I know this is a very broad/challenging question, however any responses would be helpful for my understanding.
So let say I want to model the shrinkage of a 3D object on cooling, how would computer simulations go about this? I have heard of software like ANSYS being used for such modelling, but how does it actually work under the hood?
Thank you!
In: Engineering
In your example, you start with a 3d mesh of the object, and model how the heat leaves the object. From there you can use the coefficient of thermal expansion of the material to tell you how much to shrink the distance between mesh vertices based on temperature.
For more complicated things, like injection molding, you also have to account for different areas solidifying at different times.
It probably doesn’t answer the specifics of your question, but generally all these engineering softwares do is use mathematics, based on the input parameters and show the results.
The main benefit is it will avoid any human errors of doing hand calculations (provided the input is correct), and it can display the results in an easier format to understand.
It uses millions of calculations based on user input and known data to generate a simulation. Lots of work goes into programming this type of software because it has to be incredibly powerful. The slightest error can cause a lot of problems for an engineer who is designing something. That’s why this software costs thousands of dollars.
I am a Mechanical engineer but it’s been a while since I dove deep on FEA so I may be slightly off.
Basically, with a known material, shape, size of a part/assembly, along with known external influences such as temp, gravity, other forces, etc, you can calculate the resulting displacements, forces, etc. These calculations can be done by hand with simple shapes like a bar. The tricky part is that to get a precise answer, you need to break down any complex part into tiny pieces with known volume, surface area, mass, etc. That is called “meshing”.
To get a good answer on any realistic part, it takes thousands or millions of these calculations. So the software does all the work of meshing and then all the calculations that are required once you’ve provided all the necessary variables.
We divide a big part into very many small parts. We understand how to write equations for these very small parts and also know that the boundary of a small part has to have the same result if it solved using the equations of the first small part or of the small part directly attached to it. We solve all these equations at the same time with a giant matrix so that all constraints and loads applied the bigger part are internally consistent with the equations of the small parts.
Latest Answers