UnityFS - Flight Simulation toolkit

Now that Chris has provided a fix to the wheel issues, I've migrated my project from U4.6 to U5, but it's still not working as expected:
1. Even though the terrain is smooth, the wheels seem to roll like it's rough.
[Solved] I removed and then re-added the wheel colliders and also increased their mass quite a bit.
2. The nose/tail wheel isn't steering anymore.
[Solved] Changed the SteerableNosewheel script to use steerAngle = steer

Has anyone else seen (and found a fix) for these issues?

Thanks!

 

Hi there,

I think your asset doesn't support hydroplanes if i read correctly so if not then can someone point me out on how we can implement hydroplanes with ocean system like ceto since we can get height values.

I think that drag would be the culprit and that we would need to disable buoyancy at a certain speed to simulate air lift?

Any feedback would be more than welcome thank

Regards

 

Hi all,...

Albeit taking a long break for family reasons lately, I have been playing with uFS for ages to try to create rather unconventional "fliers". Talking about aircraft weight above I have had big problems with exactly that....

For different reasons I need my fliers to be very light - 5-25 kilograms or so... I am really struggling with achieving reasonable results regarding flight maneuverability with such low weights. I kinda reach my goals setting the weights much higher but I would rather that it is more accurate.

According to Chris he has made several simulator models of radio controlled aircraft with low weights with no problems... I just can't seem to succeed - perhaps it is lack of patience and always working on different aspects of my application that is the problem...

Therefore I wonder if there is anyone who has a prefab of such a light RCT or other flier with reasonable flight abilities that I could get to elaborate on / abuse... I do not need the actual mesh model... only the uFS configuration.

I would be so grateful... Thx in advance

Jesper, Denmark

PS: On another note: Can anyone direct me to a "VERY high lift" aerofoil included in uFS??

 

You could try using Javafoil and some settings more suited to what a RC plane would be experiencing.

Things like low speed, low altitude, small etc. I forget the exact variable names, but you should be able to find them somewhere.

Another thing to consider would be to perhaps use 'wing sections' for the main fuselage of a paper plane - the vertical body bit.

 

It is not working good in Unity 5, I play scene and plane going forward slowly without even giving any kind of input. Please fix that!

 

try the breaks?

 

That is supposedly a problem with the wheels.. which should have been fixed by now? I haven't had a chance to test it though

 

Works for me, should have been in the latest update. However, you need to re-align the wheel colliders as they are off - this may fix it then for you. I found that my plane would just slowly crawl forward for no reason.

 

I've noticed a fairly odd bug






When I open the Input it hides everything else, and when I close the input it brings everything else back.
But while the input alone is open I'm getting the following errors:






And they just keep counting upwards.

Also, is it just me or has development on this otherwise awesome plugin completely halted..?

 

Well, even though progress on this plugin seems to have grinded slightly to a halt =(
(The new version isn't even on asset store, despite having been submitted some 3½ months ago)
I thought I'd just share what I've been up to lately

The runway isn't done yet, but I'm pretty happy with how the village is coming along, and the aircraft handles like a dream. =)

 

Is this engine/tool kit capable of flight physics that would satisfy a hard-core sim enthusiast, say for a WWII sim?
I'm a long time 'hard core' sim enthusiast myself, but now I'm looking to develop one, and need a suitable engine that will accommodate a large map. Any opinions on this? If this engine is not suitable at present, is it close?
Any help/insight appreciated.

 

...after reading more of this thread, I think I have my answer.
Back to hoping Outerra is ready at some point.

 

Hi @Wasp_121 I'm a longtime hardcore flightsimmer, and UnityFS meets my expectations with a few caveats. It's far and away the best implementation I've tried on the Asset Store.

I found the way the scripts interrelate (cameras etc.) too closely-coupled, so I effectively ended up modifying the core wing and control surfaces scripts to stand alone, and added a vector3 to the relative windflow calculation for environmental wind effects. It doesn't have the super convincing feel of the 777 studios flight models, but I find it more realistic than the older parameter-driven flightsims like MSFS et al.

