I've been writing a simulator - initially exclusively for R/C gliders (since that what I fly these days, when I can), but I've started adding powered planes in too, and in the last few days I've been putting in support for control-line. It's a "full" 3D simulation - aerodynamics with prop wash effects, simulation of flow through the propeller (including stalling), gyroscopic effects etc, and the C/L mode uses a max-distance constraint where control is affected if the lines go slack.

I built and flew (probably more time building and repairing than actual air time!) a handful of C/L planes many years ago, including a couple of Peacemakers. However, whilst I can remember the world whizzing past pretty quickly, I can't really remember the details in order judge how realistic what I've got is.

So this post is a request for some assistance in getting it as realistic as possible. The simulator runs under Windows (also linux with Wine, and Mac using PlayOnMac, I'm told), as well as on Android/iOS devices.

Here's the current state of the C/L mode (I suggest increasing the quality settings and viewing larger!):

Peacemaker - Control-line simulation in PicaSim (1 min 3 sec)

Please send me a message if you want to try this version (not released yet) of the sim and might be able to give me some feedback/answer a few questions.

The main website is here: http://www.rowlhouse.co.uk/PicaSim

Thanks - Danny

Looks pretty good. It flies about the same speed as my real Peacemaker!

I'm assuming it uses an RC transmitter controller though... Would be good if there were some way of using a handle. That way if you include some slower flying Stunt models people could practice their F2B pattern flying too!

nice work.

Yes it's being flown with a controller - actually an xbox controller (because it's so convenient), though any USB controller will work. You can also you a mouse, which is actually OK since there's only one axis to worry about. On the Android/iOS versions you can control it by tilting the device - though in practice I don't find that works well (touch using the on-screen joysticks is better).

I'm sure anybody who's capable of making a C/L plane could put together a little handle/pulley system that wiggles a joystick!

The lines here are probably a bit long (20m = 65 feet) - one of the things that doesn't seem right is that the plane doesn't drop out of the sky more if you try fancy stuff right up above your head - it will drop a bit, but then quickly recover. I recollect losing more control than this - and suspect it might be that it's missing the line weight/drag (which I should be able to put in).

My other concern apart from the line simulation is that if I increase the elevator throws any more, tight loops become unstable - it tends to enter a wobble where it yaws left/right which can send the plane flying inwards (i.e. lose control). I'm not sure if this is realistic - or indeed 100% sure what the cause is. It could be:

1. gyroscopic effects due to the prop - if you pitch up really hard you'd expect to see a yaw to the right - but the line tension will obviously counteract that and could cause oscillation

2. The elevator acts beyond the left wing tip (obviously!) so you can see (poss not in this video) if you pitch up, the plane rolls a little to the right. This makes inside/outside loops different. Rolling to the right sounds good - but as above, the line tension can turn this into an oscillation

3. I realised today the plane is a bit heavy - about 680g I think (24 oz). I need to check, but it might be that if you pitch up really hard the wing is stalling asymmetrically - given the yaw/roll points above - this could make it snap roll to the left.

4. It doesn't yet model the bellcrank setup - so in a real plane a yaw (assuming the handle stays fixed) will introduce some elevator movement - I think so a yaw to the right would make it pitch down. I don't know if that feedback would increase or decrease the stability (I should be able to put this in).

This is all making me want to get the old plans and engines down from the attic!

The flying simulation looks really quite good. The thing about control line flying though is how physical it is. You really do have to move your hand/wrist/arm around, as well as rotating your body and you can feel how the model is pulling and reacting to control inputs and changing wind direction. The nearest equivalent is force feedback on a joystick. The physicality and the feel of the handle are probably going to be much more difficult to try and replicate than the aerodynamics of the model flight. A Nintendo Wii controller with force feedback and attached to a variable tension spring?

If anybody wants to try it, this unreleased version is here. Google Chrome is causing me big headaches by flagging it as unsafe (simply because it's an installer that hasn't been downloaded much) - which is extremely annoying. All I can say is - it's totally virus free.

I agree about the physicality... but equally I've been surprised at how well controls transfer from one thing to another - at least for me. For example, whilst writing the first version of my R/C gliding simulator years ago, I learnt to fly by controlling the sim using the computer mouse. At the same time I was building my first glider - and as a result of using the sim with the mouse was able to fly completely securely (including inverted, on a rather non-aerobatic model!) basically from the outset. On the other hand... I agree R/C controls aren't quite the same as spinning round with a mad thing on the end of 60 feet of wire

By the way - one useful piece of info would be what's the typical weight of C/L control wire?

A little (actually - quite a lot!) more progress:

Control-line simulation in PicaSim (5 min 0 sec)

Looks great. Have you considered a program for judging F2B Aerobatics?

These days finding suitable judges can be quite difficult. A program that videos the flight, then compares it to the "perfect" pattern and scores it would make judging, easier, fairer, more consistent and perhaps solve a few issues in stunt judging like the halo effect.

