Message Boards Message Boards

How to connect data from solving equation in some loop ??


My program about solving equation , then when it finished , by looping , by another order when you ask which answer of solving equation , it plot it . i think i describe it awfully !!!! enter image description here

POSTED BY: Saleh Baradaran
14 days ago

The error in your error is in red.

POSTED BY: Sander Huisman
14 days ago


M = 100;
sols = Ω /. Table[FindRoot[Tanh[Sqrt[Ω]] - Tan[Sqrt[Ω]], {Ω, IG}], {IG, 10, 200, 50}];
f[x_, n_] := Sum[sols[[n]]^(2 i)/(4 i+2)! x^(4 i+2), {i,0,M}] - Sum[sols[[n]]^(2 i)/(4 i+2)!, {i,0,M}]/
     Sum[sols[[n]]^(2 i)/(4 i+3)!, {i,0,M}] Sum[sols[[n]]^(2 i)/(4 i+3)! x^(4 i+3), {i,0,M}];
n = Input["which mode do you want?"];
Plot[f[x, n], {x, 0, 1}]
POSTED BY: Bill Simpson
14 days ago

Thank you so much , yes your program is working , Thanks

I want to know if i want to put this program in loop or for , how can i do it? for example i want to say , ask user which beam do you want? if he say clamped free , the program choose clamped free that we define for it and do like the program above , then plot n mode for it ...

Thank you so much

POSTED BY: Saleh Baradaran
13 days ago


I would do it all in a Dynamic Module. Using loops and inputs is not very Mathematica-like.

Here is an example:

DynamicModule[{M = 100, sols, n, beam = 1},
 sols = \[CapitalOmega] /. 
     Tanh[Sqrt[\[CapitalOmega]]] - 
      Tan[Sqrt[\[CapitalOmega]]], {\[CapitalOmega], IG}], {IG, 10, 
     200, 50}];
 f[x_, n_] := 
  Sum[sols[[n]]^(2 i)/(4 i + 2)! x^(4 i + 2), {i, 0, M}] - 
   Sum[sols[[n]]^(2 i)/(4 i + 2)!, {i, 0, M}]/
     Sum[sols[[n]]^(2 i)/(4 i + 3)!, {i, 0, M}] Sum[
     sols[[n]]^(2 i)/(4 i + 3)! x^(4 i + 3), {i, 0, M}];
       Dynamic[beam], {1 -> "Free-Pinned", 2 -> "Clamped-Free"}], 
      Row[{"Mode:", SetterBar[Dynamic[n], Range[Length[sols]]]}]}],
    Dynamic[Plot[f[x, n], {x, 0, 1}, ImageSize -> Large]]}]]]

Which gives you a dynamic like this:

enter image description here

The SetterBar can be made to automatically resize based on how many modes there are in the solution. I obviously did not add any functionality to the beam type popup but you can do that. Use the dynamic variable "beam" to change your solution so the plot will change.



POSTED BY: Neil Singer
13 days ago

Dear Neil

I must say you thank you many many times , your opinion is very attractive and i want it so you give it to me , but i want to know if i have 6 different beam with 6 different function and also 6 different sols , how can i put them in this program you suggest me?

I work on your program but since the I am new in mathematica programming , I cant solve it , So please help me more How can your program extended for 6 type of beam by 6 different function and 6 different Sols

Best regards



POSTED BY: Saleh Baradaran
12 days ago

First of all, I had my bracket in the wrong place because I wanted the first row to be the Beam Type and the second row to be the mode. I fixed this in the code below.

Next, I do not understand your approach for solving for the mode shapes -- you are setting the root finder for the eigenvalues at arbitrary spacing and this will not guaranty that you get a solution. I would use the formulas in a book such as Blevins:

Blevins, R.D., "Formulas for Natural Frequency and Mode Shape", Krieger Publishing Company, 2001. ISBN 9781575241845, Page 108-109.

What I would do: He gives an approximate formula for the eigenvalues -- use this for your starting value -- its real close. solve for the actual lambdas for each mode and then use the lambda for plotting in the beam equations provided. I do not understand your "f" equation. Is that a Taylor series for the actual equation? I am curious where that comes from.