I use Edy's vehicle pack for landing gear, which gives me all the finesse I need on that front. The built-in system hasn't really worked since the Unity 5 update.

With a bit of integration into my other systems I'm still really happy with UnityFS, but I needed to do a bit of modification to get it working as I wanted. Actually it prompted me to pick up C# again for which I'm eternally grateful!

 

Hi Karearea, thank you for the reply.
As an owner of the 777 sims, I can say we're looking for at least that level of fidelity with the flight models.
We definitely don't want to take a step backwards, even a small one.

 

@Wasp_121 I meant to add earlier, I've had good results using a lot more wing surfaces at higher resolution- it's tricky to compare sims as the feel is so subjective. I know the Rise of Flight FMs take the devs as long to get right as the visual models, a big part of this would be tweaking blade elements.

Interestingly just adding some random forces a la turbulence made it feel more convincing too- I'm almost certain that it's partly the subtle weather simulation that just makes RoF feel 'right'.

 

Is there a web and/or android demo of this asset that has actual mesh terrain, buildings and clouds with textures that I can test how smooth it runs before I purchase?
I like the control and the day, night and weather options as in "Airplane Fly Hawaii" game on Google Play.
UPDATE: "Airplane New York" because of the addition of flaps, gear, and brake buttons.
Regards, Ferretface

 

Well, there is the official demo here:
http://unityfs.chris-cheetham.com/demo/

I have a web build here that you can play around with here:
https://dl.dropboxusercontent.com/u/6338532/SteamTrader/SteamTrader.html

The shop doesn't work yet, and you need to click f to enter the aircraft.
Controls are the same as for the official demo, with one exception.
The pitch doesn't stick. Rather when you let go of the pitch it returns to the neutral position.

To get the engine started you need to hold down the ignition key until it's actually running.
If you press 2 (accelerate) you will know when the engine is on as the prop mesh will be replaced by a fast spinning prop.

UFS however does not include a cloud/sky system, you'll have to get that elsewhere, and as such how heavy such a system is wholly depends on said system.
My experience of ufs is that it's quite lightweight on computers, but I've never tested it on a mobile device so I have no idea how it runs there

 

Once again, thank you for this.
I know a bit about the RoF development process, and yes the FM's take a lot of work, but also the LOD's etc eat up a lot of development time.

 

Has anyone tested the physics when a wing is missing, or part of a wing or other control surface? (i.e, damage)

 

Yes, it behaves sorta like you'd expect.

You can do it the obvious way (simply removing the whole wing), or I made a version of the Propwash script (there's a script that comes with the asset to enable 'blown' parts of the wing) that, instead of increasing airflow and lift and what not, changes the property to increase drag and reduce lift.

If you want to do something a bit more advanced like that you will have to do some scripting, but you can use the included scripts as a fairly good template.

 

I have a test script (for aircraft damage) where I simply scale down the wings during runtime as they take damage, it works exactly as one would expect, after a short while the plane goes into a spin and the pilot has to fight to get the aircraft to regain what stability can be forced out of the damned thing. ^_^

 

@Unity FS - Super Pleased with the quality of the asset.

To anyone that may know:
If I want to increase the effect of dihedral to level the plane, is the good way to do that without increasing the wing's angle / without doing it via script?

What is the "Inertia Tensor" for?

What is "Rollwise Dampening" for?

Thanks!

 

Can anyone help with getting wheel steering working again, after upgrading to Unity 5? In the above post, wwg appears to have solved it, but looking at the SteerableNoseWheel script, the quoted solution doesnt make sense, am I missing something?

Any help would be appreciated.

Thanks in advance.

 

@UnityFS:
I'll like to know if your project will be released in the asset store for Unity 5x... and when?
Any news will be appreciate.

Best regards

 

So while I am LOVING UFS I have struck a little difficulty. I am using manual input (a gui button) to raise and lower the landing gear.

My button will raise the landing gear (so its talking to the model appropriately) however when I press the button again, it doesn't trigger the landing gear to lower.

