Ivan's Workshop - AF99 Tutorial - C205 Veltro

Create the Fin Fillet

We spent a fair amount of time cutting the opening for the Fin Fillet because if we can get the locations of the vertices correct, creating the actual Fillet becomes trivial.

The first screenshot shows how trivial the parts can be. It is just a matter of connect the dots. There are only 5 Fillet Parts.

Next screenshot shows a Component Assembly Trick: The Polygons in a Component display in the order they are listed. Any Polygon in the listing is displayed in FRONT of all the preceding Polygons. We will use this to our advantage to create a CONCAVE Polygon that doesn't bleed significantly.

The basic principle is that all the Concave Parts of the Component will be listed BEFORE the Convex Parts. The Concave Parts in this case are the ones labeled 1,2 and 3. It is fairly safe to say that all of the other Parts in this component should display "In Front Of" the ones marked "1". Usually the Parts marked "3" should display in front of the Parts marked "1" and "2". There are minor exceptions to this, but they are fairly trivial and we can count on the Horizontal Stabilisers to block most of them.

All the rest of the Parts to the Component can be added after the parts labeled 1,2,3.

The result can be seen in the Simulator Screenshots. The gap between the Rudder and the Fillet is intentional. Note that the Component with defined vertices doesn't really match up well with the Fuselage. (Look for Sparklies.)

Good Night.
- Ivan.
 
Extending the Tail Cone

Some of you might have noticed that although the TailCone currently has only two rows of polygons, they are Tail02L0X and Tail03L0X. What ever happened to Tail01LXX Polygons???

They were in the plans, but the Cross Section of the fuselage at 10.50A was not so well determined. Note that there are no new techniques demonstrated here. Everything is a repeat, but hopefully a different means of illustration may make it more clear for those that don't understand but are too polite to interrupt. (If you don't understand, Ask for an explanation!)

The First part is to create the new Tail01LXX Polygons. Using the current TailCone Component and Part BulkR1050 as a reference we can easily locate the points for Tail01L01.

Create the Six new Tail Cone Parts and Add them as Left-Right Pairs to the TailCone Component. This is just a first try. Note that there are Concave areas in this Component. Hopefully the center of the Component is on the correct side of all the polygons in the component. If not, then some of the Polygons won't be visible in the Simulator.

In our check of the new Component, All the Polygons display, but there is a serious bleed. (I HATE Bleeds!)

On to the next step.....
 
Fixing the Tail Cone

Now that we know all the Parts of the Tail Cone and Fin Fillet can coexist, we can leave them as a single component and fix the bleed shown earlier.

We will be building a new TailCone component but the Parts must be entered in a certain order:

1. Enter all the Parts of the Tail Cone that may bleed through the Fin Fillet.
2. Enter all the Parts of the Fin Fillet.
3. Enter all the Parts of the Rest of the Tail Cone that can't bleed through the Fin Fillet.

Note that is is possible for the Fin Fillet Parts to bleed through some of the polygons from 2. but that is minor.

Save the resulting Component as the new TailCone.

The screenshot shows that this appears to have worked to cure the bleed shown earlier.

Time for Dinner.
- Ivan.
 
Now that we have extended the TailCone Component forward by one row, it would also make sense to delete the last row of the Fuselage Structure. This can be done by opening the Fuselage Structure in the Structures Shop and using the Backspace Key to remove the last cross section that was entered. Hopefully the last one entered was the located the furthest aft.

- Ivan.

P.S. If there are any folks following along who want scans of the working drawings I am using, PM me. I don't guarantee I will always stick to them, but they are still a pretty good reference to see how I got the numbers I am plugging in.
 
Canopy Frame Part 1

The movable (in real life) part of the Canopy is bounded by BulkR0379 and BulkR0565.
With measurements from the drawing, we add a Canopy cross section to that of the Fuselage cross section.

When we first add the new points, not all the points necessary for this Component are specified. Note that there needs to be an additional point between the top of the Canopy Frame and the first frame from the center.

