Mitsubishi A6M - Revisited

[Ivan]

The Mitsubishi A6M5 was one of my earliest design projects for Combat Flight Simulator.
It was first built in late 2001 when I was very inexperienced in making choices and gathering information for a project.
The source for the drawings was one of the books by William Green.
For those who are not familiar with these books, the illustrations about the same quality as a typical painting.
These days, I would never start a project without a "reasonably accurate" set of drawings, but back then, I didn't know any better and of course, there was the urge to get a completed aeroplane in the simulator as quickly as possible.
Unlike some of the other illustrations, the one for the A6M5 turned out to be fairly good though over the years, I have made quite a few detail corrections to the original design as I added new features. The greatest number of changes was done when the original A6M5 was modified into a A6M2 Model 21 a few years ago.

When this modification was made, there were several things that bothered me about the design but I did not have the knowledge to address them at the time.

The first, which was addressed about a week ago and is the subject of this thread, was that the Vertical Center of the visual model was located too high. At the time I started working on the model, the idea of properly locating the Horizontal Center of the model was something that was pretty obvious but the Vertical Center of Gravity of aircraft is not often discussed and very seldom found in documentation. Without knowing any better, the obvious centerline would seem to be through the thrust line / propeller shaft of a Radial Engine and that is what I used.

The two attached screenshots show the front and side views of the A6M2 with the vertical center line through the spinner.

-------

[Ivan]

.......The horizontal center of the model turned out to be fairly reasonable as suggested by the location of the Wing's Main Spar and the location of a typical suspended centerline Drop Tank.

The vertical center of the model was most probably too high....
Most of the structures on aircraft tend to be quite light despite their bulk.
The exceptions to this are the items that need to carry a heavy load such as wings and landing gear.
Other items with substantial weight are engines, engine accessories, fluid reservoirs, radiators and armament and perhaps the crew.
Fuel tanks and associated plumbing and booster pumps also weigh a bit more than their surrounding structure.
(Fuel and ammunition are disposable loads and accounted for elsewhere.)

Note also that the CoG would vary slightly depending on whether the wheels were up or down but that issue will be discussed elsewhere.

The attached screenshot shows that most of the heavy items on a typical radial engine fighter such as the A6M are below the engine thrust line.

-------

[Ivan]

The latest of my A6M projects was the A6M2 Model 21.
It also happened to be the earliest and lightest of the A6M series and probably the most well balanced.
As the aircraft got heavier, some of the added weight became more noticeable.
This was quite noticeable on the A6M3 Model 22 which I used for testing the effects of moving the CoG.

For Testing, It made sense to do as little actual work as possible but I also did not want to have any visual distractions.
The quick and dirty way to accomplish this was to shift the entire MDL file via SCASM.
The first screenshot shows the highlighted code which moved the entire MDL's display.
This is same method as used for shifting the Virtual Cockpit display for the Padlock View in Quick Combat.

The second screenshot is a view from the simulator that shows the modified Damage Profile and matching AIR file.
Note that this A6MX version has an Alpha Transparent Canopy while the original A6M3 Model 22 does not.

The third screenshot shows how the MDL file is seen by DPED which reads the offsets specified in the actual code rather than the effect of the "TransformCall" that moves everything.

- Ivan.

[Ivan]

After a bit of examination and experimenting, a shift of 0.55 feet, or 0.16764 Meters, or 6.6 inches, or 86 SCASM Units seemed most appropriate.

The SCASM edit of the MDL was good for a quick visual with little work, but it wasn't good if there was to be any further development of the series which was the original purpose of this correction.

To do it properly took a lot more effort: about 4 evenings of work.
First All the AF99 Parts needed to be moved 0.55 feet up. That was pretty easy and done in about 5-10 minutes.
Next, no Structures needed moved because there was no Longitudinal shift.
Finally for AF99, all of the Paint Mapping needed to be moved because the Parts had shifted.
This was not difficult because of the utility that I had developed earlier for the Airacobra Project.

The Animation in Aircraft Animator was very time consuming because the A6M Landing Gear has a LOT of pieces moving in multiple directions.
The SCASM corrections also took a longer than expected.
At one point, I thought I was done but one piece of the Landing Gear seemed to be displaying incorrectly.
Apparently I had corrupted a "VectorJump" call somewhere and after an hour or so of looking for the problem with no success, I just went back and generated the MDL again from Aircraft Factory 99.

The attached screenshot shows the updated model.
The corrections to the Damage Profile and obvious corrections to the AIR file were easy.
There still needs to be a fair amount of fine tuning needed for the AIR file.
The only obvious visual difference is a texture correction I made to the outline for the Flaps on the wing upper surface.
The issue with the outline is one I had known about for several years but never seen the opportunity to update until now.
It makes for a good visual distinction so that I can tell one model from the other.

- Ivan.

[Ivan]

Last Night I decided to do another minor revisit to the A6M2.
It had a minor change to the model which meant another trip through Aircraft Animator, Hex Editing and SCASM.
This time, I took photographs of the parameters used on various animated pieces so that it would not be necessary to redo everything from scratch. It turned out not to be necessary. The change was minor enough that just about all the significant animated pieces were picked up with no problems.
For some reason, the Flaps never seem to be recognized from one cycle to the next unless they are tagged "Flaps Up" in AF99.

The SCASM Edits took about 2.5 hours even though almost everything from the last SCX file was useable in the new version.
It probably didn't need to take that long but working while sleepy makes things a lot slower.

Of course, right after finishing up and while admiring the results, I found something else that I probably should have fixed.
The first Screenshot shows a view of the rear of the cockpit and canopy area.
Note that the bottom of the Antenna Mast is missing where it should be on the inside of the canopy.
Note also that this particular aircraft does not have a DF Loop which was fairly common on the A6M2. (I am NOT going to add that.)
I am also not sure that the framing is entirely correct in this area.

After the MDL edits, I took the A6M2 up for bit of fun flying and to check for any weird handling.
Below about 250-300 Knots, this is one of the best behaved birds around.
Loops can be started at ridiculously low airspeeds with only about 60 MPH over the top.
A Split-S can be done at such a low altitude that other aircraft trying to do the same would be making craters in the ground.

- Ivan.