Board index FlightGear Development Aircraft

JSBSim Piper J-3 Cub  Topic is solved

Questions and discussion about creating aircraft. Flight dynamics, 3d models, cockpits, systems, animation, textures.

Re: JSBSim Piper J-3 Cub

Postby erik » Sat Aug 05, 2017 8:03 pm

dany93 wrote in Sat Aug 05, 2017 1:50 pm:That is SH = 24.5 Sq. Ft. [horizontal stabilizer + elevator] area.
I just discovered it in the FDM when you were responding.
I also found these areas in the J3Cub Owner Manual, Piper Cub J3C-65, p.12:
Stabilizer: 15.1 Sq. Ft,
Elevator: 11.7 Sq. Ft.

In this case, see my previous message, this area is excessive when the pilot pulls the stick to pitch up.

You are making a mistake: the horizontal tail is not pushed creating drag, instead it pulls itself using lift.

Erik
erik
 
Posts: 1567
Joined: Thu Nov 01, 2007 1:41 pm

Re: JSBSim Piper J-3 Cub

Postby Thorsten » Sat Aug 05, 2017 8:14 pm

Oh good....post a link to your shuttle FDM and I'll have a look at how YOU do it, as you're not posting an example of 'good practice'...


You really do walk into these things...

I actually did post a graph of the pitching moment due to elevon deflection for precisely that purpose - I'm sure you can find it reading back in the thread (hint - page 32). :mrgreen:

So, you can see that in general there's alpha dependence, the dependence on deflection angle is non-linear and there's Mach dependence (and any FDM which doesn't have these characteristics does 'something wrong' - the question is just how badly, for instance unless you approach the transonic region Mach dependence may not matter, the moment may be rather close to linear...).

You may in general even expect that the effect of one airfoil depends on the deflection angle of another (again, question is, how much).

I suspect the most important thing right now is - understand the data you're using (disclaimer - I haven't looked at the input for the Cub), If the data is from a reputable source (say an engineering paper, a thesis,...) then you may assume the author knew what he was doing, and then you need to use the same conventions (reference areas and lengths) if you use the coefficients - you may not use hand-waving arguments or scalings of the reference areas to 'just' change them 'to make it more realistic' - you'll just make it wrong if it differs from the source.
Thorsten
 
Posts: 11470
Joined: Mon Nov 02, 2009 8:33 am

Re: JSBSim Piper J-3 Cub

Postby erik » Sat Aug 05, 2017 8:56 pm

Bomber wrote in Sat Aug 05, 2017 3:21 pm:So are you now agreeing that the way it's been done here at FG for the past 15 years could have been done better ?

Well, the f16 is one of the oldest models in FlightGear and it always had:
Code: Select all
function name="aero/coefficient/CmDh">
    <description>Pitch_moment_due_to_horizontal_tail_deflection</description>
    <product>
     <property>aero/qbar-psf</property>
     <property>metrics/Sw-sqft</property>
     <property>metrics/cbarw-ft</property>
       <table>
        <independentVar lookup="row">aero/alpha-rad</independentVar>
        <independentVar lookup="column">fcs/elevator-pos-rad</independentVar>
        <tableData>
                 -0.4360 -0.2180  0.0000  0.2180  0.4360
         -0.1750  0.2050  0.0810 -0.0460 -0.1740 -0.2590
         -0.0870  0.1680  0.0770 -0.0200 -0.1450 -0.2020
          0.0000  0.1860  0.1070 -0.0090 -0.1210 -0.1840
          0.0870  0.1960  0.1100 -0.0050 -0.1270 -0.1930
          0.1750  0.2130  0.1100 -0.0060 -0.1290 -0.1990
          0.2620  0.2510  0.1410  0.0100 -0.1020 -0.1500
          0.3490  0.2450  0.1270  0.0060 -0.0970 -0.1600
          0.4360  0.2380  0.1190 -0.0010 -0.1130 -0.1670
          0.5240  0.2520  0.1330  0.0140 -0.0870 -0.1040
          0.6110  0.2310  0.1080  0.0000 -0.0840 -0.0760
          0.6980  0.1980  0.0810 -0.0130 -0.0690 -0.0410
          0.7850  0.1920  0.0930  0.0320 -0.0060 -0.0050
        </tableData>
       </table>
    </product>
   </function>

So when the data is available it is being used.

Erik
erik
 
Posts: 1567
Joined: Thu Nov 01, 2007 1:41 pm

Re: JSBSim Piper J-3 Cub

Postby Bomber » Sat Aug 05, 2017 9:28 pm

Ok erik, I accept that there are planes that are well documented with actual data.... modern planes. The example given is one of a moving h-stab and not of a h-stab and elevator combination...

Got any data for a wwi or wwii era plane ?

