You asked for dorito, I gave you dorito. Education is good for you! Keep exploring at brilliant.org/AngeTheGreat/. Get started for free, and hurry-the first 200 people get 20% off an annual premium subscription.
huh, was gonna shit on brilliant for having expensive af annual plans only that are disguised as monthly but now they've seemed to introduced real monthly deals still fucking expensive though
I never though I would get shills from a simulation of a an engine. Its really good. Not like games that use fake "wav"s, they never made me chill in my spine, because its too fake and no dynamic response.
I feel like I'm witnessing the audio version of the millennium graphics leap in 3D racing. This dude's engine simulator is mind boggling and could take racing simulation to a whole 'nother level!
Hiring would be bad overall for the industry. Some car game license his software for their next game would be best. So that every new car game made has the change for the best sounding engines.
@@benperkins2929 some genius would be able to optimize it to run on low end CPU for the challenge of it. A company would just tell you to buy a better rig.
The true, ideal housing is an epitrochoid, actually pretty easy to calculate. There's math for the inner and outer envelopes (Wankel and Liquid Piston respectively) too, I have a Desmos setup for all of them.
The housing is calculated using the exact function (with the rotor there for visualization purposes). For the rotor, I thought it would be cool to show how the shape is optimized but I'm sure there's a closed-form solution for it.
@@DJ_Level_3 Amazing, that's a very pleasing design to look at. Has that type of rotary engine ever been made? You say it's the ideal geometry, has it been tried?
My dad would have absolutely loved to see these videos. He was a gearhead who spent years working R&D at some of the big houses, Edelbrock, NOS, and others. He engineered some revolutionary new designs for the metering blocks in carburetors, and being able to iterate on designs with today's technology would have blown his mind. Thanks for sharing these!
Add the gearbox whine to this, and then it'll be perfect. That transmission stuttering is amazing btw, at first i thought it was a bit too much but it fits perfectly (for a race engine at least).
My Rx8 reacts almost exactly like that with firm suspension and large but thin tires Well... there's also a reinforced gearbox, ultra light counterweight/flywheel combo and stage 3 clutch sooo... Might play a part too lmao
@@FuSiionCraft yeah flywheel is a big contributor to tackle stutter. having too lightweight flywheels can actually be contraproductive even kill the engine
dude the adjustments to the chassis and drivetrain flexibility in the vehicle simulation make an AMAZING difference in the end product! The engine simulator by itself is an absolute work of art, but it sounds mostly like an engine sitting on a dyno. Now it sounds like you're in the car! Bravo!
The power of finding a niche, from cereal box game to engine simulator. I'm really loving the series and the details you take in account. As someone who knows only little about engines, I didn't expect to hear any difference during the last demo, but I actually did hear it clearly. Very cool details.
While as a dorito engine enthusiast I'm excited to see and hear the Wankel engine finally being simulated correctly, I actually screamed and did the DiCaprio pointing meme at the moment you explained how the rotor and housing design were optimized using an algorithm akin to a CNC machine. That's an extremely smart solution to model the reulaeux rotors and the epitrochoid housing, and no doubt Mazda engineers also used similar techniques in their research to build the 16X range extender. Mazda also worked on different cavity designs for the rotors to optimize combustion and it turns out that by having no cavity at all meant the highest compression ratio of 16:1 can be achieved. I know you've heard this so many times but you're so cool and smart for investing into this project! Can't wait to finally mod this and test out hypothetical designs including Rolls-Royce's diesel rotary prototype in my spare time. I'm betting someone will eventually show this to Rob Dahm too.
@@cyrilio It’s certainly the most fascinating application of Wankel’s patented design. Simply put, two large rotors act as mechanical compressors and forcing exhaust gases from two small rotors to spin the large rotors, further compressing the intake air, before fuel (in this case diesel) gets injected onto the superheated air. You could say the entire operation is a hybrid of a compound turbo setup and a two-stroke diesel, except the fuel is injected directly into the small turbo, while the mechanical blower helps exhaust gases to be flushed out, all in one small package for a diesel. Engineering Explained has a great explanation of this concept.
Fantastic work. As a long time powertrain engineer with a college education in combustion I appreciate the approach you have taken. For any of the "real engineers" that have complaints about the level of accuracy or such THEY could use some more education around the concepts of optimization and fit for purpose. I also love the way this generates so much interest from other RUclipsrs who I will generously say sometimes make it their own by tweaking some of the parameters in a way that may go a bit beyond what the laws of physics would generally allow. I sadly don't get these privileges at work, but if they are learning from it while interesting countless followers in physics, I am good with it.
The intake and exaust ports overlap. Solve this and you'll have a more accurate sound. Also not to mention most rotary engines have side intake ports. Furthermore use two ignition sources and have a two step flame front propagation for more accurate sound.
He's excluding the side ports to simplify, since its still largely 2 dimensional. The flame front also seems to be faster than reality, so dunno how much adding a second spark would do anything.
@@bazinga1831 more like someone that has no idea how to drive and maintain one of these engines. You gotta drive it like a 2 stroke. Feed it pre-mix and let it sing. They are reliable engines, you just gotta let them clear their throat every time you drive it. A 50:1 2 stroke mix keeps the apex seals happy
@@roadrocket7770 they do have some problems after a while though. The Renesis series in the RX-8 burns oil, so I have to act like a BMW owner and keep some extra stored in the trunk, and even if you take good care of it the apex seals don't last forever and replacing those is annoying, especially since you most likely have to diy it because local mechanics usually don't know how, and you don't want to pay enough to make them learn.
