Hello everyone!
I conduct a series of triaxial tests for granulation with the distort domain method. when I put a measure sphere with a radius equal to 15 times of mean diameter(large enough). I find there’s an inconsistency between the volumetric strain rate and volumetric strain.
(1). As shown below(the x-axial of the curve is axial-strain while the y-axial is void ratio), as the loose sample, the granular contract during the shear process, but the measure volumetric strain rate (equal to the sum of the XX, YY, and ZZ components of the strain rate) is might not appropriate namely the value is sometimes positive(dilation) and sometimes negative(contraction).
Any suggestion will be much appreciated
should I accumulate this quantity in a short period and re-averaged to obtained the transient value?
Hello @LJYpyu, I am not sure what your question is. Could you clarify what exactly you want to know about the strain rate from measurement sphere’s?
Thanks for your reply first, what I mean is that in the triaxial test the void ratio is decreasing while the volumetric strain-rate obtained from the measurement sphere shows the dilation.
It’s difficult to guess what is going on. I would need to see how your history in the plots above are being calculated.
Thanks for your patience! The following content is my test results and the corresponding script has been added.
Export the strain_rate_xx, strain_rate_yy, and strain_rate_zz from the plot and then calculate the volumetric strain_rate((xx+yy+zz)/3.0) as shown below. There’s no observable dilatation/contraction trend indicated by the plot and the averaged value(Fig1) while the void ratio evolution(Fig2) illustrates the distinct contraction characteristics.
samplemake.dat (2.8 KB)
triaxial_test.dat (3.0 KB)
main_function.dat (288 Bytes)
Excuse me, sir. Is this question beyond the scope? It doesn’t matter. any comment/suggestion is OK
Hello @LJYpyu,
It may be best to send this to PFCsupport@itascacg.com. However, I have a hunch you just need to accumulate the quantity like you suggested. But, I’ll need to look at your data file to be sure, which take time.
yeah, thank you sir, there’s an addition observation I want to discuss with you. The following Fig is obtained from the triaxial tests with wall-servo and domain-servo respectively. As we can see, the results from ‘domain-servo’ case will shows a weird velocity field compared to the ‘wall-servo’ case. I don’t know the reason but I think it might be related to the strange strain-rate reported above.
The ball velocity vectors you show when comparing domain-servo versus wall-servo will be different. I don’t thank this is not related the issues above. These two boundary conditions are very different from each other.
yes, it is. so I think the domain distortion has an significant impact on the velocity of the particles and in that case if the velocity field is corrected the shear rate will be corrected consequently and every thing will be fine. until now, I still believe the domain-servo is available there’s just some mistake on the velocity derivation.