(* Define some beam shape equations (from Blevins)*)

freeFreeShape[l_, x_] := 
  Cosh[l*x] + 
   Cos[l*x] - (Cosh[l] - Cos[l])/(Sinh[l] - Sin[l])*(Sinh[l*x] + 
freeSlidingShape[l_, x_] := 
  Cosh[l*x] + 
   Cos[l*x] - (Sinh[l] - Sin[l])/(Cosh[l] + Cos[l])*(Sinh[l*x] + 
clampedFreeShape[l_, x_] := 
  Cosh[l*x] - 
   Cos[l*x] - (Sinh[l] - Sin[l])/(Cosh[l] + Cos[l])*(Sinh[l*x] - 
freePinnedShape[l_, x_] := 
  Cosh[l*x] + 
   Cos[l*x] - (Cosh[l] - Cos[l])/(Sinh[l] - Sin[l])*(Sinh[l*x] + 
(* Define a DynamicModule to do the interaction *)
DynamicModule[{M = 
   100, lambdas, lambdaEqns, approxLambdas, n, beam = 1, ans, 
  modeshapes, i, maxCases = 4, maxModes = 10, lamValue},
 (* Order in the following lists is Free-Free, Free-Sliding, \
Clamped-Free, Free-Pinned *)
 (* List of equations for lambda (from \
Blevins) we will be finding the lamvalue that makes them zero later *)

  lambdaEqns = {Cos[lamValue]*Cosh[lamValue] - 1, 
   Tan[lamValue] + Tanh[lamValue], Cos[lamValue]*Cosh[lamValue] + 1, 
   Tan[lamValue] - Tanh[lamValue]};
 (* List of equations for approximate lambda (from Blevins) we will \
use these as starting guesses later *)

 approxLambdas = {(2*i + 1) Pi/2., (4*i - 1) Pi/4., (2*i - 1) Pi/
     2., (4*i - 1) Pi/4.};
 (* Create a list of the solutions -- there are maxCases (I only did \
4 beam types so this is 4) lists each with the first maxModes lambdas*)

  lambdas = 
  Table[lamValue /. 
      lambdaEqns[[indx]], {lamValue, approxLambdas[[indx]]}, 
      WorkingPrecision -> 18], {i, maxModes}], {indx, maxCases}];
 (* make a list of the functions so we can choose from them when \
plotting *)

 modeshapes = {freeFreeShape, freeSlidingShape, clampedFreeShape, 
 (* make a panel and plot things up *)

 Panel[Column[{Row[{"Single-Span: ", 
       Dynamic[beam], {1 -> "Free-Free", 2 -> "Free-Sliding", 
        3 -> "Clamped-Free", 4 -> "Free-Pinned"}]}], 
    Row[{"Mode:", SetterBar[Dynamic[n], Range[maxModes]]}],
    (*Dynamically choose a modeshape equation and apply it to two \
arguements -- the corresponding lambda by indexing into the lambda \
list first by beam, then by mode number, plot it in x *)

    Dynamic[Plot[modeshapes[[beam]][lambdas[[beam, n]], x], {x, 0, 1},
       ImageSize -> Large]]}]]]

I create a list of the functions to solve to get the eigenvalues (lambdas) for each beam type. Next create a list of the approximate lambda formulas so the guesses are very close. Next create a list of actual lambdas by solving the equation from the first list with the starting values from the second list. Next I would create a list of mode shape equations. When you choose a beam type, you choose a set of lambdas for that beam type and an equation for the beam shape. I put it together for a few beams -- you will need to add the rest. Note I defined the equations for the beam shapes outside of the Dynamic -- this was not necessary -- you could paste that block inside the DynamicModule before the line modeshapes = ... and make them local functions by adding them to the local variable list -- so they are not globally defined -- its up to you.

This is what it looks like:

enter image description here



POSTED BY: Neil Singer
12 days ago

Dear Neil ,

First of all , i must thank you for your attention & also your help in solving problem , Thank you

I should describe you that which formulas i used in solving problem were come from using DTM (Differential Transform Method) on solving mode function equations . I want to program these formulas in looping that you help me very good to use them in my M.Sc thesis & also my ISI articles at the future.

I am very happy & nice to meet you

with the best regards Saleh

POSTED BY: Saleh Baradaran
6 days ago

Group Abstract Group Abstract