Does anyone have any idea what might cause this? If I switch the input mode to input manager and use my keyboard, everything works fine.

 

can you post the snippet of code you're using to do this?

also
try starting with the gear up and see if the gear will lower on that first press.

 

This is what I changed...around line 45 in SteerableNosewheel.cs:
Replace this: transform.Rotate( SteerAxis, steer );
With this: tailWheelCollider.steerAngle = steer;

 

My train of thought here (and correct me if I am way off) is that when using the input manager (keyboard 'g') all it is really doing is flicking a boolean to true when pressed.

So in my code I am doing simply that using the line:

Code (CSharp):

1. TwinEngine.LandingGearController.SetManualInputButtonPressed(true);

Now this is enough to have the gear raised, however nothing I try the lowers the gear. I have experimented with switching the bool to false (or 'not clicked') but to no avail and I feel that this is where I am coming unstuck.

 

Thanks you so much for your help wwg. Unfortunately, making the change you suggested gives me this error ; "
Assets/UnityFS/Scripts/External/SteerableNosewheel.cs(42,17): error CS0103: The name `tailWheelCollider' does not exist in the current context "

Any thoughts?

Thanks again.

 

Nevermind, I fixed it by using : GetComponent<WheelCollider>().steerAngle = steer;

Thanks again!

 

I haven't had a chance to play around with it, but could it be that it's a toggle?
So if you send true the first time, then the script evaluates whether the gear is up or down and changes it around.
Then it needs you to send false back to know that it shouldn't change it, and then next time you send true it will again evaluate whether the gear is up or down and change it...? This is merely a guess though, but it should be quick to test it.

 

You forgot one thing:

In InputController.cs

After:

Code (CSharp):

1.             case InputSource.InputManager:
2.                 {
3.                     if ("" != ButtonName)
4.                         inputValue = Input.GetButton(ButtonName);
5.                     else
6.                         Debug.LogWarning("InputController - Trying to get input from InputManager with no button defined.");
7.                 }
8.                 break;

Add:

Code (CSharp):

1.             case InputSource.Rewired:
2.                 {
3.                     if ("" != ButtonName)
4.                     {
5.                         inputValue = Rewired.ReInput.players.GetPlayer(0).GetButton(ButtonName);
6.                     }
7.                     // else
8.                     // {
9.                     // Debug.LogWarning( "InputController - Trying to get input from InputManager with no axis defined." );
10.                     // }
11.                 }
12.                 break;

 

Is there a reason why the latest update has been waiting approval in the asset store for so long? Does it keep failing? Loath to spend so much money on something that the asset store keeps telling me is for an earlier version of Unity, in case it doesn't work as advertised.

 

Bug or By Design:

Reproduce:
1) Create a wing w/ attached control surface.
2) Set the 'Root Hinge Offset' of the control surface to 100.
3) Set the 'Tip Hinge Offset' of the control surface to 50.
4) Notice that on full control inputs the updated wing segments become seriously twisted on deflection.

Is this by design? Is this an error with the flight model?

Thanks!

 

I think it's a bit of both.

If you have a square wing and your hinge is 0.5 at the tip and 1 at the base, then you end up with an angled hinge.

In reality, that would add extra forces in a lateral direction.

In UnityFS, and to be fair, in most computers you can fit on, under, or around a desk, the vast majority of the computation is in 2D slices. Each segment of the wing is sampled and rotated around.

The visual, physical equivalent I suppose would be to think of the wing with metal strakes along each segment line.

 

Got it. Thanks.

 

Are there any written docs that have the same info as the videos? Love the videos but I am much faster with a pdf

Thank you

 

I have added a blank gameobj, added the Wing component and made it a child of Dynamics in the Basic template.

I notice that my Handle seems to slide away from the center of my gameobj. Why is this happening?

Here is an image of what is going on. You can see the handle is very far from the object.
http://imgur.com/a/Kqbro

This only occurs when the gameobj is a child of Dynamics, it does not happen otherwise.