You are the reason why I got back to programming (doing Linux Admin contract right now for living). And I am so happy that I saw your first video. Of course, I am not going to mimic your work (I am more interested in embedded and automotive) but... Thanks, it is nice to write code once again :)
Simulating an engine with its sounds has been on my list of long-term coding goals for a while, it's awesome to see you did it in such a polished way! This is very inspiring to see, thanks for sharing
Not only are you a genius setting the next groundstone for engine sound simulation, you are also gifted in explaining stuff quite easy. IMHO the only video for everything you need to know about Wankel engines.
I just got done watching a few of the other videos on this simulator and i have to say im fucking blown away by it. I want to preface everything im about to say with the little tidbit that I have a degree in computer security, with 7 years of programming under my belt. This simulator combines so many different skills incredibly well, its a marvel this is a one person project. Rigid body physics simulation, fluid mechanics simulation, sound processing, data visualization, the engines being created in your own scripting language, as well as your real life knowledge of engines as the cherry on top. This is an absolutely stunning intersection of skills and the version shown is beyond what I expected when clicking on this video. Love it!
Can you imagine a racing game (or even open world sandbox!) that simulated the engines of all vehicles this way, modelling dynamic deformation of the vehicles with collisions and wear? Being able to hop in a simulated 40 yr old pickup and hear the difference in wear? Even the industrial applications for such a technology are incredible - AI running on powerfull hardware would be able to reliably predict the precise failure times of various components by simulating years of runtime in fractions of a second. This really is the future.
Another thought, I hope game developers are taking notes. If not, they should hire you to do their sound coding. This sounds better than just about any game I've ever played.
I'm here marking my position in history for the future of digital engine simulation. Cars in video games will sound so much more realistic thanks to this.
@@pikkyuukyuun4741there’s a problem though: recorded sounds only capture sound from a specific car engine at a specific acceleration from a specific direction in a specific area into a specific microphone. By instead synthesizing engine noises, we can procedurally create more specific levels of realism.
As a long time rotary enthusiast and someone who dove into the math to model one of these, this is spot on dude. Either you have also been a long time fan, or you really did your research!
It's hard to miss, since it's the only content he posts. I think it should be "software engineer" though, or programmer. Working with the programs more than the physical electronics. But I do enjoy it myself, learning C++ has been great, once you're caught up with the basics, you can easily keep expanding your knowledge forever.
Yeah, trust me he is not your average software engineer. I'm a software engineer myself, even have the "lead" title in front of it, but the thing that he does as a hobby is way beyond my pay grade professionally. If you noticed the way he was able to explain complex topics in simple terms that regular viewers can understand, that is a sign of an exceptional engineer. I don't doubt that you can do it, I'm just saying it takes a lot of work, way beyond what normal software engineer does.
@@rizkiyoist Really? How can you be a lead developer and this is above your pay grade? I do this kind stuff in my free time (simulations, not this exactly), and I can't find ANY job, not even junior dev.
I don't know much about a P-Port except delivering pizza and daily driving a friend's P-port 13B in the 90's with a Weber downdraft. Spin to 10,500 but shift around 9.5k and idle is 1450ish but wavering because of brap to 1700. Brap Brap Brap... and keep a fresh set of plugs in the center console. You might need to change them to get to work.
Of course there's already someone out there making a physically based engine noise generator (and more). Here I thought I'd have something unique to my name when I'm done with my degree. Awesome work!
I think some basic audio filters to simulate the environment could make this sound so much better. Maybe a little echo or muffling from imaginary buildings, grass, or the road would make a big difference.
One of the major problems with rotary engines is differential heating of the sides of the rotary engines. It would be interesting to simulate a rotary engine that uses one rotor for compression and feeds that into another rotor for combustion on both sides (necessitating twice the ports) and possibly a rotor for exhaust scavenging.
My god, the engine stuttering is just godly One of the biggest thing at play is fluid movement too, the biggest being the fuel But simulating it would be *extremely* difficult, not only the fuel level changing, but also the geometry of the tank itself, etc... But it's already perfect as it is, your work is amazing man !
The depression in the rotor is actually there to define a decent combustion chamber as you can see without a depression the combustion chamber shape at full compression is very weird and difficult to get a spark plug to, and is also long and flat which means you end up with a lot of non-combusted fuel. By having a pocket in the rotor, the fuel air mix tends to be concentrated in towards one spot, which is where you would locate your spark plugs. This is a really excellent series, congratulations - keep it up! If you study the maths for the peri/epitrochoid, you can generate forms with more than two lobes (the wankel uses a 2 lobe form which gives that peanut shaped housing). Be interesting to see if you could get a triple or quad lobed housing to work!
