Hi I found something wrong (memory overflow?) 回答済み

Avatar
Hyunjae Song

Hi.

I use input signal having time length 2 ms 26001 samples.

I load this input signal using data-hist and put this at func.

data
hist drv1 * water_lv1-press.dat 1.
end

func hist drv1 1.

boun
side xmin absr
side xmax absr
side ymin absr
side ymax absr
end

And I simulate during 2ms bcz input signal is 2 ms.

 

Result is well during at 0.4 ms but after then, there is wrong result I don't understand.

Here I attach my result after 0.4 ms

It just simple simulation but I can't get right result.

If I use option at data-hist 0.1, then I can get right simulation until at the end

data
hist drv1 * water_lv1-press.dat 0.1
end

simulate during 0.2 ms

 

What is my problem?

4件のコメント

  • 正式なコメント
    Avatar
    Chloe Allison

    Hi there,

     

    It looks like your model is going unstable after 0.4ms, pmin looks very high. This could be caused by the absorbing boundaries. I would try using impedance boundaries 'side xmin impd'. You could also add in hourglass supression to water in the material definition by adding the subcommand 'hrgl watr visc 0.1'.

     

    Best Regards,

    Chloe.

    コメントアクション パーマリンク
  • Avatar
    Hyunjae Song

    Thank you for your answer Chloe.

    You mean this problem is caused by high pmin right?

    Then what is allowsed maximum(minimum?) pmin value for stable simulation?

    コメントアクション パーマリンク
  • Avatar
    Chloe Allison

    The problem is the absorbing boundaries, this is what is causing the high pmin because the model is going unstable. Using impedance boundaries or hourglass suppression should fix this.

    コメントアクション パーマリンク
  • Avatar
    Hyunjae Song

    Thank for your help Chloe.

    コメントアクション パーマリンク

サインインしてコメントを残してください。

お探しのものを見つけられませんでしたか？

新規投稿