 

Has anyone been able to integrate mobile controls to the app? We're using Unity5.4, and the instructions don't all seem to apply.

 

Last update over 2 1/2 years ago? Unity version 3.5.0?
Remove your dead asset from the Asset Store.

 

My experience with this has been less than ideal. I too got taken in by the post about downloading a fix for unity 5 support whilst waiting for it to be approved on asset store.
I was unable to use the asset to model the physics of a glider. I could never get the glide to stabilise, angle of attack ranged wildly (dolphin type flying) and it never settled into a stable glide. Sent the developer an email and got a reply that it does support gliders and does work. Sent him a copy of a scene where i was having problems, got a reply saying it works. Asked for the 'working scene' back but no reply. Tried using the demo scene without an engine and had the same issues. Sent a couple of emails asking for money back.. never heard another word. Burned...be warned (your mileage might vary depending on how you use the asset).

 

I'm using it for a flight sim I'm working on.

Full disclosure - I got this I think sometime during early Unity 4.X releases, and have updated the project several times to later 4.X. For 5.0, I imported from the Asset Store, then brought my planes over via Unitypackage.

The one thing I know will probably not work right with 5.X is the wheels. That's a Unity change, and not for the better as far as I'm concerned.

In terms of the wings and engine scripts, they both seem to be working well.

I think the problem you're having is with the design of the glider maybe? Are you using dihedral/anhedral wings? Where are your CoG, CoM, CoL? Remember that all wings will have their own lift, drag and moment forces, and if they're off-axis in any way that can cause issues.

In most planes you want to keep the CoG ahead of the CoL, like a dart or arrow to keep it flying straight, but that leads to nose-down and acceleration, which, depending on the foil you're using for the tail, can cause nose-up and deceleration (porpoising). IRL you might have stability systems like auto-trim or gyro stabilization to handle things like this (even Kerbal Space Program, also on Unity, has their own SAS and Smart Automatic Stability System to handle this, which can also happen).

Are you able to post up some screenshots of the glider? iso, and ortho views with the lines visible, and the wing profile/s.

The other thing I can think of is to make sure you've got your physics set up correctly. Gravity's right, etc. I also like to bump up the physics refresh rate in Unity to >100Hz. Everything's smoother, more precise.

 

Thank Jesus for the hints,
the glider is stable in yaw and roll just not in pitch. It acts exactly as you described the issue with CoG and CoL. Unfortunately I have tried moving the CoG forward in very small increments but was unable to find 'a sweet' spot.
The porpoising was quite bad, but think the best point I found had the glider's sink rate varying between -1000 f/s and +50 f/s when flying 'hands off'. Not something that is usable in a glider game. The developer did say it is achievable after I sent him my 'test scene' but when asked to send the 'working modified scene' back to me, he disappeared from planet Earth.
I will try increase the physics refresh rate and see if that makes a difference.
In the meantime I abandoned UnityFS, tried to get my money back, failed and just de-installed it. I am now a lot more careful about what assets I purchase, after getting burned with this asset, I now only purchase assets that follow these rules: they have to be kept upto date, the developer has to be active on the forums, the reviews have to be glowing.

After failing to get UnityFS to meet my needs, I am now developing my own 'flight engine' based around 'glider polar curves' to use instead. Whilst probably not as accurate/detailed as UnityFS, at least Im approaching something that is stable and usable.

 

Here's my 5-minute ghetto-glider

Bear in mind a few things:
1 - the wing foils are the same used in the example plane, so the main wing is the same foil as that stubby twinengined wing. So, not a great choice for something that needs low drag.
2 - drag coefficient set to 0.1, mass set to 500 I think.
3 - dropped from 1.1km altitude, and it was dropped flat, so it stalled, nosedived, then smoothed itself out
4 - the CG was set to just ahead of halfway of the main spar, in the bottom pics it's what's selected (see arrows)
5 - Wind and thermals set to 0, air density and temp set to sea level.

Main wings are 16m each (total ~32m). 5 degrees dihedral, and the wing root was ~20cm below the CG. Elevator and Rudder were about 10-12m behind the CG, mounted straight. Single, 4x1m skid instead of wheels.