If I understand them correctly, most jet engines should be pretty boring to sim, since they are mostly in a steady state of gas pressures, and are mechanically just the one rotor (with lots of different fans coming off it)
Loved the ending with "flexible vehicle simulation" just like overrun, that is something missing from engine simulations, the variation in load and rpm when shifting and when going over bumps on a track.
for both the housing and the rotor there are exact mathematical curve functions, describing their shape. Implementing those would probably have been easier than your simulation method
Wowowowow! Amazing stuff as always! One small criticism: strictly speaking the driveline wouldn't be just flexible but have some play, which it sounded like you were trying to simulate by making the drivetrain more elastic than it should be. If there is a lightweight solution to simulate gear backlash and tighten up the elasticity it would improve the realism quite a bit further imo. It already sounds brilliant though, massive props my guy!
A small amount of play is incorporated into the simulation. The flex accounts for a lot of things and not just the drive train. Things like the engine mounts, wheel deformation and of course the deflection in the various shafts in the vehicle. It needs some calibration of course but this was my first attempt based on some reference footage. Thanks for watching!
@@AngeTheGreat Yeah I get that. By "driveline" I meant everything from combustion to tire contact patch, miscommunication on my part. Maybe it's the 787B being a race car, which makes me expect a tighter driveline with almost rigid engine mounts where the play sounds harsher. It could also be the lack of gearbox noises tricking my ears.
In Assetto Corsa you can simulate driving this car and shifting sounds even better, but of course the sound was sampled from real life. Old fashioned method and in no way smart like this simulation. So you have games with awesome sound already, that was not a thing. But to simulate reality details like this is really a new level. I share your exitement.
Man, you were dead on referring to the Mazda from Gran Turismo. I didn't know the significance of that car when I discovered it in the game. I just knew it got along really well with me in races. It was only after playing the game that I eventually looked up the real car, and its history. Great video! I even watched the ad.
This is incredible, and gives me so many questions/ideas. Have you considered simulating different exhaust setups? Alternatively, being able to incorporate the sound of lifters, simulating misfires, intake, etcetera would be really exciting. Great stuff.
Excellent work! The method of substracting parts of the rotor to get it's shape is exactly the method used in manufacturing. As far as I've read, there is no closed form parametric equation of a rotor's shape. The rotors shape is instead created by a "collection of curves" which describe it's shape over a given rotation. Just a fancy way of doing what you have shown. The instantaneous volumes can also be calculated for the working chambers. Look up "ROTARY ENGINE" by Kenichi Yamamoto both 1981 and 1971 editions have most equations and methods to calculate the volumes!
I like these videos because it makes me imagine the moment when someone finishes making something as complex as this, and hears it roar for the first time. These are *absurdly* amazing mechanical systems.
I've been waiting for a simulation of the Wankel engine and this is magnificent! It sounds just like the Le Mans winning Mazda 787B and braps like it too! I would imagine how difficult it was to get the geometry of this engine design correct, but I can say it was a massive success. And this is just the KKM Wankel design. I wonder if the DKM Wankel engine can be simulated and also other pistonless rotary engines like the LiquidPiston and Quasiturbine. Only time will tell and keep up the great work 👍
I've been following this project of yours for some time and I think the versatility and maturity you have achieved is astonishing! The clear benefit is that the procedural nature of the sound makes it feel very organic, it doesn't suffer the loop & blend issues of prerecorded samples. Imagining that these sound effects are simulated in real-time with approximations, while they sometimes even challenge recorded engine sounds of AAA gaming titles... Chapeau, good Sir! Maybe this might go full circle one day and provide the base for future, recorded gaming sound samples!
What are you using to make these animations or visualizations through code? I always wondered how creators made visuals to demonstrate code executions.
As many others, I'm amazed at what you have created. Your brain has successfully mixed your knowledge in computers, mathematics and cars to create this tool that nobody managed to visualize, but now we all admire. I'm thinking, would this new "rotary edition" of the engine simulator help real life engineers at Mazda, for example, to develop new and improved rotary engines? At least to figure out approximate power outputs or how would a 5 rotor engine sound (15 combustion chambers should be interesting). Maybe in an improved version, you could even accurately simulate different combustion chamber designs, flame propagation, intake and exhaust ports locations... I don't know. I just imagine that the possibilities could be very attractive. Could it even be possible in the future to use an evolved version of this tool and find a solution to the oil burning flaw that Wankel rotary engines suffer? I just feel this simulator could be much more than just a curious project. By using this software, any person could save up millions of dollars in research and special facilities by just playing with it and create new configurations of this internal combustion engine that has been ignored by our cylinder-centrical society.
the 787B also had a variable intake manifold! But the next challenge for you should be a 2 rotor with peripheral ports, bridge ports, street ports and factory ports. Just to see if the simulator can simulate what essentially prolongs intake timing like a cam, but without the cam lol
I got headphones on and listening to the simulated sound near the end made my ears ring. I'd say that's a VERY realistic rotary engine simulator. Did sound pretty accurate to a 4rotor too
People: how did you learn this stuff? AngeTheGreat: Trial and Error plus reading books. Me: *Assuming he learned it as a hobby all by himself* AngeTheGreat: I have a Degree in Computer Engineering. Me: thats a bit more time invested than imagined
Unfortunately my degree was a complete waste of money and I learned very little. I'm not saying college is useless but pretty much every useful thing I've learned and certainly everything I learned to write this simulator, I learned through self-study. Doesn't work for everyone though but that's just my experience. Thanks for watching!
something really interesting for me would be simulating the sound coming from the exhaust system, intake and mechanical sounds like the valves(?). so you could walk around a car in 3d space and hear different sounds from different perspectives. maybe something for the 3d update.
