Question about applying function using hist 回答済み

5件のコメント

• 正式なコメント

  

  Chloe Allison

  2018年10月17日 09:47

  Hi there,

  You can use your velocity information as an input signal in OnScale. You would input the time history data as you have in your other forum comment using data hist. Then you would use boun defn to define your velocity boundary. If you wish to apply your velocity input at a surface as you mentioned in your other question use the boun side command ans specify your velocity boundary. I have written some example code below to demonstrate this. Also have a look at our command reference to see the syntax of the commands I have hightlighted in italics above.

  data
        hist drv1 * impulse_signal.dat 0.1
        end

  func hist drv1 1.e7

  boun
         side ymax vel func      /* Applies velocity boundary to ymax
         end

  Alternatively, if you do not wish to apply the velocity boundary at a side of model you can switch the boundary command to use node to apply the boundary at the specified region of nodes.

  boun
         defn bnd1 vel func
         node $i1 $i2 $j2 $j3
         end

  I hope this has answered your questions!

  Best Wishes,

  Team OnScale.

   

  コメントアクション パーマリンク
• 

  Hyunjae Song

  2018年10月22日 06:44 (編集済み 2018年10月23日 01:33)

  Thank you for answering.

  I have one more question about input signal.

  I have measured signal which has 20001 data length.

  But when I use this signal, pzflex use part of input signal. (0~0.65e-4 sec data, 3081 ea) (input data length: 0~2e-3, 20001 ea)

  So I multiply 0.1 on original input signal using this function.

  - hist drv1 * impulse_signal.dat 0.1

  How can I use my total input signal?

  コメントアクション パーマリンク
• 

  Jeff Dobson

  2018年10月23日 12:31

  Hi Hyunjae,

   

  How long are you running your simulation for?

  If you can you send your model files to us and we will take a closer look for you.

   

  Thanks,

  Jeff

  コメントアクション パーマリンク
• 

  Hyunjae Song

  2018年10月24日 23:37

  Hi, jeff

  -------------------------------------------

  
  symb #get { step } timestep

  symb nloops = 10

  symb simtime = ( $x6 / $vel ) * 1
  symb nexec = $simtime / $step
  symb nrun = $nexec / $nloops

  
  symb nexec_loop = nint ( $nexec / $nloops )

  ----------------------------------------

  current definition of step ( real ) = 0.4232341E-07
  current definition of nloops ( integer ) = 10
  current definition of simtime ( real ) = 0.6666667E-04
  input- c symb simtime = 1.e-3 /* Simulation Time
  current definition of nexec ( integer ) = 1575
  current definition of nrun ( integer ) = 157
  current definition of nexec_loop ( integer ) = 157

  ----------------------------------------

  This is my simulation time code and history.

  I use vel = 1500

  help me...

  コメントアクション パーマリンク
• 

  Jeff Dobson

  2018年10月25日 13:11

  Hi Hyunjae,

   

  Your simulation time is only 0.6666667E-04 which is why only the first 0.65e-4s of your function was initially used. Scaling the time is going to change your input signal frequency. I would suggest you don't scale the time but run your simulation for longer (duration of input signal or longer) and you should see the full signal being applied.

   

  Kind Regards,

  Jeff

  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。