Ivan's Workshop - AF99 Tutorial - C205 Veltro

I

Ivan

Charter Member
Hello All,

It appears that a tutorial about how to build aircraft using Abacus' Aircraft Factory 99 would be popular. The methods to be described are what I currently use when working on a project. My own practices change as I learn new things and just about every project I have ever done has taught me something new.

The subject will be a fairly simple single engine fighter: the Macchi C205 Veltro. This aircraft would not have been my first choice, but was the most popular subject.

The construction of the visual model is what will be described here along with illustrations from the design process and lots of my own very long-winded discussion. I believe it is very unlikely this visual model will become part of a fully releasable aircraft in a short time.

- Ivan.
 
Really looking forward to it!

I believe it is very unlikely this visual model will become part of a fully releasable aircraft in a short time.
Click to expand...

This is a case where the destination is not as important as the trip itself. Good luck on your endeavor.:applause:
 
Hi Hubbabubba,

That is kind of how I see it also. Well stated. My guess is that what I cover will be pretty basic for you, but hopefully others will get something out of it.

If anyone reading along finds something that needs a better explanation, Please interrupt. That is the point of this thread.

- Ivan.
 
Abacus' Aircraft Factory 99 as a design tool has some limitations and features that look on the surface to be useful but should not be used except as a last resort. One of these is the option of setting the center of rotation of the entire assembly. Another is "Dihedral". I will explain later why these should not be used. There are also a few other features that don't really seem to do anything useful.

AF99 records dimensions for its "Parts" or Polygons and Templates to 0.01 Foot. It also has a Metric setting but I generally don't use that. I think in Feet and Inches rather than Meters and Centimeters. The SCASM code that is generated by AF99 specifies points to 1/512 Meter. Those two increments don't really line up and cause some issues. I will address that when the time comes.

The first step with any project these days is to do some basic research. Figure out WHICH version of the plane you are trying to build. I try to pick a version of the airplane which was in front line service at a time in which it did its best work. As an example, there are three distinct operational variants of the Grumman Wildcat: F4F-3, F4F-4, and FM-2. The FM-2 was easily the most potent of the three, but it was served mainly as a second line fighter on Jeep carriers and as a patrol aircraft. It wasn't normally found on fleet carriers as the F4F-3 and F4F-4 were. Thus, my choice was to build the two earlier aircraft. With the C205 Veltro, I can't really tell the difference between most of the Series of the aircraft, so this choice can be made later.

Next step is to find some good and precise drawings of the plane. The best things to find are drawings with dimensions labeled. Manufacturers' drawings or blueprints or Schematics from a maintenance or assembly manual are optimal. Keep in mind that sometimes the drawings and labeled dimensions don't line up.

These are two sites where I was able to find very good information about the Veltro:
http://stormomagazine.com/phpbb2/viewtopic.php?t=988
http://digilander.libero.it/enniotarantola/TarantolavelC205Disegni.htm

In general, the drawings found on these sites are pretty good but they ARE contradictory. Since I don't have easy access to a real Veltro, I have to make my best guess as to which drawings are correct.

Length is 8.840, 8.845, 8.850, 8.870, 8.875 or 8.877 Meters depending on which source you believe. My choice in this case is 8.877 Meters because that is consistent with one set of what appear to be factory schematics with individual sections labeled. Perhaps there ARE this many differences in the various Series of this aircraft.

Once drawings are found or created, the center of gravity of the aircraft without fuel or ammunition should be determined. This is generally easy with a maintenance drawing. There will often be a Datum point described. This Datum point is just an easily determined location on the aircraft. It is probably NOT the actual CoG. As an example, the Hawker Hurricane's Datum point is the hole in the lower cowl where the starter crank plugs in projected to the aircraft center line. The actual Center of Gravity is located almost 5 feet aft of that location.

If there is no clear labeling of the CoG, best guess is that it is about 1/4 to 1/2 of the Mean Aerodynamic Chord of the wing. It also tends to be very close to the upper wing surface on a low wing monoplane.

