So you also see the differences between AF99 and DPED....
....and you can confirm that AA agrees with DPED.
Now the next trick is to figure out which is the most correct and if the difference is constant or distorted.
I have a couple ideas as to how to do that, but until then, the Warhawk is on hold because I suspect I may be making unnecessary corrections.
Hi Ivan,
I remember that the reason I´d almost given up on the AT-9 before resuming construction was that suddenly the height started to go off somehow. Now I realize that it was when I was adjusting the AF99 wireframe to side-view photo. As the wireframe was too long, the adjustment reduced the height - it was driving me crazy.
My first impression is that both the Dped and the AA wireframes are trustworthy, but I´ll have to make some more exact measurements.
I´ll keep you posted!
Cheers,
Aleatorylamp
Here is a little comparison between Wire Frame Models in
Aircraft Factory 99
Aircraft Animator
and
DPED.
The model here is very simple. I just took one of the Workshop Projects I had and added a 12-sided Circle imported Part.
A screenshot of the saved Assembly in AF99 was then taken.
After producing the project, I then took a screenshot from AA and from DPED.
The results were interesting....
DPED
Height = 127 to 651 = 524 Pixels
Width = 250 to 774 = 524 Pixels
AA
Height = 165 to 574 = 409 Pixels
Width = 389 to 798 = 409 Pixels
So far, so good.....
AF99
Height = 111 to 695 = 584 Pixels
Width = 205 to 819 = 614 Pixels
Hey! The numbers don't match!!!
So the stretching is obviously not imagined.
There actually is just over a 5% difference.
So, now that we know which screenshot NOT to use, how does the Warhawk compare to the drawings?
The attached screenshot shows an overlay.
I believe it is pretty close but there are still a couple issues as noted by the arrows.
The worst section is on the Radiator scoop as expected.
The Carburetor scoop needs its contours adjusted a bit.
The Windscreen should also be moved back slightly. I believe the difference is under one inch.
It is interesting that the AF99 wireframe distortion now has been nailed down to a definite dimension in pixels!!
At least using DPed or AA´s wireframes allows reliable, undistorted shapes, and is a relief for trustworthy adjustments after comparisons to drawings or convenient photos. With the latter, of course, there is also the added factor of distortion which you have mentioned too, due to excessive camera proximity, and the best photos I suppose would be those taken with tele-objective from a certain distance.
I certainly miss some kind of intelligent image recognition program for airplanes, similar to those available for face-recognition in virtual photo albums, which would allow a rotation to create a 3-view set in the X, Y, Z axes! ...Wishfull thinking.
Just to check on AF99´s top and front view wireframes as well:
Comparing these to the Dped and AA ones, and it turns out that both of these are also too low in proportion.
The top view shows a shorter wingspan and the front view, a lower height.
Sometimes the degree of precision in the drawing is not matched by the degree of accuracy in the wire frame that one can get from DPED or Aircraft Animator or even Aircraft Factory 99.
Over the last few days, I tried to check the wire frame screenshots against the tech drawing.
The results were a bit less consistent than I would have wanted.
The large errors that are over about 0.07 foot are noticeable but anything less is a rather random thing as far as screenshots are concerned.
With one screenshot, I was able to get very good alignment but with offsets that were a bit different than I had calculated.
One the next screenshot, I was able to get some pretty good matches with the original alignment.
The lines themselves are around 0.05 feet wide which isn't very very bad considering that the scale means the lines are only 5 pixels wide.
I also found areas that could use a slight curve but I can't put them in because I don't have vertices there.
So.....
I just found with the latest screenshot that my model's Center of Gravity is
2.35 feet behind the Firewall (as before)
and
0.55 feet below the Fuselage Reference Line.....
After a couple hours yesterday to make sure my template drawings match up with the tech drawing, I decided to start hammering on the virtual metal....
The Canopy Frame and Glass took an hour last night and a couple hours this morning.
I believe it looks fairly reasonable and I even improved the shape a touch.
Now here is the nasty part:
Way back when I designed this aeroplane, I tried to line up as many things as was reasonable to do so.
The front of the Canopy lines up with a major partition line in the Fuselage AND THE major reference line in the Wings.
Moving the Canopy front edge from 1.21 Feet to 1.11 Feet ahead of the CoG affects how the Fuselage would line up with the Wings.
Although the reference line on the Wings doesn't need to line up wit the reference line on the Fuselage, it is / WAS nice to know that I took the care to arrange things this way when first laying out the project. Pity the dimensions were not quite right.
At this point, with a mismatch of the Wing reference and Fuselage reference, the difference needs to be made up via adjustments to the Wing Fillet.
Note that in these screenshots, the lower edge of the Windscreen has been moved back 0.10 feet and the upper edge has been moved back around 0.18 feet.
The other screenshot is a reference part I found that shows the degree of change of the Aft Fuselage between the original Wylam drawing model and the current model. Looks similar from a distance but obviously the shape is quite different.