People like you are the ones who will revolutionise (pun intended) sound design in games to be much more realistic and accurate without relying heavily on recording real engines
Only just found this channel. Love it. I'd like to see a back to back of some famous straight 6 engines and why they sound so different. The rb25, 1jz and the TVR engine would be fantastic. All are "common" engines to hear here in the UK and sound so different. The more modern BMW 6 engines would be a lovely comparison.
Nice video but wanted to add a couple points: The parametric equation for a rotor housing surface is publicly available as has already been stated. The e/R ratio and parameters which will define your curve are also available. I believe e=15mm and R=105mm for all of Mazda’s engines if memory serves. This should let you properly model the curve. The rotor flanks are in fact circular, you’re welcome to measure/roll one around in real life. The reason your optimized shape isn’t used is that the engine is knock limited before such an optimized compression ratio could yield a higher thermal efficiency. Fortunately you don’t just have to trust me on that as a random internet stranger, you can just look at the pockets in the rotor faces that vary in size across different engine models which control compression ratio, all while the flank surfaces themselves remain the same shape. Again great job. Hope this is useful to you guys. Recommend reading Kenichi Yamamoto’s book to learn a little more if you’re interested. Definitely more to find on port timing there too.
It’s silly to think that I’m this day in age we would want to go back to the original way of communicating over radio waves if we already have phones that can do everything? Well, in the event of a global crisis (like he was talking about) where either communications networks are hacked or compromised, using these radios would be the only form of communication available as you are the source of the signal, not a cell tower. Obviously having these radios that work locally is the best way to be prepared for communications during localized outages. This is such a great visual video with so much good quality information, really makes it a whole lot easier to understand the different frequencies we can use.
![SahBabii - Viking [Official Music Video]](http://i.ytimg.com/vi/D37cL92JX0I/mqdefault.jpg)
You asked for dorito, I gave you dorito.
Education is good for you! Keep exploring at brilliant.org/AngeTheGreat/. Get started for free, and hurry-the first 200 people get 20% off an annual premium subscription.
amazing!
huh, was gonna shit on brilliant for having expensive af annual plans only that are disguised as monthly but now they've seemed to introduced real monthly deals
still fucking expensive though
Had never been happier
Awesome thank you. But I still want to see if you can manage a Deltec :P
this is amazing, but i have one question. with the size of a rotary, arnt they 1.3L engines? im only asking because i see the 18CL being or something.
15:10
the rpm stutter, holy shit it's too good.
I never though I would get shills from a simulation of a an engine.
Its really good.
Not like games that use fake "wav"s, they never made me chill in my spine, because its too fake and no dynamic response.
Waiting for any game that could probably make use of this simulator
@@UshankaMaster Gran turismo and forza come to mind, they've never had the best sounding engines.
Also automation and rigs of rods
I feel like I'm witnessing the audio version of the millennium graphics leap in 3D racing.
This dude's engine simulator is mind boggling and could take racing simulation to a whole 'nother level!
@@viperfan7 you've never heard FM3 & FM4's sounds then, have you? Some of the best sounds in the series right there
The future of racing games' sound design is here! Someone hire this guy for a AAA project!
I know. Imagine this in a Forza Horizon title 🤯
I'm pretty sure a lot of game companies have contacted him and that's the reason the project went partially closed source lol
Way too computationally intensive to be run in real time
Hiring would be bad overall for the industry.
Some car game license his software for their next game would be best. So that every new car game made has the change for the best sounding engines.
@@benperkins2929 some genius would be able to optimize it to run on low end CPU for the challenge of it.
A company would just tell you to buy a better rig.
The true, ideal housing is an epitrochoid, actually pretty easy to calculate. There's math for the inner and outer envelopes (Wankel and Liquid Piston respectively) too, I have a Desmos setup for all of them.
The housing is calculated using the exact function (with the rotor there for visualization purposes). For the rotor, I thought it would be cool to show how the shape is optimized but I'm sure there's a closed-form solution for it.
@@DJ_Level_3 nice! I'll check it out
looks hella smooth
@@DJ_Level_3 Yes, yatub loves sin sore ship.
@@DJ_Level_3 Amazing, that's a very pleasing design to look at. Has that type of rotary engine ever been made? You say it's the ideal geometry, has it been tried?
My dad would have absolutely loved to see these videos. He was a gearhead who spent years working R&D at some of the big houses, Edelbrock, NOS, and others. He engineered some revolutionary new designs for the metering blocks in carburetors, and being able to iterate on designs with today's technology would have blown his mind. Thanks for sharing these!
Add the gearbox whine to this, and then it'll be perfect. That transmission stuttering is amazing btw, at first i thought it was a bit too much but it fits perfectly (for a race engine at least).
It depends on the car of course, I used an exaggerated example like a racecar to demonstrate what it can do. Thanks for watching!
787B? 😀
My Rx8 reacts almost exactly like that with firm suspension and large but thin tires
Well... there's also a reinforced gearbox, ultra light counterweight/flywheel combo and stage 3 clutch sooo...
