Hacker News new | past | comments | ask | show | jobs | submit login
How a Car Engine Works (animagraffs.com)
404 points by kercker on April 9, 2016 | hide | past | favorite | 123 comments



I'm changing my career slightly from full stack development to repairing and restoring old tractors and farming. I have a 1952 Case SC, a 1957 Fordson Major, and a 1946 Ford 2N in various conditions. I'm building a shop to work on them. Basically its an excuse to buy a sand blaster and paint room. Buy them for $500 and sell them for $3500 restored, pay for the tools and the next project tractor. Hopefully picking up a 1970 Ford 3600 this week. Old Diesels fascinate me.


I'm halfway through a similar transition to farming. It's funny how sometimes a new task can be an excuse to buy more tools. I'd be interested in reading more from you about your new work!


What's your educational background ? (I have fantasies about doing that but I am concerned I'll just end up on a repair line in a factory instead of in a tractor shop).


Out of curiosity, how many hours of work typically goes into one of those restorations?


If you were here in Ontario I'd definitely buy one of your tractors.


I learned how they worked by taking one apart, putting it back together, and got it to run (it was a junkyard car). My auto shop teacher told me he was amazed, because I was a nerd and nerds had no mechanical aptitude.


Did you take a class in high school? Wish there was a place for adults to learn without a full fledged college course or certificate program.


I did take auto shop, but didn't learn anything there, and I don't think any of the other students did, either. I learned how cars worked in my dad's driveway. This was much to the chagrin of my mother, who hated having partially disassembled wrecks in the driveway and oil stains everywhere. I passed by the house many years later, and the signs were still visible :-)

Get a car (mine arrived as a pile of junk on a flatbed), some tools, and start turning wrenches. It's not that hard, and I (and a couple friends) had a heluva good time. I suspect an account of that helped me get into engineering college.

My only regret is I ran across a '68 Chevy SS for $600 that I passed on.


You can just learn by doing. It's no different than a self taught programmer.

I learned entirely by wrenching on my own car. Engine internals and transmissions are hard simply because you often need specialized tools. But there are plenty of parts you can fix yourself with only simple tools.

The main challenge now is that newer cars have a lot more complex components. Back in the '70s you could completely disassemble and refurbish a carburator. Fuel injection systems are a lot more complex. However, a lot of the systems on newer cars are still pretty basic.

One of the benefits of knowing how to work on cars is saving money. I fixed my wifes car (emissions failure) with a $12 part from ebay when a garage wanted $150 just to diagnose. And if you do have a garage work on your car, explaining what's wrong in way that makes you sound knowledgeable goes a long way to prevent being ripped off.


I've saved a ton of money doing the simple repairs myself.


Fuel injectors are also quite dangerous. They operate at sufficient pressure to break the skin, and high-pressure fuel/oil injections can require amputations.


I managed to learn it all on my own; I rebuilt an engine too, but it wasn't some junkyard car's, it was my wife's car that she had gotten caught in a flash-flood and seized the engine on (see: "hydrolock"). Worked great for another 90k miles after that and was still running great when she sold it.

All you need is the ability to read, the desire to learn, the courage to get your hands dirty and not be afraid of breaking things, and some mechanical aptitude. And a little money to spend at Harbor Freight for some tools.


What happened? A family member's insurance totaled a car due to this recently. We guessed he bent rods. What did you find?


One of the rods was bent, and the valves in that cylinder were also bent because the piston had struck them. I still have that bent rod/piston in a box somewhere as a souvenir: the bend was pretty significant.

What's weird is my wife actually managed to get the car restarted after a while, and drove home that way!! It was making a terrible tapping noise though. I'm not sure how it was running that way; I didn't investigate too closely. This was a 2000 Acura Integra GS-R, by the way.