Most of the changes to the AF99 model are now done.
The cropped screenshot shows that the alignment is pretty close.
There still needs to be a minor shape adjustment on the Carb Scoop.
After that, the texture files need adjusted to fit all the changes to the model underneath.
The biggest issue will be the Radiator Intake which has a lot of textured pieces within.
THEN comes the fun part with SCASMing.
Note that I will need a new Canopy Interior because the pieces below the texture have changed shape.
Most of the Canopy Frame interior is adjusted to be narrower, so this will need to be done again.
There are also a few animation changes from the version that Aircraft Animator generates.
This is the correct screenshot. The differences are on the bottom of the Radiator Intake.
Note that the vertical mismatch of the Cowl is due to the incorrect offset of the Thrust Line in the tech drawing which I assembled using the notes on the drawing itself.
This drawing listed 3.75 inch from FRL.
The factory drawing lists 3.078 inch....
....which is what started off this latest rebuild cycle.
Luckily none of the textures required any significant remapping.
Some of them needed a bit of re-drawing though, so one more step is finished.... Maybe.
There was also a minor adjustment in assembly sequence and adjustment of Glue to reduce a bleed.
It was only reduced, not cured.
Next is another dimensions check. I found something a bi strange on the lower part of the Cowl.
Hopefully I won't find anything significantly wrong.
The lower part of the Cowl was re shaped to make the sides more parallel at the opening.
One very interesting thing was that I first tried to draw some of the contour lines for the Cowl and could not get them the way I wanted.
I then decided to try copying the contour of the lower edge of the Cowl / Radiator Housing and by rotating it 30 degrees, I came up with a reasonably pleasing line that was shaped pretty close to the way I wanted.
There was a moment of panic when rebuilding when the resulting model showed a coloured stripe extending off to the horizon.
This was strange because I was only using 1127 Parts of 1200 allowed. Usually this is enough margin to avoid AF99 failures.
My development machine crashed shortly after and strangely enough, after a reboot, the subsequent builds all were clean.
Here is the current status. I still see a minor shape problem at the lower rear of the Cowl by the Radiator Flaps.
Hi Ivan,
Interesting to follow!
Glad the improvement is turning out the way you want!
I had recently found a much clearer, larger and more detailed 3-view drawing for the P-3 Orion that I thought looked wonderful, and promptly widened the fuselage width by about 1.5 ft accordingly, but alas!, the result was disgusting because the plane got all fat and ugly!!
So I put it all back to what it was, which was not hard to do as the fuselage was all structures.
This just shows how difficult it can get as regards reliability of some expectedly reputed sources.
The obvious question is: How reliable were the drawings you found? If the dimensions are correct, perhaps the aeroplane really is not as nice and sleek as you might want to build. If so, then you have the choice of building it accurately or to appear the way you want.
I say use the correct dimensions and the shapes should fall into place. If they do not, then your shapes need to be reworked until they DO fall into place.
I ran into a similar problem with the drawings by William Wylam which is what started this thread a while back. The dimensions were all labeled out and were "derived from Curtiss factory drawings", but the shapes in the drawings did not match the numbers sometimes. Also, I found out later that a rather large portion of the labeled dimensions were simply wrong.
Correct dimensions and shapes is what started the last couple rebuild cycles for the Warhawk.... Which incidentally are not done yet.
Reworking the lower rear of the Cowl was not terribly painful, but it involved changing a LOT of pieces.
After making changes, I found a bleed with the Carburetor Scoop........
THAT required some Glue Part changes and upon closer examination, I found another bleed which required more changes.
A few hours later, I finally arrived at a shape I was satisfied with... and then noticed another mismatch of vertices.
I had missed a couple Parts that connected with those I had changed.
The untextured model is what I often use to check on bleeds. A simple dark colour overall paint scheme can hide a great number of sins.
The other screenshots show some of the changes. If you know where to look, there ARE differences though some are not so obvious.
Time to flip it around in the simulator to make sure there are not bleeds or shape issues I do not yet know about.
Hi Ivan,
Given dimensions are usually no more than wingspan, overall length and height.
Fuselage width is not really mentioned very often, unless the info is on a publicity pamphlet for a passenger plane. I´ll have to do some serious photo-studying, because on some drawings, the Orion looks fat, but not on the photos.
Anyway, slowly we are progressing.
Cheers,
Aleatorylamp
After adjusting the front of the Canopy / Canopy Frame backwards by around 0.10 foot, the Internal Canopy Frame would need to be changed also.
The problem is that I could not find it....
I know I had called it "IFrame-Something" but could not find any of the related files.
As I commented earlier on Post #18 http://www.sim-outhouse.com/sohforum...?80890-Warhawk
Most of the pieces of the Internal Canopy Frame were narrowed a bit and I did not want to have to make the same edits since the majority of the Parts had not changed.
