Hello

I need the steps to simulate fatigue under abaqus (2D)

thank you.

Hi friends,

I am modelling a pressure vessel and i am unsure how to use Low Cycle Fatigue Analysis. Is there any good tutorial that may be of some help to model fatigue using this step in ABAQUS?

Also can you tell me what c1,c2,c3 and c4 mean while defining damage initiation and propagartion in low cycle fatigue analysis as abaqus manula doesnt talk about it? Your help in this regard will be highly apprecited

Waqas

Dear all,

Can anyone offer a reasonable ball-park figure for the number of cycles to achieve elastic shakedown for an Aluminium alloy? Assuming the conditions are within the statical and kinematical bounds; and no fatigue failure.

Or any other advices.

Many thanks.

Yours sincerely,

BC

thank you for the ansver Stephane

I have now more elements for discuss my problem. My principal problem is to check on the behaviour of a steel riveted connection under cyclic loading. I have some results from experimental tests to compare with. i know from this tests that my speciemen reach a break after 300000 cycles under a cyclic load of 800kN amplitude. i know that it isn't a good idea to simulate a so big number of cycles then I first made a "model for cyclic loadings" for iso/kinematic hardening for 100 cycles, and so I can observe the elasto-plastic behaviour and watch when my specimen reach a cyclic stability. now I have to do a fatigue analysis for reach a sperimental results.how can I do this??

thank you very much

regards

Bojan

Hi,

I agree with Michele, cyclic direct method is a very efficient method. I'm not sure but i think that it's Nguyen Quoc Son (with Akel his PhD student) who was at the origin of this method. It's true that later Habibou Maitournam had made the major developments.

To answer to the question of bojan, it's important to precise that cyclic direct method give the stabilized state (elastic shakedown, plastic shakedown) reached by the structure after cycling. The method will not give you the number of cycles.

Fatigue analysis is generally based on post-treatment of fem results.

If you have elastic deformation after one cycle, use this stress state with fatigue criteria to obtain the life (don't use cyclic direct method)

If you obtain elastic shakedown, use the stress state reached after stabilization to perform fatigue calculation with the same criteria like previously ( u can add the residual stress given by cyclic direct method to mean stress)

HCF include this 2 cases.

If you obtain plastic shakedown or rachetting use LCF approach.

Sorry for my english but if u give me more informations about your problem, i can help you.

Regards

Stephane

i have a similar problem. i have a constant cyclic load (higher limit 900kN and lower 100kN) and with a miner's damage law i must to find out the number of cycles i need for the crack.how can i do this???

thanks

ABAQUS introduced recently a new analysis procedure named Direct Cyclic procedure. For cases in which it is known that the response to cyclic loading is a stabilized cyclic response, this rocedure computes directly a stabilized cycle, without having to compute a number of sequential cycles that would lead to such stabilized cycle. The procedure is based on using Fourier xpansions for the (cyclic) displacement responses, and iteratively computing the coefficients of the Fourier series (Abaqus 6.4 User’s Manual). During every iteration the displacement is used to compute force residuals based on the material behavior, and these residuals are integrated over the time period of a cycle to obtain residuals corresponding to each Fourier coefficient. Corrections to the Fourier coefficients are then obtained using an elastic tangent stiffness matrix. This method is very economical, especially for larger models, because the stiffness matrix is factorized only once;
moreover, it does not compute multiple cycles until a stabilized cycle is obtained. All material onlinearities are taken into account during the residual computations. Material history evolution is somewhat picked up through the iterative computation of the Fourier coefficients. Although any revious history of the model corresponding to earlier steps in the simulation is completely retained, such as simulation of manufacturing and/or deployment, and has no limitations with respect to large deformations or material behavior, the additional deformations during the cyclic procedure are assumed to be small strain.

In fact it is taken from Maitournam.

A direct method for the solution of evolution problems
Comptes Rendus Mécanique, Volume 334, Issue 5, May 2006, Pages 317-322
François Comte, Habibou Maitournam, Pierre Burry, T. Mac Lan Nguyen

H. Maitournam, B. Pommier and J.J. Thomas, Détermination de la réponse asymptotique d'une structure anélastique sous chargement thermomécanique cyclique, C. R. Mecanique 330 (2002), pp. 703–708.

http://henry2.hpc.ncsu.edu:2080/v6.4/books/stm/ch02s02ath16.html

Obviously it is not feasible to simulate 25000 cycles. You need some data for cyclic hardening and softening (for instance from incremental step test). Then you can use cyclic loading models as described in the manual (11.2.2 in version 6.5).

Hope this helps.