Learning Forum

[louiee1907]

Hello,

I try to do aircraft wing fatigue life analysis. I have acceleration-time data (1G-1.5G) more than 20000 datas. I determined boundary conditions for wing (acceleration). I converted acceleration-time graph to PSD to add fatigue tool history data. However, I have a problem. How can I calculate the life cycle as hours and infinite life?  How can I use the Rainflow Cycle to calculate it? 

 

[Ashish Khemka]

Hi Louie,

Please see if the following link is of help to you:

 

How to plot hysteresis and Rainflow matrix using Fatigue Tool in ANSYS Mechanical - Ansys Knowledge

 

Thanks and Regards,

Ashish Khemka

[louiee1907]

Thank you for answer, however I have a different problem. I have almost 98700 acceleration datas for aircraft's wing. I will pit them the history data to calculate fatigue life. I want to find 1 block is how many hours? How can I calculate it? 

[louiee1907]

 

I want to give detail. It is going to more than 2000 datas. 

I have 2 questions. 

1- How can I find the 1 cycle/block is equal to how many hours? 

2- Should I convert acceleration-time to PSD to put history data or must I put as acceleration-time?

	time	acceleation
	0.1	9,80
	0.2	10,7
	0.3	9,8
	0.4	9,8
	0.5	9,79
	0.6	10,8
	0.7	9,8
	…	...

 

[Ashish Khemka]

Hi Louie,

Thanks for the update. I do have an answer to this query. Please wait for other members to comment.

Regards,

Ashish Kumar

[danielshaw]

Louie:

I assume that you are using the Mechanical Fatigue Tool (FT) with time history loading (non-constant amplitude, proportional loading).  If so, the FT outputs life in repeats of the loading cycle.  For example, if your loading cycle represents 1 hour of operation then reported life is repeats of 1 hour of operation.

The FT internally uses rainflow counting to organize the stresses into “bins or blocks” of data that represent stress ranges and mean stresses.  You can block the rainflow matrix, but the FT will not report the fatigue damage per bin.

If you have the time history loading, you should not need to use a random vibration approach.

Dan

[louiee1907]

Thank you so much sir. I have changed the fatigue analysis program as nCode DesignLife. 

You said ''If you have the time history loading, you should not need to use a random vibration approach'' I thought Since I use nCode, I need to transfer this acceleration-time data to the system. So I thought of doing an analysis using vibration generator. Am I right? 

Workflow ;
Static Structural - nCode DesignLife (Time Series SN) - Excel Input - Vibration Generator - Analysis - Result

Is it true or not? In addition, should I add acceleration condition on model in static structural? I have moment condition too. I know the fatigue, however I can't realize the thought in my head. I need your help sir. 

[danielshaw]

Are the data in the acceleration table multipliers on the applied acceleration in the FE model?  If so, why can’t you use a Time Series Load Mapper in Ansys nCode DesignLife? DesignLife time series load mapping uses one set of FE results and scale factors from a time history file to create time varying stress.  If you can convert your time base acceleration data to a PSD spectrum, you can run a DesignLife vibration fatigue analysis using that spectrum, but I do not see why it is necessary (or preferred).

Is the moment load constant or does it vary with time?  If it is a constant load, then the stress it produces does not directly contribute to the alternating stress.  It acts a mean stress.  The DesignLife time series load mapper has an option to specify a “static” result set.  The stresses from the static result set will not vary over time, but they will be used in any mean stress calculations.

[louiee1907]

 

My acceleration-time datas are in Excel table. There is acceleration(1G) condition in FE Model. My datas must effect my wing acceleration from 1g to 1.5g with time, and I want to see after 2000 sec, my fatigue life in FE Model

There is moment which is constant in FE model.

I want to analyze that performing fatigue analysis of my wing, whose acceleration changes during the flight between 1-1.5g.

What should my workflow be?

 

[louiee1907]

In addition you are right vibration analysis is not necessary, but I want to add acceleration-time datas on acceleration which is in FE Model. I hope I explained my problem

[danielshaw]

Walking you through a complete Ansys nCode DesignLife fatigue analysis is beyond the scope of this forum.  I can only provide some guidance.  In addition, there are multiple acceptable approaches.  I cannot know the best approach without obtaining more details.  I recommend searching AIS for the “Introduction to Ansys nCode DesignLife “training course.  That training course will explain how to set up an Ansys nCode DesignLife time series analysis.  I also recommend reviewing the first 2 Worked Examples in the Ansys DesignLife Help (see below).

If you are using 22R1 or later, I recommend using the Ansys nCode DesignLife Mechanical Add-On (WE#1).  If you are using an early release, I recommend using the Ansys nCode DesignLife Workbench Integrated systems (WE#2).  You can also use the HBK standalone Glyphworks interface, but I do believe that approach is required and it requires more DesignLife knowledge.

A quick summary of the required steps.

1.       Run Mechanical model with 2 load steps:  static moment and unit acceleration

2.       Store the acceleration load multipliers as a single column time history file in the appropriate format (ASCII for embedded Add-On and .s3t for the Integrated systems)

3.       Choose the appropriate set-up for a time-series analysis (depends on the interface)

4.       Choose time-series load mapper and read in Mechanical result sets

5.       Set the moment result in the load mapper to be “static”

6.       Associate the acceleration load step in the load mapper with the time history file.

7.       Specify the appropriate mean stress method

8.       Solve.

 

[louiee1907]

 

Hi,

I searhed many times, however I cannot add my excel datas(acceleration-time) in nCode DesignLife. How can I convert excel data to s3t? Time series input does not see excel datas on workspace

EDIT : I solved the problem

[louiee1907]

Hello Sir, 

I hope you are good today :) 

We have talked before about calculating from life cycle to hours. I confused a bit. You know I have acceleration-time data and time finishes at 9700 second. I analyzed my model life which is 389000 cycles, so should I (9700x389000)/(3600) to calculate as hours life?

[louiee1907]

Sir, you are the best. Thank you so much :) I'll do it and I'll watch every video.