Time for bed.
'Til next time.
- Ivan.
 
this will be interesting.
i look forward to it,
along with hubba's ...fun stuff... thread.

at the risk of sounding negative,
it's to bad this wasn't done years ago
when there was a bigger audience
and, to be completely honest,
others, including myself,
were more into CFS.
i, for one, am still interested,
but find that i have branched out.
 
Hi Smilo,

I agree with you about the timing, but part of the issue is knowing enough to do such a tutorial. A few years ago, I didn't think I knew enough. I am still not entirely certain that I do, but will give it a shot anyway. Besides, a lot of the techniques here are general enough to be used with just about any 3D design package.

I did try a Painting thread, but I wasn't certain there really was an audience even back then.

- Ivan.
 
well, all i can say is,
so it goes.
hopefully, we will be able to learn something.

i vaguely remember the painting thread.
i can't remember,
did you complete it?
i should go find it and bump it.
:bump:

note to self...
DON'T HIJACK THIS THREAD!!!
 
Make sure you have the "Patch" installed for AF99.

To create a new project, first think of a good and somewhat unique name for the aircraft project. There is a commonly available download called "Veltro" that stopped me from using that name. Think also of a good description for the aircraft that is different from all the others you have installed on your machine. The obvious choice would have been "Macchi C.205 Veltro", but there is already an aircraft by that name. I could have just renamed Bruno Duffort's version and taken the same name for my own installation, but I figure he put his together first so gets to keep that honour. Besides, other folks are likely to have that aircraft installed as well, so try not to create a conflict that the user needs to resolve.

My project will be MC205Veltro, and the description will be "Macchi C.205V Veltro". The C.205V was the version of the Veltro with the greatest production numbers. I could have also put a "Serie" designation into the name which would have done as well. The description can be changed at any time. Changing the project is more difficult.

Execute the following steps with the AF99 Wizard:

1. Start the Project From Scratch
2. Select a Folder Name (MC205Veltro) and follow steps to create
3. Select a Description (Macchi C.205V Veltro)
4. Select an existing Aircraft to copy (Stock P51D)
5. Confirm.

After doing this, A new project will exist in a new folder. A new folder will have been created in your CFS aircraft directory. The next start of AF99 will use this new project but of course there won't be anything actually in the project yet.

'Til next time.
- Ivan.
 
At some point, we need drawings to work from. I posted a couple links earlier about where some good dimensional references for this plane can be found. The problem with those links is that there are no complete profiles of good quality even though there are dimensions. I found some drawings by a fellow named Brioschi which look to be of very good quality. There are some obvious problems in them which will need to be corrected or ignored (the propeller disk is too far forward), but it is a start. To be useful for development, I will try to resize the drawings to a scale of 1 pixel = 0.01 foot. This will result in a VERY large drawing. At that point, the locations of various pieces will need to be confirmed against the dimensional drawings and most likely pieces of the drawing will be moved around to match the dimensions. There is also the possibility that the drawing will need to be stretched in different scales vertically as compared to horizontally to match known dimensions.

- Ivan.

Attached is the original profile.
 
Here is the first cut at a reference drawing for the plane. It LOOKS very similar to the one in the previous post, but isn't.

We know from the factory drawings that the centerline of the propeller spinner is the datum line. We also know that the datum line meets the bottom of the tail cone about in line with the rudder hinge line. The Brioschi drawing needs to be rotated roughly 1.2 degrees counter clockwise to get this alignment. The actual rotation is by the pixel using GIMP, so is much more precise than that. MY drawing has also been scaled to 1 pixel = 1 foot by counting pixels from the prop spinner to the tip of the tail cone and scaling the drawing until it is very near 2912 pixels long at that point. The "Very Near" rather than Exact is because when the drawing is rescaled, the typical line is around 4 or 5 pixels wide.

I also added a Crosshair at the location *I* decided should be the center of gravity of the plane (Our Model's Center of Rotation). After doing that, I changed the front and top margin of the drawing so that the CoG is at coordinates (1000,1000). The CoG is 0.70 foot below the datum line and just a bit behind the firewall.

The next step is to create a spreadsheet to calculate offsets from the CoG from coordinates in the drawing. Yes, this isn't hard to do in your head but I prefer the spreadsheet.