Anyway, I left everything alone in the other 3 cylinders because that all looked fine, and I replaced the piston, rod, and journal bearings on the bent one. I also replaced all 4 valves. It wasn't really that hard; I took the head and new valves and valve seals to a shop that specialized in rebuilding cylinder heads and they cleaned it up (it already had close to 100k miles at that point, so there was a lot of carbon build-up) and replaced the valves. They also put together the piston and con-rod for me as I didn't have a press for that. Then I just did the standard bore honing and put it all back together (and cleaned up the domes of the other pistons while I was at it). Worked great after that. The total cost wasn't much either; the bare parts only came out to a few hundred IIRC, from one of those online OEM parts sellers, and the shop work was $250 IIRC. The only specialized tools I needed were torque wrenches (which you should have anyway, but they're absolutely critical for engine work), and an air impact wrench (the crankshaft pulley bolt on that car was a real bear), and a drill-operated bore hone (cheap).


To get suspension nuts off once I had to get a 3 foot breaker bar and put a 2 foot pipe on it to extend it. Finally cracked that sucker loose!

The manual showed a picture of a guy using a foot long breaker bar. What a joke.


When the car's brand-new, you probably can use a foot-long breaker bar. Over time, though, they can get seized.

I have a breaker bar, but I haven't used it ever since I got an impact wrench. You can put a lot more torque on a bolt with a handheld impact wrench than you can with a giant breaker bar; they're really a quite remarkable invention when you think about it. The one I have (1/2in drive Ingersoll-Rand titanium body) has 1000 ft-lb IIRC.


Slight correction: the valves were bent because of the water pressure most likely, not the piston hitting them.


Get a VW Bug or an ATV with an air-cooled, four-stroke engine. Buy a good torque-wrench, a socket set, some allen wrenches, and a proper disassembly/reassembly manual and get to work!

I have removed every rolling part of my ATV. It's quite strange to go 65MPH knowing that you put all the parts together... I am still alive so far, ha. :)


Last few cars I've owned seem to be designed to prevent the owner from even looking under the hood. They don't seem to design them for easy user access anymore. But then again, I came across a computer from ASUS the other day that said you void the warranty if you open up your case. WTF???


On the plus side, you have far more to work with these days than a Haynes manual. You'll have Youtube videos, discussion forum posts on every job, 'borrow a tool' programs from stores/libraries, a range of parts-suppliers competing for your dollars that will ship to your door...


Yeah... Haynes or Chilton. Or, my preferred avenue, Factory Service Manual (the price is like $60 vs $100, ex. my Jeep Cherokee).


Indeed. After looking in the right places I was able to get the 300-page shop manual for my motorcycle. Everything is very well diagramed and OEM parts are readily available online for not much money.


In my limited experience, OEM is horrible, unless it is the only option.


This is funny because there is a brand called "OEM" at parts stores that sells chinese junk.

I am all about finding the original parts that the dealer will sell, but on the aftermarket at reasonable prices.


I meant to say: "horrible" from a value perspective. You can often get >= "OEM quality" parts from 3rd-party vendors at a much lower price than OEM.


It has gotten easier for DIY auto mechanics. I spent a year in a community college auto repair program. I took just the classes, I felt I needed. I didn't want to work as a mechanic, but open a shop. Well, that never took place. I am glad I took those classes though.

What still troubles me is passing my bi-annual smog checks. It's not just my vechicle, it's family members vechicles.

I can check most of the smog components, except the catalytic converter.(I found away around checking the catalytic converter, but won't divulge, because it might be illegial.)

The one tool I might buy if the price was reasonable is a device that measures the emmissions out of the tailpipe. I think there's a big need for portable emmission testers. The price point would need to be around $300 for my taste. I don't know if it's even possible at that price point? I'm basically just concerned with HC's.

In order to pass emmission standards, I make sure the engine is running reasonably well. I change the oil. I check all ERG. I make sure that cat is hot. I then check the voltage off one of the O2 sensors(gotta pick the right on, and be quick with the voltage measurement. The computer will throw things off pretty quick with the wrong O2 sensor dissabled.) With the 0-5 volts you read you can get a good idea of the stoichiometric burn of the gasses. You can detect wether the vechicle is running rich, or lean, but it's not fool proof.

Even with preparation, it's hit it miss whether a vechicle passes smog.

I would love to gave a home devise that measured these gasses.

I'll pass this along. All Smog shops are in CA are required to have onsite one Emission manual to show the customer.

Good shops have two references. They look at this information when they do the visual examination. Good shops usuall have Motor Emmissions for the current year, and a subscription to Mitchell Manuals. A subscription to Mitchell manuals(OnDemand5) is more money than the Motor Emission publication, but it's hardly ever Wrong.