Note that after adding that point, the curve isn't quite right. The Top point should not be adjusted. It is a dimension that WE did not deduce. That is a specification from the Factory Drawings. The points we have scaled from drawings may be moved as needed....
 
Canopy Frame Part 2

Once we have finished the Canopy Frame Templates at 3.79A and 5.65A, we can create the actual parts for the Canopy Frame.

From Drawings, I scaled the first Transverse Frame at 3.79A to 3.88A. The second is narrower at 4.37A to 4.43A. The third is at 4.96A to 5.02A. The last is at 5.56A to 5.65A. I made a couple adjustments to make the frame widths consistent.

The next question is how to determine the locations of the frames in between. If the two endpoints are determined, we can find a point in between at a given longitudinal location just by inserting points using the point editor until we have one at the correct longitudinal location. With this canopy that leve of detail isn't needed. The slope is very gradual from front to rear and a row of points may go fore or aft for quite a ways before moving vertically or transversely 0.01 foot.

The rest of the Parts for the Canopy Frame can be created by using reference parts we create by just picking the end points of a line and then inserting points until the numbers match.

Good Night.
- Ivan.
 
By inserting points in between two existing ones, the theory is that the slope will be preserved. You will notice at some point that the two points straddling the location you are trying to reach will have their other dimensions the same. When this happens, just use the point editor to change the value to what you are looking for.

The following screenshots show the resulting Component in the Components Shop and in the Paint / Special Effects shop for final assembly and then in the simulator itself. Note that I have shortened the back end of the Fuselage Structure as discussed earlier.

- Ivan.
 
Please notice that we have not tried to build the Wings or the Wing Fillets up to this point. This is because the wings' cross sections are smooth curves and there are many line segments that can be used to approximate the curves. Ideally they should come close to matching the locations of the fuselage cross sections we have chosen but we have not completed those yet.

In earlier models, I built the Wings and Fuselage in any order without considering how they would fit together. The Wing Fillets would end up as complicated components for that reason.

Good Night.
- Ivan.
 
The Fuselage of the Veltro isn't really a constant oval cross section but it is a good approximation for a while. The Tailcone is pretty close. When we hit the back of the Cockpit (Canopy) the cross section changes to be very irregular. This would be at location 5.65A or BulkR0565.

Note that there are a bunch of new Reference Parts Y01-Y07 which are stringers lining up equivalent vertices at each Bulkhead. Each one should be a nice smooth curve unless there is a specific reason it should not be a gradual curve.

Note that the new BulkR0565 does not match the structure cross section. It DOES need to match the outline of the Canopy Frame.

- Ivan.
 
If we put nothing between 10.50A and 5.65A, there would be a long triangular fillet between those two stations. The problem is that there really isn't such a thing at that location, so we add a new bulkhead at 8.05A to make the transition to the fillet more abrupt. BulkR0805 is actually very easy to create. Just add a point on each stringer between 5.65A and 10.50A. The result will be at location 8.08A, but this can be adjusted forward with the Point Editor. Examine each point thus moved to make sure it does not create a concave section on the Aft Fuselage. Just about every point needs to be move outward by 0.01 foot.

Once the two additional bulkheads are created at 5.65A and 8.05A, we also use BulkR1050 as a reference and create all the polygons determined by these reference parts.

The next bulkhead at 3.79A has a large section (underside of the fuselage) that needs to line up with the underside of the wings. To figure out angles, I use the templates I created earlier for Wing Dihedral. This omits the thickness of the wing but a guess here is good enough for now. We are certain to be back to adjust the dimensions. Note that the Wing Root Fillet, the Fuselage, and the Radiator all come together around this location.

After creating BulkR0379 (which by the way should not have completely vertical sides, note that the angle for the side of BulkR0565 doesn't quite follow the contour. Adjust the angle of that as well.

Next step is to create a Wing Root Template. I still haven't figured out how this will all fit together yet, so it will have to wait until next time.

