0
$\begingroup$

Recently I have obtained a csv value form and experiment I have computed. However, I was trying to understand how each individual component being calculated. The image attached shows sample point deformation with time as each row represent a time step (I am not sure if I can upload the csv file here as there is no upload option). I have managed to calculate the deformations (U,V,W) which is for example U = Xi - Xo. However, I am trying to calculate the lagrangian strains using the equations provided by the software it is not giving me similar results. I would appreciate if someone can show me how the strains are being computed at each time steps using the provided data.

Deformation Gradient Data

$\endgroup$
3
  • 1
    $\begingroup$ The strains would depend on the distribution of points and the interpolation used. The latter depends on the method. Were the data obtained using a finite element or finite difference method? $\endgroup$
    – nicoguaro
    Oct 25, 2022 at 12:09
  • 1
    $\begingroup$ If each line corresponds to a time step, then you track a single point from which you cannot compute strains. Or do you have further tables for more points? Or does the program process simulated/measured surrounding points not listed in the table? $\endgroup$ Oct 27, 2022 at 9:08
  • $\begingroup$ @DominikKern Yes, this is just a point in a larger dataset. The csv file that the data came from is too big (0.5 GB) so I am not sure if I can share that. The reason why I want to know how these being computed is that I want to compute the strain in the z direction (Ezz). Would this be possible for the dataset given that they all have the location (X,Y,Z) at each time step. $\endgroup$
    – Abdullah
    Nov 2, 2022 at 3:10

1 Answer 1

2
$\begingroup$

If you have $x$,$y$,$z$-coordinates and displacements $u$,$v$,$w$ for a set of points, then you could approximate the displacement gradient by finite differences (or possibly other interpolations). So for derivatives with respect to $x$, you would sort the points with respect to the $x$-coordinate, such that they are neighbors in this direction and then compute ($n$ corresponds to spatial position on a row in $x$-direction) \begin{align} \frac{\text{d}u}{\text{d}x} &\approx \frac{u_{n}-u_{n-1}}{x_{n}-x_{n-1}} ,\\ \frac{\text{d}v}{\text{d}x} &\approx \frac{v_{n}-v_{n-1}}{x_{n}-x_{n-1}} ,\\ \frac{\text{d}w}{\text{d}x} &\approx \frac{w_{n}-w_{n-1}}{x_{n}-x_{n-1}}. \end{align} Then sort the rows with respect to $y$, repeat the steps for the derivatives with respect to $y$.

If you want to calculate $\varepsilon_{zz}$ then do the same for $z$, required that there are same neighbored points with offset $z_{n}-z_{n-1}\ne 0$, i.e. the points are not complanar, which may be, since the shown formulas are for two dimensions. In 3D you would have \begin{align} \varepsilon_{xx} &= \frac{\text{d}u}{\text{d}x} +\frac{1}{2}\left( \left(\frac{\text{d}u}{\text{d}x}\right)^2 + \left(\frac{\text{d}v}{\text{d}x}\right)^2 + \left(\frac{\text{d}w}{\text{d}x}\right)^2 \right)\\ \dots& \\ \varepsilon_{zz}&= \frac{\text{d}w}{\text{d}z} +\frac{1}{2}\left( \left(\frac{\text{d}u}{\text{d}z}\right)^2 + \left(\frac{\text{d}v}{\text{d}z}\right)^2 + \left(\frac{\text{d}w}{\text{d}z}\right)^2 \right) \end{align} Note that with geometric linearization you neglect the quadratic terms in the strains, but if they are relevant for your problem, account for them.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.