Update Updating TI Pegasus Ares

[vchamp]

I'm currently updating Total Immersion vessels to make them compatible with Orbiter 2016 and D3D9Client. Pegasus Ares is the first to get the update.

I'm going to make this vessel as advanced as I can, instead of copying again the behavior of the original Pegasus Ares addon from francisdrake.

Due to time restrictions I don't know how long will it take to finish this project, so I decided to start this thread to post updates, questions and get a feedback.

If you want to try this update, first you have to install the original addon from francisdrake and the VC extension for Orbiter 2010. Links:
[ame="https://www.orbithangar.com/searchid.php?ID=2717"]Pegasus Ares 01[/ame]
[ame="https://www.orbithangar.com/searchid.php?ID=3816"]Total Immersion - Pegasus Ares[/ame]
Finally install the build from https://gitlab.com/vchamp/ti/blob/master/TI_PegasusAres/build/build.zip.

What's new compared to the old addon:
- Works in Orbiter 2016
- Works with D3D9Client
- Removed UMMu because sadly it is not updated for 2016
- New touchdown points specs
- Added exhaust stream for atmospheric flight
- Altitude indicator follows terrain, i.e. ground relative
- Disengage thrust auto hold when 'Full thrust' or 'Kill thrust' is pressed
- More complex aerodynamics, cross sections change according to the status of heat shield and parachute, variable drag element for parachute, auto stabilization in the direction of airflow
- Empty mass is updated according to attachment status of parachute, heat shield and gear
- Screens readability improvements
- Damage model: parachute failure from high dynamic pressure, disintegration from high temperatures on atmospheric entry, gear damage from high vertical speed on touchdown. Functional heat shield. Temperatures display and damage check button.

Planned next changes:
- Instruments for choosing landing target and displaying relative position to it
- Indicator of terrain slope below vessel, maybe some instrument to find the best place to land (most horizontal and flat, in crater or outside), if Orbiter SDK allows it
- Update all necessary parameters depending on the heat shield and parachute status
- Autopilots

 

[4throck]

It's better, but chute aerodynamics need work.
It's on your planned list, and it will make a difference.
Also, some external cameras would be great, specially downward facing!

But great job, keep working on it, I appreciate it!

 

[vchamp]

Trying to figure out how to make the capsule orient its z axis along the airflow during reentry.

The first question is: Do real capsules have stable orientation so that their heat shield is perpendicular to the speed vector just by having that characteristic shape? Or this orientation must be kept forcibly with thrusters? Would real PAres rotate back to "z-axis along airflow" position if rotated with RCS given enough dynamic pressure? I assume it would because the air will press against the side walls.

So this orientation shouldn't be possible in a free fall:

The second question is how to simulate it. It seems that I need two things for this: create an airfoil with vertical lift and add a drag element, like I did with parachute. I'd like to understand how airfoils are related to drag elements and why adding only an airfoil or a drag element is not enough for my purpose.

 

[Urwumpe]

The first question is: Do real capsules have stable orientation so that their heat shield is perpendicular to the speed vector just by having that characteristic shape? Or this orientation must be kept forcibly with thrusters? Would real PAres rotate back to "z-axis along airflow" position if rotated with RCS given enough dynamic pressure? I assume it would because the air will press against the side walls.

Both. First of all, the coarse orientation is not kept by shape, but rather by moving the CoG away from the Center of Pressure, where the aerodynamic forces act: For a rotation symmetric capsule, this means the CoG is further outside and not on the centerline of the capsule. That is also why the lack of moon rocks in Apollo 13 was a small issue, the CoG needed to be trimmed, because it was near the limit.



Next, there are ALWAYS small oscillations, so Yaw and Pitch thrusters are always firing a bit to keep the orientation within limits. The CoG "trim" just ensures that less fuel is needed.



If you just want to do a ballistic reentry, remember also: The aerodynamic forces on a capsule are not turning you around into proper orientation automatically and instantly. If you fly nose first, like Soyuz 5 did, it would be stable in that orientation. It started to swing around once it oscillated far enough away from the other stable orientation that the correct orientation was reached. But the shape is also ensuring that the correct orientation has a bigger zone of stability.



The Space Shuttle for example would actually prefer to fly engines first during reentry, if the FCS would not prevent this by keeping the AOA right in a tiny zone where only small control inputs can keep the nose up at one end and prevent cartwheeling at the other end.

 

[vchamp]

If I understand correctly capsules tend to orient themselves to the needed orientation, however not because of their shapes but because of their CoG position. It seems that this is not simulated in Orbiter and the only way to achieve this is by changing aerodynamic properties. Please correct me if I am wrong.

I think the right way is to change cm parameter in the AirfoilCoeffFunc depending on aoa and M.

Drag element could also be used for self-orientation in airflow but it would add a constant drag which is not desired.

 

[Urwumpe]

If I understand correctly capsules tend to orient themselves to the needed orientation, however not because of their shapes but because of their CoG position. It seems that this is not simulated in Orbiter and the only way to achieve this is by changing aerodynamic properties. Please correct me if I am wrong.

I think the right way is to change cm parameter in the AirfoilCoeffFunc depending on aoa and M.

It is. And you write already how. cm describes essentially the effect of the airfoil / capsule itself, if it would have no mass.

And then, you could also simply place the airfoil reference to a different location. That way, you could start with a symmetric capsule lift function (f(aoa) = -f(-aoa)) and reuse it for different capsule or different mass distributions (For example, think of the Curiosity rover, which ejected metal weights during reentry for different trim angles)

 

[vchamp]

Update:
- Made it self-orient in the direction of airflow by using a variable drag element. Now Ares tends to orient with heat shield first during atmospheric entry. But it is also less stable during high-speed ascent in atmosphere.
- Heat shield produces lift, it is possible now to change vertical speed and slightly direction of atmospheric entry by specific orientation relative to airflow.
- Empty mass is updated depending on whether parachute, heat shield or gear is attached or not.
- Horizontal speed text is not going out of the screen.

I hope the behavior of the vessel in atmosphere and during entry is not totally unrealistic At least it is playable, I successfully tested it in Earth and Mars entry. I will stick to this model for now, unless someone points to any inaccuracies in it.

Next is probably some damage model, first of all for entry loads and heat.

 

[vchamp]

It took me a while to finish the damage model, and finally here's the update.

What's new:
- Parachute may fail if dynamic pressure is above 12 kPa (depends on the current vessel mass). I don't know if the airspeed should also be considered, it's now possible to open parachute at the beginning of atmospheric entry on orbital speed.
- Hull and heat shield temperatures are calculated depending on the dynamic pressure and airspeed. New display that shows temperatures (in Celsius) and button to check the damage status.
- Vessel's hull and heat shield strength will degrade if any of the temperatures exceeds max acceptable value (527 Celsius for the hull and 1827 Celsius for the heat shield) to the point of complete vessel disintegration.
- Gear damage if vertical speed on touchdown is higher than 5 m/s.
- Slightly increased lift from the heat shield for more control.
- New scenarios with Earth and Mars aerocapture.

Now there are changes also to other files other than dll, so updates will be released as a zip now. The link is in the first post.

Next is navigation, landing target selection, terrain scanner. Will try to add external camera views. Also going to upgrade other TI vessels.

 

[4throck]

Gave it a quick try (not much time these days) and the damage model seems to work
Great work!

 

[francisdrake]

Hi, nice to see the old Pegasus Ares is still around and you are taking care of it!

As far as I understand, you are making a completely new coded vessel. In this case, feel free to include my meshes in your distribution package, so users only have to download a single package.

Regarding an unstable behaviour during ascent, because the vessel is aerodynamically stable in a heatshield-first attitude:
You may consider having the vessels coded in 2 'modes'.
- A descent mode, with all the aerodynamic surfaces active.
- An ascent mode, where all aerodynamic surfaces are deleted.

These different modes make it easier for vessel which undergo significant structural changes. I had no chance to test-fly the new Ares yet, but will do in April.

 

[vchamp]

As far as I understand, you are making a completely new coded vessel. In this case, feel free to include my meshes in your distribution package, so users only have to download a single package.

Thank you for the kind permission and for making this lovely spaceship, really appreciate it.

Regarding an unstable behaviour during ascent, because the vessel is aerodynamically stable in a heatshield-first attitude:
You may consider having the vessels coded in 2 'modes'.
- A descent mode, with all the aerodynamic surfaces active.
- An ascent mode, where all aerodynamic surfaces are deleted.

So far in my testing the instability wasn't critical. If some problems occur during further testing then I'll probably follow your advice.