Good Night.
- Ivan.
 
Oddities

When I do research for a new aircraft project, there is generally something unusual I find that isn't readily obvious or is non-intuitive. With the Macchi C.205, there have been several odd things:

1. The canopy is not sealed in the back. The sides of the headrest are much narrower than the canopy and the pilot can see to the aft quarters without any glass in the way.

2. The Left Wing is 200 mm (almost 8 inches) longer than the Right Wing.

3. The last fact became obvious when trying to build a Wing Root Template to help with doing building the underside of the Fuselage and the Wing Root Fillet: The Wing on the Macchi Fighters has an VERY thick section. Most fighters of the era had wings whose thickness was 12% of the Chord. The Root Chord on the Macchi C.205 (and earlier models) is 1924 mm. The thickness at the Root is 349 mm. This works out to be slightly over 18%. For a while I thought I had made a mistake because the shape looked wrong in AF99 when using the correct numbers, but then I saw a side view of the plane with the root section visible. It IS a very thick and almost symmetric cross section with very little camber.

- Ivan.
 
The first screenshot shows a quick mock up of the Wing Root Template. The shape is all I am after here. The Leading and Trailing edges are fixed, but nothing else is besides the highest point and the lowest point to set the thickness of the section.

The second screenshot shows how the Fuselage Bulkheads are extended to meet the Wing Root Template. This is a check to make sure that the angles between the Fuselage Keel and the Wings form a continuous curve without any concave areas.
Note that the straight line between the Keel and the Wing Root Template is a bit too shallow. This means that the lower edge of the Wing Root needs to be raised.
Note also that the lines between the other Bulkheads and the Wing Root appear to be much closer together from top to bottom as they reach the Wing Root. This is because the Wing Section is thicker toward the middle than at the Trailing Edge.

The third screenshot shows what happens when the Lower Edge of the Wing Root Template is raised and the points are Longitudinally lined up a bit more. Note that there are many more points on the Wing Root than on the Fuselage. The Wing Fillet will need to formed to match the various pieces.

This is just a first cut, so I will most likely make adjustments before all is done. Although the idea is presented here, the actual work is to adjust the fitting and relationship of the various parts shown here until the builder is satisfiedl. Consider this the "Rawhide Mallet" stage of development.

Good Night.
- Ivan.
 
Windscreen Preparation

This screenshot shows the remaining Bulkheads created by looking at the Fuselage cross sections from a book and stretching and moving some vertices of the Bulkheads created earlier from the Nose STRUCTURE.

Note that an additional vertex has been created with an offset of 0.34 Feet Left. This dimension was gathered by measuring a plan view of the aircraft. This vertex has been added to all of the Bulkheads leading up to the rear of the Spinner. The contours around the Oil Coolers isn't all that well determined yet and I suspect that there will need to be some additional bends between the Aft end of the Oil Coolers and the Front of the Wing Fillet.

One step that will need to be done is to adjust the stringers that we used earlier on the Tail Cone and Rear of the Fuselage to match the new vertices and see how things line up.

- Ivan.
 
Pilot

I believe that some of you might interpret this post as cheating. Consider that I have been building these planes for quite a few years now and when possible, I will try to re use pieces that I can. I have also developed a few tools to make things quicker.

Now for some History....
My first AF99 aircraft did not have a pilot. At the time, it seemed a bit difficult to create a nice looking Pilot's Head and Shoulders that one could see above the cockpit sill.

My first attempt at creating a pilot was done with a piece of closed cell styrofoam and a pocket knife. I goofed up a few times, but eventually finished a shape with all flat sides that was shaped pretty well in my opinion.

I did my best to translate this to two AF99 Components and put it into my P-51A Mustang. I thought it looked good. I added it to other aircraft and I thought it still looked pretty good. These were all WW2 Fighters.

