fsdsv2 aileron animation

[Guest jlwoodward]

Problem: the aileron is split into 3 parts for better polygon smoothing. For the parent part, the axis is moved and rotated about 2 axes to align with the hinge point. The remaining 2 parts are linked to this part, r_airleron.1 etc.Result: the parent part stays in place and animates ok. The other 2 parts end up way out of position as soon as they are linked.I tried relocating the axis of each part to the same place before linking the parts. They still shift out of position, just to a different place. I tried the reset vertex rotation command but I don't understand its purpose, and it also just shifted the parts more.I'm lost. I thought I had better ask for help now that I spent 2 hours getting nowhere.Thanks,John Woodward

[Guest]

The parts rotate about their pivot point, not the center point of the object's axis you see on the gizmo. Just in case you are not aware of pivot points although it sounds as if you are talking about the pivot point.The aileron parts should all animate the same using stock animation if you give them names like r_aileron_1, r_aileron_2, r_aileron_3 and so on. All my flap hinge arms move the same amount this way. I oriented all their pivot points the same. FS ignores the part after the underscore, so you can have several objects move at once.I also tried an experiment where I GROUPED the parts I wanted to move all at once the same way. It appeared to work.I never link any parts (so they will move as a unit) unless I am doing my own keyframe animation. Maybe I should be?The center pivot point is a very useful tool. It would be difficult to find the exact center manually. And you often need to (re)center the pivot after modifying the dimensions and shape of an object.Steve

[Guest jlwoodward]

Steve,I think I am doing it the way you describe but there must be something I am overlooking. It sounds like you don't have to use the parent/child relationship as long as each component is named r_airleron_ etc (I was using a dot instead of underscore, maybe that was wrong).I am confused on setting the pivot point. For the r_airleron main part I aligned the x axis with the hinge line. Do you do that with the other components or leave them as is?My last attempt was to join all the parts, align the axis, then split the parts. However, when I linked them they scattered. Animation worked ok for the joined part.John Woodward

[Guest]

>Steve, >>I think I am doing it the way you describe but there must be >something I am overlooking. It sounds like you don't have to >use the parent/child relationship as long as each component >is named r_airleron_ etc (I was using a dot instead of >underscore, maybe that was wrong). >>I am confused on setting the pivot point. For the r_airleron >main part I aligned the x axis with the hinge line. Do you >do that with the other components or leave them as is? Although this image is from GMax, it does illustrate how the pivot point is aligned at the center of rotation (in this case, the middle of the leading edge of the aileron) and how the pivot axis angle should match the dihedral of the wing.I don't know how "link" is implemented with FSDS2, but in GMax the part(s) are effectively "glued" to the parent, and will not change relative position to the parent whenever it's moved or rotated.

[Guest]

I am designing in GMAX (frustrating but free---I was not going to design a/c until my curiosity got to me after discovering it on the CD). GMAX is so complicated I forget as much as I learn as I move thorugh the project. But as you do more you do get an intuitive understanding of how things work.Looking at the properties for objects, it appears that any object without a parent object is a child of the parent object (a virtual root object consisting of everything in the scene).In GMAX, you can open a dialog window called Select Objects, which acts as kind of object browser. It can show the object hierarchy as a tree view.

