Board index FlightGear Development Aircraft

Hydrodynamics hydro/float-beam-ft

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

Re: Hydrodynamics hydro/float-beam-ft

Postby AndersG » Wed Mar 15, 2017 5:19 am

wlbragg wrote in Wed Mar 15, 2017 1:50 am:Why are there the amount of tables used for hydro/float/height-agl-ft?
What are the relationships, as in, height-agl-ft = 4 or 5 or 6?
Why does it start at 4 and end at 8?
Is any of this documented anywhere? If so I apologize for not doing the legwork to find it.
Code: Select all
<table>
   <independentVar lookup="row">hydro/float/pitch-deg</independentVar>
   <independentVar lookup="column">hydro/float/roll-deg</independentVar>
   <independentVar lookup="table">hydro/float/height-agl-ft</independentVar>
      <tableData breakPoint="4.0">
        ~~~~~~~~~
        ~~~~~~~~~
      </tableData>
      <tableData breakPoint="5.0">
        ~~~~~~~~~
        ~~~~~~~~~                           
      </tableData>
      <tableData breakPoint="6.0">
        ~~~~~~~~~
        ~~~~~~~~~                       
      </tableData>
      <tableData breakPoint="7.0">
        ~~~~~~~~~
        ~~~~~~~~~                           
      </tableData>
      <tableData breakPoint="8.0">
                         
      </tableData>



These guys hydro/float/pitch-deg, hydro/float/roll-deg and hydro/float/height-agl-ft describe the position and orientation of the vessel with respect to the water surface. The table encodes that in this position and orientation the vessel has this amount of buoyancy (or generates this moment w.r.t. the hydrodynamic reference point). This scheme is used for all the hydrostatic "coefficients". In the MFI-9B example roll is dropped since the effect of rolling one float about its reference point is much less than rolling two floats around the center point between them (thus ignoring that the float bodies themself is not rolled).

wlbragg wrote in Wed Mar 15, 2017 1:50 am:Also, could you explain what this means, what offset are you referring to?
Code: Select all
<fcs_function name="hydro/float/height-agl-ft">
  <description>
    The float height with waves and squat due to the transverse wave applied.
  </description>
    <function>
       <sum>
          <property>hydro/height-agl-ft</property>
            <!-- Account for the 6.5cm vertical offset of the float model. -->
             <value>-0.2133</value>


Which aircraft is this from? I suppose it is due to differences between the float model as run in Gerris and how it is incorporated in the aircraft. I usually try to avoid such offsets but it has sometimes been needed when working with existing 3d models. It can be pretty messy.
Callsign: SE-AG
Aircraft (uhm...): Submarine Scout, Zeppelin NT, ZF Navy free balloon, Nordstern, Hindenburg, Short Empire flying-boat, ZNP-K, North Sea class, MTB T21 class, U.S.S. Monitor, MFI-9B, Type UB I submarine, Gokstad ship, Renault FT.
AndersG
 
Posts: 2454
Joined: Wed Nov 29, 2006 9:20 am
Location: Göteborg, Sweden
Callsign: SE-AG
OS: Debian GNU Linux

Re: Hydrodynamics hydro/float-beam-ft

Postby wlbragg » Wed Mar 15, 2017 5:11 pm

Thanks Anders,

These guys hydro/float/pitch-deg, hydro/float/roll-deg and hydro/float/height-agl-ft describe the position and orientation of the vessel with respect to the water surface. The table encodes that in this position and orientation the vessel has this amount of buoyancy (or generates this moment w.r.t. the hydrodynamic reference point). This scheme is used for all the hydrostatic "coefficients". In the MFI-9B example roll is dropped since the effect of rolling one float about its reference point is much less than rolling two floats around the center point between them (thus ignoring that the float bodies themself is not rolled).

I understood that, but specifically what do the hydro/float/height-agl-ft 4 through 8 in this particular table refer to. I can't wrap my head around 4 foot, 5 foot, 6 foot. 7 foot and 8 foot above ground level as having any bearing on floats that are sitting on or below the water level. Those breakpoints for the different agl 4-8 are a reference to something, it can't be "feet above ground". Even an arbitrary reference point doesn't account for the difference between 4' and 8', that is a 5 foot difference. Anything affecting the floats is going to happen in a matter of inches, not feet. So I obviously don't understand what the breakpoint is?

