User: unknown || Login || HOT NEWS: Gallery, 2nd PUM - free lectures, PragTic v.0.2betaH, PragTicA
PragTic Software
Registered users can be informed about release of any new version of PragTic thanks to the Notification system. The proposition to include them to the e-mail list is shown at this place whenever they are logged in.
Commentary to Actual Versions
Notice
Oct 23, 2008 - PragTic v. 0.2 betaH
The new version is released in order to present the most up-to-date state of PragTic without bugs to
participants of the users' meeting.
The most important changes (except for removed bugs) are:
- Show in coordinate system function should really do the right transformation to the coordinate system chosen by the user.
- There is a new data_vector item Setups of Analysis. It allows saving of a chosen combinations of load regimes, computational methods and selected place on the FE-model. This solution should quicken the start of the calculation for various variants of inputs.
- Help was re-written to include the current changes.
The transformation of location, vectors and tensors between various coordinate systems should already work, but some next or previous steps can make still problems. I know already now that the transformation need not be correct for shell elements, if the result output uses some of their native local coordinates system. Please, in case you find any marks of non- or disfunction, write about that to Bugs section.
Sep 05, 2008 - PragTic v. 0.2 betaG
This is a preliminary release of the promised betaG version in order to suppress an annoying
bug affecting results of all uniaxial strain-based methods.
This means that reporting of various parameters in different coordinate systems is not completed
and will be fully implemented in the next beta.
The major changes now are:
- Content of the item Coordinate Systems was completed in order to allow the work with them (mutual shifts of CSs, definition of their origin, CS used for definition etc.).
- The new coordinate systems can be defined even within PragTic. They can be defined also in coordinates of other CSs than the global rectangular CS.
- The node, element or IP sets from now on need not be sorted from minimum to maximum value. Manual definition of a new set triggers a query if the sorting procedure should be run. The automated import function through Import imports what you prepare - there is no sorting.
- View On Set function was elaborated in the same way, which means that you can let PragTic printout e.g. the stress tensors along some defined sequence of nodes.
- Keyboard shortcuts were introduced to the most of active buttons in the interface. Use Alt + ?, where ? is replaced by the character that is underlined in the name of the button.
- You can change order of the load regimes (right-hand click on the load regime list on the left side of the LR window)
- The bugs discovered during testing of betaE or betaF were removed, the older ones unfortunately remain.
The work on the incorporation of coordinate systems into post-processing by PragTic is not finished yet - the transformations of read result files to other selected CS above all.
Jul 03, 2008 - PragTic v. 0.2 betaF Some more important changes from the previous version concern:
- Equivalent strain - the previous version was replaced by the commonly accepted formula used e.g. in
BONACUSE, P. J.; KALLURI, S.: Elevated Temperature Axial and Torsional Fatigue Behavior of Haynes 188. [Technical Memorandum 105396], NASA 1992.
The change concerns the division of the final product by 1+nu, where nu is Poisson's ratio. - There was a bug in the final part of the rain-flow procedure that could under some circumstances wrongly analyze the full cycles from the reminder of the load history left after the first run through it.
- Very important change concerns work with elastic-plastic conversion methods used over elastic FE-data (Neuber, Glinka,...). Up to now, the solution has not included the Bauschinger's effect and each data point was plasticized as being loaded from the zero (isotropic hardening in fact). This procedure is nevertheless more appropriate for cases, where the load history effect are neglected (e.g. analysis from rain-flow matrices or spectra). Here the solution was preserved even in betaF, any other load history is nevertheless converted with full use of kinematic hardening only. (well, it is not in Help still, so if you do not understand, ask.)
- The set of nodes or elements created automatically during the computation if the computation is ran over a set, for which some stress/strain data are missing is now correct.
- Remove Rows command started from the right-hand click pop-up menu in the Edit dialogue works corrently from now on.
- Load Regime definition - command Compose Local Load History could ask you for material parameters or trigger a warning, although they have been already defined.
- Tools->Material->S-N curve command starts a dialogue, which allows you to retrieve the best fit S-N curve from the input experimental data.
There still are left the uncertainties in results of Spagnoli method and the MCEM solution. Any help in their solution is welcomed. The expected implementations of Jing Li's method or of that used in NASALIFE have to wait till the next beta - the localization of the problems in the rain-flow took me too much time...
Unfortunately, some of the changes done in betaE or betaF are not still reflected in the Help.
Well, ask for that by an e-mail, if you need that...
Mar 03 & Feb 20, 2008 - PragTic v. 0.2 betaE The new beta version is once again closer to the full alpha. I'm releasing the betaE once more on Mar 03, since there were several minor bugs that could be removed quickly and do not deserve its own release. Overall in betaE, there were done quite extensive changes including:
- Revision of the critical error mentioned one week ago.
