Hello everyone,

I am currently having trouble with an explicit simulation at a µm scale.

I have a very simple model (see attachment) which is a cube 5x5 µm² with a support at the bottom and and a displacement boundary condition at the top (smooth step).

I have chosen the following units:

Length: µm
Mass: ng
Time: µs

Consequently, the other units that follow are:
Stress: MPa
Density: ng/µm³ (=kg/m³ divided by 1e06)

So the values that I have used:
Time: 1e06 (=1s)
Density: 0.00114 (=1140kg/m³)
Youngs Modulus: 1000 (=1GPa)

Plastic: 40,0; 48,0.001; 52,0.002

If I want to apply the load over 1s (eventually I want 30-60s to be quasi static), so I need a time of 1e06. This seems to be too much as the following error occurs:

THE ANALYSIS MAY NEED A LARGE NUMBER OF INCREMENTS (MORE THAN 20,000,000), AND IT MAY BE AFFECTED BY ROUND-OFF ERRORS. FOR ACCURACY, RUNNING DOUBLE PRECISION EXECUTABLE IS REQUIRED

Now this is just a test and ultimately I want to work with larger geometries (200µm*200µm) and use the Gurson model - Porous metal (which requires and explicit calculation).

Even when I switch to ms the same message occurs. Now I could activate Double precision, but is this the best practice?

Your help and suggestions will be greatly appreciated

Kind regards

Mike