Use mulgrid() to read existing TOUGH input file

[jxalfoudqt]

Hi Adrian,

How can I visualize the grid of radial model example by using 'mulgrid()'? There is 'dat.grid' but no 'geo' as specified in Hydrostatics 3D model. I think this method can be used for grid visualization of all existing TOUGH input files.

Thanks. Tao

[acroucher]

It depends whether it is a 1-D radial model or a 2-D radial model (i.e. r-z cylindrical coordinates).

For a 1-D radial model, you can do the visualisation with simple line plots vs. radius, using e.g. matplotlib. There is an example of this in the PyTOUGH wiki.

For a 2-D radial model in r-z coordinates, you can create a dummy 2-D rectangular mulgrid object representing the r-z geometry, purely for visualisation purposes. Give it the same radial and vertical block sizes that you pass to t2grid.radial() to create your TOUGH2 grid, and a dummy (arbitrary) block size in the y-direction. Then you can do visualisation of your results with e.g. mulgrid.slice_plot().

[jxalfoudqt]

Many thanks Adrian. Actually, I take your idea to generate a mulgrid object and write it to a .vtu file for visualization.

[chenmingzhang]

just a comment for others who may face similar situations https://tough.forumbee.com/t/18hx69

[jousheinfo]

Hi @acroucher, one question please. I have one radial grid very similar to the example in the link shared by @chenmingzhang, but I would like to know if it is possible to make a 3D plot for visualization. I have been reading other questions related to this one, but I could not find any useful so far. I would very much appreciate your help!

[acroucher]

hi @jousheinfo , just to clarify, are you wanting to make a 3D plot of a radial grid? I generally don't do that, because a radial grid is really only 2D (r, z), so it can be visualized better using 2D plots, as explained above.

If you are wanting to make a 3D plot of a 3D grid, then you can do that in PyTOUGH by using the write_vtk() method of a mulgrid object (or a t2listing object if you are wanting to plot results).

[jousheinfo]

@acroucher thank you for your response. I got another question, please. I have the following radial grid made in PYTOUGH:

from t2data import * from math import log10

mesh parameters:

well_radius = (6 / 2) * 0.0254 reservoir_radius = 1000. well_depth = 2500 thickness = 50 num_blocks = int((well_depth / thickness) + 1)

layer_thickness = np.repeat(thickness,num_blocks-1)

dr = np.logspace(log10(well_radius), log10(reservoir_radius), num_blocks) dz = layer_thickness

create TOUGH2 input data:

dat = t2data() dat.title = 'model_injection' dat.grid = t2grid().radial(dr, layer_thickness, origin = [0.,0.,0.])

Even tough I defined the max desired radius 1000 m, in the input file, I can see the last block has a radius of about 10 times bigger (5.288e+3*2 m). Why is this happenning? The desired number of cells is 50, and the thickness of each cell is constant equal to 50m.

I do not really understand how pytough computes the logarithmic increment and how can I find the increment factor?

I am trying to replicate the mesh file with meshmaker, but I'm struggling. I would very much appreciate your help. Thank you in advance!

[acroucher]

Note that the dr values are radial block thicknesses, not radii of the centres. Your last dr value is 1000 which means the last block will have a radial thickness of 1000 m. The outer radius of the grid will be the sum of the dr values which in your case is 5788.5 m at the moment.

Also, I'm not sure how many layers you were meaning to create, but if you just want 50 blocks (i.e. one layer) then your layer_thickness value should have just one element. You don't need to specify thicknesses for each column. At the moment you are creating a grid with 50 * 51 = 2550 blocks.