Screenshots taken every 50m alt, in 250m it changed less than a degree of pitch and 0.1m of sink rate. Airspeed in m/s, same with Rate Of Climb, pitch in degrees. Continued with those numbers until is hit the ground. And, that's with no scripted trim / stability control at all. Just good design of a balanced aircraft.

I'll be honest, this seems more like user error on your part than anything wrong with UnityFS. Post up some screenshots of what you had, let's see what's wrong with it maybe?

 

At the moment Im working on other things to pay the bills, but took a couple of screen shots. Model is just some open source place holder, and far too many polygons. First image is just the unityfs stuff (+wheels for some reason). Second shows the model in place too.
I did try moving the cg up above the wing root and also turning on one of the physics settings that defaults to disabled in v5 and above (enabled in v4). The glider stabilizes, just the sink rate is poor.
Thanks for you help on this. Loved to see a copy of your glider to examine.

 

In terms of PC (or whatever) performance, quick tip - reduce the number of sections per wing element' where possible. The tail section on yours, seems like it could just be 1 section. Only need to use more sections when you've got more things like blown surfaces (prop wash) or control surfaces that only affect part of the element.

As for glider performance (re sinkrate, etc), the best advice would be to look at the drag, set in the main rigidbody - bring that to something reasonable. From there, check your airfoils. The ones I used, the ones from the example plane, are horrible. The main wing is designed to carry 6 passengers, the fuselage and 2 engines with limited wing surface area. WAAAAAAAAAAAYYYYYYY too draggy and too much lift for when you've only got a 1-man glider to carry. Also the tail foils are NACA 0900s I think; neutral (symmetrical) but really not too far off what the Red Bull air racers used, from memory. Too thick, too much drag.

Obvious tips - make sure everything is lined up. From static, there should be no AoA on any of the tail surfaces, that's just drag. And if there's more than a degree of static AoA on the main wings, that's draggy too. In your case, check the wing endplates, if they're pointing in or out, that's a lot of drag.

 

Real developers and artists have no need for constant updates on an asset that works fine throughout. Unitrolls, however, require constant updates as it makes them feel like they're opening a birthday present when the updates come.

 

Even real developers have a birthday once a year and hope to get presents at Christmas as well, not have to wait over 2 1/2 years for one, especially if the last one no longer works unless you mod it with a patch that isn't officially approved.
Actually real developers wouldn't use an asset, they would code their own.

 

I can see all these bugs on my computer with Unity 5.4.4f1. The input device bug #469, moving handle #489 and of course the incompatible undo function. When I find a bug, I'm testing with the original fixed code (done by the unity5 import function) and a drastic modified version by me.

I have removed the input manager and the undo manager. I have also reduced the editor scripts to the necessary parts. So I have fixed bugs #469 and #489.

Now I cannot import own airfoils because the uFS Airfoil Importer throws out only errors instead of the curves. Has someone fixed this already?

I'm very angry about Chris (developer of UnityFS) because he replies only when I don't report a bug. It seems like we will never ever get a Unity5 compatible version of uFS from him. I have purchased uFS only a few months ago.

And be careful with the simulation result! uFS is not able to simulate aerobatic figures like knife edge, hoover, snap roll and so on.

The center of gravity is also not simulated because no matter where it is, it does not simulate the important trim function. Normally, you can move the CoG backwards a bit and then you don't have to pull elevator for flying horizontal. I think this is also described in post #495.

 

Yeah the input system is pretty strange, not too hard to bumble through and get working, but def something that needs to be updated. Same with the wheels, I think.

I can say for sure though that you CAN use it to handle knife edge and other 3d movements, but you will have to manually give the body of the plane some sort of lifting body effect. Otherwise it's like expecting a plane to take off when the only wing you've got it your right elevator.

The CG IS simulated. I'd been moving it and got it to roughly the same spot. From memory, you can play and in the inspector adjust the transform bit by bit and it'll update the handling. So if you want to trim that way you should be fine...