- Accelerated read of data to the Edit window.
- View on Set command created in the right-hand click popup menu in the database tree. It allows you viewing of a selected part of too large block of data only and thus it can be seen as a possible solution of the problem of viewing too large data_vectors.
- Apply to Set button can help you to revise properties of nodes or elements. You can assign selected element type, real constants, materials, properties of locality or coordinate systems to a set of nodes or elements by it.
- Help was substantially extended and finally covers also the low-cycle methods. It is still in development nevertheless - the most of the changes introduced today are not reflected in it unfortunately.
- Minimum circumscribed ellipse method (MCEM) was implemented for derivation of the shear stress/strain amplitudes and mean values. It can be today selected as an another choice to the minimum circumscribed circle method (MCCM). It is still not perfect, so you could encounter cases, where the MCEM algorithm does not converge. Any your help is welcomed here.
- There is a new section Tools->Loads->Decompose, where you can apply the MCEM or MCCM to a selected record of the load history on a plane
- The command Tools->Arrays->Superpose allows superposition of similar data_vectors.
- The structure of the data_vector describing methods was significantly changed so that any my further change in choices of methods would allow use of older tasks. Note, please, that the new version uses slightly different composition of the method data_vector - it thus notifies you about necessary update of the task. It asks for the name of the new task, to which it has to copy the updated task. Check then please any method item, if all the necessary options are filled in, otherwise the computation could crash.
- The options of methods were extended to cover also the type of damage accummulation. It is prepared for later implementations.
- Default pre-set values were assigned to all methods so that the users needn't to fill in all parameters.
- There is the new option of an export of fatigue results back to ANSYS.
Feb 12, 2008 - A serious bug have been detected in the iterative scheme for retrieval of number of cycles from Basquin and Manson-Coffin formulas. The results of all methods based on the e-N curve (SWT and its clones, Morrow, Feltner, Socie, Findley v. Socie, Wang & Brown) are likely to be shifted to the non-conservative side. Please, wait for the next patch in the PragTic v.0.2 betaE, which should be released during one week from now.
Dec 21, 2007 - PragTic v. 0.2 betaD There are several bugs in interface removed. One important revision concerns the computation process - the Glinka's formula for conversion to elastic-plastic values was bad. There is also one new nice feature - you can scale the content of some items in the database tree by a right-hand click on the item and selection of Scale By... command.
Dec 4, 2007 - PragTic v. 0.2 betaC Several bugs were found in the previous beta and a chaining of separate result files one after another to form the transient analysis input was implemented. Options of the automated universal importer were also extended by possibility to define precise positions of each column (hard separators). What you could still miss:
- Finalizing of the LESA method
- Finalizing Help and its link to the executable (see the remark from Oct 1)
- The chaining of the result files is enabled in the Load Regime menu. If you would like to perform a fatigue analysis on it, first let the command "Compose Local Load Histories" be executed and create a new load regime with the prepared local load history - this one can be used as an input for the fatigue analysis. The use of the result files chain directly in the analysis is being implemented now (a temporary file will be created).
Oct 1, 2007 - PragTic v. 0.2 betaB I have found a relatively important error, which was triggered whenever a new task was defined. The second substantial patch concerns the data import, where the definition of separators could lead to wrong diagnosis. The help was improved, although there are still some themes missing - the content of previous manuals is nevertheless fully integrated. Unfortunately, the relation between help and PragTic makes me unhappy - the beta B version released now lost even the start of the help pages by a click on the Help buttons.
August 31, 2007 - PragTic v. 0.2 beta Well, I am coming with the new beta version one year after the previous one. Yes, it is once again signed as the beta version - I think it is already too long time from the previous version and the new features implemented in the v.0.2 are worth of release, although it is not perfect.
I have lost too much of my time trying to cope with the way of analysis used within FemFat fatigue solver. It seems that it will take a bit more of my time, so please do not use the LESA method until the final version is released.
The other unfinished point is again the manual to the program. The pages concerning the methods itself are missing there above all. The most of the new features are described within it, so do not worry that you will not know how to use the program. The methods themselves are the same as in the previous version with exception of methods by Ninic, Sines and Crossland, the LESA method and several pieces of my methods being in development.
I will try to build the manual step by step - please visit this site from time to time - or I can inform you about the changes in the subject, if you write me...
Please, be so kind to inform me about any problems you encounter. The changes in PragTic are really substantial and it is possible that I omit to finish some commands. This problem should not concern the computation itself, but the preparation of the task, viewing or editing the data, dependencies of data_vectors, etc. I would also welcome if you be so helpful to write me on problems concerning the readability and clarity of the manual - maybe I'll be able to find some time and prepare the FAQs page a bit more interactive.