Also, because we have some factory dimensions, it would be a good idea to confirm that the locations specified actually match those in this rescaled drawing. If NOT, depending on how bad things are, we may need to find a different drawing or cut and move pieces of this one to line things up properly. The spreadsheet mentioned above (if properly built) also let us do a dimension check along with metric conversion VERY quickly and use the factory datum lines instead of our offsets.

Attached is a 1/2 scale version of the drawing I am using. Don't worry if the idea of a spreadsheet is a bit confusing at this point. It will become clear once you see it.

Until Next Time,
Good Night.
- Ivan.
 
This spreadsheet is what I use to translate pixel coordinates into AF99 offsets. The pixel coordinates are entered in B12, and C12.

If this is a profile drawing, the offsets are calculated in B4, and C4.
If this is a plan view, the offsets are calculated in B6, and C6.
If this is a front view, the offsets are calculated in B8, and C8.

Each set of calculations uses it own offsets in columns D and E.
Columns F and G normally are not present. They are used in this case because I want to see the Metric measurements in mm. For the metric measurements, my X offset is even with the firewall, and my Y offset is at the centerline of the propeller shaft (955, 930).

For the Veltro, there are only three basic length (Offset) dimensions that I trust:
The Spinner is 590 mm in length (with a diameter of 640 mm).
The Cowl from Firewall to Spinner is 2190 mm.
The fuselage from Firewall to Tip of Tailcone is 6097 mm.

The offset of the upper rear of the spinner is (243, 825) which means the length of the cowl is only 2170.18 mm though the spinner is correct at 320.04 mm in radius.

The tip of the spinner is at (58, 930) which means the spinner is 2731-2170 mm or length 561 mm instead of the 590 mm that it should be.

The tip of the tailcone is at (2968, 850) which means it is at -6135.62 mm instead of 6097 mm as it should be.

Seems like these length offsets are around an inch or so off. I will be taking a bunch more measurements, but there is no need for a description because it will just be more of the same.

Hopefully you all get the idea of how a spreadsheet helps. The next question is "How badly do we want the dimensions to be accurate?"

Good Evening All.
- Ivan.
 
To be useful for development, I will try to resize the drawings to a scale of 1 pixel = 0.01 foot. This will result in a VERY large drawing.
Click to expand...
This is one point that needs clarification, at least for me. When enlarging the scale of a drawing, lines tend to get thicker. How do you decide where to place vertices' coordinates? Do you redo the lines? Do you use "mid way" approximation? Knowing you, approximation is not in your genes:icon_lol:...
 
Hi Hubbabubba,

Sorry to disappoint you, but the dimensions we get from the drawings are approximate. This is what I meant by "Very Near rather than Exact". In this case, the dark centers of the lines are about 4-5 pixels wide with a shadow line of pixels on each side. This drawing is pretty good. With worse drawings, the lines are often about 8 pixels wide, and the shapes may be much worse.

In places, some of the lines will need to be redrawn and specific points will be marked. When doing one of these drawings, many times, vertical lines such as a Rudder Hinge or a Canopy Frame will end up slightly slanted. A point is picked and those lines are redrawn.

Folks might be wondering if the Tail Cone is 39 mm too long and the Nose is an equivalent length too short, then why not just shift the datum line at the firewall back by 39 mm? The reason is that there are lots of things that are at that datum line. There is a major wing panel at that location and the back side of the main gear lines up there also.

The issues with Horizontal locations were listed in the last message. Vertical location issues observed thus far are the following:

The lower edge of the spinner is 6 mm too low.
The bottom propeller blade tip is 1 mm too high. (That's pretty good!)
The bottom edge of the oil coolers are 6 mm too high.
The belly of the fuselage is 0.24 mm too low. (To me, that is exact!)
The radiator bottom is 4 mm too high.
The top of the rudder is 10 mm too high
The top of the canopy frame is 28 mm too high but that might be error in interpretation.
The fuselage deck in front of the canopy frame is 9.5 mm too high.

Take a ruler and look at the measurements. That isn't bad for something the size of an airplane.