If you fail the visual on a smog test, ask to see the refrence material they used to fail you. If it's a Motor Publication--the information might be wrong. It's filled with many errors. Most shops cheap it out and only buying the cheaper Emission manual. The Motor Emission manual is joked about among Smog Techs. They know it's filled with errors.

I kinda went on, but frustrated from dealing with a recent smog check.


At least on European cars you can read the O2 sensor voltages over the OBD2 port. $30 bluetooth dongle from ebay and you get the readings graphed on your phone. I just did that yesterday on my '02 Peugeot 307.

When you say the "right" O2 sensor I presume you're talking about the one after the cat? Doesn't that mainly tell you whether the cat is working or not? If the engine is running a bit rich and there is a bit of excess fuel coming into the cat, the cat should burn it up, I think.

And you probably know this, but there is an important difference in the voltage readings depending on whether you have wide- or narrow-band O2 sensors. Wideband is typically for performance cars.


I can recommend a odb2 dongle as well. Very interesting to plug in every once in a while. Do mind most dongles are read only though. You need a good cable to write. Be carefull if you want to tweak de ECU! Theres lotsnof info on the internet on how to tweak stuff.

$30 is way too expensive btw. On aliexpress i bought a wifi (because of iphone) dongle for $9,99 incl shipment.


They're designed to not look scary or confusing when the prospective buyer looks at the engine bay and knows nothing about cars. Extreme example: https://iedei.files.wordpress.com/2011/04/2010-lexus-ls460-s...

Plus it just looks more like an Apple product and less like their mom's shitty Camry they grew up being to school in.


I don't really get the point of having a "second bonnet" - there is already one covering the engine and everything else.

Lexus has been known to make things deliberately difficult to reach, like putting the starter motor in a place where getting to it involves disassembling most of what's on top of the engine (requiring the replacement of lots of auxillary "soft" parts like gaskets and o-rings in the process):

http://www.lextreme.com/starter.html

In comparison, most other makes and models mount the starter motor somewhere on the side of the engine or transmission where getting to it doesn't require disassembling much else.


I doubt their goal here was to make repair more difficult -- after all, some must have failed under warrantee and were repaired on their dime. Sticking the starter in the valley instead of on the side just makes the engine package size a bit smaller (and the SC was a small car).


Holy cow this is terrible. Especially given toyota starters 'click of death' when their brushes get worn.


Every failing starter I have encountered exhibited the "click of death" (turn the key and get "clicks" rather than a turning engine). I doubt it's a problem exclusive to Toyota.

PS - I drove a manual transmission Mazda truck that had no starter for 8 months. Thankfully, I live in a mountainous region, so "push starting" the vehicle by using a hill was "easy" (except when it wasn't).


You rock!


> Last few cars I've owned seem to be designed to prevent the owner from even looking under the hood.

I found I can't even rotate the tires on my current car (Sonata) because it's got no central jack point. Seems to be designed just for shops that can lift all 4 jack points at once. :-(


They are designed to look sexy in the dealership.

Most engine covers pop off without even using any tools.


Plus one for the air-cooled VW. We just bought a 1981 VW Westfalia, which has the same 2.0 liter engine as the Porsche 914 I owned thirty years ago. It's like coming home again. Dead simple, and the instructions for dropping the engine don't cover even two pages. I've always considered the Bosch fuel injection of VWs and Audis, no matter what version, to be problematic and a general pain in the ass, but it mostly works most of the time and is the only potential sticky part for a newbie.

Parts are still readily available, though anything that moves under its own power is likely to sell for more than it did new (which wasn't much in today's dollars). But you'll get smiles and waves anywhere you go whether you end up in a Bug or a Bus.


Absolutely this. The VW is the MOS 6502 of engines.



If you have a car it probably needs some type of work, somewhere. The big cost is in tools, but just start simple and work your way up to something like an engine rebuild. Even doing simple but time consuming tasks like replacing leaking gaskets or hoses teaches you a lot and saves a ton of money.


That's a wrong attitude for a shop teacher - nerdy types are usually very good at mechanical work.

You're right in that the best way to learn is to do.


