Mostly correct, but 'constant attitude' isn't a requirement. The angular accelerations, p-dot, q-dot, r-dot need to be 0, but the angular velocities, p, q, r don't need to be 0. So you can trim with a specific roll rate, e.g. p=5deg/s.

See for more details, 1 == tFull:

And

Thank you for the comment. I misunderstand the trim.
I understand that trim itself means stay with initial conditions, and trim mode means which acceleration component should be zero. Is this correct?

The comment should be # 1 means zero recti-linear and angular accelerations, usually for straight-and-level flight?
(while level have to be set by initial conditions.)

I am confused.
Someday, I would like to create examples of changeable property lists, trim mode lists, optimal usage situations, and so on.

I think there is often a common misconception that trim implies that the angular velocities need to all be 0. Just like the translational velocities don't need to be 0, if they did have to be 0 then the only trim solutions would be stationary trim solutions on the ground, there is no requirement for the angular velocities to be 0, they can be 0, but they don't have to be, the only constraint for both the translational velocities and angular velocities is that they be constant, i.e. that their accelerations are 0.

Hence my comment that your original comment claiming that trim requires a constant attitude isn't accurate.

Take a look at the following paper for some more details on trim.

A General Solution to the Aircraft Trim Problem

Note that there are multiple ways to solve the trim problem, the one described in the paper I linked to is not currently available in JSBSim. The current trim algorithm in JSBSim is based on a different algorithm.

Thank you very much. I read this paper, and found it very easy to understand. I realized that trim is actually very profound.
I understand that trim refers to determining a constant control values that brings the aircraft to a steady state where linear/angular accelerations are zero and linear/angular velocities are constant. And this condition is satisfied not only when flying straight, but also when performing steady turns or pull-up.
Also, I understand that level flight was achieved if the user set fdm['ic/h-sl-ft'].

The comment should be # 1 means straight flight by using all changeable control variables. right?

You can thank @agodemar regarding the paper since he's one of the authors and also a JSBSim developer 😉

Also, I understand that level flight was achieved if the user set fdm['ic/h-sl-ft'].

No, what will indicate that you want level flight is setting gamma $\gamma$, flight path angle to 0.

fdm['ic/gamma-deg'] = gamma

You need to set fdm[ic/h-sl-ft'] to some value even if you're requesting a trim that involves climbing or descending flight. So setting it doesn't request level (non changing altitude) trim. Gamma controls that, pretty sure it defaults to 0 if you don't explicitly set it.

1 - tFull - Longitudinal and lateral trim.

@yuki-2000 yes, trimming a dynamical system is a subtle subject, especially when the system is a flying vehicle in atmosphere.
One example of subtlelty is this:
When you trim for a flight along a rectilinear trajectory with a nonzero flight path angle ($\gamma\ne0$), you want the true airspeed to be maintained constant with a constant regulation of aerodynamic control surfaces. This my be approximately true for a limited time interval. If you let the state propagate, you'll realize that the altitude is changing (because of nonzero $\gamma$), hence air density $\rho$ is changing, hence external forces and moments are changing.

So, rigourously, to keep a steady airspeed magnitude (and steady angle of attack $\alpha$ and a steady sideslip $\beta$) along a climbing or descending trajectory, one should continuously change flight control deflections and throttle setting. This is why sometimes you ear the term "instantaneous trim".
A similar reasoning is valid for a pull-up condition.

You run into this issue of the trim only being valid in the near term even for level trim cases, i.e. $\gamma = 0$. For example given fuel burn both the mass and the moments of inertia will be changing after the trim has been computed.

Now 'near term' is larger for this case than the climbing/descending case.

JSBSim offers a property to prevent the fuel burn affecting the mass and moments of inertia.

propulsion/fuel_freeze

I understand that trim the plane just once doesn't mean it will stay the state forever. Thank you for your lecture.
So, does real commercial air plane trim itself at regular intervals repeatedly ?

In effect yes, either the pilot if they are hand-flying or the autopilot when it's engaged, which for commercial airliners is for a large percentage of the overall flight.

So for example the autopilot would be configured to maintain some altitude, e.g. 33,000ft at some speed, e.g. Mach 0.82 and it will then continuously modify the control surfaces and the engine throttles to maintain that.

Out of interest, what is your background and what are you using JSBSim for?

Out of interest, what is your background and what are you using JSBSim for?

Well, I studied aerospace engineering in the university, so I could somehow keep up with your discussions. However, my research field in the university was structure. I wanted to learn the flight dynamics and control for a step up or make flight simulator using UE5 as a part of my hobby (I do not know whether I can do it by the way)
I just find JSBSim just a few month ago from UE5`s antoinette project website, and tried to build in my environment and find it very hard. Therefore, I learn JSBSim little by little.
I might dissapear if I lose interest in the simulator or my job gets busy, but I try to continue studying.

Detail that speed can be Mach number if it's < 1 as you have, but also mention that > 1 specific KCAS (Knots Calibrated Air Speed).

Mention it's to track/record the minimum thrust and the angle at which the minimum occurs.

Start sentence with capital letter, i.e. # Update the....

Mention for the min and max speeds that it's CAS

See earlier comment about comment about 1 - tFull trim comment.