Which aircraft is this from? I suppose it is due to differences between the float model as run in Gerris and how it is incorporated in the aircraft.

It is from the c172p, I want to know what the measurement 6.5cm is to and from?
Kansas(2-27-15)/Ohio/Midwest scenery development.
KEQA (2-27-15), 3AU, KRCP Airport Layout
Intel i5 3570K AMDRX480
User avatar
wlbragg
 
Posts: 5194
Joined: Sat Aug 25, 2012 11:31 pm
Location: Kansas (Tornado Alley), USA
Callsign: WC2020
Version: next
OS: Win10/Linux/AMDRX480

Re: Hydrodynamics hydro/float-beam-ft

Postby AndersG » Wed Mar 15, 2017 10:03 pm

wlbragg wrote in Wed Mar 15, 2017 5:11 pm:Thanks Anders,

These guys hydro/float/pitch-deg, hydro/float/roll-deg and hydro/float/height-agl-ft describe the position and orientation of the vessel with respect to the water surface. The table encodes that in this position and orientation the vessel has this amount of buoyancy (or generates this moment w.r.t. the hydrodynamic reference point). This scheme is used for all the hydrostatic "coefficients". In the MFI-9B example roll is dropped since the effect of rolling one float about its reference point is much less than rolling two floats around the center point between them (thus ignoring that the float bodies themself is not rolled).

I understood that, but specifically what do the hydro/float/height-agl-ft 4 through 8 in this particular table refer to. I can't wrap my head around 4 foot, 5 foot, 6 foot. 7 foot and 8 foot above ground level as having any bearing on floats that are sitting on or below the water level. Those breakpoints for the different agl 4-8 are a reference to something, it can't be "feet above ground". Even an arbitrary reference point doesn't account for the difference between 4' and 8', that is a 5 foot difference. Anything affecting the floats is going to happen in a matter of inches, not feet. So I obviously don't understand what the breakpoint is?


The relevant range of height-agl-ft depends on where the hydrodynamic reference point is located. For the c172p it seems to be quite high, 38.5 in over the FDM X-Y plane.
The resolution of the static coefficients is also quite low but it is interpolated between the tables. The larger height-agl-ft values are needed for extreme attitudes, but perhaps not as much as 8 ft (if you look at the buoyancy coefficient you can see when the values goes very small).

wlbragg wrote in Wed Mar 15, 2017 5:11 pm:
Which aircraft is this from? I suppose it is due to differences between the float model as run in Gerris and how it is incorporated in the aircraft.

It is from the c172p, I want to know what the measurement 6.5cm is to and from?


The c172p FDM + 3d model setup is very messy. I don't remember exactly where those 6.5cm came from but there are offsets (IIRC both translations and rotations) at several levels in the model files making the relation between the FDM structural frame and the 3d model complicated. I probably missed one of those when setting up the 3d model of the floats for the Gerris runs and had to compensate for that later. I don't recommend starting from the c172p files and if you have any possibility to affect the 3d model - FDM relation make sure to make it as easy as possible. I prefer to use the same origin and no rotations between them.

[EDIT]
I know this is not a popular suggestion given the amount of tuning you have put in already but I'd still recommend starting over from the MFI-9B hydrodynamics file.

1. Position the hydrodynamic reference point 0.5m above the point of the step and set metrics/hydro-beam-ft to the distance between the center lines of the floats and hydro/float-rp-arm-y-in[x] to -/+ half of that.
2. Measure the angle between the float and the X-axis of the structural frame and set this offset in hydro/floats/pitch-deg[X] (it is 3 deg for the MFI-9).
3. Fix the contact points and the pointmasses in the Enable/disable channel, or, as I'm sure you have something similar or more advanced in the J3 Cub, connect hydro/active-norm to your existing logic.
4. You should now have the static part of water operations done for floats shaped like those of the MFI-9B.
5. Go through the file and replace any function with planing in its name with the corresponding hack from the old j3cub hydrodynamics file. Each function there will now be two - one for each float - so add [0] and [1] to the properties.
6. Check that it floats ok - it should if the step of your floats is positioned as IRL for this type of float (the MFI-9B usually uses the same size of floats as the J3 Cub according to the information I've found).
7. Some retuning is very likely to be needed.

