Questions regarding applications of physics/math to automotive eng

[1992camrysedan]

As you can tell from my username, I am OBSESSED with cars. This obsession is what drew me to apply to a university to study mechanical engineering. I had struggled in high school phyiscs, it wasn't until I got the hang of it in college in my statics and dynamics course that I started to really LOVE physics. Its amazing, the applications of it, and the power it holds in engineering.

At Colorado State University, I appreciate my classes for their rigor and teaching me everything I need to know, but sometimes professors, even really good professors, teach a lot of theory with complex equations and math without really showing its applications that much. It wasn't really until Thermodynamics that the professor, who was beyond brilliant, would tell us applications and even took us out to the local coal plant to see how thermodynamics is used. I finally could appreciate how some of the beautiful theory is applied to real engineering. My ordinary Differential Equations instructor also mentioned modeling internal combustion engines with it.

I feel like cars are very underestimated by the general public. My generalized impression is that many people are somewhat ignorant that cars are these marvels of modern engineering and there is a crazy amount of math and physics involved in making these. And even though they have downfalls, such as emissions, I love internal combustion engines. Not to mention with modern emission controls like EGR and Cold Start Catalysts, and advanced computer systems and tuning, cars are becoming amazingly environmentally friendly compared to their predecessors, although still not perfect.

But I wanted to ask if you guys could provide any specific applications of advanced physics of math being used in automotive engineering, and perhaps tell me if these applications bellow are infact correct? Thanks

-Vector/Multivariable Calculus in computational fluid mechanics/ aerodynamics for modeling flow such as Stoke's Theorem, Green's, Divergence, ETC as well as riemman sums

-Precise definition of a limit in component tolerances

-obviously Statics and Dynamics for chassis design, unibodies, gear systems such as teeth curvature in the differential and transmission, most classical mechanics applications

-Ordinary/Partial Differential Equations used to design camshafts and lobes

-Laplace Transforms for modern engine management systems and NVH (Noise, Vibration, Harshness) and cruise control/self driving technologies?

-EMF and Faraday's law in alternators

Can somebody more knowledgeable provide some other examples? I'd be very interested to know, thanks.

What's crazy is my university is supposed to be a very applied program relative to other schools and it still seems very theoretical at that. But I've only taken a few 300 level classes and the ones coming up should be much more applied.

 

[jlefevre76]

Well, in terms or aerodynamics and fluid flows, they are governed by multi variable equations, specifically PDE's in all three directions (x,y, and z). The Navier Stokes equations (with modifications for turbulent flow) in three dimensions can be solved using CFD methods. Since computing power has dramatically increased, CFD models of flow over the vehicle, for instance, are preferable to expensive wind tunnel testing or other such physical tests, as long as they are high fidelity.

System dynamics comes into play when looking at the interactions of components, as you mentioned. Anything from struts and suspension to transmission, drive shafts, etc. Their interactions are governed by system dynamics and kinematics.

Alternators are basically just electric motors running in reverse. They take mechanical energy and convert it to an electric potential (a current and a voltage). The fluxuation of the magnetic field through the copper coils is what induces the voltage/current.

I hope that answers at least some of your questions on how what you're learning applies to actual automotive systems. One thing to note, however, is that the problems you solve as you design or improve real systems will be far more complicated than the problems you solve with pencil and paper. Setting up a CFD simulation of an entire car can take dozens of hours (to accurately model and apply relevant conditions). And depending on the resolution of the meshed fluid space that you require, it can take several hours just to solve for the fluid flow (which ultimately tells you the drag force on the car, any vertical forces acting on the car, and any moments on the car). Basically, the demand for solving simple engineering problems just isn't there, because they've already been solved. However, you'll use those general principles to solve the more complex problems, like the one I mentioned above.