Maybe science nerds tend to be good at mechanical work, but surely it's not the case for language nerds, history nerds or any of the many other possible areas of nerdiness.


I found a middle ground in engine geometry http://mechanicaldesign.asmedigitalcollection.asme.org/artic...


Interesting. How can I learn more about this stuff?


No idea, I just search from anything related to 'closed' engines https://en.wikipedia.org/wiki/Wankel_engine


There are a lot of tech nerds in motorsports.


How do you make a small fortune in motorsport racing? Start with a large one.


don't tell gaben.


This was in 1974 or so. Things were different then.


The animations are nice, but it wouldn't hurt to step up the level of detail: there's no mention of ECU, MAF sensor, lambda sensors, injector pulse duration, throttle body, crankshaft sensor, flywheel, turbocharging, connection to accelerator pedal, etc..

But the most annoying detail is showing a distributor on a fuel injected engine. Come on, this isn't 1986, at least show us a dual coil pack wasted spark system.


That's nice and all, but until one understands the four stroke cycle, the rest of what you mention is noise. Suck, squeeze, bang, blow: until you know that, knowing what a lambda sensor is just uses up space that could be used learning Rust instead.

I mean, get real, anyone capable of understanding injector dwell is probably not clicking that link. It's not for the likes of you and me. What's next, go leave comments on the "Idiot's Guide to How Computers Work" about how it's not complete if they don't explain pointers?


My biggest gripe is not including the link to the accelerator pedal. If you don't know how pushing the pedal makes the engine spin faster, do you even have a basic understanding of how engines work?


Yeah, I'll give you that one. I think it does open a can of complexity, though, what with explaining intake vacuum and all. OTOH, it would solve the mystery for those that know what the accelerator cable attaches to, but have no idea how opening a plate makes "vroom vroom" noises.


This.


I think understanding the principles of each system here is key. For example, if you understand how a motor works, it's easy to see you could either have a mechanical contraption with cams responsible for timing and controlling fuel injection or simply a computer-controlled system with some electrically actuated valve. The benefit of the computer controlled system is of course you have more flexibility and have the ability to integrate data from various sensors to achieve some kind of "optimal" injection volume. Most systems in a car are easy to generalize once you understand the fundamentals.

It's also a good way to work because since it's so easy to grasp those advanced concepts from the basics, you don't have to waste precious resources trying to illustrate everything. A wikipedia page, or a journal article description will be enough, which is just quick and cheap text.


>I think understanding the principles of each system here is key. For example, if you understand how a motor works, it's easy to see you could either have a mechanical contraption with cams responsible for timing and controlling fuel injection or simply a computer-controlled system with some electrically actuated valve.

Are you trying to teach theory or real-world mechanics?

No engine in any car in the world has electrically actuated valves, except for a handful of prototype test mules such as one by Koenigsegg in Sweden. Everyone's still using "mechanical contraptions with cams", so it's pretty important to teach that, because any engine you look at will be made that way, and the design of the camshafts (and any support components, such as a variable-vale-timing mechanism) is critical to the engine performance.

>The benefit of the computer controlled system is of course you have more flexibility and have the ability to integrate data from various sensors to achieve some kind of "optimal" injection volume.

Yeah, that's great, but you might as well be talking about Moon colonies. It's sci-fi at this point. It may come in the near future, it may not; we might end up just skipping it in favor of EVs.

However, distributors by now are truly obsolete. I don't think any cars have them any more; they've all gone to coil-on-plug systems. Distributors aren't quite that old; there were still cars being made with them about 15 years ago, but those were the last holdouts. So I don't see the point in showing the distributor as is done on this site: it's simpler and easier to just show the spark plugs by themselves firing at the appropriate time, perhaps with some wires shown connected to the ECU.

It's a bit odd too, because aside from the distributor, this guy's illustrations seem to show a fairly modern engine, though he's omitted the variable-valve timing system that most modern cars have now. Notice that the fuel injector is squirting fuel directly into the cylinder: that's called "GDI", or gasoline direct injection. It's become pretty common these days (I think some Kias and Hyundais have been holding out on it), but go back just 5 years and not that many cars had it, and go back 10 years and it was pretty much none. Of course, like distributorless ignition, GDI is simpler in concept, though technically more difficult to accomplish reliably (the fuel pressure is much higher for one thing), as the old way had injectors shooting fuel into the intake plenum before it got sucked into the engine.

