Discussion Forum

You should be able to specify the time steps in sequential sequences, doing the first 10 microseconds in (for example) steps of 0.1 microseconds, followed by 10 seconds in 0.1 second steps (for example). The details of how well this will work likely depend on the details of your model. But I know that I have created models that start with short time steps and later switch to longer time steps. (Note that you will almost certainly have to set the time steps manually, rather than let Comsol try to figure them out for you.) I suggest you choose the "manual" option for your time stepping, although I suppose that "strict" might also work.

I have done a similar thing to what Robert describes. Also note that you can apply a function to the step values. This makes it possible, for example, to have steps that depend exponentially on time, useful if you have some process that is roughly logarithmic in time.

Be sure to use Strict stepping.

You should be able to specify the time steps in sequential sequences, doing the first 10 microseconds in (for example) steps of 0.1 microseconds, followed by 10 seconds in 0.1 second steps (for example). The details of how well this will work likely depend on the details of your model. But I know that I have created models that start with short time steps and later switch to longer time steps. (Note that you will almost certainly have to set the time steps manually, rather than let Comsol try to figure them out for you.) I suggest you choose the "manual" option for your time stepping, although I suppose that "strict" might also work.

Thank you.But I do not konw how to choose the "manual" option for my time stepping, and Where to choose strict?

I have done a similar thing to what Robert describes. Also note that you can apply a function to the step values. This makes it possible, for example, to have steps that depend exponentially on time, useful if you have some process that is roughly logarithmic in time.

Be sure to use Strict stepping.

Thank you. My voltage is the AC signal, the sin function，V=50sin(2pi200t), and the viscosity of the fluid is the e index,0.05823+(1.27386-0.05823)*exp(-t[1/s]/23.022). Both functions contain time. The specific details of the time step are not very clear.

I've attached a picture. This example is unrelated to fluid dynamics, so do not assume any other settings shown apply in your case. Anyway, I encourage you to study Comsol's provided application examples and familiarize yourself with the best settings for various time-domain problems of relevance to you.

I've attached a picture. This example is unrelated to fluid dynamics, so do not assume any other settings shown apply in your case. Anyway, I encourage you to study Comsol's provided application examples and familiarize yourself with the best settings for various time-domain problems of relevance to you.

Thank you very much. Do you need to divide the research into two steps, the first steady state and the second transient state? My research questions: the first part is the transient change of the double electric layer, the second part is the velocity change of the fluid, both of which are related to time.