From my many years of experience with realistic online simulators, I would say that at least ninety nine out of one hundred aerial combat simulator users have no idea how trim actually works on real aircraft. This is mostly due to the impossibility of modelling trim completely correctly in simulators due to limitations in gaming joysticks, but also due to a few very popular simulators (such as the IL-2 Sturmovik series) making it worse by not modelling trim as well as it can be done. Indeed, IL-2 modelled trim very badly; changing the trim, in that simulator, caused the elevator to move but the virtual stick stayed stationary! In reality, for most aircraft, such an occurence would mean that the control cables were broken.

Simulators such as Aces High II have gotten it as correct as it is possible to do without precise force-feedback joysticks which are capable of a two hundred pound pull and which match your virtual stick's position at all times. And those obviously don't exist for the commercial gaming market. Existing force-feedback sticks available commercially are rubbish; they are much less precise than a non-force-feedback stick and they do not necessarily match the position of your virtual stick. Indeed, if they tried to always match virtual stick position, they would be snapped off when you tried to fight the stick with a strong pull.

Aces High II getting it as accurate as possible without these non-existent special joysticks still doesn't eliminate the confusion. I therefore drew up a little diagram with my humble Microsoft Paint skills to demonstrate how real trim works in an effort to remove as much of the confusion concerning trim as can be done. Don't laugh.



This diagram was orginally part of my reply on another forum to the question, "What do you think of the effect of trimming on flight attitude and top speed?" My reply accompanying the diagram was as follows:



Dante, I'm sure you know how real aircraft trim works already. But in the case that you do not, I beg you earnestly to drive to your local airfield, find a pilot who is waiting for someone, and ask him to show you how trim works on an airplane. It is not at all how it is portrayed in P.C. simulators.

Hi Benny Moore, here is a reply from our FM expert, John Cagle:

"Benny,

The very first thing one should do in determining trim conditions is research the trim operation for a specific aircraft. As an example, the A-4 the horizontal stabilizer is a de facto stabilator where the trim control adjusts the stab angle of attack and the pilot stick controls the elevator deflection angle. Obviously, this is treated completely different from a stabilizer with a trim tab.

For the more traditional trim tab, such as the one you have illustrated, the tab functions as secondary control surface added onto a primary control surface. They are treated compoundly and use nearly identical math as a single control surface, just more of it. As far as physical function goes, I quote from my textbook, "The trim tab of a mechanical control system is held in place by an irreversible mechanism, such as a worm gear or jackscrew, that can be adjusted by the pilot but requires no pilot input to maintain...........While a trim tab can be used to zero the stick force at trim, it does not significantly affect the force needed to deflect the elevator away from its neutral position." The latter portion of the quote is derived from the fact that a trim tab usually has a small planform area, meaning a low reaction force; but a large moment arm which leads to a significant moment. So, we are on top of trim function, but I am not quite sure what you were meaning or criticising when referencing control forces and current force feedback joysticks.

Kind Regards,
John"

I'm glad to hear that you guys know what you're doing yet again. I knew that some aircraft, such as the Messershmitt Me-109, used the stabilizer trim instead of trim tabs, but I didn't know about the A-4. Most American aircraft use trim tabs. Simulators tend to get it wrong; for instance, in IL-2, all aircraft use the stabilizer system (you can tell because the virtual stick does not move with trimming), which is ridiculous because most of the fighters portrayed in that game used the trim tab system, historically. IL-2 also got the brake action wrong, portraying all aircraft as having the Russian lever-and-rudder brake system instead of the Western toe brakes. I'm not even going to talk about how badly they messed up on brake strength.



In all simulators I have ever examined, physical gaming joystick input directly controls virtual pilot effort. It does not directly control virtual stick position. Physical joystick full back does not necessarily mean that the virtual stick will be full back; it only means that the virtual pilot is exerting maximum back pressure on the stick. When stick forces are high (such as in a dive), the virtual stick might only be 5% back even though the physical stick is 100% back.

This isn't an easy problem to avoid. Simply causing the virtual stick to always match the physical stick would require negating important factors such as trim and stick forces. That would not be a realistic portrayal; while it's always possible to easily move a gaming joystick from neutral to full deflection, in the real aircraft you can't always do that. The only way I can think of to correctly portray this would be to have a gaming joystick that is capable of withstanding a two hundred pound pull, and is equipped with a precise force-feedback mechanism (not the shaky rubbish commercially available) which causes the physical stick to exactly match the virtual stick position at all times. Furthermore, this special joystick would require the player to exert the true stick forces to move the joystick, so that the player would have to put real pressure into it. This would remove the problem of "virtual pilots" and virtual stick forces in a realistic manner.