My only other complaint about this site is that they left out the V-12 configuration. There's still plenty of Ferraris that have those.


> No engine in any car in the world has electrically actuated valves

It was just an example, I don't actually know what I'm talking about :) But it was my understanding that electronic fuel injection involves solenoids to, well, control fuel injection (as seen here https://en.wikipedia.org/wiki/Fuel_injection#EFI_gasoline_en... ).

The point is that conveying the operating principles should be the main goal, if you call that "theory" or "practical info", it doesn't matter -- if you don't understand the basics well the practical (or theoretical) details will escape you.

Your suggestion about getting rid of distributors sounds reasonable.


I'm pretty sure he meant that no car has electrically actuated intake and exhaust valves, which is mostly true, but the Fiat/Chrysler MultiAir (https://en.wikipedia.org/wiki/MultiAir) line of engines does have electro-hydraulic valve actuation which allows for variable duration and lift, and I think they don't even have a throttle plate.


I'm having certain difficulties with the use of 'plenty' in relation to Ferraris.


There's probably as many Ferraris sold per year as, say, a Chevy Volt. They're not super-high-volume cars by any means, but they're not as rare as a Maybach or even a Rolls.


This is a basic presentation so people who have a limited understanding of what's under the bonnet can get an idea. Basic means not adding extraneous details regarding items that are considered 'secondary', for e.g turbocharging. Fuel, oil, air, exhaust, cooling.. these things have been the building blocks of car engines for the past 100 years and that wont change.

I swear some people just cant appreciate simplicity.


I agree to some extent that simplicity is nice, I'm just saying I think the level of simplicity is inconsistent: including some sensors but not others, having an explanation of hybrids but nothing of turbos, etc. And that not including the link between the accelerator pedal and the engine leaves a pretty big hole in the layman's understanding.


But thats exactly my point, you cant include everything unless you want to turn a basic how-to into an instruction manual. Cover the fundamentals of an engine and give little interesting facts like that regarding hybrids. I think your really having a hard time grasping what basic is when you mention things like 'they covered this, but not that', yes thats cause its BASIC.


Sure, but again: "how accelerator pedal tells engine to make higher vroom noise" is a very very basic part of understanding an engine.

Omitting it leads to a big hole in the reader's understanding, since the accelerator pedal is likely their only method of interacting with an engine.


Read the 'about me' page before criticizing please.


Seems to support the criticism...

"I’m obsessed with solid, hiqh-quality research and design presentation. If you’re gonna do something you might as well do it RIGHT. Right?"


The timing chain and the camshaft seems really inefficient; I'm looking forward to when camless designs become more mainstream. Koenigsegg (among others, I assume) has been working on this tech [0][1].

0: http://blog.caranddriver.com/koenigseggs-camshaft-less-engin...

1: https://youtu.be/Bch5B23_pu0


It is really inefficient. A lot of energy is wasted compressing the valve springs to open the valves. This is a problem if you try to make a high-revving engine, for instance, in order to get better fuel efficiency at low speeds and good power at high rpms: to allow high-rpm operation, you need stiffer valve springs, but this significantly affects your fuel economy.

They've been working on camless valve actuation for decades. I read an article about a prototype system back in 1992. The article then said that the power requirements of the solenoids was a big problem. I'm not sure what the problem is these days. As you noted, Koenigsegg has a working prototype in an actual car (not on a test stand, connected to mains power, like the one in the article in 1992), and it seems to work fine. There's probably some kind of problems with reliability.


> It is really inefficient. A lot of energy is wasted compressing the valve springs to open the valves.

This isn't as bad as you might think, because of the nature of springs. When the valve closes again, most of the energy spent compressing the spring is returned to the engine!


In the motorcycle world Ducati has been shipping desmodromic[0] valve trains for years. Not camless, but springless.

[0] https://en.wikipedia.org/wiki/Desmodromic_valve


Yeah, and they're the only ones using it. As your link notes, there are significant disadvantages to that system (as well as advantages), namely complexity, greater moving valvetrain mass (you need two cam lobes per valve), and inability to use hydraulic valve lash adjusters (which means you have to periodically adjust them; not a problem on a racing or high-performance motorcycle, but definitely a problem for a passenger car where these days you want to be able to go 100k miles without any service other than oil changes and tire rotations).


This is a problem if you try to make a high-revving engine, for instance, in order to get better fuel efficiency at low speeds and good power at high rpms: to allow high-rpm operation, you need stiffer valve springs, but this significantly affects your fuel economy.

The alternative is to just not design for high RPM; the most efficient internal combustion engines are huge low-speed two-stroke diesels used for ships and stationary generators.


Large marine diesels have very low RPM. Some of them red-line at 102 RPM, and normally operate much less than that. A typical passenger cars idles at 1200RPM and redlines around 6-7K RPM.

Then again a single piston in the marine engine weighs 2x of an entire car so they can easily do things that one wouldn't bother to do on a smaller engine.


>A typical passenger cars idles at 1200RPM

No, they don't. 750rpm is pretty typical these days. They'll idle at higher speeds when cold, though, to warm up faster.


Huge engines aren't feasible in cars. More mass to carry around means lower fuel efficiency. Stationary plants don't care much about mass, and with ships it doesn't have that much of an effect.


I'd wager that the advancements (since 1992) in batteries and electric motors would solve many of the power problems you referred to.

Yes, reliability is probably the issue. However, the same could have been said about EFI back in the day.

Full camless engine design is something we may only see in high-end racing and supercar applications for while yet.


Well, power electronics have really come a long way since 1992, and that might make a big difference. For instance back in 1992, DC-DC converters were shit, most power supplies were still linear instead of switching, and the electronics needed for, for instance, driving a 3-phase brushless motor were big and expensive. These days, that stuff has gotten much smaller and cheaper. We now have power MOSFETs able to handle huge currents with ridiculously low losses, in rather tiny packages. A chip the size of your fingernail can handle over 100A with losses in the single milliohms, and the main problem becomes handling the waste heat (with 100 amps, even 4 milliohms means 40W!); usually these things are used for high-frequency switched applications, so the total on-time isn't that much.

Back in 1992, making an electric car like the Tesla probably wouldn't have worked out too well. Instead of an electronically driven induction motor (or a 3-phase brushless motor as some other vehicles probably use, also electronically commutated), they would have had to use a brushed DC motor which isn't so great for longevity or efficiency, because the electronics needed for the better motors would have been too expensive, bulky, and inefficient. Now, DC brushed motors are all but obsolete except for small, cheap applications like <$50 power tools.

The same might have been an issue with those camless valves: actuating those solenoids precisely probably required some serious power electronics capable of supplying and switching high currents at very high speeds. Modern power electronics are good at that now.

I wonder if one problem isn't control of the valve speed. If the solenoid is moving too fast, that'll wear the valve seat much faster (in a regular cam engine, the valves don't shut that quickly at low rpms because the valve profile prevents it). In a cam-less engine, the solenoids would effectively be going full speed all the time, meaning more valve seat wear from slamming shut so hard. This makes tuning engines easier perhaps (more digital-like operation: on/off), but increased wear is bad for something you want to last for hundreds of thousands of miles with no service. So they might be looking for a way to control the actuation speed of the valves, and vary it with engine speed as cams do now. That means even more complex power electronics.


BMW have had VANOS for at least a decade.


As well as Valvetronic. They're also doing research into camless engines, since at least the late 90's / early 2000's: http://www.nytimes.com/2003/08/21/technology/circuits/21next...


That's a partial solution, yes. It still requires cams.

I was referring to full-on solenoid control of valves without needing cams at all.


Actually two decades now.


The Cartoon Network version.[1] Not bad, actually.

And, inevitably, the Jam Handy / Chevrolet film on engines.[2] There's a whole series of these, with ones for suspensions, transmissions, differentials, lubrication, and frames.

[1] https://www.youtube.com/watch?v=IQj00NTFXew

[2] https://www.youtube.com/watch?v=fKbSloJ5o7o


Those videos aged really, really well (the 1930s ones)


If you like designing engines and cars then look at this game http://store.steampowered.com/app/293760


What software do they use to make these animations? Does anyone know?