Please don't use the p51.... it's not a good example.
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1934
Joined: Fri Dec 14, 2007 7:06 pm
OS: Windows XP and 10

Re: JSBSim Piper J-3 Cub

Postby Bomber » Sat Aug 05, 2017 9:37 pm

Thorsten...... what dihedral does the shuttle wing have ?
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1934
Joined: Fri Dec 14, 2007 7:06 pm
OS: Windows XP and 10

Re: JSBSim Piper J-3 Cub

Postby dany93 » Sat Aug 05, 2017 9:57 pm

erik wrote in Sat Aug 05, 2017 8:03 pm:You are making a mistake: the horizontal tail is not pushed creating drag, instead it pulls itself using lift.

Sorry, I don't understand this phrase at all. Can you explain with more detail?

What I was saying is that, when the elevator is deflected (in fact, up or down), the aircraft attitude pitches up or down, hence an alpha angle appears. The area which counteracts this pitch up or down of the aircraft as a function of alpha (responsible for the pitch_alpha moment) is the [horizontal stabilizer] area. Not the [horizontal stabilizer + elevator] area.
dany93
 
Posts: 769
Joined: Mon Sep 07, 2009 3:43 pm
Location: France (Paris region)
Version: 2018.4.0
OS: Linux Mint 18 (64 b)

Re: JSBSim Piper J-3 Cub

Postby Alant » Sat Aug 05, 2017 10:36 pm

To put things in perspective.

The number of aircraft within Flightgear that have detailed aerodynamic data available can probably be counted on your fingers.

The science/art of estimating aerodynamic properties and weigh/moments of inertia is not trivial and can not be picked up overnight.

For conventional subsonic aircraft (e.g. the Cub), the best next step after Aeromatic is probably Datcom. You will still need some knowledge of aerodynamics and be able to accurately estimate weights, c.g. position and moments of inertia.

Flightgear with JSBSim is very capable. As an example my own TSR2 simulation has an FDM with over 7000 lines of JSBSim code derived from original manufacturer´s reports. I have checked the simulation against trim curves and time response traces over the full flight envelope at all weights and c.g s.

When you make changes, please add comments to the code so that others can see what you have done.

Don´t let me put you off. Flightgear has far too many "toy" models and needs more developers who can produce accurate simulations.

Alan
Alant
 
Posts: 940
Joined: Wed Jun 23, 2010 5:58 am
Location: Portugal
Callsign: Tarnish99
Version: from Git
OS: Windows 10

Re: JSBSim Piper J-3 Cub

Postby Thorsten » Sun Aug 06, 2017 4:07 am

I see the red herring has been pretty successful. It's fairly easy to grab a random plane and criticize it's flaws - but Simon, before you do so - any progress on the equilibrium glide of the ASK-13? The sink is way too high - on what should be the optimum glidepath (around 90 km/h) I sink with more than 2 m/s (!) - much faster if the speedbrakes are deployed, so that's not the problem - the vspeed ought to be less than half of that (compute the glide ratio, it comes out lousy). The bird flies like a brick - nothing like reality. So please don't make us believe everyone else has problems modeling FDMs and that you can fix them...

Otherwise I would suggest again to study how the data are organized. Remember that at the end it's a parametrization of the physics, not the physics what is in coefficient functions. So while reference lengths and areas can change in real physics, the coefficients may not be organized that way (they usually are not) and it is quite possible to capture the effects of hstab and elevator in one deflection-angle independent term and one term that depends on alpha and deflection angle (and I believe that is the usual way to organize things).

Got any data for a wwi or wwii era plane ?


I'm not going to do your research for you, but I am fairly sure people have done papers on them. If not, you can still try CFD.
Thorsten
 
Posts: 11470
Joined: Mon Nov 02, 2009 8:33 am

Re: JSBSim Piper J-3 Cub

Postby Thorsten » Sun Aug 06, 2017 5:28 am

The number of aircraft within Flightgear that have detailed aerodynamic data available can probably be counted on your fingers.

The science/art of estimating aerodynamic properties and weigh/moments of inertia is not trivial and can not be picked up overnight.


There would seem to be a relation though - studying the cases where there is detailed data give you insight into how certain effects play out. Having seen a couple of functions of elevator action against alpha allows you to see a pattern and will make your guess better if you have to guess.

As for the number of planes for which is data, I'm not sure either way, but for instance all NASA data is public domain, so any craft NASA ever put into a wind tunnel ought to be found on the document server. Also I would suspect that at least academic researchers tend to be co-operative and will send you copies of their papers if you ask nicely.

Otherwise, well, planes come in degrees of realism. For high realism we generally lack the resources to do it ourselves (a wind tunnel or high-performance computing cluster) so we need to get lucky and have someone else done it for us and published it. The next lower level would be low-res CFD for the full body which some of us might be able to do, and then we come to data-driven cooking-recipies like scaling laws from similar planes etc, followed even further down the scale by cooking-recipies based on 'the feel'.

In the absence of good data a data-driven cooking recipe might be the best one can do, it just shouldn't be confused with the higher realism levels.

