What aerodynamics simulator would be best for modelling a helicopter?
Torrentable/free only please.
how is this even remotely /g/?
>>54762734
/sci/ doesn't know of any such software.
Microsoft Flight Simulator 2004
>>54762352
How is this a maze? The correct way is almost a straight line from the entrance to the exit
>>54763132
I mean something that will demonstrate whether or not a flying contraption will work.
>>54762734
Helecopters are technology, as are programs to simulate fluid dynamics. Get fuck outta /g/ you basic faggot .
>>54763580
Xplane uses blade element theory
You can get pretty accurate results in it but there is a lot of work involved that most people cant/wont do
Not sure how accurate helicopters are since I'm not a helo guy, but most other sims use lookup tables of the like for things
Flight gear supports a similar aero model to xplane but I've never used it
>>54763640
I'm not looking for something which gets fast results for familiar designs in aerodynamics, I mean something that will let me toss any 3D models in, specify that some parts are attached to motors, and then cell-based fluid computation from there I guess.
>>54763871
Solidworks fluid motion?
>>54763901
matlab or GNU Octave?
>>54763901
>>54763951
Here's a simple test of whether they're what I need or not:
If you stick a frisbee in it, can it simulate the motion of that frisbee, or will it only tell you how passing wind will be affected by that shape?
>>54763606
Stoves are technology. So are woks. Let's discuss stir frying methods on /g/.
>>54763973
i mean what i suggested was a programming language so i sort of misunderstood your reqs
google found this tho: http://www.illinoisrocstar.com/news/rocstar-simulation-suite-open-source
>>54763973
>I want an all in one cfd solution
lol
You can try freefoam
>>54764051
If you want a good stir-frying experience and only have an electric stove, I suggest getting a gas bottle and camping stove. You really need gas to get the wok up to that searing heat.
>>54764201
Everything except stresses on the solid, since it's small scale.
>>54764260
Explain more?
>>54763871
I think blender can do something like this but i don't know how accurate it is.
>>54764391
It's a free as in freedom CFD solver which is also used at a professional level.
I've never tried it though and I don't know how accurately you can simulate the behaviour of the helicopter (with all the flight mechanics).
>>54762352
>modelling a helicopter
>modelling a whole fucking helicopter
Serious, you are going to have to be so much more specific than that otherwise this guy >>54763132 has given you the answer you deserve.
math
yw
>>54764412
It seems accurate at least, but I can only simulate how the aircraft affects the air, whereas I'm more interested in simulating things that fly efficiently.
>>54764753
If you can do CFD for a wing, why not an entire aircraft?
To me it just sounds like a question of processing power.
>>54762734
You'd rather 100 browser and phone threads?
>>54765361
What you are asking is very advanced stuff. If you have to ask what software can do this, you are very likely simply not familiar enough with the matter to actually do anything with that software.
>>54765361
Well you're wrong... it's far more complicated than that.
Modelling something like that would be impossibly complicated. When modeling something like a wing all kinds of assumptions and a simplifications are made. For example, maybe you only model a 2D cross section section of the wing, or maybe a full wing but you assume the airflow is smooth across it, or more technical stuff like that you only need to model the Reynolds flow neglecting Stokes flows, etc.
Now you come along and propose modeling something in full detail and simply put this is not possible — it isn't possible to construct a simulation model that complicated and even if you could the world's most powerful computers wouldn't come close to running it.
With assumption you can make this tractable, like say I want to model 4 rotating blades, or I want to model this shape of helicopter as the air flows over the body. Saying "I want to CFD a whole aircraft" is daft. Which part under what assumptions is how this is done irl.
Catia ? Auto CAD ?
>>54765506
Is it possible to model 4 rotating blades, plus the body of the craft, getting some sort of feedback that I can use to tell how much lift it's getting and which way that lift will tip the aircraft?
Also, what do you mean by "full detail"?
For me, full detail means "All major moving components (rotors), plus the body, with those acting on the air and the air acting on those".
>>54765509
Wow gee thanks for magically transporting knowledge of whether those programs are suitable straight into my head
I didn't even know that Katia could do CFD
![forces[1].jpg](https://i.imgur.com/BbQLpiG.jpg "forces[1].jpg")
X-Plane
>>54765557
Catia can do almost everything. It's not the most simple software to learn modelling though.
OP, I believe you are GREATLY underestimating the complexity and computation involved in a problem like that. There is no program that can just simulate some random 3d model in a bit of air with a few clicks. Stuff like that is highly advanced stuff that people are occupied with for a good amount of time. You need expert knowledge to properly simplify the models enough to even get results in your lifetime. You can simulate aspects of it, but the pieces put together? NO, never. It is too hard. That's why people are still paying for wind tunnels. Because you can't realistically simulate every aspect of this problem and if you can't, you might just as well not simulate it at all.