Unfortunately, in the case of aircraft without power-boosed controls, that would mean that only the players with strong arms would be able to perform some maneuvers. But that was true of the real aircraft as well. Anyway, I'm not seriously suggesting that these hypothetical joysticks be manufactured. I'm just saying that it's the only way that trim can be modelled correctly. All P.C. simulators out there simply use a virtual pilot.

More from our FM expert John Cagle:

"Benny,

Federal Aviation Regulations (FAR) part 23 require that no elevator stick force exceed 60lbs or any yoke force to exceed 75lbs. Also, prolonged forces cannot exceed 10lbs. These are modern rules, so they do not effect all historic aircraft. However, I believe they cover the scope of the aircraft presented in Jet Thunder (my aircraft specific knowledge has been very focused on the A-4 to this point, so I can't make the claim for all of the aircraft). What would be a possiblty is a user selectable exertion force in the options menu. Those looking for more complete control can select a higher strength and those looking for more tempered characteristics can select one to their own abilities."

6.6 lbs per G in the Mirage III in Autocommand Pitch mode, which is Analog pilot selectable FBW, with Autotrim

Thank you very much, everyone. You've allayed my fears about the misrepresentation of trim in Jet Thunder, reassured me that Thunder Works is indeed dedicated to making a realistic simulator (not just using "realism" as a buzzword like most of the other companies out there), and even provided me with a most excellent idea for my imaginary force joystick (the scalable exertion option).

As an interesting historical side note, here's a comment from Jimmy Doolittle from the 1920's report "Accelerations In Flight" on airframe G-loads, and pilot stick forces:

"In the case of the pursuit airplane the speed range is great, and stabilizing and damping forces are reduced to a minimum consistent with easy handling in combat. The ability of the pilot to impose large dynamic loads therefore depends largely upon the ease with which he can move the elevators when traveling at speeds in excess of the maximum horizontal speed. If the elevators are perfectly balanced, there seems to be no reason why he can not impose loads very close to the theoretical values, and in the above tests, made with an airplane having elevators almost perfectly balanced, the actual loads obtained were but 3.5 per cent less than the theoretical. This airplane was designed to support a dynamic load of 8.5. Actually it would probably support about 10, judging by a static test of an airplane exactly like this one except that the wings were fabric covered. It would follow that if the airplane were suddenly pulled out of a dive at a speed in excess of 185 mph (and which would frequently occur in actual combat with this airplane) the wings would fail. It was this consideration which caused the engineering division of the Air Service to require a factor of 12 at high angles of attack for pursuit airplanes and to recommend against the use of balanced controls on that type."

The tested aircraft from which these early conclusions were drawn, was the high-wing monoplane Fokker PW7 model with wood veneer-covered wings, some of which peeled back as a result of Doolittle's rigorous flight maneuvers.

What is interesting to note in the quoted passage, is the conclusion by the U.S. Air Service's engineering division that aerodynamically balanced control surfaces should be avoided, so that greater stick pressures would be necessary to affect (pitch) input, thereby reducing the likelihood of a pursuit pilot being able to over-stress his plane's airframe in maneuvers such as the diving pull-out. This engineering recommendation also spoke directly to that day's state of aeronautical technology, where airframes could suffer catastrophic failure during expected combat maneuvering, and this fact encouraged the thinking to maintain (impose) physical limitations on a pilot's ability to move the controls under high-speed/high-load conditions (this to be incorporated into original aircraft design considerations).

Of course pertaining to today's era of military aviation, with the use of boosted/servo assisted surfaces, harmonized controls, and fly by wire technology, coupled to stout airframes made of exotic materials, the need to impose "force feedback" is far less of an issue, but it's taken aeronautical engineers (and test pilots) decades of research to reach this point where control ergonomics is such a highly-managed state of technology.

Interesting! I've had the wonderful opportunity to sit in a Republic P-47, a Bell P-39, and a Messershmitt Me-109, and all three had perfectly balanced controls. I wonder if the increase in structural integrity by the late 1930s caused the U.S.A.A.F. to reverse its policy on mass-balanced elevators.

I also recall the story of the Lockheed P-38's external mass balances on the elevator. Kelly Johnson said that the only thing they ever did was kill pilots during bail-out; the P-38 had internal balances that were perfectly good, but one of the test pilots (Army, not Lockheed) who encountered compressibility insisted that it was "tail flutter" and so the U.S.A.A.F. decreed that all P-38s manufactured, no matter what kind of improvements were made to the airframe, have useless external mass balances. The source for this is Warren Bodie's P-38 book.

The F.A.A. is specific to the United States of America, is it not? Or is it international?