r_gear_strut  r_axle_housing	 r_tire	 r_tire_blurred	 r_tire_still

This shows how objects are linked to one another. I was incorrect to say that I only used linked objects with keyframe animation. It helps to create a link between the parts of landing gear for two reasons. You need a link for any future keyframe animation for retractable landing gear. I assume you do not necessarily need a link for steerable gear. But it probably helps because all parts can steer without having to give them weird names (I have not created a steerable gear yet, but will have to for my a/c). I like to create a tree of objects just so I can see how they relate.Looking at my a/c, each flap hinge is a group of objects named [flap_actuatorNN] starting at 00 (the programmer in me!). Each group contains a leading and trailing arm and a hinge cover. Opening one of the groups, I see that the front arm is called front_armNN (the NN is just to give them a unique number, it does not have anything to do with animation) and the trailing arm is called l_flap_NN. This does have importance in the FS a/c model.In the FS2002 MakeMDL SDK, it states that MakeMDL recognizes certain part names and FS handles these parts in various special ways, such as applying stock animation (usually simple rotation) to them. A note states that the part name is the portion that remains after any TRAILING numbers are removed. The numbers are of the form _### (underscore digit digit digit) up to an implied limit of three digits.Any part with a stock animation name automatically uses the corresponding animation. As far as I know, you may have several parts using the same name but differentiated by the trailing number and all will move the same in response to that sim value.My flap arms seem to bear this out because there are no links to be found between the trailing arms (three on each wing) and any other part and they are named l_flap_00, l_flap_01, l_flap_03 and so on as well as r_flap for the other wing.[Okay, I just re read the note. It states that parts with the part name and parts with a numeric suffix are considered the same aircraft part. So l_flap is the same part as l_flap_7 and this implies to me they will all respond the same, i.e. be animated the same, which is not necessarily true, but a good guess and it turns out to be correct AFAIK.]They all move in synchrony in FS.This was an easy way of animating all those flap actuators. You see that it does not matter what group the part I want to animate belongs to, they move because of the label they are given.However, to get correct flap movement I may decided to use a different arrangement. Although right now the flap actuators are logically grouped (trailing arm, flap, hinge cover) into a single actuator group (that can be treated as an object in its own right), there is a glitch. The only reason I can get away with this arrangement is that the flap trailing arm has it's pivot point below the flap. As the flap rotates the arm rotates the same amount appearing to support the movement. But it really does not and at a certain flap angle the arm will poke through the top of the wing.My proposed solution for this will be to group the arms with the flap into a single grouped object. The arms and the flap will rotate back together the arms staying "attached" to the bottom of the flap. I have already tested this and giving the group a name of l_flap moves the whole group around its composite axis.I have not given my wings their dihedral yet. I am unfamiliar with how the dihedral is applied to a wing. I hope that I can just angle the wings up a bit (the dihedral is small, about 1.7 etc.). Thanks for reminding me to set the pivot angle to match (and I suppose all the wing parts like aileron, flap, trim will have to match).Because I made up my flap parts independent of the a/c model, I failed to realize their pivot points would not be at the correct orientation once on the a/c. So I had to manually move them down to the bottom of the arm and rotate them so they move in the correct direction. This was tricky because of their angled orientation on the a/c. It required moving the pivot down by eye and they are probably not all exact or matching as they would be had I set the pivot before adding to the a/c.Also, I have not closely inspected my control surface pivot points to see if they hinge properly. I think the rudder is to one side plane of the object and will have to be centered so it visually hinges properly. But they mostly look correct.Steve

[Guest]

Hi Steve!I'm not going to comment on everything, but only a selection for which I have something (hopefully useful!) to say:>Looking at the properties for objects, it appears that any >object without a parent object is a child of the parent >object (a virtual root object consisting of everything in >the scene). That is absolutely correct!>This shows how objects are linked to one another. I was >incorrect to say that I only used linked objects with >keyframe animation. It helps to create a link between the >parts of landing gear for two reasons. You need a link for >any future keyframe animation for retractable landing gear. >I assume you do not necessarily need a link for steerable >gear. But it probably helps because all parts can steer >without having to give them weird names (I have not created >a steerable gear yet, but will have to for my a/c). I like >to create a tree of objects just so I can see how they >relate. It is useful to think of 'linking' as if you were using Duco Cement to glue a plastic model together! Each part, when linked to a parent, becomes a part_of_that_part, as though you had 'glued' them together. When the parent is moved, the 'linked parts' will follow along and maintain the exact relative position to the parent they had when first linked... [Okay, I just re read the note. It states that parts with >the part name and parts with a numeric suffix are considered >the same aircraft part. So l_flap is the same part as >l_flap_7 and this implies to me they will all respond the >same, i.e. be animated the same, which is not necessarily >true, but a good guess and it turns out to be correct >AFAIK.]This is a very important point, because the improper use of 'stock animation names' can lead to some behavior which is undesired! (as several have already found out...)!As you may note from the nosegear example above, all my parts begin with the same unique identification, c_gear_xxx such that when still ungrouped, they will all appear listed together in the alphabetical listing in the "Pick List."However, we can't use that exact naming convention with 'stock animation names' without regard to the corresponding action that will be applied!For example, if I had a 'hubcap' to add to c_tire_still, I could name it c_tire_still_hubcap, link it to c_tire_still and it would (a) spin in syncronicity with the tire, and (:( move in synchronicity with the tire during gear retraction.Now, applying that same principle to a flap assembly, as you know in order to 'stock animate' the left flap you would simply tag the part with l_flap. However, if you have a bellcrank attached to the flap, you cannot name that part l_flap_bellcrank!!! Doing so will cause the bellcrank to move according to the axis point of l_flap, which is not what we want!Instead, you could name the part l_flapbellcrank (which would preserve the convention mostly), so that the part is separated from the 'stock animation' and instead simply link that part as a child of l_flap. In effect, you have 'welded' that part to the flap assembly.>I have not given my wings their dihedral yet. I am >unfamiliar with how the dihedral is applied to a wing. I >hope that I can just angle the wings up a bit (the dihedral >is small, about 1.7 etc.). Thanks for reminding me to set >the pivot angle to match (and I suppose all the wing parts >like aileron, flap, trim will have to match). After you have set all the pivot points properly for your wing assembly, move the pivot point for the wing to the center of rotation (i.e, the root of the wing at the central axis of the a/c!), select all the parts that make up that wing, and use the rotation tool to angle the entire wing assembly to the correct dihedral angle. All of the existing defined pivot points for flaps, ailerons, etc. will maintain their existing relationship with respect to the wing! :)I hope that this helps someone, somewhere!