Might play a part too lmao
@@FuSiionCraft yeah flywheel is a big contributor to tackle stutter. having too lightweight flywheels can actually be contraproductive even kill the engine
@@AngeTheGreat bmw m3 gtr have high transmission noise
That new vehicle physics and load simulation is top notch. It sounds so unbelievably real. Great job. I'm blown away.
Thanks!
@@AngeTheGreat I'd love to see some power numbers
So happy you got a sponsorship ! Your channel is really growing now !
15:40 WOW i didnt realise how much it was missing till you added it in. That is absolutly amazing how that contributes so much to a realistic sound.
The gods gave us everything we asked for
YESSIR!!!!!!!!
God did 🙌🏿
AMEN!
Can Bluetooth this while driving- easiest four rotor install. 🙏
facts
dude the adjustments to the chassis and drivetrain flexibility in the vehicle simulation make an AMAZING difference in the end product! The engine simulator by itself is an absolute work of art, but it sounds mostly like an engine sitting on a dyno. Now it sounds like you're in the car! Bravo!
That drivetrain flex makes all the best noises happen!
The power of finding a niche, from cereal box game to engine simulator. I'm really loving the series and the details you take in account. As someone who knows only little about engines, I didn't expect to hear any difference during the last demo, but I actually did hear it clearly. Very cool details.
Just got a Toblerone ad on the video. The triangle gods are watching over.
i clicked for the math and braps, i stayed for the thicc rotary that started fartin
A man with taste
that shit FAARDED
She gotta poopr on er
that thang do be farting for real
9:20 You truly are fortunate to have such a... special research division
While as a dorito engine enthusiast I'm excited to see and hear the Wankel engine finally being simulated correctly, I actually screamed and did the DiCaprio pointing meme at the moment you explained how the rotor and housing design were optimized using an algorithm akin to a CNC machine. That's an extremely smart solution to model the reulaeux rotors and the epitrochoid housing, and no doubt Mazda engineers also used similar techniques in their research to build the 16X range extender. Mazda also worked on different cavity designs for the rotors to optimize combustion and it turns out that by having no cavity at all meant the highest compression ratio of 16:1 can be achieved.
I know you've heard this so many times but you're so cool and smart for investing into this project! Can't wait to finally mod this and test out hypothetical designs including Rolls-Royce's diesel rotary prototype in my spare time. I'm betting someone will eventually show this to Rob Dahm too.
This Rolls-Royce diesel concept sounds fascinating. Never heard of it before.
@@cyrilio It’s certainly the most fascinating application of Wankel’s patented design. Simply put, two large rotors act as mechanical compressors and forcing exhaust gases from two small rotors to spin the large rotors, further compressing the intake air, before fuel (in this case diesel) gets injected onto the superheated air. You could say the entire operation is a hybrid of a compound turbo setup and a two-stroke diesel, except the fuel is injected directly into the small turbo, while the mechanical blower helps exhaust gases to be flushed out, all in one small package for a diesel. Engineering Explained has a great explanation of this concept.
already happend. robs fan base is very close and personally, there is no way our god will miss anything
the fully simulated sound at the end is ridiculously good, you could tell me it was recorded in a real car and id believe you
oh my god, after so many people attempted to make a "rotary" engine with this sim, and you go and build the damn thing proper in the simulator hahaha
Fantastic work. As a long time powertrain engineer with a college education in combustion I appreciate the approach you have taken. For any of the "real engineers" that have complaints about the level of accuracy or such THEY could use some more education around the concepts of optimization and fit for purpose.
I also love the way this generates so much interest from other RUclipsrs who I will generously say sometimes make it their own by tweaking some of the parameters in a way that may go a bit beyond what the laws of physics would generally allow. I sadly don't get these privileges at work, but if they are learning from it while interesting countless followers in physics, I am good with it.
The intake and exaust ports overlap. Solve this and you'll have a more accurate sound. Also not to mention most rotary engines have side intake ports. Furthermore use two ignition sources and have a two step flame front propagation for more accurate sound.
more evidence of a trash design lol
He's excluding the side ports to simplify, since its still largely 2 dimensional. The flame front also seems to be faster than reality, so dunno how much adding a second spark would do anything.
@@bazinga1831 more like someone that has no idea how to drive and maintain one of these engines. You gotta drive it like a 2 stroke. Feed it pre-mix and let it sing. They are reliable engines, you just gotta let them clear their throat every time you drive it. A 50:1 2 stroke mix keeps the apex seals happy
@@roadrocket7770 yup
@@roadrocket7770 they do have some problems after a while though. The Renesis series in the RX-8 burns oil, so I have to act like a BMW owner and keep some extra stored in the trunk, and even if you take good care of it the apex seals don't last forever and replacing those is annoying, especially since you most likely have to diy it because local mechanics usually don't know how, and you don't want to pay enough to make them learn.
I wasn't expecting this to come so soon!
Thank you for your efforts and for listenning to our requests. You are truly the great.
You are the reason why I got back to programming (doing Linux Admin contract right now for living). And I am so happy that I saw your first video.
Of course, I am not going to mimic your work (I am more interested in embedded and automotive) but... Thanks, it is nice to write code once again :)
Simulating an engine with its sounds has been on my list of long-term coding goals for a while, it's awesome to see you did it in such a polished way! This is very inspiring to see, thanks for sharing
Not only are you a genius setting the next groundstone for engine sound simulation, you are also gifted in explaining stuff quite easy. IMHO the only video for everything you need to know about Wankel engines.