Eventually I built a Fokker E.III Eindecker. When I added the Pilot, I found that he looked a bit goofy. He was HUGE in this tiny little plane. I found a book on Anatomical proportions (Look for "The Measure of Man") and found that the Pilot I was using had a head and shoulders that were appropriately sized for a 7.5 Foot man. I had been wondering why his shoulders barely cleared small cockpits like on the Lavochkin La-5FN I had built.

After this, I pretty much started from scratch again and built another pilot with dimensions more appropriate for a 5 Foot 8 Inch person or somewhere close. Since this was the second try at a Pilot, all the Parts of him are named PilotB-something to distinguish from the original version.

I have been using the same Pilot ever since and these screenshots show how I put him into each new airplane.

The Canopy Frame was created before this because it locates the Cockpit Sill which I call the "Cockpit Floor". The base of the Pilot's Shoulders will be adjusted to rest on the "Cockpit Floor".

The First Screenshot shows the current Macchi C.205V with the Parts for PilotB copied from my A6M2 Zero. Note that he is nowhere near where he should be.

The Second Screenshot shows a simple AF99 Part created by using the PilotB Parts as a reference. As stated before, this is easy. Just click near each reference vertex to create each new vertex and then use "Rework / On Screen" and <SpaceBar> "P" to snap to each reference vertex.

The Third Screenshot shows the Pilot Template Part being moved into alignment with the Canopy Frame.

To figure out WHERE to put him, find photographs that show a PILOT in the cockpit. The emphasis is to ignore photographs of Girl Friends, Infantry, and Ground Crew. Most aeroplanes of this era used Parachutes that also served as seat cushions. A Pilot in the cockpit will likely be wearing a Parachute and his head will be at an appropriate height. Others will likely be sitting way too low or back too far without the Parachute.

AF99 Parts are simple text files with a structure that is quite easy to decode. Way back when I first started, I wrote some C Programs that I intend to use here:

Mirror: Copy a Part from Left to Right or Right to Left.
MoveIt: Adjust the Location of a Part Left-Right, Fore-Aft, Up-Down.
CMoveIt: Adjust ALL the Parts in a Component using calls to MoveIt.

None of these programs is absolutely essential to AF99 Construction. They just make things move a bit faster. The most labour saving is CMoveIt. The equivalent can be done in the Parts Shop by using Control-Click to adjust the location of each original Part.

If any of you want these programs, Email me at Ivan1GFP@yahoo.com.

Good Day!
- Ivan.
 
Pilot Moved

My Program CMoveIt isn't all that smart. It goes into a Component file and moves every Part in the same direction. It doesn't check if a part is Reversed or if the same Part is included as a Left-Right pair, so the easiest way to use this tool is to create a new Component (which I call X1) containing all the Pilot Parts and C Move It.

The First Screenshot shows a call with parameters.

The Second Screenshot shows the result.

- Ivan.
 
These screenshots show why we use stringers to check the contours of the Pieces we build. As you can see, from Stringer Y14 and Y15, the vertices between the different Bulkheads hardly line up well. Even from a Side view, it is obvious that the curves between the Bulkheads are hardly smooth.

In this Rawhide Mallet stage, there will be a fair amount of time spent in smoothing out the reference parts before any panels are constructed. Consider this the equivalent of setting up jigs and dies to make the actual parts.

Note that you haven't seen any screenshots from the simulator in quite some time. The reason is because there aren't any changes there. These are all reference or template parts. None of them go into the real plane.

- Ivan.
 
Nose Templates Rebuilt

I won't bother showing the specifics of what needs to get tweaked to adjust the contours, but here are the basic principles:

The Stringers should be fairly smooth curves. There should not be abrupt changes of direction unless the real aircraft has them (such as around the Oil Cooler or Wing Fillet). Consecutive Bulkheads should have only gradual changes unless there is some feature of the aircraft that forces a more abrupt change.

The vertices of the Bulkheads and Stringers MUST meet.

