Learning Forum

[arekbed]

hello.

I create simulation of compressor blade with notch. Later I used Fatigue Tool - e-N anlysis to get information about fatigue life.

What exactly mean "fatigue life"? It is a number to load cycles to crack initiation or something? In experimental results is a number of cycles to initiating a crack with (i.e.) 1 mm size. But how it work in Ansys Mechanical? Is it a number to load cycle to fracture of element or there were used a special criterion of fracture?

I will be glad if someone will explain me this.

With best regards,

Arek

[peteroznewman]

Fatigue Life comes from experimental data.

Samples of a specific material are prepared and put in a fatigue testing machine. There are many shapes of samples, from flat plate "dog bone" shapes to cylindrical rod shapes. There are several kinds of fatigue testing machines from tensile testing machines that can take a variety of sample shapes to Rotary testing machines that only take cylindrical rod shaped samples. The Rotary testing machines are, by design, a fully reversed load case. The tensile test machines could be a zero-based load cycle that does tension only.

You get to define failure for the experiment. You mentioned a 1 mm size crack.  It is more typical to define failure as the complete fracture of the part into two pieces.

You get to design the experiment. You have to choose how many samples to test at each load setting. If you have fifty samples, you could put ten samples at each of five load settings or you could put five samples at each of ten load settings. You put each sample in the machine and run it at its load setting until it fractures. Then you average all the samples run at the same load to get the average life at that load.

The average life vs the load (either Stress or Strain) is the data that gets entered into Engineering Data for ANSYS to use.

What did you learn from your research into this question?

Regards,
Peter

[Sandeep Medikonda]

Arek,

Fatigue can be studied with or without crack. It is very well explained in these resources:

Resource 1

Resource 2

Resource 3

Resource 4

 

Regards,
Sandeep
Guidelines on the Student Community

[arekbed]

So, maybe I explain my situation. I create several experimental tests to get information about the number of cycles to crack initiation in compressor blade (crack length equal to 1 mm). After this, I try to compare numerical and experimental results. To prepare strain-life analysis, I designate data (based on models: Mansons, Mitchell, Roessle-Fatemi, Baumell-Seger, Ong, etc.) to Manson-Coffin-Basquin equation and data connected to work hardening. After loading on the blade I found information about stress distribution in the top of the notch. Based on stress, I try to achieve information about fatigue life.

My supervisor told me (and show in MSC Fatigue documentation) that numerical results of fatigue life correspond to crack with 1 mm, and suggested me to find information about to what kind of damage correspond the numerical result of e-N analysis or maybe exist some kind of criterion what is implemented in ansys code.

So my question is: what kind of damage I theoretically will have in my element after the obtained number of load cycles (crack appearance, crack with 1 mm length or total failure - dividing to two separate elements)?

We need to remember, a numerical model of material is based on some assumptions. Additionally,  we need to remember about that blade have complex geometry and material parameters inside the blade are different than in the surface (because of treatment during production).

Resources what I found above do not give me information what I'm looking for. Part of them I known for long.

I'm appreciated that you try to help me and I hope so with your help I will find informations to finish my PhD work. 

 

with best regards,

Arek

[arekbed]

@SandeepMedikonda I know this publication but there I cannot find answer for my question. Please read what I write above.