I just got done watching a few of the other videos on this simulator and i have to say im fucking blown away by it. I want to preface everything im about to say with the little tidbit that I have a degree in computer security, with 7 years of programming under my belt.
This simulator combines so many different skills incredibly well, its a marvel this is a one person project. Rigid body physics simulation, fluid mechanics simulation, sound processing, data visualization, the engines being created in your own scripting language, as well as your real life knowledge of engines as the cherry on top. This is an absolutely stunning intersection of skills and the version shown is beyond what I expected when clicking on this video. Love it!
IT BRAPS, IT REALLY BRAPS, IT DOES THE BRAP, BRAP DORITO
Can you imagine a racing game (or even open world sandbox!) that simulated the engines of all vehicles this way, modelling dynamic deformation of the vehicles with collisions and wear? Being able to hop in a simulated 40 yr old pickup and hear the difference in wear? Even the industrial applications for such a technology are incredible - AI running on powerfull hardware would be able to reliably predict the precise failure times of various components by simulating years of runtime in fractions of a second. This really is the future.
You are who we all aspire to be. Keep it up!
Wow the flex really brought the whole experience to life. Seems a bit underdamped but what a transformative feature. I love this project so much!
It's a work in progress of course :)
Another thought, I hope game developers are taking notes. If not, they should hire you to do their sound coding. This sounds better than just about any game I've ever played.
Even better, he can just start his own company and license out his sound generation engine to game companies.
@@jacobm2625 That sounds like a great idea. There are so many parts of games now that are outsourced tools.
I'm here marking my position in history for the future of digital engine simulation. Cars in video games will sound so much more realistic thanks to this.
car games already have realistic sounds via recording the sound of the actual cars
@@pikkyuukyuun4741there’s a problem though: recorded sounds only capture sound from a specific car engine at a specific acceleration from a specific direction in a specific area into a specific microphone.
By instead synthesizing engine noises, we can procedurally create more specific levels of realism.
omg that flex simulation is AMAZING. you sir are a genius.
I love how the audio is so accurate for what's happening
As a long time rotary enthusiast and someone who dove into the math to model one of these, this is spot on dude. Either you have also been a long time fan, or you really did your research!
You know it’s good when it gives you goosebumps. And this one (with flex simulation) surely does. Wow, again.
love seeing the amount of progress and I can't wait to be able to use it sometime in the future!
I'm surprised and so happy that you actually did the rotary simulation, this whole physics engine and engine simulator is so so freaking cool.
I had no idea you are a computer engineer! That's exactly the career I'm looking to pursue.
he's not an average computer engineer
It's hard to miss, since it's the only content he posts. I think it should be "software engineer" though, or programmer. Working with the programs more than the physical electronics. But I do enjoy it myself, learning C++ has been great, once you're caught up with the basics, you can easily keep expanding your knowledge forever.
@@aneeshprasobhan he's not your average every day computer engineer, he is...
advanced computer engineer.
Yeah, trust me he is not your average software engineer.
I'm a software engineer myself, even have the "lead" title in front of it, but the thing that he does as a hobby is way beyond my pay grade professionally.
If you noticed the way he was able to explain complex topics in simple terms that regular viewers can understand, that is a sign of an exceptional engineer.
I don't doubt that you can do it, I'm just saying it takes a lot of work, way beyond what normal software engineer does.
@@rizkiyoist Really? How can you be a lead developer and this is above your pay grade? I do this kind stuff in my free time (simulations, not this exactly), and I can't find ANY job, not even junior dev.
I don't know much about a P-Port except delivering pizza and daily driving a friend's P-port 13B in the 90's with a Weber downdraft. Spin to 10,500 but shift around 9.5k and idle is 1450ish but wavering because of brap to 1700. Brap Brap Brap... and keep a fresh set of plugs in the center console. You might need to change them to get to work.
the next racing games will be amazing, thank you very much for the study and tests you've been doing
Of course there's already someone out there making a physically based engine noise generator (and more). Here I thought I'd have something unique to my name when I'm done with my degree. Awesome work!
I think some basic audio filters to simulate the environment could make this sound so much better. Maybe a little echo or muffling from imaginary buildings, grass, or the road would make a big difference.
I'm actually so happy you got a sponsor for the vid! You definitely deserve more recognition!
I love it. You're amazing Ange! Can't wait for the public release.
Not often am i glad people are sponsored. But here we are. Finally someone well deserved of one. Your content is amazing
One of the major problems with rotary engines is differential heating of the sides of the rotary engines. It would be interesting to simulate a rotary engine that uses one rotor for compression and feeds that into another rotor for combustion on both sides (necessitating twice the ports) and possibly a rotor for exhaust scavenging.
My god, the engine stuttering is just godly
One of the biggest thing at play is fluid movement too, the biggest being the fuel
But simulating it would be *extremely* difficult, not only the fuel level changing, but also the geometry of the tank itself, etc...
But it's already perfect as it is, your work is amazing man !
