Hacker News new | past | comments | ask | show | jobs | submit login
Animated Engines (animatedengines.com)
215 points by d4vlx on Jan 27, 2014 | hide | past | favorite | 34 comments



Which of the petrol powered engines scales best (those used in cars)? This is an open question to all. I'll take a swing at it.

I've worked with 2 strokes, 4 strokes (including Miller Cycle versions), and rotaries.

The two strokes hit a wall pretty quickly. They are very rev-happy, but their combustion leaves a lot to be desired. Most lack conventional valves and rely on ports to move the mixture around. You can supercharge them, but I've never been able to do so properly. Once the engine size starts to increase, their main advantage (RPM capabilities) start to diminish. Plus their combustion is not very clean, due to the lubrication being part of the fuel. They are fun when modified, but noisy. Ask my neighbor (who gets to enjoy the glorious sound of the modified 2 stroke in my weed whacker).

The 4 strokes suffer from harmonics, weight, and packaging issues. But they can be supercharged without much problems. You can get at least 100 HP/Liter on a 4 stroke engine without much issue (when you build it properly, of course). Building a 1,000 HP 5 liter engine can be done by buying off the shelf parts. The problem is that when harmonics start dancing to the devils fiddle, your valves start to drop, and your pistons shatter like the dreams of those who bought ENRON stock.

The rotary is my favorite racing engine. It doesn't drop valves (it lacks them). You can scale them by bolting on more rotors (12 rotor beast here: http://i436.photobucket.com/albums/qq83/Adam-Blackshaw/12rot...). They rev like crazy and feature few moving parts. But they are hard to tune properly, and tend to run a bit hot if not maintained correctly. Mazda built, raced, and won Lemans in 1991 with a 4-rotor equipped car called the 787B. Which sort of proves that they are viable engines. The real problem is emissions. The rotary has not received as much R&D as the 4-stroke, and thus it is a bit behind. Further developments by Mazda have improved it, but I don't see it lasting too much with the future CAFE standards looming over the horizon. It is also not very fuel efficient, due to it using ports rather than valves. Though I'm curious as to how a direct injected rotary might improve all of these issues. Not sure if I will ever see it happen.

Before you call me biased towards the rotary, let me tell you that I do not, or have ever owned one. I have worked with them, and got to enjoy it very much. My current favorite engine is the Porsche flat 6, because the damn things will take any boost you send their way.


> The two strokes hit a wall pretty quickly. They are very rev-happy

People think that, but in reality they only SOUND like they are rev happy.

I raced motocross for years (still ride dirtbikes, though mostly endurocross style riding now) and only recently did these bikes switch over to two strokes. It previously wasn't possible because four strokes didn't provide the needed power to weight ratio.

A 125-150cc two stroke competes against a 250cc four stroke in most clubs. When listening to the bikes go by, the 125cc two strokes sound like they are screaming, when in reality the 250 four strokes rev to 14000 RPM and the two strokes only hit 12000.

The two strokes just sound like they are revving 24000 because they fire every stroke, not every second stroke.

Two strokes are really fun, and I still ride one quite a bit (2008 GasGas EC200) but the four stroke power-to-weight ratio is getting so high I can't see them lasting. That and emissions.. they burn oil by design like you said, which is quite dirty.


The definition of rev-happy may have different meanings in different locales. In my locale, rev-happy means that they are very fast to rev, not that they have an RPM of X.

Not to the point at all, but I collect and restore old Elsinores. Lovely old hardware.


Fair point, but the two strokes still aren't more rev-happy using this definition either. Direct injected four strokes have nearly instantaneous throttle response, whereas because the air-fuel mixture in a two stroke has to flow through the reed valves, the crank case, and then up to the combustion chamber, they have a delay.

This is made worse by the un-clean burn inherent in the gas/oil mixture a two stroke burns.

A 125 might rev slightly faster from 6000-12000 than a 250 four stroke, but it definitely doesn't right from idle. And also note that the four stroke I'm comparing it to is TWICE the displacement. A high performance 125cc four stroke (like European cafe racers use) would certainly rev quicker.

Still.. I ride a two stroke all the time like I said. I grew up on them, so the sound, the vibration, and the powerband just feels like home. I'm sure you feel similarly which would explain why you restore old bikes (I love the Elsinores too, even though they are before my time). But it's impossible to say they are "better" than a four stroke in any quantifiable way other than raw power-to-weight ratio.


I never understood why Wankel engines aren't paired with continuously variable transmissions so that you can keep the engine at a constant rotational rate. Many of the issues with Wankels (gas mileage primarily, as well as emissions) are eliminated when tuned to a specific RPM.


The thing to remember about a CVT is that it's held together by a rubber band. Materials engineering has gotten very good at making stronger rubber bands that don't snap under hard use, but the application you have in mind sounds like an awful lot of wear and tear for a rubber band to undergo.

And, if the whole setup is tuned very specifically on the CVT working properly and the rubber band goes snap, that could be a bad thing.


CVT's in automobiles rely primarily upon metal chains for drive, not rubber bands.


They are sort of metallic rubber bands.


There's also these toroidal Centroid continuously variable transmissions. But they rely on friction and have small contact surfaces. http://auto.howstuffworks.com/cvt3.htm


CVTs have only improved in the last decade (and only so much). The real big push for the rotary started during the late 60's. Transmission technology back then was lacking. Plus you did not have effective electronic engine controls.


I think they have all had their day, the future is micro gas turbine generators[1] and electric motors. Look at the specs of the (admittedly vaporish) Jaguar C-X75[2].