Blender, Flash professional, Photoshop and Illustrator. https://www.behance.net/gallery/10749619/How-A-Car-Engine-Wo...




Most of my guy friends really like cars. They know good amounts about different parts of the cars.

I've never enjoyed reading about them. So, I've never tried learning much about them. It also takes alot of time.

I wish there were more of these type of animated illustration. I've scanned it for a few minutes and I grasped the concept it was trying to show me relatively faster than anything I've ever encountered.


Theres a whole site full of them if you click the title.


I have a question: exactly what causes the piston to come back after the intake/power stroke? Let me write what I guess I have understood -

1. The intake stroke piston movement may be considered to be caused by gravity and/or atmospheric pressure

2. The power stroke piston movement may be considered to be caused by gravity and/or atmospheric pressure and/or the expansion due to explosion all taken together

Correct me if I am wrong.

But what baffles me is this: what causes the piston movement during the compression stroke and exhaust stroke? Any expert here to enlighten me on this?

edit: typo


If you can find a pull-start gas engine lawn mower, you can learn lots from it. These things generally have a single piston. The pull string that you use to start it gets it moving in the right direction, and once it starts running on its own, it's momentum that brings it back around for another go.

Also keep in mind that this single piston engine is two-stroke and not four as in the article.

Back to four stroke engines: Momentum and the firing of other pistons keeps the whole thing cycling back around. The more pistons you have, the more there are firing between each revolution of a given piston (and thus overcoming resistance and supplementing that momentum) and thus the more power available to drive the vehicle.


This is a great explanation. I'd add one small clarification:

> Back to four stroke engines: Momentum and the firing of other pistons keeps the whole thing cycling back around

The firing of other pistons helps, but really, momentum alone is sufficient. There are tons of single-cylinder, 4 stroke motorcycle, scooter, and off-highway engines out there.


Yup, two stroke model engines from cars/planes are also great for learning. Pull start, cheap, easy to disassemble, and you'll need to learn about tuning too - which is another whole load of fun!


The kinetic energy of the flywheel keeps things moving when burning gases aren't moving the pistons. The evidence of this is demonstrated on the two flavors of the motorcycle I own. One version (BMW R1200GS Adventure) has a heavier flywheel than the other (non-Adventure). I've ridden both. With the light flywheel it's easier to kill it pulling from a dead stop because there isn't as much mass to overcome the lack of subtlety in my clutch hand. On the Adventure version that I own, sloppy clutch work that often comes with trying to get a 600 lb. "dirt bike" out of a mud hole is ignored by the more massive flywheel. Even when the cylinders are barely able to fire because I'm one sniff of clutch away from killing the engine, the bigger flywheel will keep things turning for a wee bit longer before I kill it and fall over in the mud.

As for the answers parallel to mine that say it's the mass of the other pistons, the mass of a piston is trivial in comparison to that of the flywheel. Even lawn mowers mentioned in one example have flywheels (at least the ones I owned always did). The mass of a tiny lawn mower piston is going to have a hard time against the relatively large mass of a lawn mower blade with no flywheel.

To speak to your two points: 1. My motorcycle's pistons lie on their sides, gravity has no practical effect.

2. Power stroke movement is caused by the expansion of burning gases, period. All other factors, such as flywheel inertia, contribute relatively minimally.


Momentum from the flywheel/flexplate keeps it spinning


The flexplate has very little mass to provide momentum. The fact that the flexplate is bolted to the (heavier) torque converter is what's really going to provide the momentum.


None of the intake, compression, power, or exhaust strokes have anything to do with gravity or atmospheric pressure.

Initially, a secondary motor causes the rotating components to rotate, which causes the pistons to move in the cylinder. Later, either inertia or the power stroke from another cylinder maintains the rotation.

The piston movement itself is caused by the piston arm, which is offset from the axis of the crank shaft.


There's usually more than 1 piston on the crankshaft. So while one is coming up, the power for that is coming from another going down


Not an expert, but I always assumed that it was caused by the rotational inertia of the Flywheel[1].

