TAGGED: #mechanical#workbench, fatigueanalysis


October 16, 2024 at 3:50 amben.mcpherson12Subscriber
Hi there,
I'm currently using ANSYS2023R2 to complete FEA analysis on a reclaimer structure. More specifically, I'm investigating the fatigue life of the structure based on different loading scenarios. A disclaimer: the model is very large (100 m long, 60 m high)  and I have a beefy enough computer system to handle it.
Within my static structural, I am inputting three forces in the x,y,z directions respectively. Each force is input as a sin wave, all of which have a nonzero mean and have different phases, amplitudes and frequencies.
I have put a fatigue tool into the simulation, and have adjusted the ratio to (min stress / max stress) to account for the nonzero mean. I am using the Goodman correction theory.
My question is about my fatigue life results. Despite a few stress concentrations that are just due to meshing, my fatigue lives range from 1e5 to 1e6 seconds. This seems fine, but the structure I am simulating has been operating under the same force conditions for over 2 years without structural failure  which does not reflect the 1e6 seconds FEA predicts.
Could this small predicted fatigue life be due to one of my input parameters?
Thanks in advance,
Ben 
October 16, 2024 at 1:19 pmKaushal VadnereAnsys Employee
Hi Ben,Â
The load that you are applying to structure, represents one loading cycle right? So, in the life, the value you are getting is equivalent to 1e06 cycles (does not necessarily mean 1e06 seconds or minutes or days, etc.) For example, if the load history represents one month of loading in one cycle, and the life found is arount 120, the expected model life would be 120 months. In addition, for this case, the application imposes a limit of 1e06 cycles if the Linear SN curve definition is used. Hence, as a general rule, if your model has a life of 1e06 cycles, it suggests infinite life.Â
You can refer to these articles in ansys help document for fatigue analysis: 19.11.5. Fatigue Results (ansys.com), 19.11.4. Fatigue Analysis Application (ansys.com)
Let me know if this answers your question.
To learn few tips and tricks about fatigue analysis, you can refer to this free tidbit from Ansys: Topics in Fatigue Analysis in Ansys Mechanical  Innovation Space.Â
Also, if you have ALH subscription, you can refer to 'Durability' learning room in structures building: Structures  Ansys Learning Hub
Thanks!
October 16, 2024 at 7:28 pmben.mcpherson12Subscriber
Hi Kaushal,Â
Â
Thanks for your reply. The loads I am applying in the x,y,z directions are the cyclical loading profile of the structure. I currently have my simulation to solve for t=1.5 seconds. The period of the three loads range between 0.3 and 1 second, so I simulated 1.5 seconds in order to capture at least one cycle from each respective load.Â
Would this mean my life output would be quoting me cycles of 1.5 seconds? 1e5 * 1.5 seconds lifespan for example?Â
The SN curve is not overly applicable to my model given the nonzero mean stress, hence my thinking in using the Goodman correction theory. Have I applied it correctly in the screenshot in my previous message?

October 17, 2024 at 2:52 pmKaushal VadnereAnsys Employee
You are solving the simulation for 1.5 seconds, but is that the same time frame 'in real life' when these cyclic loads are being applied to the structure?Â



 You must be logged in to reply to this topic.

791

427

265

201

168
Â© 2024 Copyright ANSYS, Inc. All rights reserved.