Impact Loading Analysis in Abaqus Very Different From Analytical Solution

[lemslems]

Hi everyone. I am new to Abaqus. I am currently modeling a simple impact loading analysis and I am comparing an analytical solution to the Abaqus simulation. However, it seems like not only that the solution I got from Abaqus is wrong, it differs by several magnitudes. I have checked my units over and over and over again to make sure my units are consistent (in my case I am using mm, everything else is in SI units) but no matter how much I tweak it, nothing seems to change.

It is also worth noting that the analytical model assumes all energy from the impactor is transferred onto the beam and sticks with it (perfectly inelastic). As far as I have read, there is no way for me to simulate that in Abaqus, so that might be an error point. However, I hypothesised that the beam deformation resulting from the impact would be less than the analytical one, since less energy is transferred onto the beam. This was not what I obtained in the simulation unfortunately. The deformation I got in the simulation is more than two-folds bigger than the calculated solution.

Quick comparison of solutions:
1. max deformation:
Analytical: -1.831mm
Abaqus: -36.704mm

2. max tensile stress (S33 in Abaqus):
Analytical: 41029.9 kg/mms
Abaqus: 4249.4 kg/mms

In the attached files I have included my hand working for the setup, the Excel calculation result, and the .CAE file for comparison. Please point me in any direction I got wrong, it is very much appreciated. Any and all help is deeply appreciated as I need to finish this ASAP as part of my coursework. Thank you.

 

[FEA way]

The density unit in Abaqus is wrong. It should be tonne/mm^3 instead of kg/mm^3 so e-9 instead of e-6.

 

[lemslems]

Hi FEA Way,

Thank you for your reply! But no, that was not the issue. I stayed consistent with kg and that should still work because Abaqus does not have any default mass units. As long as your units are consistent, they should compute to the same answer. I figured out the issue shortly after posting this but I forgot to update it. The issue was from my Young’s Modulus. I previously converted from Pa to N/mm^2, whereas N itself is kgm/s^2 which, in obvious sense, relates to length. The actual correct full conversion of Pa from m to mm is 1e-3 kg/mms. After I made this fix, the results I obtained align with my hypothesis.

 

[FEA way]

Indeed, any consistent units are sufficient. But with kg instead of tonnes, the other units become much less confusing. We definitely prefer (and advise) standard unit systems like SI(m) and SI(mm) with mass in tonnes and stress in MPa (those are also suggested in Abaqus documentation) to avoid issues. As you can see, it's really easy to make a mistake with such unusual units.