Random Fatigue Analysis

Danilo_21605

Dear all,

I should perform a random fatigue analysis.

They provide me a PSD as input data and the target is that the component must survive to 40 hours with a 0.2 of damage.

How can I convert the 40h into cycles?

Or can I input in optistruct the hours number instead of the number of cycles into the FATSEQ card?

 

Thank you

Abe_22283

Hi Danilo,

I think it might be possible. Please try following method.

In Card Edit of FATSEQ, check "Real" to change [N(1)] into [T(1)], and click [T(1)] to make [T(1)] active.

Then you can enter the number of time.

 

Adriano Koga_20259

one possible apporach would be to calculate your structure modal participation masses, and understand what are the main modes of your structure and the correspondent frequency. Once you know the frequency, you will have the number of cycles per second (Hz). So if you have 40h*3600s each, and multiply by your frequency, you will have the number of cycles. It would be an estimative.

Danilo_21605

Adriano Koga_20259 said:

one possible apporach would be to calculate your structure modal participation masses, and understand what are the main modes of your structure and the correspondent frequency. Once you know the frequency, you will have the number of cycles per second (Hz). So if you have 40h*3600s each, and multiply by your frequency, you will have the number of cycles. It would be an estimative.

Thank You Adriano,

I understand what you said, but maybe I was wrong when I posted the answer.

I have this requirement for my component "Fatigue analysis of installed component under random vibration for 40 hours per axis - Cumulated Damage <0.2"

In the optistruct analysis where and how should I set the "40 hours"?

Should I convert them in other? Where should I set them? 

I supposed to set them into FATSEQ Card, but I'm not sure.

 

Thank you

Danilo_21605

Thank You Adriano,

I understand what you said, but maybe I was wrong when I posted the answer.

I have this requirement for my component "Fatigue analysis of installed component under random vibration for 40 hours per axis - Cumulated Damage <0.2"

In the optistruct analysis where and how should I set the "40 hours"?

Should I convert them in other? Where should I set them? 

I supposed to set them into FATSEQ Card, but I'm not sure.

 

Thank you

Adriano Koga_20259

Danilo_21605 said:

Thank You Adriano,

I understand what you said, but maybe I was wrong when I posted the answer.

I have this requirement for my component "Fatigue analysis of installed component under random vibration for 40 hours per axis - Cumulated Damage <0.2"

In the optistruct analysis where and how should I set the "40 hours"?

Should I convert them in other? Where should I set them? 

I supposed to set them into FATSEQ Card, but I'm not sure.

 

Thank you

there is not a filed for time.

You need to input it as number of cycles. So you need to convert it into number of cycles based on a certain assumption, which in general considering the most participating frequency would be a good approximation.

Danilo_21605

Thank you,

in a previous analysis they provided me this indication: 

"Fatigue analysis of installed component under random vibration for 16 hours (200000 design life) - Cumulated Damage <0.2"

According your experience, could I take this value (200000) as number of cycles?

In that case shoud I input this value into the FATSEQ Card?

FATSEQ  45
+             44  200000

Adriano Koga_20259

Danilo_21605 said:

Thank you,

in a previous analysis they provided me this indication: 

"Fatigue analysis of installed component under random vibration for 16 hours (200000 design life) - Cumulated Damage <0.2"

According your experience, could I take this value (200000) as number of cycles?

In that case shoud I input this value into the FATSEQ Card?

FATSEQ  45
+             44  200000

As an approximation, if you have a dominant frequency of 10Hz, and you need to calculate cycles for 40h test, then it should be:

Example:

#cycles = 10Hz * 40h * 3600s = 1.440.000 cycles

 

This could be added to your FAT cards for fatigue.

Danilo_21605

Adriano Koga_20259 said:

As an approximation, if you have a dominant frequency of 10Hz, and you need to calculate cycles for 40h test, then it should be:

Example:

#cycles = 10Hz * 40h * 3600s = 1.440.000 cycles

 

This could be added to your FAT cards for fatigue.

I understand what you said.

If I don't have any indication I could use this approximation method.

I was asking before if you would use that value (200000 design life) as an indication of the number of cycles and if it is correct to insert it in the FATSEQ card.

 

Thanks

Adriano Koga_20259

Danilo_21605 said:

I understand what you said.

If I don't have any indication I could use this approximation method.

I was asking before if you would use that value (200000 design life) as an indication of the number of cycles and if it is correct to insert it in the FATSEQ card.

 

Thanks

i think 200.000 in 16h is just someting around 3.5Hz dominant frequency.

Abe_22283

Hi Danilo,

I think it might be possible. Please try following method.

In Card Edit of FATSEQ, check "Real" to change [N(1)] into [T(1)], and click [T(1)] to make [T(1)] active.

Then you can enter the number of time.

 

Danilo_21605

Abe_22283 said:

Hi Danilo,

I think it might be possible. Please try following method.

In Card Edit of FATSEQ, check "Real" to change [N(1)] into [T(1)], and click [T(1)] to make [T(1)] active.

Then you can enter the number of time.

 

Thank you Abe,

this was the answer I'm hoping to have.

 

Thanks