1: http://www.bladonjets.com/technology/gas-turbines/

2: http://www.jaguar.com/about-jaguar/concept-cars/cx75.html


Ferrari, Porsche, and McLaren have high performance engines coupled to electric motors already. I think that technology will keep improving hybrid powertrains, but we won't likely see an electric car for the masses for one or two decades more. Due to the amount of politics that surround petroleum.

Now, what would like to see get more R&D is biofuel from a simple to produce source (like algae). Couple a small engine running biofuel to a generator and put a nice 200HP electric motor to the backwheels and we will have a nice every day appliance on wheels (what most people need/want).

What will keep happening is the reduction in complexity of the 4-stroke engine. We removed the hydraulic power steering pump in favor of an electric one, then comes the A/C compressor driven by an electric motor (some use it already). A bit later we will get an electric water pump. And then we will forgo camshafts and all related parts in exchange for electric actuators. Which will turn the intake manifold into a sort of thing in the past. Imagine having a 12-cylinder engine that's smaller than a 4 cylinder, but makes 3 times the power with the same efficiency. Now imagine a 4 cylinder engine that is about the size of a small box making upwards of 300HP, while getting more than 40MPG. It is very close.


The strength of Mazda rotary engines, high horsepower at very high RPMs, doesn't translate all that well to normal driving behavior compared to 4-strokes, and the low torque makes them less fun to those who just want to blast out a quick 0-70.


Well, people buy Hondas. They pack high-rev low torque engines. They just use a higher final ratio and a lower final overdrive gear to compensate. After all, its just (torque X RPM) / 5252.


Interesting question! I think it's two-strokes that scale best. The company Wärtsilä builds 14 cylinder diesel two-strokes that generate 80,080 kW of power. They are used in cargo ships.

EDIT: Oh, "used in cars". I misunderderstood the question.


I did not know of those engines. Very interesting. That does win the question, given how you just can out scale a cargo ship. :)


A two stroke marine diesel is very different from a gasoline two stroke. Only intake is done through the cylinder wall ports and exhaust is through poppet valves at the top. (Uniflow) Also lubrication is done like in four strokes. There is no crankcase compression, instead turbochargers are used.


As someone who has been to the 12 Hours of Sebring many times... The biggest problem with the Wankel is that it sounds awful. It is positively pain inducing to listen to those Mazdas go by for hours on end. :-)


Have you ever heard any in drag racing? The annoying buzzer sound is turned into a glorious shriek.


> (12 rotor beast here: http://i436.photobucket.com/albums/qq83/Adam-Blackshaw/12rot...).

WTF... how many HP here?


I only linked to that image to prove a point. It seems that 12 rotor monster is simply someone having fun. However, I have seen 3, 4, and 6 rotor engines with my own eyes. I have also see the tail lights of a turbo 3 rotor (a story for another day). The rotary is simply the engine that scales closest to how you would scale a computer system. It even has ports. :)


4-strokes inline 6 engines are well balanced harmonically.


I love this, as well as the sister-site, 507 Mechanical Movements [1].

I'm surprised at how laborious the process is [2]. Is there not a way to automate this in a CAD program and export it (e.g. SolidWorks [3]); or is that limited to video, not a more interactive simulation? Are there any open source CAD programs that would support generation of such simulations?

[1] http://507movements.com/

[2] http://www.animatedengines.com/howto.html

[3] http://www.solidworks.com/sw/products/3d-cad/cad-animation.h...


507 Mechanical Movements has animations with canvas that seem to have a higher frame rate than the animated GIFs on this site. However, all those diagrams are something that SVG would be a perfect fit for and you can actually animate SVG in a more natural way by saying »This object should rotate« or »this object should move there« instead of drawing every frame by hand (with the occasional pixel errors). Plus, it'd be vector graphics.

(Edit: Just read the explanation how they do it, and notice that even the CAD program doesn't make it any easier, if they have to manipulate everything by hand, but at least the SVG process would be pretty much identical while allowing for arbitrary framerates and higher quality output.)


I was rather disappointed to see that they were animated gifs. I would much rather see them in a format where you can run them with independent speeds and framerates.

I figure it would be better to represent the image as a bunch of shapes with changing position/rotation. In fact using these animations as a base It might not be too hard to automate the process to identify the shapes and the connections between them.


At least on the site liked in the OP they are sprite maps* saved in gif format not animated gifs.

For example: http://www.animatedengines.com/jets.html

The custom player takes the class of the div container (http://www.animatedengines.com/img/tprop_fl.gif) "aeplayer aepframes_6" to determine the number of frames (6) and offsets the background-position vertically for each frame by the amount of the individual frame height which equals the div container height, thus appearing animated.

*see this example with a sprite map in png format http://jsfiddle.net/simurai/CGmCe/


Given that the site has been around since the year 2000, assuming the images haven't changed, the tools widely available / understood at that were rather limited when compared with today.


Sketchup has animation built in and is trivially scriptable in Ruby. Interactive controls are created in HTML/JS. It's not open source, but it doesn't cost anything for the regular version.


Fantastic. However, it needs the Napier Deltic engine ...

https://upload.wikimedia.org/wikipedia/commons/thumb/e/e4/Na...


They should include the Tesla Turbine. Only one moving part.


No 2-stroke diesel?

(They make the most awesome sounds, IMHO. Especially the Detroits.)


The rocket engine section could be made more interesting by going into engines with turbopumps and gas generators.


Am I the only one who sees the images slowly floating upwards?


This is a wonderful education tool. Nice work.




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

Search: