P-47 Progress Thread

[NachtPiloten]

prop pitch - airfile

Do you have the information on the prop pitch range for the P47? Found a document at ww2 that had numerous props and this type of data.

 

[Frosty]

Do you have the information on the prop pitch range for the P47? Found a document at ww2 that had numerous props and this type of data.

Thanks, I think I got that document - and the pitch range is stencilled in yellow on most blades. You haven't found a way to animate props using the cockpit controls, have you? Otherwise I am not sure how to use that data other than rotating the blades to a position somewhere in that range. In the end, it isn't all that important other than that I would like to have found a decent drawing of basic lengt/width dimensions of the different blades so I could approximate their different shapes as best as possible. You'll only see 'em with the engine turned off anyway...
(And I do have various smallish drawings of the different P-47 blades but how can I trust such sketchy work?)

 

[gecko]

We have animated prop pitch in the SJ Spitfires, but it requires the RSM code to make them move correctly. Though I'm guessing the Curtiss Electric props had a manual mode, which could perhaps be managed by stock animations, but the operation of the prop will not match the animation. Alternatively, you could set them at the lowest setting, because that is the position the rpm governor will move them to at low RPM or when the engine isn't running.

(I still have to see if I can get your latest - well, maybe not so latest by now - thingie to work but I am up to my neck in P-47 stuff. With the new HD, I had to reinstall some stuff so maybe it does work this time...)

I realized not long ago why that thingie wasn't working. I've not been able to do any work on CFS3 proper lately, but hopefully soon that will change.

 

[NachtPiloten]

Frosty

Have you been able to get the turbosuperchrager to work? Or has anyone. What is this RSM code?

 

[Frosty]

Have you been able to get the turbosuperchrager to work? Or has anyone. What is this RSM code?

@ Q1: I'll quote myself here :

I haven't tried, sorry.

And I don't think I will. I'll stick to modeling only because at the present that is what I do best. With the aid of Daniel I'm sure I can get the additional animation parts working (clickable cockpit) but doing the airfile etc. is way beyond my capabilities. ...

@ Q2: RSM is Real Systems Module, the extra code that makes the additional animations and functions possible and is 'injected' into the game while running it (or something to that effect). It came with the latest SJ Spits - if you have those installed, look into the CFS3 'systems' folder for all those extra 'Systems.*.exe' scripts. You'll have to ask Daniel and Steve about the specifics because I don't have the faintest clue how to implement that myself...

 

[Frosty]

We have animated prop pitch in the SJ Spitfires, but it requires the RSM code to make them move correctly. Though I'm guessing the Curtiss Electric props had a manual mode, which could perhaps be managed by stock animations, but the operation of the prop will not match the animation. Alternatively, you could set them at the lowest setting, because that is the position the rpm governor will move them to at low RPM or when the engine isn't running.



I realized not long ago why that thingie wasn't working. I've not been able to do any work on CFS3 proper lately, but hopefully soon that will change.

Did I miss that?? . I am not sure I understand you correctly; Should the prop blades of the aircraft model change pitch when I move the appropriate lever in the VC? The lever itself, of course, is animated but I have not seen the blades move on the Mk.II, Mk.VA and Mk.VB models (I just tried) when manipulating it in the VC or using the rotary I have assigned that function to. The rest of the RSM action seems to work, so please tell me what I am doing wrong!



But the thingie worked for you, didn't it? But if you have found why it didn't work for me, I wasn't a certain long-eared member of the Equidae familiy after all!

 

[gecko]

I don't think you are missing anything, since the movement of the blades is not directly tied to the movement of the prop control lever. I'll explain.

There are basically three type of propellers - fixed pitch, variable pitch, and constant speed. CFS3 doesn't have the ability to model a variable pitch, but I can make one with the RSM.

With a fixed pitch prop, RPM is a function of engine power setting, and airspeed (mainly). Go faster, or increase throttle, and you get higher RPM. There is nothing to keep it from getting too high or too low.

A variable pitch propeller can change the pitch of its blades, adding a third factor to RPM. Here, the pilot directly controls the blade pitch, and can use that to adjust RPM manually. There is still no automatic system to keep RPM within safe limits, but the pilot does have greater control over RPM than with a fixed pitch prop. As far as I know, all variable pitch propellers are electrically operated, usually by an inker-dinker (increase-decrease) switch, so blade pitch can be adjusted at any time, including when the engine is off.

Constant speed propellers have a governor between the pilot's propeller control and the propeller itself. Instead of setting a blade pitch to try to control RPM indirectly, the pilot sets the desired RPM with the cockpit control and the governor sets the blade angle to maintain that RPM. In most conditions it does this quite precisely. Most constant speed propellers are hydraulically operated. Don't think about the aircraft's hydraulic system though, hydraulic here simply means fluid, and that fluid in this case is oil from the engine. This means that when the engine is not running, or you are losing oil pressure, the blade pitch cannot be controlled by the pilot. Other constant speed props are electrically operated, and usually still have a manual control mode which allows them to function as a variable pitch prop, using a separate control, as desired. With either type of constant speed prop, the governor will do whatever it can to maintain the pilot's chosen RPM, but eventually extremes of airspeed and engine power will cause the blades reach their rotational limit, at which point the governor just holds them there until airspeed and engine power allow for normal operation to resume. This means that when the engine is idling on the ground, airspeed and engine power are very low, not enough for the governor to maintain the pilot's selected RPM, so the propellers blades are held at their smallest angle by the governor.

The SJ Spits and your Hamilton Standard P-47 have hydraulically operated constant speed props, so direct control of prop pitch isn't correct for them. But if you look carefully at the Spitfire, you will see that the thickness of the prop disk is animated and reflects the actual pitch of the propeller blades. It is also possible in certain conditions to get the prop to stop in a position other than fully fine pitch. Is suspect your Curtiss Electric P-47 has a manual mode the pilot can select, in which case direct control of the blade angle would be possible. But none of it is going to be truly correct without external code. CFS3 just doesn't have that capability on its own.

And yep! The thingie failed through no fault of your own. Which I guess gives me the long ears and stubborn temperament!

 

[Frosty]

I don't think you are missing anything, since the movement of the blades is not directly tied to the movement of the prop control lever. I'll explain. ...

Thanks for the explanation, mate!!
This all came up because I remember a Spit (I think it was) in a civvie sim (FSX, I guess) where one could see the blades rotate when the prop lever was moved (sitting in the VC with the engine off).

 

[gecko]

Ah, yeah, the real airplane doesn't work like that.

 

[Frosty]

Wanting to complete the animations for the tail wheel doors, I have been struggling with the final set-up for the landing gear animations; I have found a drawing with the exact positions of the LG fully extended, in the static position and fully compressed. I have also studied the various tutorials on how to make landing gear animations look convincingly real. Unfortunately, I can't seem to get the hang of it. Even though I have modeled the wheels at the proper locations at the correct keyframes (50, 75 and 100) in gmax, they do not seem to be in the correct position in game (either too high above the ground or to deep in the terrain with often a difference between the main LG and the TW) and do not move in a realistic manner. In addition, when coming to a full stop, the halted compression animation sometimes makes the tires seem to float above the ground. Not really life like or convincing and downright annoying. So, it is question time again:

1. At what exact keyframe are the .cfg file contact points for the landing gear determined? I know the aircraft is supposed to rest on the runway at frame 75, so should I determine the XYZ positions for the .cfg at that frame number, at frame 50 (fully extended) or at frame 100 (fully compressed)? The P-47 has different compression distances for the mains and the tailwheel, so this might be one of the reasons for things not matching up (see values below).

2. Could it be that the center of the aircraft in gmax (SDK: 25% chord at XY=0,0) is slightly off compared to the value in the .cfg file and that this results in incorrect contact points?

3. How can I avoid the jackhammer effect when taxiing on the ground? I have the oleos take the full weight of the aircraft at frame 51 (so there's no additional compression between frame 51 and 75). When I tried to leave frame 51 out (so there's just frames 50 and 75), the landing gears jackhammered the distance between the fully extended and static positions when taxiing over rough terrain. This looked utterly ridiculous. With frame 51 in, there's just the minimal compression between the static and fully compressed values which looks much more natural.


If anyone wants to do the .cfg math, the values are as follows:

Mains: maximum oleo extension is 9 inches. The oleo compresses 7.4in to the static position. From static to fully compressed, it compresses a further 1.6 inches.
Tailwheel: maximum extension for the TW shock absorber is also 9 inches. It compresses 7 inches to the static position. From static to fully compressed takes another 2 inches.
Static angle is 12°.

To get an idea how little the LG compressed when taxiing off-road and at combat weight, watch the 30 seconds from 12:00 min in this well-known 1947 film (Thunderbolt!) https://www.youtube.com/watch?v=tWD4ITJGdTw. You can judge the stiffness of the oleo's to some degree by looking at movement of the belly tank in relation to the ground.

Thanks for your help!

 

[hairyspin]

My own findings were that the contact points in aircraft.cfg and the suspension travel in the model are more imtimately linked than I first thought, especially the relationship between maximum and static compression as tied to the animation keyframes. This link is for FSX modelling, which has a 200-frame animation instead of CFS3's 100 frames, but the principle is just the same. BUT be sure to follow the sdk's animation for correct keyframes at each stage.

https://www.fsdeveloper.com/wiki/index.php/Contact_Points


edit: there should be movement between frames 50 and 75 – that's half of the compression animation range. Looking at the Thunderbolt video footage, I'd say, like the Mossie, the Jug depends on compliant tyres as much or more than the oleos!

 

[Frosty]

Thanks for the reply Tom!

I've used that tutorial, I've used AirWrench to check some data, but I find the results of both disappointing: it leads to jackhammering oleos which I utterly dislike. You may be right in that the tires absorb a lot of the forces encountered but I don't think it is possible to replicate that in CFS3. So, until I (or someone else) come up with a better solution, I'll stick with the present setup. I need to tweak it a bit but for now it seems the best I can do. Too bad things aren't straightforward enough to directly copy real-life data but that probably also has to with the way the simulator is programmed...

 

[NachtPiloten]

Compression on oleos

I fear that the game is coded for linear compression, that is from 50-75 and 76-100 is the same in terms of resistance, not visual movement. Attempting to model the non-linear compression of the oleo may not really feasible in CFS3. If someone can prove me wrong that would be great.

Prop pitch is another issue. I am not convinced that if you model say 30 degrees of pitch (ignoring feathered), that the game models this well has you coarsen or move to fine pitch. The Germans had these real nice prop pitch gauges that I have been able to animate as per actual specs, but a mere 6 degrees of pitch reduces engine rpm by several hundred. I could be wrong and if this is accurate then great, as I have no experience with how much a change in pitch translates to changes in rpm. Of course this depends on the engine, prop, boost etc.

Prop disk hmm that is interesting. I wonder if you scale animate the slow and blurred disks (just like you scale animate dust covers for the oleos), call them lever_prop_pitch, and then link them to a dummy helper called prop_slow, and prop_blurred, and in turn link these to the prop node, that whilst in the cockpit one could "see" the animation effect.... prop_slow and prop_blurred are visual checks activated by engine rpm...

Oh so much to consider and so little time......

Oh, have you ever been able to activate a turbosupercharger lever and effect?

 

[hairyspin]

Tyre compression

Tyres can be made to apparently compress just by making the suspension travel in the contact points a little longer than the animated movement. The tyre sinks slightly into the ground and looks like it's just squashing with the full fuel load, ammo, drop tanks/1,000lb bombs or a heavy landing.

BTW, your figures for suspension travel to static compared to full deflection are 7.4" versus 9" – that's a max:static ratio of 1.216. I think that's much too low and your sources might be making a mistake somewhere, but it leaves the sim with very little resistance to compression beyond static deflection.

 

[Frosty]

I fear that the game is coded for linear compression, that is from 50-75 and 76-100 is the same in terms of resistance, not visual movement. Attempting to model the non-linear compression of the oleo may not really feasible in CFS3. If someone can prove me wrong that would be great.

Prop pitch is another issue. I am not convinced that if you model say 30 degrees of pitch (ignoring feathered), that the game models this well has you coarsen or move to fine pitch. The Germans had these real nice prop pitch gauges that I have been able to animate as per actual specs, but a mere 6 degrees of pitch reduces engine rpm by several hundred. I could be wrong and if this is accurate then great, as I have no experience with how much a change in pitch translates to changes in rpm. Of course this depends on the engine, prop, boost etc.

Prop disk hmm that is interesting. I wonder if you scale animate the slow and blurred disks (just like you scale animate dust covers for the oleos), call them lever_prop_pitch, and then link them to a dummy helper called prop_slow, and prop_blurred, and in turn link these to the prop node, that whilst in the cockpit one could "see" the animation effect.... prop_slow and prop_blurred are visual checks activated by engine rpm...

Oh so much to consider and so little time......

Oh, have you ever been able to activate a turbosupercharger lever and effect?

All interesting ideas and suggestions Ted! I 'll have to think them over and see what works for me. Thanks!

I know about the lack of time (and no, I have not been able to make the supercharger lever animation work, let alone its effects on the FM. I am counting on Daniel's external scripts to fix that for us some day )

 

[Frosty]

[Arrgh, what is wrong with the quote button ]

@ Tom: Yes and yes.

I know about that trick but I think Ted hit the nail on the head with a single word: linear. But come and think of it, you can set a non-linear animation in gmax. I just don't know if CFS3 buys that. Hmm, something I'll have to look into.

The other 'yes' concerns the compression values and the resulting Max/Static ratio: I got the same low ratio as you (and AirWrench too). You might be right in the data being wrong but then please tell me whether I interpreted this factory drawing incorrectly (see attachment):

View attachment 93X10200 AIRPLANE - GEN'L. - ASSEM. THREE VIEW (2).pdf

 

[hairyspin]

Compare the movement of the P47 aircraft in the video you linked earlier with these figures. The maximum deflection specified seems plausible but the movement visible in the video seems more than 1.6”. I’d ignore that 7.4” static figure, work with the maximum deflection given and specify the contact points to suit CFS3.

 

[Frosty]

Thanks for the suggestion, I'll take that into consideration!

Another question, though, about that tutorial you posted: Why calculate a keyframe number to determine the Mac/Static ratio? Why isn't this simply done at eg. frame 75 (static position)?


BTW, I take it the static position of 7.4 inch compression is at rest. The P-47s in the video are taxiing which I assume would lead to oscillations around that number and not only in the remaining 1.6 inch region that is left to max. compression. The terrain (or speed/weight) would have be very rough (or high) to reach max. compression while taxiing - or so I imagine.

 

[hairyspin]

With a high Max:Static ratio, the undercarriage is not going to be at the rest position at frame 75, but before it - by quite a way if the ratio is especially high. OTOH, a low ratio means the rest position will be beyond keyframe 75. In fact frame 75 would only be correct for a ratio of 2, which almost no model uses.

I realised this while puzzling out why some models would load in the sim just sitting perfectly while others fidgeted around - including mine. There had to be a reason and setting static height and pitch for the correct resting keyframe was what I came up with. Gratifyingly, it worked.

Without manufacturer’s explanation, I’d be unsure that “static” position means the same as it does for CFS3.

 

[Frosty]

Here's one of the (hopefully) last LARGE modelling tasks I had set myself for the external model: a new Pratt & Whitney R-2800 engine!

As I don't have access to any blueprints of the R-2800 and I couldn't find any serious schematics on the web, the engine model is mostly built by studying a lot of photos, even more guestimation, some free hand doodling and copious amounts of fantasy. But since a significant portion will be hidden in the cowling, I hope the glaring mistakes are not going to be that obvious .

Anyway, to me it looks reasonably convincing and much better than the placeholder I made at the start of this project (maybe I'll keep that for a lower LOD). It took me a while to be able to distinguish between the many R-2800 versions but I think the GE magnetos and the blanked off locations for a governor and distributor are correct for the R-2800-59 with a Hamilton prop (if not, feel free to correct me). There are a few small details that need some work, but it's nearly ready for the unwrapping and paint shed.
The model needs some cleaning up (don't ask me about the poly count, especially not that of those horribly complicated GE magnetos ) and texturing. But being a radial engine, I can copy a lot (eg. I didn't create the different cylinders in the front and back bank, I just copied and mirrored them) after unwrapping and painting. Enjoy!

Old engine:



New engine:



In the cowling:





Oh, and I lost some time dabbling with visual effects: I think I have cobbled together an acceptable cold start smoke effect that fits the soundpack I use.