- Ivan.
 
Looks like the same drawing doesn't it? It is a bit dirtier with Red and Green lines in strange places. Some detail has been deleted. Things have been moved around a bit to match the known dimensions within about 2 mm. Sometimes it is a bit closer. That is as accurate as AF99 is capable of handling, so it will have to do.

This is now suitable for use as a reference drawing. I am sure there are still inaccuracies, but I have fixed everything I know about.

(In the full sized version, it is a 14 Megabyte Bitmap.)

Way Past Time for Bed.
Good Night.
- Ivan.
 
Consider the time spent thus far as working on Blueprints for our virtual plane. Some folks may be impatient and wonder when we will cut some virtual metal. I tend to be impatient also, so here goes.

The back of the Spinner in the drawing is 7.63 feet Forward from our CoG. This works out to the Cowl section in our drawing being 2188.46 mm in length whereas it should really be 2190 mm. This is about as close as we are going to get. (Remember the comment about within 2 mm or so?) The front of the Spinner should be at 9.57 feet Forward.

To save on typing, I will be using some pretty obvious abbreviations:
F-A == Forward - Aft
U-D == Up - Down
L-R == Left - Right

Aircraft Factory 99 has a construct called a Structure which can be used for pieces of an aircraft. To use a Structure, a Side View of the piece and a Top View is needed. At each transition within this piece a Cross Section is picked from a selection list. AF99 then attempts to join the Cross Sections in a reasonable manner to create the piece of our aircraft.

Some Cross Sections are "Restricted Use" which means that AF99 doesn't know how to join them to Different Cross Sections. It has no trouble joining them to the same Cross Section even if the sizes are different.

As an example of a creative use of a AF99 Structure, look at the Machine Gun on my Fokker Eindecker. That piece came out surprisingly well.

Thus the Points on the Spinner we know about are at
7.63F 1.75U
7.63F 0.35D
9.57F 0.81U
9.57F 0.59U

I plotted one more point at 8.92F 1.40U.
The Tip of the Bottom Propeller Blade is at 8.30F 4.30D

From those points, we can create a Part for AF99 called "SpinnerSide" which serves as a Structure Side Template. We need to add a couple points in between the given points to adjust the shape to something reasonable. One of the points must be at 8.30F because that is where the propeller will be located.

After SpinnerSide is created, we can rotate it (Bank 90 degrees), Change its type to Structure Top Template and save it as SpinnerTop.

I chose to use 12 sided Circles as the Cross Section when creating the Structure. The resulting piece looks fairly nice, but just cost us around 50 polygons of the 1200 we are allowed for the project.

Next comes a propeller.
- Ivan.
 
Propellers are Easy! (Part 1)

Reasonably good looking Propellers are quite easy to make. In the course of doing this we will be learning a couple cool features of AF99.

As stated before, the Tip of the Bottom Blade of the Propeller is at 8.30F 4.30D. Since the CoG is 0.70 feet below the Propeller Centerline, The Propeller is 10.00 feet in Diameter. The references state the Propeller is 3050 mm in diameter. 10.00 feet == 3048 mm.

There is a little display at the upper right of the AF99 Parts Screen. That shows the location of the cursor. Move the cursor to location 8.30F and Right Click. This centers the screen at 8.30F. Change to a Front View of the workspace. When this is done, all the points that are then drawn with the cursor may vary L-R and U-D, but will all be at 8.30F.

To get the shape of a single propeller blade, I brought up a slightly oblique picture of the actual plane and traced the image on the screen onto a piece of graph paper. Draw this shape in the AF99 Parts Shop.

There are four VERY important points in this drawing: The first point drawn should be at 8.30F 0.70U. That is the center of the propeller / spinner. Next point is where the propeller blade meets the spinner. (It should be slightly lower than the point of intersection because the spinner has flats and isn't really a smooth circle. Also a little overlap is less distracting than an air gap.) Next point is the tip which I put at 8.30F 5.70U. Last important point is the intersection of the back of the prop blade and the spinner.

Add points and shape as appropriate. The resulting part is a template for your propeller blade. Don't get too carried away.

