sheet named “csv” in the selected Excel file


Action buttons – Remove

Allows the user to remove a selected test group while keeping the data sheet links OK

Removing the data sheet manually by the Excel functions causes, that the MARK listbox (C3 cell in form) will not reflect it and the application can break down when this item is asked Can be solved today by closing

the FinLiv application and opening it again


Action buttons – APDL Specimen

Information about specimens used for selected curves is processed a text file is generated it serves as an input for ANSYS

APDL macro based on this input

that macrocreates the FE-modelof the specimen, itsboundary conditionsand runs the analysis


Action buttons – Input for PragTic

An extensive form for automated defining of the analysis in PragTic fatigue solver is opened

Functionality of this command is described later on


Working with FinLiv

The data can be input through form but also directly into the data sheets

Be careful – there is no identity between the number of your choice in a selected listbox and the number in appropriate cell of the data sheet sometimes it is identical, sometimes not it takes some time before you grasp the right ID

codes of the mostly used items test and check first by using the form


Data sheets

Each curve of a particular test group is recorded to a separate column

A header related to all curves ends by TREAT4 line

The exceptional ID_REF record is placed to the line 3

It is expected that most of the records could be placed to another line, there is no defined order Exception: Records of

Px_MARK, Dx_y, Ix_y and D_COMMENTx should be placed consecutively, without any interruption


Naming conventions

The name of the data set should correspond to abbreviation of the surnames of the team members: Hudson, C. M. : A study of fatigue and fracture in 7075-T6 aluminum

alloy in vacuum and air environments . [Technical report NASA-TN-D-7262, L-8829 ]. Washington, 1973

Results in the test set „Hud“ 1 author only no previous similar record three letters optimum

Grover, H. J.; Bishop, S. M.; Jackson, L. R.: Fatigue strengths of aircraft materials axial-load fatigue tests on notched sheet specimens of 24S-T3 and 75S-T6 aluminum alloys and of SAE 4130 steel with a stress-concentration factors of 2.0 and 4.0. [Technical report NACA-TN-2389]. Washington, NACA 1951.

Results in the test set „GBA“, „GBB“, „GBC“ more authors, but also more materials used three letters optimum first letters of the first two authors used, an alphabet marker added on the 3rd position


Numbering rules – S-N curves The experiments on smooth unnotched specimens should be numbered

within IDs 01-39, if possible Special curves

fully reversed axial loading ID30 repeated axial loading ID31 fully reversed torsion ID32 (repeated torsion ID33) – preferable, but not strictly required

Experiments on other notched geometries set to next decades, e.g. specimen with a hole through (40~reversed push-pull, 41~repeated

tension, 42~reversed torsion, 43-49~ other combinations) specimen with a circumferential U-notch (50~reversed push-pull,

51~repeated tension, 52~reversed torsion, 53-59 other combinations) …


Numbering rules – S-N curves II Also the mode of axial loading matters – say e.g. the most of

experiments is done while combining tension and torsion, but also combinations of torsion with plane bending occur: 01-29, 33-39 – any more complex experiments covering

tension and or torsion 30-32 – basic uniaxial load modes for tension and torsion 40-41 – plane bending experiments 42 – torsion (can be related to a different specimen type

then 32) 43-49 – any combination of plane bending and/or torsion

on this specimen type


Numbering rules – Static data

In order to quickly orient in the data sheet a prefix S is added to any static tests, while being reported in „form“ sheet

The datasheet doesn’t include this prefix and you can recognize the static curve by CUR_TYPE=3

Special IDs: averaged tension ~ 30 averaged compression ~ 31 averaged torsion ~ 32 averaged bending ~ 33


Static data – averaged vs. experimental

Static tests are usually done on more samples (3-5), but only the average values are mostly used

Experimental curves those with complete experimental data (force vs. displacement) will be evaluated after Show (not ready in this version) averaged curve will be completed in SumUP a link between the experimental curve and the averaged curve will be

saved to csv Averaged curves

do not contain any experimental data, just the statistics of the regression analysis

have unique IDs (see the previous slide) a link between them and all experimental curves used for their

deriving will be saved to csv


Input of static data

Mostly the same dialogues, but load modes, load types and end of the tests differ

Other description of the datapoint – it relates to one time instant of the static test

Substantially more time instants can be recorded than 60 Parameter ROWS allows to change the dimension of the

record (limited only by the number of rows in Excel) ROWS is hard set in fatigue curves processing to 60


New items

New test group: item “new…” in the MARK listbox of the form

asks for a name of the test group, from which the content is inherited (curves can be then removed manually in the data sheet)

import of an existing csv file copying an existing sheet to a new location and renaming

it copying the csv file from another file

New curve: item “new…” in the CURVE_ID listbox of the form copying some existing record to appropriate column of the

data sheet – the CURVE_ID listbox is updated, when another test group is selected and then the right one chosen again

Push the button

in the form


Show procedure


Show procedure

For selected curves of a selected test group creates a file in the same directory the file is named “TestGroupName.xls”, where the

TestGroupName is replaced by the test group abbreviation data of each curve is placed into a separate sheet named

“TestGroupNameCurve” Two additional sheets are created:

SumUP – covers results of all separate regression analyses done for individual curves

csv – contains the original data sheet itself

The operation can be repeated additively, i.e. a part of curves can be processed first, then another one, etc.


Curve sheetParameters of regressions

Experimental data points

Buttons for starting the regression analyses

Graph for visual interpretation

Regression curve according to Basquin formula

Regression curve according to Kohout-Věchet formula


Initial state after Show operation

Only experimental data and related points in the graph are shown


Running S-N curve button…

S-N curve shown in the graph derived material parameters reported in the left table regression data points in the central table below the table with

experimental data points


…and check the SumUP sheet Contains

functional buttons Write to csv: writes the regression parameters to

the csv sheet (can be then imported to FinLiv via Import button)

Dummy Model: prepares input for Dummy model application (APDL script for pre-processing multiaxial fatigue limit experiments in ANSYS)

description of regression parameters proposal on credibility coefficients master curve description

Here shown only S-N curve data Below that also the same output for

Kohout-Věchet model (K-V) is provided


Running K&V curve button…


K&V regression

More complex necessitates the use of Solver add-on to Excel (a

non-linear solver utility) check its existence in your MS Excel before

running it Run the initial estimate procedure?

a query box asking if to run the initial estimate or continue from the last saved estimated parameters

useful when the regression analysis collapses


K&V regression collapse

1. Run the analysis from the initial estimate on the set of data points omitting the outward-lying points

2. If succeeded – include all data points, deny to use the initial estimate and run the analysis from the last saved regression parameters

3. If failed – go to item 1 and repeat it again with another set of data points

Work with FinLiv v. N

Documents

test groupsp

papugasn curves

description of input

load channels

finliv databasedecember

selected test groupdecember

particular test grouptreatment

colour convention valid