I think that'd be it.
[/EDIT]
Callsign: SE-AG
Aircraft (uhm...): Submarine Scout, Zeppelin NT, ZF Navy free balloon, Nordstern, Hindenburg, Short Empire flying-boat, ZNP-K, North Sea class, MTB T21 class, U.S.S. Monitor, MFI-9B, Type UB I submarine, Gokstad ship, Renault FT.
AndersG
 
Posts: 2454
Joined: Wed Nov 29, 2006 9:20 am
Location: Göteborg, Sweden
Callsign: SE-AG
OS: Debian GNU Linux

Re: Hydrodynamics hydro/float-beam-ft

Postby AndersG » Sat Mar 18, 2017 4:01 pm

I have update the MFI-9B repository on my GitHub with higher resolution hydrostatic tables. There seems to be some wires crossed in the MFI-9 hydrodynamics, though since the float wakes are crossed (lift the left float and the wakes on the right one stops and vice versa) - and this swap goes further into the FDM - it is not just a mistake in the wakes.
Callsign: SE-AG
Aircraft (uhm...): Submarine Scout, Zeppelin NT, ZF Navy free balloon, Nordstern, Hindenburg, Short Empire flying-boat, ZNP-K, North Sea class, MTB T21 class, U.S.S. Monitor, MFI-9B, Type UB I submarine, Gokstad ship, Renault FT.
AndersG
 
Posts: 2454
Joined: Wed Nov 29, 2006 9:20 am
Location: Göteborg, Sweden
Callsign: SE-AG
OS: Debian GNU Linux

Re: Hydrodynamics hydro/float-beam-ft

Postby wlbragg » Sat Mar 18, 2017 4:56 pm

the float wakes are crossed


Oops, are you planning on tracing that down anytime soon? If so, I think I will wait until you sort it out to use it for the Cub.

You think the higher resolution will help with the drop off of in speed due to friction and the settling into the water over a shorter time period. Or, do you think the tweak-factor/rpm table I used is a viable alternative?
Kansas(2-27-15)/Ohio/Midwest scenery development.
KEQA (2-27-15), 3AU, KRCP Airport Layout
Intel i5 3570K AMDRX480
User avatar
wlbragg
 
Posts: 5194
Joined: Sat Aug 25, 2012 11:31 pm
Location: Kansas (Tornado Alley), USA
Callsign: WC2020
Version: next
OS: Win10/Linux/AMDRX480

Re: Hydrodynamics hydro/float-beam-ft

Postby AndersG » Sat Mar 18, 2017 6:28 pm

wlbragg wrote in Sat Mar 18, 2017 4:56 pm:
the float wakes are crossed


Oops, are you planning on tracing that down anytime soon? If so, I think I will wait until you sort it out to use it for the Cub.


I'll try to do that ASAP. I think you can safely start experimenting based on the MFI-9 hydrodynamics but maybe wait a bit before switching over on the public branch.
It'll take a while to retune it to get the behaviour you like too. The swap/cross over of the sides is not obvious to me in the handling (meaning the moments probably come out the right direction by double sign error).

wlbragg wrote in Sat Mar 18, 2017 4:56 pm:You think the higher resolution will help with the drop off of in speed due to friction and the settling into the water over a shorter time period. Or, do you think the tweak-factor/rpm table I used is a viable alternative?


Using the MFI-9 approach as is? No, since those tables are only concerned with the hydrostatic behaviour, but when reusing the static tables for the drag possibly so.

For the tweak factor table, I think I had rather indexed the table using water speed or the amount of lift from the water or pitch - engine RPM would seem have a rather loose coupling to the water behaviour. The hydrodynamic functions I have in the files are very basic so tuning will have to be applied, see e.g. the wiki piece on the MTB drag. That one goes from 50 knot to below 10 knot in about 15 seconds when the engines are cut. That does seem to be a bit long but the drag curve matches the tow test rather well - OTOH maybe the trim/pitch ought to be different while slowing down compared to when accelerating and that could give higher drag.
Callsign: SE-AG
Aircraft (uhm...): Submarine Scout, Zeppelin NT, ZF Navy free balloon, Nordstern, Hindenburg, Short Empire flying-boat, ZNP-K, North Sea class, MTB T21 class, U.S.S. Monitor, MFI-9B, Type UB I submarine, Gokstad ship, Renault FT.
AndersG
 
Posts: 2454
Joined: Wed Nov 29, 2006 9:20 am
Location: Göteborg, Sweden
Callsign: SE-AG
OS: Debian GNU Linux

Re: Hydrodynamics hydro/float-beam-ft

Postby AndersG » Sat Mar 18, 2017 11:37 pm

I think I've solved the float cross over/swap problem. Now the per-float height-agl-ft properties and the resulting static and dynamic lift makes sense (in terms of which float it happens to). I did have to change the signs in the roll channel too. For the MFI-9B you need the latest GitHub version (I'll push to FGAddon ASAP) + the latest hydrodynamic-planing-floats system from fgdata. For the Cub development you can ignore the latter.

