[varsha solanki]

Here are a few straightforward and candid thoughts based on experience on things to consider as you move forward with CFD and CFD consulting. I hope some of them can help you avoid bumps along the way.

1. Software Capability:
If your application involves more complex physics such as multiphase flows, dynamic fluid-body interactions, compressible flows, electrodynamics, and combustion, you’ll definitely want to be looking into a higher-end product such as STAR-CCM+ or ANSYS Fluent, or the range of software packages that fall somewhere in between.

2. Workflow & Ease-of-Use:
A smooth workflow makes it easy and efficient to move from idea to 3D concept to simulations and to presentation of results. Remember, your time is always worth more than you think it is, so don’t get scared off by a software’s sticker price if it means hundreds of hours of time saved each year.


3. Quality:
Your software is your gateway to high quality. The old saying, “you are only as good as your tools” is nowhere more applicable than when talking about computational simulation.

[LaminarFlow]

And the most important: Understand your problem to avoid errors. The valuable content is not the program, but the idea how to solve and at best reduce the physical problem. If you have not before I would start first with the basics of simple fluid mechanics (hydrostatic, aerostatic, bernoulli, viscous/friction). Afterwards understanding tensor calculus (e.g. notation, summation convention, Kronecker-Delta, Levi-Civita...) and the Navier-Stokes equations. As soon as you are familiar with that you can start to understand discretization with FDM/FVM and numerical solving (damping, stability, convergence, computational cost). Also learn which boundary and inital conditions are necessery for certain types of simulation. If your problem is more complex (e.g. compressible, multiphase) you should also look at specific literature. For all of this tasks I can highly recommend to do some tutorials/examples and repeat/applicate this. First with analytical calculations by hand (simple fluid mechanics) and afterwards programming your own simulation of easy problems can help to go further.

If you know almost nothing of the points above, you should invest some time to learn this, before you work at such a problem. Simulating without knowledge is dangerous and will cost your company much more money than hiring a person with appropriate knowledge. Also look at the short advices here: How to learn CFD (For Absolute Beginners)

If you know the most things of the points above. You can really start with specific literature and working on CFD. In companies working on commercial code is definitly the way to go. You get faster results because you don't have to implement everything by yourself (your time is expensive) and also it is easier to work in groups. If you leave one day it is easier to reuse your results if you worked with commercial code because reading another person's code is not always easy. On the other side: For complex problems, cooperation with a university or service provider can be useful. In this case programming can be better than using commercial code since you have more degrees of freedom.