(And of course a wrong cooking recipe which double-counts or drops some effect altogether doesn't translate into realism...)
Thorsten
 
Posts: 11470
Joined: Mon Nov 02, 2009 8:33 am

Re: JSBSim Piper J-3 Cub

Postby erik » Sun Aug 06, 2017 7:21 am

dany93 wrote in Sat Aug 05, 2017 9:57 pm:What I was saying is that, when the elevator is deflected (in fact, up or down), the aircraft attitude pitches up or down, hence an alpha angle appears. The area which counteracts this pitch up or down of the aircraft as a function of alpha (responsible for the pitch_alpha moment) is the [horizontal stabilizer] area. Not the [horizontal stabilizer + elevator] area.

It's the airflow which generates the lift (up or downwards) which makes the horizontal tail want to go up or down. What you are describing is an airflow which instead forces the horizontal tail the other way. At some value of alpha the sum of the airflow generated lift and the pitching moment caused by the fuselage and main wing will become zero. But it will never go negative (like you are suggesting) unless something changes, like the pilot moves the elevator again. But then again it is change in lift which causes the difference.

Erik
erik
 
Posts: 1567
Joined: Thu Nov 01, 2007 1:41 pm

Re: JSBSim Piper J-3 Cub

Postby Bomber » Sun Aug 06, 2017 9:55 am

Thorsten wrote in Sun Aug 06, 2017 4:07 am:I see the red herring has been pretty successful. It's fairly easy to grab a random plane and criticize it's flaws -


Alan has explained that there are only a handfull of planes in FG of the 400 fdms that use published data.

Understand if you have data then use it.... but the 390 other planes in FG don't have published data.

So you're telling us we can see patterns using these handfull of planes to extrapolate to be used on others. Yes if they're of similar design and era but you can't use data for a supersonic fighters h-stab on a 1917 cloth covered bi-plane.


Thorsten wrote in Sun Aug 06, 2017 4:07 am: but Simon, before you do so - any progress on the equilibrium glide of the ASK-13? The sink is way too high - on what should be the optimum glidepath (around 90 km/h) I sink with more than 2 m/s (!) - much faster if the speedbrakes are deployed, so that's not the problem - the vspeed ought to be less than half of that (compute the glide ratio, it comes out lousy). The bird flies like a brick - nothing like reality. So please don't make us believe everyone else has problems modeling FDMs and that you can fix .


You're going to need an up to date version.... so you're going to need to go to FGMEMBERS and download the non-GPL fdm add-on before you critic my work...

But I ask again what's the dihedral on the shuttles wings, maybe someone else knows if you're not prepared to answer this simple question.

Simon.
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1934
Joined: Fri Dec 14, 2007 7:06 pm
OS: Windows XP and 10

Re: JSBSim Piper J-3 Cub

Postby Thorsten » Sun Aug 06, 2017 10:28 am

You're going to need an up to date version.... so you're going to need to go to FGMEMBERS and download the non-GPL fdm add-on before you critic my work...


No, I'm critiquing specifically the version you complained for ages you can't get committed to FGAddon and you confirmed I should commit even after I queried about potential problems with the FDM (the version 99% of the user-base is going to identify with your FDM-creation skill). A version you clearly judged as ready at some past point.

For all your Shuttle questions, I may refer you to Aerodynamic design of the space shuttle orbiter and references therein - I'm not interested in playing this game with you.
Thorsten
 
Posts: 11470
Joined: Mon Nov 02, 2009 8:33 am

Re: JSBSim Piper J-3 Cub

Postby Bomber » Sun Aug 06, 2017 10:37 am

Sorry.... are you saying that only fdm's that are 100% correct are deemed 'ready' to be added into FGdata...

I thought it was a repository of planes for people to have access too, download, try and critic.

So it's wrong with its sink rate, thanks for pointing it out and in the intervening 3 years since it was authored I've resolved it.

You know you can't make an omlete without breaking eggs, especially if you're doing things differently. There comes a time when you have to bite the bullet and publish a WIP version for people to review.

And as FG had decided to restrict the publishing of non GPL content to the widest audience 99% of its user base I've had to publish somewhere else..
Last edited by Bomber on Sun Aug 06, 2017 10:46 am, edited 1 time in total.
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1934
Joined: Fri Dec 14, 2007 7:06 pm
OS: Windows XP and 10

Re: JSBSim Piper J-3 Cub

Postby Bomber » Sun Aug 06, 2017 10:41 am

Thorsten wrote in Sun Aug 06, 2017 10:28 am:

For all your Shuttle questions, I may refer you to Aerodynamic design of the space shuttle orbiter and references therein - I'm not interested in playing this game with you.


What you're not interested in is answering a simple question....

Yet in arrogance you expect others to answer yours.

Who put you above everyone else ?
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1934
Joined: Fri Dec 14, 2007 7:06 pm
OS: Windows XP and 10

Re: JSBSim Piper J-3 Cub

Postby Bomber » Sun Aug 06, 2017 10:45 am

Well can someone tell me what the dihedral for this plane the Piper J-3 Cub is ?

It's a simple question.
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1934
Joined: Fri Dec 14, 2007 7:06 pm
OS: Windows XP and 10

PreviousNext

Return to Aircraft

Who is online

Users browsing this forum: AhrefsBot [Bot] and 4 guests