Unfortunately, the J-3 Cub seems to be too heavy for my ancient system (8+ years old; Intel Core2 Q9400, 8GB RAM, GeForce 9800GT 512MB) - FG got stuck in paging.
Callsign: SE-AG
Aircraft (uhm...): Submarine Scout, Zeppelin NT, ZF Navy free balloon, Nordstern, Hindenburg, Short Empire flying-boat, ZNP-K, North Sea class, MTB T21 class, U.S.S. Monitor, MFI-9B, Type UB I submarine, Gokstad ship, Renault FT.
AndersG
 
Posts: 2454
Joined: Wed Nov 29, 2006 9:20 am
Location: Göteborg, Sweden
Callsign: SE-AG
OS: Debian GNU Linux

Re: Hydrodynamics hydro/float-beam-ft

Postby wlbragg » Sun Mar 19, 2017 4:35 am

@AndersG

I'm sure you know when your froze or not, but just in case, I just timed how long it took to load the J-3 on my core duo 6GB GeForce 640GT and it took 1:47 to load.
Kansas(2-27-15)/Ohio/Midwest scenery development.
KEQA (2-27-15), 3AU, KRCP Airport Layout
Intel i5 3570K AMDRX480
User avatar
wlbragg
 
Posts: 5194
Joined: Sat Aug 25, 2012 11:31 pm
Location: Kansas (Tornado Alley), USA
Callsign: WC2020
Version: next
OS: Win10/Linux/AMDRX480

Re: Hydrodynamics hydro/float-beam-ft

Postby AndersG » Sun Mar 19, 2017 11:19 am

Ok, in this case the IO wait state and the freezing mouse pointer told the story of badly out of main memory. One my second try fgfs was killed by the kernel Out Of Memory reaper... I have 8BG RAM + a 15 GB paging partition so memory use must have been rather significant. I'll update fgfs to the very latest (its a week old or so) and try again.
Callsign: SE-AG
Aircraft (uhm...): Submarine Scout, Zeppelin NT, ZF Navy free balloon, Nordstern, Hindenburg, Short Empire flying-boat, ZNP-K, North Sea class, MTB T21 class, U.S.S. Monitor, MFI-9B, Type UB I submarine, Gokstad ship, Renault FT.
AndersG
 
Posts: 2454
Joined: Wed Nov 29, 2006 9:20 am
Location: Göteborg, Sweden
Callsign: SE-AG
OS: Debian GNU Linux

Re: Hydrodynamics hydro/float-beam-ft

Postby Thorsten » Sun Mar 19, 2017 2:32 pm

Unfortunately, the J-3 Cub seems to be too heavy for my ancient system


I think you could simply use a graphics program of your choice and scale down the textures - I bet they're the expensive bit here.
Thorsten
 
Posts: 11375
Joined: Mon Nov 02, 2009 8:33 am

Re: Hydrodynamics hydro/float-beam-ft

Postby wlbragg » Sun Mar 19, 2017 5:07 pm

you could simply use a graphics program of your choice and scale down the textures

Actually, Erik and I are considering doing this in the main repository. Erik just made a normals map and it is getting even more expensive.

I'm spending my vacation working on that today before I start back on my "real" job! :)
Kansas(2-27-15)/Ohio/Midwest scenery development.
KEQA (2-27-15), 3AU, KRCP Airport Layout
Intel i5 3570K AMDRX480
User avatar
wlbragg
 
Posts: 5194
Joined: Sat Aug 25, 2012 11:31 pm
Location: Kansas (Tornado Alley), USA
Callsign: WC2020
Version: next
OS: Win10/Linux/AMDRX480

Previous

Return to Aircraft

Who is online

Users browsing this forum: AhrefsBot [Bot], YandexBot [Bot] and 5 guests