>>>54765632
All the aerodynamics software I've come across so far do nothing but show the flow of air for single components, without giving a more noob-appropriate indication of how this in turn affects the craft.
Is Catia any better than that?
>>54765646
>You can't realistically simulate every aspect of this problem and if you can't, you might just as well not simulate it at all.
I'm not some high-flying engineer, any parts I make that are not reusable across different aircraft designs will be 3D printed at a material cost of about $25 a kilogram.
So I really don't care if it doesn't capture every little detail.
>>54765547
Okay this is better because now you have at least defined what you want from your model.
I am not aware of any off-the-self solver that allows for modelling such feedback but they may well exist. You will easily find solvers that can model the fluid (air) flow around the body and solvers that could simulate the flow coming from the blades. Coupling these together would be non-trivial but could potentially give you what you are looking for.
Planes are far easier to model because they move through clean air, at least it is clean compared to the downdraught a helicopter body sits under. The helicopter would have some horrendously complicated non-linear behaviour between the body and the rotors.
>>54765677
>So I really don't care if it doesn't capture every little detail.
It doesn't work like that. With turbulent streams you have very chaotic behaviour. To even get anything close to what you want you need to do some programming work yourself, there's no solution out there that works ootb. It's a lot of work and then needs some insane computation and it probably won't give you any useful information.
Just try it out in the physical world, a simulation is not useful at all here and definitely not worth the struggle.
>>54765677
You're missing the point here. You don't need to capture every detail indeed, but you do need to adequately capture every important feature. For example, perhaps the spinning g rotor assembly introduces some instability in the airflow from the rotors. If this is true then you would need to also incorporate the rotor assembly into your model which would again be horrendously complicated.
What you are asking for is simply not achievable for the given amount of effort you want to put into this.
>>54765752
>>54765750
>>54765677
i want to add to this that you can prototype a bit more and achieve what you want this way. Just buy more more material and test stuff out, it will take less time (and therefor money) this way.
>>54765700
Why does it need to be "solved" though?
You don't do any crazy maths for a 3-body problem to solve it since we don't know how yet, you just simulate it step by step. Not as accurate, but it works, and even with a thousand bodies it will still work.
If it's mathematically too difficult to calculate how the airflow will settle if you have all sorts of moving parts, might as well just give up on that and do it step-by-step, producing a series of frames that can be analysed.
>>54765786
A simulation helps with optimisation though if you lack a wind tunnel.
Plus it'd be really difficult in real life to test a cross between a glider and a helicopter.
Again, xplane is your best bet
It will do what you want well enough
>>54765901
What can XPlane do though?
>>54765993
http://www.x-plane.com/desktop/how-x-plane-works/
>>54766157
Oh, that looks like it actually might be what I'm looking for!
But by "geometric shape" does it mean I can toss in any 3D model?
And will it accept any shape to be used as a propeller?
>>54765830
>You don't do any crazy maths for a 3-body problem to solve it since we don't know how yet, you just simulate it step by step. Not as accurate, but it works, and even with a thousand bodies it will still work.
Oh you have no fucking clue do you? Jesus Christ, I wasn't sure before, but now I am: Stay the fuck away from this, you don't know what kind of heavy shit is involved here. Simulating NS equations is among most fucked up thing a computer can do. You will calculate this shit for days and get shitty results that won't help you in any way. Just prototype more, you will get far more useful results than a fucking simulation.
>>54766221
You have to make the 3d shape of the fuselage in planemaker, the software that comes with it to make plane/copters
Props can be curved. Not sure what you mean by any shape
>>54766404
Guess I'll have to try it out!
>>54766397
Don't just use 2-letter acronyms like people magically know what they mean, because it is impossible to look them up.
And I still don't see why Blender can (on a powerful computer) do fluid dynamics simulation in real time at low quality and a frame every several seconds at high qualty, yet translating that fluid movement back into force on the object should multiply the computation time by a factor of fuck you.
>>54765830
Yeah, just follow this guy's advice >>54765830. Not to be a dick or anything, but you really don't have a breeze just how complicated this is to do. It needs to be solved because at each of the steps you talk about everything must satisfy the Navier-Stokes equation — i.e. you are solving for a valid solution of the equation.