The depression in the rotor is actually there to define a decent combustion chamber as you can see without a depression the combustion chamber shape at full compression is very weird and difficult to get a spark plug to, and is also long and flat which means you end up with a lot of non-combusted fuel. By having a pocket in the rotor, the fuel air mix tends to be concentrated in towards one spot, which is where you would locate your spark plugs. This is a really excellent series, congratulations - keep it up! If you study the maths for the peri/epitrochoid, you can generate forms with more than two lobes (the wankel uses a 2 lobe form which gives that peanut shaped housing). Be interesting to see if you could get a triple or quad lobed housing to work!
I know it sounds silly talking about a simulation of a thing that exists in real life, but it does give me goose bumps hearing it.
I guess, the next step would be to simulate jet engines :) Or just turbines in general.
If I understand them correctly, most jet engines should be pretty boring to sim, since they are mostly in a steady state of gas pressures, and are mechanically just the one rotor (with lots of different fans coming off it)
So cool seeing the different rotor configs results simulated. Thanks!
This is genuinely everything we have asked for. Thank you
The torque/load simulation is unbelievably great! Keep up the great work my friend!!
That RPM flutter at the end. I could feel it through my desk xD
Loved the ending with "flexible vehicle simulation" just like overrun, that is something missing from engine simulations, the variation in load and rpm when shifting and when going over bumps on a track.
Very cool! I'd love to see an opposed piston engine as well, like the new Achates engines.
There’s lots of them on RUclips. The simulator already supported it
@@Roomsaver They are still modeled with valves, which aren't present in the real design
When clutching is involved it feels like you're literally on the race track dashing through the wind. Incredible!
for both the housing and the rotor there are exact mathematical curve functions, describing their shape. Implementing those would probably have been easier than your simulation method
I didn't know about the closed form solution, thanks for sharing!
How do i get the formulas?
Wowowowow! Amazing stuff as always!
One small criticism: strictly speaking the driveline wouldn't be just flexible but have some play, which it sounded like you were trying to simulate by making the drivetrain more elastic than it should be. If there is a lightweight solution to simulate gear backlash and tighten up the elasticity it would improve the realism quite a bit further imo.
It already sounds brilliant though, massive props my guy!
A small amount of play is incorporated into the simulation. The flex accounts for a lot of things and not just the drive train. Things like the engine mounts, wheel deformation and of course the deflection in the various shafts in the vehicle. It needs some calibration of course but this was my first attempt based on some reference footage. Thanks for watching!
@@AngeTheGreat Yeah I get that. By "driveline" I meant everything from combustion to tire contact patch, miscommunication on my part.
Maybe it's the 787B being a race car, which makes me expect a tighter driveline with almost rigid engine mounts where the play sounds harsher. It could also be the lack of gearbox noises tricking my ears.
I know this must have been said a thousand times by now, but we need a game with this engine at its core.
In Assetto Corsa you can simulate driving this car and shifting sounds even better, but of course the sound was sampled from real life. Old fashioned method and in no way smart like this simulation. So you have games with awesome sound already, that was not a thing. But to simulate reality details like this is really a new level. I share your exitement.
This is by far one of my favorite projects I’ve seen on RUclips
This is incredible!
You should do reverb / echo in a tunnel next, with the Ferrari F2004.
Man, you were dead on referring to the Mazda from Gran Turismo. I didn't know the significance of that car when I discovered it in the game. I just knew it got along really well with me in races. It was only after playing the game that I eventually looked up the real car, and its history. Great video! I even watched the ad.
That drivetrain flex is super sweet!
This is incredible, and gives me so many questions/ideas. Have you considered simulating different exhaust setups? Alternatively, being able to incorporate the sound of lifters, simulating misfires, intake, etcetera would be really exciting. Great stuff.
Reads title:
Simulating a Rotary Engine
*Meh...*
(yes, it braps)
*HELL YEAH*
Excellent work! The method of substracting parts of the rotor to get it's shape is exactly the method used in manufacturing. As far as I've read, there is no closed form parametric equation of a rotor's shape. The rotors shape is instead created by a "collection of curves" which describe it's shape over a given rotation. Just a fancy way of doing what you have shown. The instantaneous volumes can also be calculated for the working chambers. Look up "ROTARY ENGINE" by Kenichi Yamamoto both 1981 and 1971 editions have most equations and methods to calculate the volumes!
Amazing stuff!
I would really like to see this simulation handle turbochargers
I like these videos because it makes me imagine the moment when someone finishes making something as complex as this, and hears it roar for the first time. These are *absurdly* amazing mechanical systems.
this is a certified dorito moment
This video is so underrated. The sound of it running the gears is... 🙌🙌🙌
Would like to have 2-stroke support in Engine Sim as well.
I've been waiting for a simulation of the Wankel engine and this is magnificent! It sounds just like the Le Mans winning Mazda 787B and braps like it too! I would imagine how difficult it was to get the geometry of this engine design correct, but I can say it was a massive success. And this is just the KKM Wankel design. I wonder if the DKM Wankel engine can be simulated and also other pistonless rotary engines like the LiquidPiston and Quasiturbine. Only time will tell and keep up the great work 👍
using this would you also be able to simulate a "liquid piston engine" or would that require remakeing the whole engine?
So happy you got a Sponsorship!!