After a while, I found it (In the directory for the P-40K Project) and was able to do the edits pretty quickly.
Most of the SCASMing was done last night and I believe all that is left is to reset the Collision Bubble and change the Animations a bit.
I was debating whether or not to release a new copy of the P-40E. The differences are not even visible to ME and I know where I made the changes. The differences are so little that I don't think even the Flight Model has to change.
Other than the model, nothing else is different yet.
I needed this model as accurate as possible mainly to serve as a starting point for other versions.
Only a couple days after I thought the P-40E Warhawk was done, it is heading back into the Workshop again.
In looking at reference drawings for a P-40F, I found that my interpretation of Drawings / Photographs for the E model was not correct.
I noted earlier that the Carb Scoop was bothersome because the shape which was done from Photographs a while back did not line up well with the Factory Drawings. Most notable was that the Scoop extended much further forward on the actual aeroplanes than it did in the drawings.
Now I know why: Apparently the longer Scoop was a characteristic of the later P-40M/N. There is also more of an arch on the later aircraft.
The P-40E/K had a Scoop Inlet that nearly lined up with the Radiator Scoop and also had less of an arch.
This explains why I kept seeing something different every time I looked.
The most clear and largest resolution photographs are of modern aircraft and I suspect that the pieces are interchangeable between the E and the M/N models. Perhaps some of the modern P-40Es are wearing later pieces.
So.... Photographic evidence is great but knowing what is original is important and in this case I clearly did not know.
Hello Ivan,
Philosophically speaking, if the real aircraft of the different editions that you mention could wear either type of pieces, you´d be correct both times!
One thing would be a pristine model just run out of the factory, and another, a unit that has already had quite a few flying and combat hours.
It just occurred to me, now that you´ve discovered what´s going on: If you were to depict a certain unit with a certain registration number and a certain type of scoop seen on a photo, it could be an interesting detail to be able to mention something to that effect in the readme!
Cheers,
Aleatorylamp
My preference is to build them as they were when they made their reputations.
They should be as close to fresh from the factory as possible.
I don't really care to represent aircraft that are full of replacement parts, worn or damaged.
I figure the Virtual Pilots deserve aeroplanes in the best condition consistent with the time period.
I know there are plenty of P-40E that served until they were war weary or have even lasted to modern times, but I am not interested in depicting those.
What sense does it make to build a P-40E with Pieces from late wartime P-40Ns? By that time they would most likely serving as Trainers instead of Fighters. The ones then in combat would be P-40Ns. To use the mixed parts P-40E earlier in the was would be an anachronism and I do not like those. To create those accidentally is an indication of poor research.
I will correct the P-40E eventually. It is a simple fix but the re-SCASM may take a while.
I have a new set of reference drawings to work on first.
Hello Ivan,
I see your point. It could be seen as an excuse for skiving and not working the correct part into position. Pacience then, to put it right!
I´m trying not to bend the parts too much to force them into position on the Orion. I have all the Fuselage Station points lined up correctly now, all nicely measured out from the nose, but some of the Waterline ones are giving me a hard time now. Pacience here too!
Cheers,
Aleatorylamp
In doing a lot more poking around, I am finding that there were at least TWO different length Carb Scoops on the wartime P-40E on the assumption that the photograph captions are reliable. Often they are not.
The very short Scoop appears to have been standard on the early aircraft. There are even a couple surviving P-40Es with that scoop today.
This scoop can be recognized by the large gap between the Spinner and the opening of the scoop. It appears to be nearly inline with the Radiator Opening. (It is actually slightly ahead.) It appears to be slightly under 6 inches from the Spinner line.
Next there was the long Scoop which with an opening about 1-3 inches behind the Spinner line. This can be recognized in photographs as being distinctly ahead of the Radiator Opening. I have found a few war time photographs of Short Tail P-40s (captioned as P-40E) with this scoop. One photograph shows several in a flight with the same scoop so it is unlikely to have been retro fitted.
The USAF Museum's P-40E has this scoop and I would imagine that they have the resources to put it the way it should be.
The Smithsonian's P-40E (Lope's Hope) also has this length Scoop and again, I expect they would have the resources to make it the way it should be even if they needed to fabricate pieces
There may be variations on the longer Scoop. Some look like they extend all the way to the Spinner while others have a noticeable gap there. There really should not be a gap between the Spinner and Cowl because on the real aircraft, there was actually a bevel in the Cowl and a slight overlap but it is not easy to be subtle with AF99.
So.... It appears that the current Carb Scoop on my P-40E isn't necessarily incorrect for a War Time P-40E.
This is not to say there are not other errors because I know there are at least a couple which I have chosen not to fix because of resource limitations in AF99.
I wonder if any of my other projects would hold up at this level of examination.
Here is a little comparison between the way it is today and the outline of the way it was a couple years ago.
Although overall dimensions are the same, many things are obviously shifted.
Members who have read this thread: 6
Posting Permissions
Bookmarks