[Guest]

The nice thing is that it helps both the gmax and FSDS2 community!

[Guest jlwoodward]

Hi, I'm back. As you can see from the picture, I am completely missing the point on something.The middle part is selected from the stock parts as "rudder" and the axis was moved to the hinge point. It seems to work. The remaining parts are named rudder_top, rudder_bottom, and so on. I did not move the axis for each of these parts to the hinge point as I understand that you don't have to (right or wrong?). For the top part I selected rudder as the parent, and for the other parts I left the line blank. It did not make any difference.I don't dare work on the landing gear until I learn how to do the simple stuff.Thanks,John Woodward

[Guest]

>Hi, I'm back. As you can see from the picture, I am >completely missing the point on something. >>The middle part is selected from the stock parts as "rudder" >and the axis was moved to the hinge point. It seems to work. >The remaining parts are named rudder_top, rudder_bottom, and >so on. I did not move the axis for each of these parts to >the hinge point as I understand that you don't have to >(right or wrong?). For the top part I selected rudder as the >parent, and for the other parts I left the line blank. It >did not make any difference. Ah! You said aileron in the first post. Now I see you are working on a rudder...Name the major part that will be animated as:rudderName the other two parts as you wish, but name them you must! Then, link them to the parent like this:rudder ruddertop rudderbottomNotice that I used the root name rudder, but 'hid' it from FS2k2's stock animation by masking the name! This way, the parts will still be listed together in any alphabetical listing, but will not respond to the animation.Now, set the pivot point for rudder to the leading edge of the rudder, with the pivot axis aligned with the leading edge, and centered on the width of the rudder.When you move the rudder in FS2k2, the linked parts should perfectly mimic the parent part's movement!

[Guest jlwoodward]

So far I got the ailerons, rudder, flaps, and elevator working by using Bill's naming convention. Still having difficulty with the axis. I can move and rotate the axis for the main part to get the correct animation, but it seems that if you have attached parts you have to go to each part and move that axis to the same point. If they are not exact the parts will separate slightly during movement. In some cases I just joined the parts back together and turned off the polygon smoothing.This still doesn't seem right to me, as my understanding is you can link a number of parts and they will all move about a common hinge line. At least that is what I think the tutorial on gear animation says.Gear animation is next (cringe!).John Woodward

[Guest]

>I can move and rotate the axis for >the main part to get the correct animation, but it seems >that if you have attached parts you have to go to each part >and move that axis to the same point. If they are not exact >the parts will separate slightly during movement. In some >cases I just joined the parts back together and turned off >the polygon smoothing. Well, not having FSDS2 myself, I'm not sure how it was designed to work. I can only speak of how GMax works. Linking in GMax works as I've described; perhaps it's different in FSDS2?Again, as far as gear animation works in FSDS2, I'm clueless. I would suspect that it's not that much different than GMax in that regard.Have fun!

[Guest]

That's a nice looking aircraft.I wonder about making the rudder of three parts. In GMAX, although I have found it very tricky, you should be able to join three parts together into a single part. For the counterbalance piece I did end up extruding it from a front rudder face polygon because I could not find a way to completely attach (not Atttach) the verticies to form one single object. I did succeed using Boolean add to make the top piece and the rudder wing one object. I was just not satisifed with how they connected.Is there some easy and reliable way to connect two objects into a single object in GMAX? Every time I try, some vertices will connect and I can draw polys between some, but not all.It would be nice to make your rudder parts then "merge" them into a single object for animation.Steve

[Guest]

ufo has a good point.Once you have your rudder parts set, select the three parts and join into the main rudder part. (Parts--->Join Selected) or j keyThat way you just worry about the alignment of the one pivot/axis.

[Guest jlwoodward]

Steve,Originally I made my rudders, airlerons, etc. as one piece. However, applying polygon smoothing rounds off edges that should be sharp, so I split them into components. Unfortunately, my approach just falls apart when I try to animate them. I found that by joining them back together and turning off the smoothing that the coarseness is not so noticeable if a detailed texture is applied.Don't know about Gmax, but in FSDS you just select the components then click on "join selected parts". Some parts were created separately and then joined later. In that case, I checked all the points at the mating surfaces to make sure they had the same coordinates.John Woodward

[Guest]

I had exactly the same problem with multiple parts & moving the axis.You can join all the bits together & then use smoothing groups to smooth it how you want.Select your polys & 'Current item properties'. Give a different group number for Top/Bottom/Front (or whatever) surfaces & you'll get your sharp edge.Smoothing groups is one of the new features in FSDS2.For some reason I have not had the same problem with the gear. I think it is only a problem if the axis is rotated, but I'm not sure.David