Cheers

Originally Posted by warrenlead

Looks great. Have you considered a program for judging F2B Aerobatics?

No, I haven't! It's been a long time since I've flown C/L (and even then not a huge amount) - having worked on the simulation I'm quite tempted to have another go... but I think it's a lot harder finding people/places to fly these days. When I was young I could get my dad to launch... now the tables have turned I'm not sure my 5 year old is quite up to launching something sporting my trusty PAW 19

I've now released a version of PicaSim with the C/L simulation - at least on Android and Windows - download (free) from here*:

http://rowlhouse.co.uk/PicaSim/download.html

Feedback would be very much appreciated. It will be in the iOS version in a week or two, I expect.

Thanks - Danny

* Google Chrome (and possibly IE) are flagging up the Windows download - all I can say is that its completely safe and the warnings are bogus, but cause me a big headache. It's a combination of it being an installer and not yet downloaded much (so this hits whenever I make a new release, for a couple of weeks). A lot of other small-scale developers have a similar problem,

I'll give it a go and I look forward to seeing the iOS version so I can fly on my iPad Mini! Will let you know how it goes.

Cheers

The view without the trees works better for me. The lack of color on the plane makes it hard to distinguish from the trees. That may be a real world problem at some flying sites, but a beginner needs all the help they can get. I'm way past senior citizen status and my eyesite may not be the best, but I now have a Brodak Smoothie and Evolution 36 in my shop!

McD

I am very impressed with your sim, I was looking for a control line sim myself, actually an electric RTP sim as that seems the only option for me now, in my limited space. Flyline RC is a good system for me but too expensive at the moment.

How on earth did you write it? To make it run on WINE as well, which I use for CRRCSIM. What is the language you used?

Some comments on the flight model: the graphics are fine, however there are some effects that seem to be not modelled. See CRRCSIM and try it out for realism. Particularly the moments of inertia are not modelled well, the model pitches up and down like it had no weight at all: It should be slow to start pitching up, and then once it pitches up, slow to pitch down again, relatively. Also, did not try it out, but what happens when the model gets too high, the lines go slack and it heads towards its pilot at full speed?

Can you simulate a racetrack pattern - I can think of several ways to do this with RTP as well as C/L.

Again: amazing achievement!

Originally Posted by knlever

How on earth did you write it? To make it run on WINE as well, which I use for CRRCSIM. What is the language you used?

Glad you like it! It's written in C++ everything from scratch, except that it uses the Marmalade SDK for some cross-platform functionality (mainly for the user-interface, but it also lets me develop for iOS without needing a Mac!).

there are some effects that seem to be not modelled. See CRRCSIM and try it out for realism. Particularly the moments of inertia are not modelled well, the model pitches up and down like it had no weight at all...

Which plane(s) were you trying? There are two parts to getting the behaviour right - one is the core simulation itself, and the other is setting up the planes (most of which, to be honest, are at least a little outside my direct experience). I'm pretty sure that the moments of inertia are pretty close for the majority of the planes, but other things could give the effect you noticed - e.g. the servo speed, CG position or tail area... The Extra CL plane is (if I remember right) quite lightweight - probably more so that anybody would really build a C/L plane!

Also, did not try it out, but what happens when the model gets too high, the lines go slack and it heads towards its pilot at full speed?

It's a full 3D simulation, including using "rope" physics for the control line - the plane will drop out of the sky if it doesn't have enough speed to keep the lines taught, and the elevator control depends on there being tension in the line. It only simulates one line, so it doesn't represent twist - I might add that...

Can you simulate a racetrack pattern - I can think of several ways to do this with RTP as well as C/L.

My plan is to make a separate app/program for C/L that would include combat and aerobatics, and maybe some other "game" modes. I think it could really suit hand-held devices, and be quite accessible since there's only one control, and none of the orientation problems of full R/C.

Anyway - thanks for trying it!

Cheers - Danny

Originally Posted by kingsflyer

The view without the trees works better for me.

Yes - I agree really. I think I'll add another panoramic site that has less clutter in the background.

Cheers - Danny

You have a very good looking sim in the works here. Way over my head how you did it!

Suggestion:

For the best simulation (and thus aid us with muscle memory), it really needs to be linked to a USB joystick. That way pulling back on the joy stick replicates pulling "Up" control on a handle, etc.

Quick Note: Downloaded and installed, (WINE, Ubuntu 10.04 LTS - the Lucid Lynx ) however:

"This build does not support SW rendering no HW support" or something to that effect

and some other memory error.

Does the fact that I am running a Dell D800 with a 32 MB Nvidia N96 type card anything to do with it - but hang on if it will run on am Android tab and phone maybe...

Will install an Android emulator and get back to you... 420 MB seems not worth it: any options?