Hacker News Reader

Comment by sixQuarks | Hacker News Reader

Another reason why Tesla is so far ahead of the game. Elon Musk had the engineers design the vehicle to be as safe as possible, period.

They didn't design the safety of the car with the crash tests in mind. They passed all the crash tests with flying colors as a side-effect of actually designing for safety.

View on HN

A lot of people think that airliner safety comes from the FAA hammering on Boeing. That's not true at all. It's the attitude of Boeing and Boeing engineers that make it safe. (I used to be a Boeing engineer on flight critical systems.) The worst nightmare of a Boeing engineer is an accident resulting from his design mistake.

The worst for me was when the Alaska 717 went down because the stabilizer jackscrew failed. I worked on the 757 jackscrew. It was a while before I found out it wasn't the 757 design.

Safety came from the engineers who knew what they were doing and would refuse to put their name on anything that they weren't sure about.

I also know of instances where Boeing produced safer designs and had to lobby the FAA to get them to accept them. One example is using twin engines on overwater flights, rather than 3 or 4. The twin engines are statistically safer.

I know of cases where Boeing designs were much safer than that required by the FAA, and the FAA later adopted the Boeing designs as a requirement.

I've actually felt a lot safer flying after my time at Boeing and knowing the nuts and bolts of how they were made and the dedication and competence of the engineers in charge.

> I also know of instances where Boeing produced safer designs and had to lobby the FAA to get them to accept them. One example is using twin engines on overwater flights, rather than 3 or 4. The twin engines are statistically safer.

Interesting. Why do you think it is that twin engines are statistically safer? I thought redundancy was a normal part of the safety toolkit for planes.

I was told that more engines mean more possible failures, and that outweighed the advantage of redundancy. It's counterintuitive, but when you work the numbers for engines, it's true. It took Boeing a few years to convince the FAA.

I think it's even true for single vs. twin engine planes. A lot of it has to do with - are the systems redundant, or are they both needed to properly fly the plane. Also whether the probability adds up in a serial or parallel fashion. I know a friend that pilots a single engine Cessna. Obviously one engine, but it also has no electronic ignition and four magnetos. https://en.wikipedia.org/wiki/Ignition_magneto The engine operates without the use of a battery because of this. I think the Cessna has a battery for auxiliary systems, but by removing the battery, spark module, and electronic timing/distributor from the critical path of engine operation, the chances of failure are lower. The four magnetos are for the four cylinders in the engine, but any one magneto can run the entire engine. In this case, the four items are redundant.

I think the first twin engine commercial airliner to be certified for over-water flights was the 777. It can take off, fly 6000 miles, and land safely with one engine and a full load.

I'm interested in how these redundancies have accumulated over the life of building aircraft. They're obviously for good reason. However the recent news pieces comparing electric aircraft, I wonder if there's a lot of efficiencies on the table for conventional aircraft if they could just drop in a modern engine like you might find in a race car.

That's a good point. Gasoline engines have come a long way since the days of magneto operation, with variable timing, ignition, and direct injection (of which some piston driven aircraft may already use). Turboprops and turbofans (jets) are already very efficient compared to automotive piston engine aircraft. A 747 for instance gets around 91 mpg/person while buses average around 200 mpg/person - but, the drag forces on an aircraft are I think at least an order of magnitude greater.

Minor point: it's two magnetos, each magneto driving a separate ignition harness and spark plug in each cylinder.

Each system is tested in the pre-flight by grounding out (shutting down) the left magneto and making sure the right magneto can run the engine and all those plugs are firing, then repeating the test with the right magneto grounded (shut down) and making sure the left mag, harness, and plugs are capable of running the engine.

If you're further trying to wrap your head around it an example would be: https://en.wikipedia.org/wiki/United_Airlines_Flight_232

An engine failure manages to disable all hydraulic links and redundancies in the aircraft.

The FAA hammers on pilots and airlines too though, which might otherwise try to cut corners to save costs on safety.

I am very impressed with US Aerospace engineering. Having read a book or two containing personal accounts, it makes me feel "proud" of American engineering, even though I'm not American.

I've written software for 9-1-1 and the first few months in operation, I barely slept and would often awake terrified. I can only imagine how it must be for someone that might actually actively kill many people due to an engineering flaw.

Tesla Model S is interesting example actually. Initially it was imported into EU with something of a waiver from full ECE regs due to few number. Some of the earliest ones in EU have 8 airbags just as in US (larger dual stage front airbags), then more with ECE front (dual layer) airbags. By MY15 all had only six airbags, the knee bolsters were removed. The reason for this is that in US models with dual stage airbags the knee bolsters deploy in unison with the seatbelt pretensioners to keep the front occupants positioned safely when the front airbag deploys an since they are dual stage units it is very important that happen. The Model S is a big enough car that with the smaller less forceful ECE airbags it is not so much of an issue if the occupant slides under a bit and has abdomen closer to airbag as it deploys. This was likely done for cost as it did not provide enough ROI in terms of safety in EU markets. There is a digit in the VIN in fact that codifies these differences. In addition there are a number of other safety systems that differ between US and EU models of the S.

Tesla did make a very safe car, but they did not do so blind to varying regulations in different markets as well as the demands or running a business profitably.

When you remove the huge chunk of metal in front of the cabin that is the engine, you might expect better performance in a front impact test. The structure of the car doesn't have to circumvent the engine, gearbox and a lot of other complicated and heavy bits that have to sit in a specific place.

None of the Tesla cars are as "safe as possible". Trivial proof: Reducing the max speed to 1kph would make it safer. Just like after 11-Sept the USG said they will prevent another attack "at all costs". Why not shut down all flights permanently then?

Tesla might design cars to a very high safety standard, but it's certainly a tradeoff with performance, style, comfort, etc.

Apologies in advance if this was pointlessly pedantic.

Reply on news.ycombinator.com