- Ivan.
 
Propellers are Easy! (Part 2)

Now that we have a propeller blade template, change to a Top View and rotate it. I generally rotate the blade 20 - 25 degrees because it should be at the low pitch range. It will only be visible either stopped or at low RPM. When RPMs go higher, the blur will show, not the blade.

After rotating, Select the Point Editor and Mod every point. The original rotation will show the points EXACTLY as they are rotated. When saved (or Mod-ed) the points will snap to their final location within a 0.01 foot grid. You may notice as I did that when viewed from the tip, the blade isn't straight any more. The rotation moved some of the points out of line. Using the Point Editor, correct this by shifting points as needed.

Check the Center Point of the Propeller Blade. You may find as I did that the rotation also moved the Center Point. Select Rework, On Screen. This will highlight the current point that will be moved. Put your Cursor on the location you wish the point to be: 8.30F 0.70U and while holding down the Control Key, Click on the new location. The entire part will shift with the currently selected point at the new location.

Save this as PropBlade1.

Select Rework, Rotate, Bank and enter 120 degrees. (On a 3 Bladed Propeller, the blades are 120 degrees apart.) After Rotating, Select Rework, On Screen, and Control-Click to move the new blade to its proper location. I number the blades according to the order they pass in front of the pilot, so this became PropBlade3.

Follow the same method to create PropBlade2. I prefer to rework PropBlade1 to get this because there was some interpolating of points to get PropBlade3 from PropBlade1 and to use it to get PropBlade2 is about the equivalent of Photocopying a Photocopy. The more times removed you go, the worse the copies get.

Now bring up each Propeller Blade and delete the Center Point. It was only needed to get the alignment of the Blades to each other.

Now you have a bunch of Propeller Blades!

Good Evening.
- Ivan.
 
Here is what the full set of propeller blades looks like. Generally, I try to put them all into a AF99 Component to get the shading the same across all the blades. If they are not all in the same component, CFS tries to shade each blade by where it is located. This sometimes produces differences which make sense if the Propeller Blades don't move, but if they do move and the shading of each blade does not change and the blades appear to be different colours as they rotate.

- Ivan.
 
Now that we have all these bits and pieces, it's time to assemble them into something we can actually look at in the simulator.

To assemble the Propeller, go to the Components Shop and Build, Add Part, and Select all the propeller blades.

Save this Component as "Propeller".

Go to the Structures Shop, (Which we used to build the Spinner initially.) and RE Build the Spinner. First though, we need to build a second Structure so we can split the Spinner into two pieces.

Open "Spinner"
Select Manufacture and then Reject the build. That allows you to redo the cross sections. Only put 12 sided circles in the last two positions.
Manufacture again.
Save as "SpinnerBack"

Open "Spinner" again. (Since we saved off the last structure as "SpinnerBack" we have not affected the original Spinner structure.
Manufacture / Reject again.
Put 12 sided circles again in all positions but the last.
Manufacture again.
Save.

- Ivan.
 
Time to actually add something to the Aircraft Assembly.
Go to the Paint / Special Effects Shop.
Add Structure Spinner.
Since there are no other pieces, it will add to Body,Main.
Move it to the "Nose".
Select Spinner in the Nose Assembly.
Add Component Propeller.
Select Propeller.
Add Structure SpinnerBack.

Change Colours to be more appropriate. You can see what I picked. I like to have adjacent piece to have different colours so that we can tell where one piece ends and the next begins.

Exit the Paint / Special Effects Shop.
Save the project.
Produce.

When I did this, I found that the AIR file was missing. I went to the actual CFS Aircraft directory for the MC205Veltro and renamed the P51D AIR file and DP file to MS205Veltro.AIR and MS205Veltro.DP. Redo the build.

After that, you can go into CFS and admire your new pieces. Note that there is a slight bleed between the pieces. For CFS, it doesn't get much better than that. We can fix things slightly, but it will cost a few more Parts and we should leave that for the very end of the project when we will know if we have any Parts left over of the 1200 Polygon Limit.

Time to get ready for Work.
- Ivan.
 
You must log in or register to reply here.
Back
Top