{{ post.title }}
글 편집
글 편집 (이전 에디터)
{{ post.author.name }}
Posted on
| Version | {{ post.target_version }} | Product |
{{ product.name }}
|
|---|---|---|---|
| Tutorial/Manual | {{ post.manual_title }} | Attached File | {{ post.file.upload_filename }} |
CAE (Computer Aided Engineering) usually consists of 3 steps, Pre-Solve-Post.
Solver is in charge of 'Solve' process.
There are many parameters for the solver.
It is very useful to understand what they are to use the solver more efficiently and accurately.
Even if there are various parameters, in this article, Maximum Time Step will be explained in detail which is very useful.
Maximum Time Step
Purpose: It controls the maximum size of each time step of the solver. The smaller value is used, the more accurate the result would be. (But the calculation time may get longer.)
Recommended Value: 0.001 (= 1.e-3). The optimal value must be decided depending on the time scale of the behavior of the target model. If you are interested in the behavior varies at a level of 0.0001 sec, the maximum Time Step must be smaller than 0.0001.
Description
RecurDyn solver uses the 'variable step size'. It means when End Time = 2 sec, steps = 100 are used, the step size is not a fixed value, 0.02 (=2/100). Solver checks the appropriate step size according to the status of the calculation, which makes simulation speed fast.
The Maximum Time Step can be compared to the speed limit of a car.
In other words, the default value of the Maximum Time Step, 0.01 means, the step size of the solver cannot exceed 0.01.
So, if the Maximum Time Step is set to 0.01, even if the solver can use 0.013 or 0.02, it uses the step size equal to or less than 0.01
Even if the default value of Maximum Time Step is 0.01, 0.001 is recommended to the users who want more accurate result.
Of course, the optimal value can vary depending on the model.
If you perform a very high-frequency model, then you should use very small Maximum Time Step.
The Step size of the solver is displayed during simulation on the Message window as below.
This information shows which time instant is solved and what the current step size is.
- There is a parameter called 'Initial Time Step'. (The default is 0.000001), This value determines the size of the first step.
- Usually, the small enough value is used for the initial value.
Tips
- If the Step Size is not decreased, and is similar to maximum time step, it means the model is being solved without a big problem. It can be compared to the car which runs at a fast and seady speed on a highway..
- On the other hands, if the step size is decreased suddenly, it means there is some big change of the status of the system, such as contact or big reaction force. It can be compared to the car which runs on a winding road with the frequent use of the break. Unless the big change of the step size was expected in advance, the model needs to be investigated carefully. In this kind of case, it is recommended to 'Pause the simulation', and review the animation or plot if there is any unexpected problem in the model.
- If you want to decrease Maximum time step for the higher accuracy, you can refer to the step size on the message window. For example, even though you used 0.01 as maximum time step, if 0.00001 is used in most of the period, it is no use changing it to 0.001 or 0.0001. (because 0.00001 is already used internally.) so you should use the smaller value than 0.00001.
For the other parameters, please refer to the below articles.
- Solver – Becoming an Experienced User #1, First Episode: Learn about Maximum Time Step!
- Solver – Becoming an Experienced User #2 – Dynamic/Kinematic Analysis Parameters (General)
- Solver – Becoming an Experienced User #3 – Dynamic/Kinematic Analysis Parameters (Parameters)
- Solver – Becoming an Experienced User #4 – Simulation Parameter #1
- Solver – Becoming an Experienced User #5 – Simulation Parameter #2
