Flow Simulation questions

Some of these questions can be answered by checking the help documentation in the software.

1. Travel time: this is a certain unit of time that is defined by your mesh density and initial conditions, check the documentation.

2. There is no way to know exactly how long a simulation will take to converge, the value the software gives you is a best estimate based on current convergence trends of your stated goals.

3. For domain size, consider how you would test this model in a wind tunnel. How far do the walls need to be in order not to interfere with your model? Treat the software like a virtual experiment, and take the same considerations and cares as you would performing a real experiment.

4. Global goals will be calculated based on global values, that is to say the min/max/average value of every cell in the domain. Surface goals are based on integral values calculated on a specified surface in your model.

5. This is a more in-depth question that depends on the specifics of your simulation, maybe open a new thread for this.

6. CPU value means how many processing threads the simulation will make use of. The default is the maximum detected by the program. I'm guessing you either have 8 cores or 4 hyper-threaded cores on your motherboard.

1. One travel in unit of iterations is the average number of iterations for a fluid particle from the entry point (inlet) to the exit point (outlet) of the computational domain. The largest average  is taken if more than one inlet/outlet.

2. See Amit's answer.

3. Smaller domain size for external flows, below certain limits which are model-geometry and Re dependent,  normally result in larger computational time to converge (if at all the model converged). This is especially true for the location of the outlet which needs to be far enough so that the pressure is almost constant across the entire outlet. Some experimentations are needed if there is no knowledge of the wind tunnel size. The result will slightly vary for domains smaller than needed.

4. See Amit's answer

5. The choice of the convergence goals depends on which result values are  of the highest interest for the Analysis- namely those values (or their parent values if the values of interest are not in the goal lists or cannot be created as equation goals). Forces convergence is the slowest, then comes pressure , then velocities (not considering the turbulence intensity and dissipation - dissipation convergence is also slow).

6. See Amit's answer. Please note that most of the CPUs perform slower with hyper-threading ON regardless that the hyper-threading double the number of the processors. Also, if two parallel runs are launched use the Affinity (Task Manager/Performance tab/right click on the Solver Process) to distribute the different Solvers on separate Processors (in the case of Dual or Quad Processors). This will speed the computations by eliminating the time for inter-processor communications. I wish Flow Simulation does it automatically but it does not.

Firas Helou wrote:

 

hello guys,
i am new to solidworks flow simulation but not new to fluid mechanics, i started learning the flow simulation using the SolidProfessor package about flow simulation
so i would like to ask some questions about the information window we see while running the calculation
1 - there is an line that says : "iterations per 1 travel : 193" what does this mean exactly ? is it that it needs 193 iterations to complete the simulation ?
2 - how to know the exact solving time it needs to finish ? because the time keeps jumping which i think is due to change in geometry surface

3 - i am simulating an F1 model of 1.4 meter in length and 0.5m in width which i placed it on a rectangle that represents the track, so i defined a domain around that is not much big and in contact with the road but not englobing the track rectangle, so my question is : Defining a less computational domain would result in less time to solve the calculations ?

4 - i did not understand the difference between "Global Goals" and "Surface Goals", can anyone please explain it ?

5 - i chose the surface goals because from a search on youtube and google i found that it has been chosen for cars simulations, so i chose the lift force or "force in y" and the drag force or "force in z" in solidworks so the results i will get are for the average force on all the car ? if i want to calculate the Drag coefficient using the formula of the drag force which i got in the simulation is that possible taking the highest force for example ?

6 - what does CPU value 8 means ? i keept it as the default value which was 8 and i chose accuracy of 1 the very low because if not it will take me ages to finish the calculation and now the main goal for me is to show that i understand the formulas i am using that is all

 

thank you everyone in advance