>>54766397
>Simulating NS equations is among most fucked up thing a computer can do
Pleb
>>54766462
>do fluid dynamics simulation in real time at low quality and a frame every several seconds at high qualty, yet translating that fluid movement back into force on the object
Because how lift is generated by a wing is much more complex than that gross simplification
>>54766157
>>54766462
This looks like a nice program. It will not work for a helicopter so don't even bother.
>>54766483
How so?
Lift is just pressure, and all the software out there for aerodynamics already does pressure.
>>54766492
It won't?
Can't you just stick some propellers on the top?
>>54764753
my favorite part about the Gau-8 is how when firing it produces more thrust than one of the a-10s engines.
I would love to see a plane flying just by shooting it's gun.
>>54766572
>Lift is just pressure, and all the software out there for aerodynamics already does pressure.
Not when you have turbulent streams, then the whole issue becomes significantly more difficult. But anyway, let's say it's just pressure. Then you take the pressure at the surface, multiply it with the surface normal, add it all together and have a force acting on the whole object (for this you have to assume that the body is absolutely rigid, which is of course not at all the case). Now you just use some simple Euler method and move the plane a bit along the direction of force for the next time step.
What happens with the air where the object once was? Pure vacuum? What happens to the air where the plane now moves into? These are difficult problems. Usually they are solved by not moving the objects through the air, but by moving the air around the object. When now the object is not in one piece but violently moving you can't do that and have to find a way to approximate above problems. I have honestly no idea how well you can do that. But adding those seemingly simple pieces together makes the physical simulation exponentially harder as you can make less and less simplifications.
>>54766492
But it does
>>54766676
Turbulent streams wouldn't matter if it's cell-based simulation like Blender.
Since, the pressure would be calculated for each cell touching the wing, and there cannot be turbulence within a cell as the turbulence is a composite of interactions between the cells.
Problems like what happens to the air where the object was are actually already solved by Blender; even smoke that is not directly touching an object will experience a tugging if that object moves away.
All the pieces already exist in Blender except for having the air act on objects, which is easy to do in cell-based simulation, it doesn't make any sense.
>>54766794
You are taking a lot of simplifications for granted here. Blender is also not a scientific tool, blender just simulates close enough to look convincing. There's no guarantee that any of that will translate into the real world and if you don't have that, why bother?
>>54766602
It wouldn't have enough ammo to liftoff.
And if you added more ammo, it would have too much mass to lift off.
>>54765646
don't listen to this wanker
i got u, OP. google "KSP Helicopter Mods"
>>54764753
>>54765506
>>54765700
>>54765752
>>54766474
>>54766492
Okay OP, I am this guy. I have a master's in simulation science and I also have done large scale fluid simulations that have run on supercomputers (albeit low Reynolds number Stokes flows, i.e. not governed by the full NS equations, rather the linear Stoke equation). Anyway, I am telling you what you want is not simple and not achieveable with free off-the-self tools without a significant amount of detailed coding by someone who knows what they are doing (x-plane looks interesting, but I doubt it will achieve what you want). Go ahead and waste your time one though — I would advise you not to bother but I would also say you shouldn't listen to someone who just wasted an hour of his time arguing with some random bloke on /g/. Best of luck with your endeavours.
>>54768378
>trying to help retards on /g/
You're a good and knowledgeable anon, but next time, don't even bother helping idiots on this forum.
>>54768413
Cheers — every now and then you get some food advice so I like to try contributing too. But perhaps sticking to one post max is the sensible balance!
>>54768459
good*... bollox
>>54768459
You're better off sticking to physics/maths stackexange in the future
>>54768493
>muh internet points
fuck off this thread is perfectly fine, /g/ is just some salty-ass bitches is all
>>54768493
>>54768566
Not me.
>>54768493
You're probably right.
>>54768653
shit, wrong pic...
Should have been this one.
>>54768378
>I have a master's in simulation science
Background? What made you want to go in to that specialism?
>>54768833
I did my undergraduate in theoretical physics and there was a scholarship to do the Sim Sci MSc at the same Uni. So I figured I'd use it no matter what I ended up doing (I was not wrong about that, it has been extremely useful). It's was called simulation science but that was just a fancy rebranding of a computational science (not to be confused with computer science).
openFOAM
I am not memeing
http://wiki.flightgear.org/Howto:Make_an_aircraft
>>54762352
You can run aerodynamics simulations with onshape. It should be free although It might have been an offer special to me or something like that as promotion.
You cant simulate whole helicopter though, you do blades, engines, etc.
>>54762352
what results do you want to get from the simulation?
a full, realistic 3D simulation around the whole geometry is not a good place to start for 95% of the answers to the above
