4 ms·
I'm still not convinced about the product life. We have 20 year old ICE vehicles running absolutely perfect (well-maintained and designed ones), is the same pos
by ameen 11y ago
I'm still not convinced about the product life. We have 20 year old ICE vehicles running absolutely perfect (well-maintained and designed ones), is the same possible with an electric car? Also are the byproducts of manufacturing an electric worse than an ICE vehicle. We need to think about the environment as a whole. I'm still convinced electric mass transportation would be better than investing in individual carmakers.
- thaumasiotes 11y ago> We need to think about the environment as a whole. Actually, I'm not so sure about this. It's easy to see the case for "a little more pollution, but far away" over "a little less pollution, constantly hanging over your head".
- toomuchtodo 11y ago> I'm still not convinced about the product life. We have 20 year old ICE vehicles running absolutely perfect (well-maintained and designed ones), is the same possible with an electric car? An electric car has no more than 50 moving parts. For the powertrain, you'd only ever need to swap the battery pack or the electric motor (weighs about 70lbs). Electric vehicles by their nature embody longevity. > Also are the byproducts of manufacturing an electric worse than an ICE vehicle. No. > We need to think about the environment as a whole. Yes. Electric vehicles are only one part of the mobility equation. > I'm still convinced electric mass transportation would be better than investing in individual carmakers. You're going to cover every American suburb with electric mass transit? If so, the cheapest way to do so is with electrified self-driving vehicles.
- chongli 11y agoElectric motors are much simpler and easier to maintain than an ICE. There's no reason you couldn't keep an EV running optimally for 20 years, assuming you replace the battery at regular intervals.
- vlehto 11y agoSorry to nitpick, but it's not about "simplicity". It's mostly about lubrication and temperature gradients. Wear and fatigue have nothing to do with out brains ability to comprehend the system. But it's usually easier to design reliable machine that is somehow simple. So it appears to layman as if "reliability = low number of moving parts". But it's bit like saying "lots of megahertz makes good computer".
- chongli 11y agoAll other things being equal, a CPU that runs at a higher clock rate is faster. We're also talking about machines that are maintained by human mechanics throughout their lifespans, not Mars rovers. If a machine is simpler in design, it ought to be easier for a mechanic to replace a part, no?
- vlehto 11y agoIn real life all things are seldom equal. In many cases less moving parts is more reliable. Thats if the parts we're there to give you more features. But sometimes they are there to give you more reliability. My favorite counter example is car timing belt. Replace it with chain and you add about 200 moving parts, while increasing reliability and maintainability. Now add roller sleeves around the pins of the chain, you doubled the amount of moving parts and again increased the reliability and reduced friction. The thing with chains is that they typically start to make noise long before failure, saving you from complete engine overhaul. Belt might just snap. (Though for noncritical applications opposite is true. Belt drive requires no lubrication and they are silent and light weight. Id love to replace messy bike chains with belt drive.)
- chongli 11y agoIn real life all things are seldom equal. In many cases less moving parts is more reliable. Thats if the parts we're there to give you more features. But sometimes they are there to give you more reliability. You state this as if it were a dichotomy: that parts only provide features or reliability. That's not true. In many cases, parts exist only because they are essential to a particular design. A fundamentally different design may have fewer parts, better features, and better reliability all at once.