Benny, the report was not referring to mass-balanced elevators, but to aerodynamically-balanced control surfaces, sometimes referred to as "horn-balanced" controls (ref: the Fokker Dr.I or D.VII's rudder, elevator, and ailerons).

Aircraft design (and any design effort for that matter) is not always a forward-moving art/science (sometimes there are lateral or even backwards movements). The thought to impose greater stick force requirements as a limiter on a pilot's potential control input in the 1920's was an illustration of one of these "lateral" design developments, where airframe integrity dictated that potential g-loads be controlled. Therefore the recommendation to avoid aerodynamically balanced elevators in subsequent designs as a means to regulate a pilot's ability to physically exceed his pursuit plane's structural limits during high-speed/high-g maneuvers was one particular example of an aeronautical dictate being a trade-off decision.

Now in reference to WW2-era aircraft controls, the imposition of higher force thresholds required to deflect control surfaces during high-speed/high-g manuevers could be affected by more deliberate means, but the effort to keep the pilot from bending or breaking his bird was still the goal, as it had been with early aviation aircraft (from "America's Hundred-Thousand"):

In some airplanes, noteably the P-51, pilots were sometimes overstressing the airplane during particularly violent maneuvers with structural failures resulting. Designers modified flight controls for both elevators and rudder to alleviate the situation. They added a 20 pound metal bobweight in the elevator control system to make it harder for the pilot to move them. When he moved the stick he had to swing the bobweight, which action opposed his applied force. Many pilots did not like exerting the extra stick force, but less airplane structure got broken.

Now (historical) flight sims are just snapshot treatments actually, that aim to portray the state and art of technology that might have existed in a certain time frame, but beyond their entertainment value, they shouldn't really be considered as true source material.

Of course we all recognize the goal of wanting "greater realism" in our flight sim products, but comparing historical developments that occurred during different times, and then in the context of contrasting different flight simulations, made in different years, by different design teams, with different imperatives, and different budgets, with different entertainment focuses, and then to an up-coming flight sim which has yet to be released.........(is an interesting endeavor to pursue to say the least).

Still, nothing wrong with discussing aircraft design and performance issues here (of course), and how these issues might relate to hopes of higher fidelity in future flight sim products. I think we've all enjoyed your questions and comments Benny, and those of the Thunder Works design team too (I'd just suggest that maintaining context in framing these discussions is also important).

Just wanted to point out that this is an inaccurate claim. IL-2 actually simulates all aircraft with trim tab systems, not movable stabilizers. Check the external views - the stabilizer is fixed, but when trimming, the control surface itself (in our case the elevator) moves as if affected by the trim tabs (aren't modeled). However, the joystick doesn't move accordingly (which doesn't really bother me as there's nothing interesting to see down there in flight). The other error in modeling is that IIRC the control surfaces move when trimmed even if the aircraft is stationary, something they shouldn't do as there's no airflow over the trim tabs.

The brake system is correctly implemented for not only Russian, but British aircraft of the time, which used the lever-and-rudder system as well. American aircraft are a late addition to IL-2.

Don't want to hijack the thread though.

Can't wait for JT. It's great seeing all the work being put into this project.

Thank you for the correction; I should have said "IL-2 does not properly model trim tabs." There are several glaring errors, as we've both pointed out. I believe (but did not extensively test) that changing trim also affects the gimbal limit of the elevator in IL-2, which it should not, regardless of airspeed.

Historically, German aircraft also used toe brakes. At least the Me-109 and FW-190 did. Those aircraft have been in IL-2 since the first release (as has the P-39, which is American), so there's really no excuse for them using the wrong braking system except that Oleg Maddox does not research things well before hammering them crudely into his dubious simulator. Speaking of incorrect brakes, did I mention that it is truly impossible to pivot on a wheel in IL-2, as you can in the real world (in World War Two planes, even)?

Fortunately, Thunder Works looks to be taking a very different approach, one that I admire very much. While I'm still ready to be disappointed (so many times in the past, "realism" has been used as only a buzzword for hyping and selling a shoddy product), I am allowing my hopes to be raised once more by the knowledge and dedication that the Thunder Works team has displayed.

Look out, the ole straw man is being set up only to be knocked down in the future (after release).

Most sims look promising before they're ultimately released, that is until their newness wears off and the nit-picking urges begin (we all succumb to it in one degree or another).

We're all also hoping and wishing well that Thunder Works can get as much "right" as possible with their yet to be released sim here, and then the process inevitably will repeat itself, and all eyes and hopes will begin to focused on the next unreleased promise.

You are right, of course, that no simulator will ever be perfect, and that there will always be improvements to point out. However, it is quite possible to create a simulator that has few enough major errors that I don't complain about them. Lock On: Modern Air Combat is a good example of this; there are only a tiny handful of significant flight model errors which I've found, so I don't heap abuse upon Lock On: Modern Air Combat. Aces High II is even better; I have almost nothing but praise for Aces High II's flight model.

But the IL-2 series has been so shoddy and had so many major errors, and for so long, that I cannot help but maintain a hostile attitude towards it and towards anyone who claims that it's a "realistic simulation" when it is, in fact, nothing of the sort. I once created a page-and-a-half list of major physics errors in that game (including the hilarious ability to fly in circles indefinitely, gear up, while dragging a wingtip on the ground). It's a seven-year-old engine, and people seem to think that this excuses the problems. Ha! If the engine's that old, and if it's impossible to do certain important things correctly with it as Maddox says, then write a new one.

However, I am fairly certain from talking to the very knowledgeable Mr. Cagle that Jet Thunder will be at least as realistic as Lock On: Modern Air Combat (and probably quite a bit moreso), and if that is the case I will have few complaints. I realize that perfection is impossible, but we do have standards that have been set by a tiny handful of simulators (and IL-2 is not one of them).

Well without name dropping, pro or con, I do agree with your overall hopes and desires for future sim releases Benny.

I do believe that most new combat flight sims begin as a clean sheet of paper, and attain what they will based on the abilities, focus, and time available to their respective developers. None, unless they're sequels, or projects that use an established flight engine platform really copy other sims as far as their faults or attainments, so generally each product's merits rise or fall independently.

You're right though Benny, the time to discuss wants and desires is always best pre-release, when significant changes if possible would be easier to implement (supposedly).

"But the IL-2 series has been so shoddy and had so many major errors" .... thats a pretty sweeping statement for a Combat Sim that has been going pretty strong for 6 or so years and you bring Aces High II into the argument !

Now I will just run off to Aces High II and enable "Auto combat trim" ! .... just like pushing in the Autocommand plunger on the Mirage III !

When maximum realism settings are enabled in both simulators, Aces High II has a far, far better flight model than IL-2 has. The only issues I have concerning Aces High II's flight model are one or two slightly questionable points relating to the stalls. But that pales in comparison to the host of major physics errors in IL-2, which does not even model the kind of accelerated stalls where the aircraft stalls straight forward. And don't tell me that World War Two era fighters couldn't do this, as many IL-2 players incorrectly claim in defense of the game, because quite a few World War Two fighters could.

Just because the IL-2 series has been "going strong for six years" (I don't call fifty players playing online per day "going strong, but that's irrelevant) doesn't mean that it's realistic. The Half-Life game series has been going a lot stronger than IL-2 and for a few years longer, you know. Does that make Half-Life a better simulator than IL-2?

Benny Moore, I don't know what you found wrong in the IL-2 FM, so it's impossible to argue with your claims, but I don't want to and will definitely not do it on this thread.

What I'd like to mention instead, especially for the developers of JT, is one of the main reasons (IMO) for IL-2's success for 7 years now.

Network code.

It's one of the greatest things about IL-2. I've never, NEVER, in 4-5 years of using it online, had a glitch in IL-2 multiplayer.

Edit: saw your reply to IvanK. 500 would be a more realistic number of players every evening on Hyper Lobby, many times more than 900. The importance of multiplayer is huge. [/Edit]

Now if we're talking about Lock-On, I think it's not a great example of a good combat sim. It could have been mind-blowing (remember how people anticipated it?)...HOWEVER... It's not brilliant in any department, but the part that kills it is stability - especially online. I've never, in several months of trying Lock-On: Flaming Cliffs, had a completely stable online session.

Thunder Works must, IMO, keep in mind that the community can handle a less-then-perfect product in some areas as long as it's a stable engine, online as well, which is supported by the developers.

This means: a good dedicated server version of the sim, stable online play (and offline of course), appropriate session modes (dogfight, coop, campaign etc.) and functionality.

I really hope we'll be seeing all this in JT, as it can make or break a sim (at least for me).

It's true. In spite of Lock On doing a much better job with flight model than IL-2, I played IL-2 online for years and only played Lock On online for a few weeks, because of the netcode. IL-2 had a very good netcode. Aces High II's is even better, being an M.M.O. with each server holding three hundred players.

Have to differ on this one - on-line stability is low on my list of essentials, I'm interested in historical realism, the 'what was it like out there' factor. We already have one Air-Quake........
I cant see a lot of paying customers wanting to spend a lot of time on-line replaying what was in truth a small conflict between Britain and Argentina.

For me IL-2 is 20% brilliant and 80% 'WTF were they thinking'. The only thing really wrong with LOMAC is they didn't do an AFM for a western jet.

So I'd be much happier if JT follows the LOMAC mould.

Cheers,
Keith