Right, it seems correct (sometimes it can be really confusing). Then if you use uniform distribution of the pressure, specify amplitude as Newtons divided by surface area in m^2 and it should be fine. If the results are still incorrect, this means that error is somewhere else.
Are you sure that you fixed all errors that I described yesterday ? I can share the input file if you want.
Which results did you compare with the article and what is the difference between them ?
Here are my ideas of what can cause the differences:
- there is no contact between embankment and vertical sides of one instance of „tie” part
- as I mentioned previously I changed contact between geotextile and ballast to tie constraint because it was causing problems, however it should be modeled as contact interaction if you want to have full agreement with article (so this contact should be resumed and checked for errors)
- what is the source for the dimensions of area on which pressure load is applied ? I haven’t seen it in the article (maybe I missed it though). Did you take it from other resource ?
- if I understand the article correctly, they use 1 Hz load (variable magnitude of 10 to 60 kN) on simplified model with steel plate and then proceed to analyze full model called parametric study with 16 Hz load (constant magnitude of 200 kN). Your ampltidue looks like a mix of the first and second case: 1Hz load but with constant magnitude of 10 kN (with some slight discrepancy due to surface area value). It should also include more cycles.
One more thing I would check: make sure that all contacts and ties work properly and refine the mesh a bit.