I wonder if this could be used for something like the strange soviet pentagon rotary engine
I've been following this project of yours for some time and I think the versatility and maturity you have achieved is astonishing! The clear benefit is that the procedural nature of the sound makes it feel very organic, it doesn't suffer the loop & blend issues of prerecorded samples. Imagining that these sound effects are simulated in real-time with approximations, while they sometimes even challenge recorded engine sounds of AAA gaming titles... Chapeau, good Sir! Maybe this might go full circle one day and provide the base for future, recorded gaming sound samples!
What are you using to make these animations or visualizations through code? I always wondered how creators made visuals to demonstrate code executions.
He made his own program.
@@ethanmonat yes and he's true chad
As many others, I'm amazed at what you have created. Your brain has successfully mixed your knowledge in computers, mathematics and cars to create this tool that nobody managed to visualize, but now we all admire.
I'm thinking, would this new "rotary edition" of the engine simulator help real life engineers at Mazda, for example, to develop new and improved rotary engines? At least to figure out approximate power outputs or how would a 5 rotor engine sound (15 combustion chambers should be interesting). Maybe in an improved version, you could even accurately simulate different combustion chamber designs, flame propagation, intake and exhaust ports locations... I don't know. I just imagine that the possibilities could be very attractive. Could it even be possible in the future to use an evolved version of this tool and find a solution to the oil burning flaw that Wankel rotary engines suffer?
I just feel this simulator could be much more than just a curious project. By using this software, any person could save up millions of dollars in research and special facilities by just playing with it and create new configurations of this internal combustion engine that has been ignored by our cylinder-centrical society.
that's pretty cool!! I gotta put this in a beamng car, and we can race :P
I'm ready!
the 787B also had a variable intake manifold!
But the next challenge for you should be a 2 rotor with peripheral ports, bridge ports, street ports and factory ports. Just to see if the simulator can simulate what essentially prolongs intake timing like a cam, but without the cam lol
yes
Mah boi, this rotary simulation is what all true dorito fans strive for!
I got headphones on and listening to the simulated sound near the end made my ears ring.
I'd say that's a VERY realistic rotary engine simulator. Did sound pretty accurate to a 4rotor too
I'm here before Rob Dahm found this video..
My favorite engine ever is the 4-rotor R26b engine found in the 787b, so this is a neat find. Good job.
A dorito engine! :)
People: how did you learn this stuff?
AngeTheGreat: Trial and Error plus reading books.
Me: *Assuming he learned it as a hobby all by himself*
AngeTheGreat: I have a Degree in Computer Engineering.
Me: thats a bit more time invested than imagined
Unfortunately my degree was a complete waste of money and I learned very little. I'm not saying college is useless but pretty much every useful thing I've learned and certainly everything I learned to write this simulator, I learned through self-study. Doesn't work for everyone though but that's just my experience. Thanks for watching!
awesome stuff man
so glad you caught yourself a sponsor too
Thanks! Glad to have your continued support on my channel since the early days!
@@AngeTheGreat no one, then or since, has matched the indepth explanations and amazing visuals that you consistenetly put out
13:00 sounds like Rob Dahm's 4 rotor in the early stages of the build with all the gremlins
What a beautiful early Christmas gift
something really interesting for me would be simulating the sound coming from the exhaust system, intake and mechanical sounds like the valves(?). so you could walk around a car in 3d space and hear different sounds from different perspectives. maybe something for the 3d update.
People like you are the ones who will revolutionise (pun intended) sound design in games to be much more realistic and accurate without relying heavily on recording real engines
Only just found this channel. Love it. I'd like to see a back to back of some famous straight 6 engines and why they sound so different.
The rb25, 1jz and the TVR engine would be fantastic. All are "common" engines to hear here in the UK and sound so different.
The more modern BMW 6 engines would be a lovely comparison.
Those upshifts with the enhanced sound sim sound *exactly* like the upshifts from that one 787b onboard Le Mans lap. Well freaking done, man.
Nice video but wanted to add a couple points:
The parametric equation for a rotor housing surface is publicly available as has already been stated. The e/R ratio and parameters which will define your curve are also available. I believe e=15mm and R=105mm for all of Mazda’s engines if memory serves. This should let you properly model the curve.
The rotor flanks are in fact circular, you’re welcome to measure/roll one around in real life. The reason your optimized shape isn’t used is that the engine is knock limited before such an optimized compression ratio could yield a higher thermal efficiency. Fortunately you don’t just have to trust me on that as a random internet stranger, you can just look at the pockets in the rotor faces that vary in size across different engine models which control compression ratio, all while the flank surfaces themselves remain the same shape.
Again great job. Hope this is useful to you guys. Recommend reading Kenichi Yamamoto’s book to learn a little more if you’re interested. Definitely more to find on port timing there too.
That sounds remarkable. Im amazed at how well this simulation is working.
Subscribed for the engine simulations. Please don't stop these!
I cant help but let my jaw drop every time I watch one of your videos. Amazing stuff
It’s silly to think that I’m this day in age we would want to go back to the original way of communicating over radio waves if we already have phones that can do everything?
Well, in the event of a global crisis (like he was talking about) where either communications networks are hacked or compromised, using these radios would be the only form of communication available as you are the source of the signal, not a cell tower. Obviously having these radios that work locally is the best way to be prepared for communications during localized outages.
This is such a great visual video with so much good quality information, really makes it a whole lot easier to understand the different frequencies we can use.