Time increment in Dynamic Analysis

[Gianrebus]

Hi,
I'm a new user of Abaqus, I'm making some dynamic analysis with Abaqus-Explicit in an adaptive domain with ALE technique, I have two Eulerian boundary regions with material inflowing and outflowing.
My question regards the choice of the time increment for integration in this kind of analysis. Is it possible that if I don't input manually the maximum time increment (I leave blank the space after the line *Dynamic), so I leave Abaqus to change it during analysis, I obtain a different result from the case when I input it manually choosing a lower value (to respect the condition of stability discussed in the manual)? I specify that I don't use mass scaling.
Hope you can help.

 

[JSDaniel]

In theory, the same answer is obtained for any time increment lower than the stable time increment. In practice, anything is possible, but you shouldn't ever need to specify a maximum time increment. Abaqus has a very well-developed algorithm for computing the necessary increment which is quite a bit more involved than what's given in the manual. If you suspect the computed time increment is too large for a particular problem, it's also possible to ask for a more conservative estimate by setting "IMPROVED DT METHOD=NO" in the parameter line (for 3D continuum and plane stress elements), or you can use "SCALE FACTOR = X" with X < 1 to scale down the default time increment by a fixed amount.

If it does turn out that you are getting a bad estimate on the time increment for that particular application, I suspect that it's something Simulia would want to hear about.