If there is any artwork at all in the design of a AF99 model, this is really where it happens. The designer gets to figure out how the aircraft will be shaped. It won't be completely faithful to the original because it really can't be. We are using flat polygons to represent curves. There will be divisions and breaks somewhere. The trick is choosing the optimal places for these breaks. To some extent, the exact location doesn't really matter all that much as long as the pieces of the plane fit together.

The First Screenshot shows the Nose Bulheads. The Starboard Side of the aircraft shows mostly the templates as they were originally generated from the Nose Structure. The Port Side of the aircraft shows the modified templates shaped to match my interpretation of the cross sections of the aircraft. Note that some of the Bulkheads shown here don't really meet up with any places on the Structure because they were created to address specific issues.

The Second Screenshot shows the Bulkheads, Stringers and Oil Cooler. Note the differences between this picture and the one shown in a prior post. There isn't MUCH movement of the vertices, but there is enough to smooth out the curves.

The Third Screenshot shows the basic shape of the Fuselage.... Almost. What isn't visible from this picture is the contour change that is needed on BulkR0565 (Back of the Cockpit) for the Wing Fillet.

In reworking the nose section, the Canopy Frame was affected as well. It will need to be modified to meet up with the adjusted Fuselage shapes. There still needs ot be a Wing Fillet Template created because it will be a reference for many of the lower Fuselage Parts.

Y'all having fun yet?
- Ivan.
 
This Screenshot shows a rather strange part. It is a Bulkhead at Location 3.35F. There is no radical transition in contour at this location and it is very close to the Bulkheads Fore and Aft of this location..... So WHY did we put it here?

The explanation is that the Macchi C.205V has Wing Cannon which extend quite a bit ahead of the Wing Leading Edge. If the Cannon did not exist, the Bulkhead Aft of this location could serve as the divider between the AF99 Nose Group and Body,Main Group. The Nose Group Pieces should be ahead of all Body,Main Pieces or Wing Pieces. If they are not, then from certain views the Nose will bleed through the cannon. It is possible also to put the Cannon into Nose,Left and Nose,Right Groups, but then the separation between the Cannon and Wing will be less well determined.

- Ivan.
 
The Wing Fillet Template isn't technically all that difficult to make, but it is somewhat tedious because getting the shape right may take some time. First draw an Airfoil shape. Locate points at each Bulkhead and adjust the height as appropriate. Insert points between tbe Bulkheads as necessary to smooth out the curve and to somewhat match the Wing Root Template. The Point Editor is the easiest way to do this.

When the shape looks right, change to a vertical view and adjust the points to meet the side of the fuselage. In the Second Screenshot, the bottom of the Wing Fillet Template has been adjusted to align with the Top, and while this is very neat looking, it is really not necessary.

Note also that I added Radiator Templates. These were created with dimensions from the aircraft drawings.

We just spent a LOT of time getting all the contours of the Fuselage as smooth as possible and now we are about to cut holes in the underside of the Fuselage and bend things to install a Radiator.

This is the appropriate time to introduce another concept:
The Radiators have rectangular cross sections for the most part. The bottom of the Radiator is flat. Like most parts attached to the Fuselage, the Radiators will be textured Left-Right. In order to properly texture a Piece Left-Right, we should make the bottom a shallow V shape with a slight point on the bottom. If this is not done, the bottom will only show the texture from one side. Any variations within that texture will be seen as streaks across the piece.

The V shape does not have to be very obvious. My Fokker Eindecker has points on both the Top and Bottom of the Fuselage but the angles are so shallow that the sections still look square.

Good Evening.
- Ivan.
 
Radiator Template Pieces

A Radiator Structure can be made from the crude Side and Top view that was created from the aircraft drawings. The shape will not be correct, but it WILL give us some a reference to create Parts for a Radiator Component.

The First Screenshot shows a simple Radiator Structure. In hindsight, this Structure should have extended further up into the Fuselage.

The Second Screenshot shows where the Radiator is in relation to the rest of the aircraft. Note that it starts behind BulkR0274 and in front of BulkR0805.
 
You must log in or register to reply here.
Back
Top