August 5, 2006 - PragTic v. 0.1.2 beta The reason, why the new version is released, is related to the new process of data import. I suppose that the change is so important, that it is deserved. Nevertheless, the works on it prolonged too much. I was unable to finish the Help in most of the points mentioned in the May 14 sections here below. Well, it seems that I am not able to continue with the smooth development in my current situation. I'll try to fix the problem later. I do not want to promise anything as regards programming - I have to continue also with my research work (not mentioning Evektor...).
At least one minor change can be expect during August - I am preparing functions enabling to create and save local load path in a selected point or even in a selected direction.
May 14, 2006 - PragTic v. 0.1.1 beta I spent every free moment of last three months with completing the promised Help manual to PragTic so that it could be available on the occassion of Fatigue 2006 conference held in Atlanta this May. Well, I have not finished it as I would like. Thus there are features, which I do not recommend to be used:
- uniaxial methods
- low-cycle multiaxial methods
- transient analysis data import
The help manual is not finished for these points and the import of data for the transient analysis needs some more check (although maybe you can be succesful in their upload).
The other problem is, that I wrote linearly and have not enough time to reread the text. It is very well possible, that there are many phrases in "czenglish" or erroneous at all. Please, be tolerant. Or - if you are a native speaker and have some spare time I invite you to help me with proof reading or building the theory manual.
I finally decided to release PragTic in the current state. There are already people from outside, who would like to test PragTic. I tried to postpone the release time, but now at least the testing can start. Though it is a beta version I believe that the basic functionality of PragTic is clear from the help.
Download & Install
Conditions of use
PragTic is a freeware intended just for research purposes. It is not expected to be used for any commercial computation and if it is used in such a way, it is on your own responsibility.
Please, if you find a bug, just write me, how it occured. If you are a registered user, you can use the on-line system of bugs reporting in Bugs section. Well, this is a bit funny, but I have to work to earn some money, so it is very well probable, that you will have to evade the same problem for some time before I mend PragTic.
I cannot promise you, that everything is working correctly. I would really appreciate your help in validation of implemented methods. I checked nearly all of them towards analytical solution or results of their appropriate authors, but this needn't to be enough.
If you would like to use PragTic, you are required to become an official user (either privileged user or anonymous user). You can choose the form of your account after you click on the Download Entire Program button below.
Present State of PragTic
What it can do:
- Full import of MSC.NASTRAN data through *.dat, *.nas and *.pch files.
- Reading of nearly any ASCII FE-data observing relatively common formatting rules - this feature is usable for solid elements only.
- Output of solution to a spreadsheet form (readable e.g. in MS Excel) or to the FEMAP neutral file (*.neu)
- Performing a fatigue calculation either on FE-results (at nodes, nodes of elements, elements) or at isolated points without any relation to the FE-model.
- Preparing the multiaxial solution, i.e. relating more simultaneous load channels.
- Loads given as load spectra, load histories, transient analysis (results from non-linear FE-calculation solved in more steps for a load sequence)
- Accessing (viewing/editing) the input/output data of calculation directly inside the program.
- Filter for selection of data desired for viewing.
- PragTic can be used as a base for an implementation of other methods processing the FE-data (creep, limit states computation, FE-data check, etc.)
- Tools section producing details on FE-models topology, results, etc.
- Computation by many high-cycle fatigue methods (Findley, Papadopoulos, McDiarmid, Zenner & Liu, Crossland, Sines, etc.) and low-cycle methods (8 uniaxial ones, Wang-Brown, Socie et al. from the multiaxial methods).
- Unique solution by PCr method designed by myself concerning the computation towards the fatigue limit - check its interesting results in the FatLim database.
- S-N curve based LESA method derived from the Eichlseder's gradient method and FKM guideline as implemented in FemFat.
- Control over the broad range of all the methods' parameters.
- Ability to run the computation by several different methods simultaneously.
What it cannot:
- There is no way how to introduce plasticity to an elastic FE-calculation except for the uniaxial methods (where the Neuber's and Glinka's methods are available). The only solution here is to prepare the transient analysis using the non-linear FEM-results in different time instants.
- I expect, that you will strongly miss a possibility to count in size, stress gradient, surface quality or statistical effects for the most of the methods. This is still ahead. Similar solution is implemented now only within the LESA method, but the derived effects are not linked up to the rest of the methods.
- No intern visualiser - upload of result data back into the FE-utility is expected.
- Input of finite elements was tested only on several element types of ANSYS, ABAQUS and COSMOS. The only nearly fully implemented FE-package import is that one from MSC.NASTRAN. Further element types (or even upload of particular FE-data) can be implemented on your request.
Be aware, that it is a one man's work. I'm proud of the product, but it still lacks more working hours.
Help for PragTic
Preview