[1] https://en.wikipedia.org/wiki/Flywheel


Wow! What a straightforwardly cool and enriching post. I've had a vague idea of how engines work (as well as turbochargers) but somehow I've never bothered to get it straight.


Does anyone know what software was used to render the animations? His interview with Adobe suggests that he makes 3D models in Blender, but I'm not sure about the renderer.


How a gun works is also an amazing piece of artwork: http://animagraffs.com/how-a-handgun-works-1911-45/


Single piston engine technology looks pretty promising as it has proven to be 30% more efficient than 4 piston combustible engine.

https://youtu.be/JoQkTIfAB2U


The video you link to doesn't describe a single piston engine, it describes a 2-stroke opposed piston engine. The animation shows 6 pistons in 3 cylinders.

I think we'll move fairly smoothly from ICEs to electric vehicles without significant commercialization of alternate designs.


The one in your video appears to be a 2-stroke opposed-piston diesel, and that was tried several decades ago with not much success:

https://en.wikipedia.org/wiki/Commer_TS3

Here's an animation of one:

https://www.youtube.com/watch?v=pGaISFg_ZIw

There's videos of these running on YouTube, they have a very distinctive sound like other 2-stroke diesels.


Doesn't TFA say that opposed pistons were used in the classic VW bug ("Flat Four"), suggesting there's a good reason they aren't used elsewhere?


As the other poster said, they have advantages and disadvantages. Subarus and Posches use them (though in the Porsches, they're flat-6 engines, not flat-4).

The big advantage is center of gravity: their shape lets them sit basically flat, near the ground, and since engines are the heaviest part of a car, this gives the car a lower center of gravity, which is good for handling.

The main disadvantage is mechanical complexity: instead of all the cylinders lined up in a row, with the crankshaft on one side and the valves and dual cams on the other, you have two banks of pistons, each with their own camshafts. So now instead of two camshafts, two camshaft gears, and one timing belt/chain, you need double all those. This of course increases cost too.

Flat-fours also tend to have rather distinctive exhaust notes, which not everyone finds pleasing.

Another notable place flat-4 engines are used is in small aircraft. Lycoming and Continental engines for small (e.g. Cessna) airplanes and helicopters are all flat-4.


Porsche now makes a flat 4 engine again- the new Boxster/Cayman engines are turbo 4.

On the cost issue, Porsche famously tried to cut costs for the 996/Boxster gen 1 by using a common cylinder head casting for both sides. This was done by flipping it over for the other side. The only problem with this was providing the drive for the camshaft from the rear of the engine. To do this, they created an intermediate shaft operating at the rear of the engine. This worked ok but it couldn't have an oil fed bearing, so they used a roller bearing. This caused a failure rate if something like 5% of engines, and caused a class action suit that Porsche had to settle.

The other advantage of boxer engines is their suitability for air cooling and packaging. In fact, Porsche produced an aircraft version of the flat 6 but it ended up being too exprensive to build and run, despite outperforming a lycoming by a long way.


The grandparent isn't talking about quite the same thing. The Beetle's engine is called a Boxer engine, and you can find modern cars using it (mostly Porsches and Subarus). They have advantages and disadvantages, just like V shaped and inline engines. The differences in geometry don't have much of an effect in fuel efficiency though.


Just FYI, you linked to a 3 piston engine (opposed pistons, which is a neat idea I hadn't heard of before), but think you may have dropped the wrong link.


I like the animation on this site: http://auto.howstuffworks.com/engine1.htm


It may be karma-bait, but it's very nice looking karma-bait.

OTOH, no inline six, which is the platonic ideal of the ICE.

As such, I can't up vote this.


The world is flat :P

Although flat 6 is missing too...


No rotary either!


My previous car was a Mazda RX-8, and my current car is a Nissan LEAF. I feel under-represented.


Watercooling on an enthusiast/gaming PC and a car's radiator + waterpump definitely has its similarities.


When I click and drag the mouse, it doesn't rotate or zoom the model. Am I using it wrong? ;)


Yes.


Suck, squish, bang, blow.


Brilliant post;


Second that. We should thank Jake for all those amazing Animagraffs on his site. Those are a lot of work to produce.


It was only worth a few points each time its submitted. If you're going to post something a few years old, then put the date in the title.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: