elevation from contour lines

Hi everybody, I am a recent user of FEFLOW. Actually my company is thinking about buying FEFLOW but first we would like to explore the software potentialities and check whether it meets our needs. This is why I am becoming familiary with the software through the DEMO version.
I am struggling with the elevation assignment to the different layers of my model.
I have a dxf map representing the contour lines of let's say, layer number 2.
I would like to use the contour lines to assign the elevation, either assigning the elevation to the elements between the lines or assigning the elevation value directly to the contour lines. The latter method presents the problem that some parts between two lines could not get the elevation value. I know that this depends on the snap distance but if I set the snap distance too high I noticed that some elevation values "invade" areas which belong to other values of elevation.

Does anyone have a suggestion about that?

Do you think it is better to use a different format map rather than a dxf file?

I figured that I can transform the contour lines in a set of points and then creating a simple triplet-text with the coordinates of the point and the elevation. Then I would use the regionalization for interpolating the values. But I think there still is the problem that some nodes will not get a values, for example the nodes really far from contour lines.

I appreciate any help.

Thank you.

I would convert the lines to points and then interpolate. Any interpolation method will assign values all over the domain. The problem is simply that the farther the nodes from your data, the more unreliable will be the result.
Bye,

Giovanni

Thanks. I have already tried that with different interpolation methods but as you said I got many artifacts due to the lack of data in some areas.

In situations like this, I use to add reasonably invented (I know it doesn't sound very scientific!) data, consistent with the geological information I have, to constrain the interpolation. There's also the possibility to make different datasets and check the sensitivity of the model to the unavailable data.

Giovanni

I tried that but again the output is characterized by many artifacts! Thanks anyway.
Actually I came up with a method.
I convert my dxf file with the contour lines into a shape file with polygon features. Then in ArcGis I attribute to each polygon (representing an area between two lines) an elevation.

Then I go to FEFLOW and use the parameter association tool for associating that elevation to the supermesh.

It seems an elegant way to work around the problem but still I don't know whether there are better and faster methods. I hope so because converting the contour lines into a polygon feature could be quite tricky.

I'd support Giovanni's method, breaking up the lines into points and interpolating between the points. If necessary, additional lines will have to be added in advance. With the polygon option - if I understand your approach correctly - you will get zones with spatially constant values, which is probably not in accordance with elevation in reality (unless we think about man-made structures), where elevation will typically transition smoothly between the contour lines.

Thanks a lot Peter, but which software would you suggest to me for undertaking the interpolation?
I have a really complex contour shape. Please see the attachment.
I tried to use Surfer (with all its interpolation methods) but although I add points or other lines I always got artifacts. This I think because the contours are really close each others and have irregular shapes.

Anyway, you think that in order to implement the information on FEFLOW I should create a shape file (with point feature in this case) and then linking the parameters of that file to the supermesh by using the parameter association tool?