When will "evolved structures" make their way into automotive engineering and design?

[Mark S.]

My inner gearhead is fascinated by NASA's foray into "generative design," a process involving software-simulated evolution. A design engineer specificizes a broad outline for a given part—dimensions, attachment system, etc.—then tells the software what kind of stresses and from what directions the part must withstand. The software dreams up something, then stress tests it against the defined parameters. Portions of a part that aren't strong enough get beefed up, while portions that are stronger than required are pared down. Eventually you get something that looks organic, much like the skeletal structures you might find in living creatures that evolved over eons to adapt to a given environment. Truly astounding.

Sponsored

 

[Meanderthal]

The systems I have seen demonstrated also give you a good number of variations that meet the input criteria. So you can pick one that maybe makes more sense or looks better. I have seen examples that included designing a chair. It's interesting to see what the systems come up with. You are pretty much stuck with a 3D printing process to make any the generative designs. That is possible for very low volume products but I don't know when it will actually be a better part and be less expensive.

 

[Mark S.]

. You are pretty much stuck with a 3D printing process to make any the generative designs. That is possible for very low volume products but I don't know when it will actually be a better part and be less expensive.

I figured that was the limiting factor for automotive use.

 

[patmat2350]

The three most important characteristics an automotive designer must pay attention to are: Cost, cost, and cost. Oh, and can we make it a little cheaper? But also on the list are functionality and package (it has to fit, and by the way we're taking away some of the space you thought you had).

So the parts have to be makeable with the typical high volume production processes: plastic molding, metal stamping, metal casting. Meaning requirements for draft angles, casting wall thickness, minimal use of slides in injection molding tools, bending radii, etc etc, all have to be adhered to.

And you can be sure that within those restrictions, every part on a modern car has been subjected to intensive computer analysis, for strength, fatigue resistance, air flow, weight reduction, etc etc.

Now sometimes there may be big picture issues that have been missed while we were focusing on the details. Example, Jim Farley's admission that the Mustang Mach-E has a mile too much copper wire, compared to a Tesla. This might come down to decisions such as using discrete control wires to components instead of a control bus, or placing a battery too far from the motors it's powering.

It'll get better though!

 

[wiyeti]

Organic designs like that are definitely gorgeous to look at. I remember the old Britten V1000 motorcycle when it first appeared. The exhaust reminded me of something living not mechanical.

 

[Alfa 2407]

I am interested to see where this heads as well, especially with the recent introduction of metal 3d printers, though my only concern, how will we tell the machine, hey, this part needs to be easily replaced for maintenance, filters and such for gas powered cars, tires and various sensors in those electric ones

 

[patmat2350]

Ah, but your concerns are primarily for production cars, and metal 3d printing will never be a high volume production method. Just ain't happening, too slow and energy intensive.
That said, Ford opened a big research center for 3d printing in the past few years. I think it's partly to impress shareholders, and partly on Fear of Missing Out. I wouldn't expect the processes to be used for anything more than personalized custom order trinkets though.
Don't get me wrong though, I love 3d printing, have 2 printers at home, and have used the 3dp services while at Ford for plastic and metal prototype parts. It's just not a high volume production tool.

 

[Alfa 2407]

Ah, but your concerns are primarily for production cars, and metal 3d printing will never be a high volume production method. Just ain't happening, too slow and energy intensive.

I agree, with what you say in part, yes they are slow and energy intensive, but remember, this is new tech, relatively speaking, and like with everything, the more demand for them, the more money will be invested in researching and further developing these processes, this is true with anything and everything, from the concrete foundation of your home, the phone or computer you are using, to the welder used to assemble your car, welders 20-30 years ago often pulled more than 2x the power modern welders draw, and achieved arguably worse or less reliable results

 

[Bucko]

Gee, I thought they (Ford engineers) used this when they created the BS.

 

[patmat2350]

Nah, they just draped a sheet over two cardboard boxes and said "That's it!"

Sponsored