On the topic of CH47D, Tandem Flight Models and SU15...
A lot of discussion has been happening in the MSFS Forums over the past few months since the initial release of CH47D regarding the topic of Tandem Flight Models.
It quickly became clear that the MSFS Flight Model wasn't well-suited for a tandem rotor setup. While it was possible to set up two rotors in "Tandem" mode and get the aircraft to lift off, the limited physics simulation made it difficult to get the helicopter to yaw or maneuver correctly. It soon became obvious why no other developer had tackled the challenge of creating this iconic helicopter.
However - if you know us well, you might know that we like to challenge the limitations of the sim. So, we took on the challenge, just like we did with the MV22 and Supercarrier Pro. The end result was a solid flight model that closely mirrored the handling of a CH47D (with SAS active). To fine-tune the flight model, we worked closely with CH47 pilots and the support team, who helped validate that it accurately simulated the stability, maneuverability, and capabilities of a Tandem Rotor configuration. However, the underlying mechanics weren't exactly "true to life." We ultimately decided on a workaround, using an overpowered main rotor and tail rotor setup to mimic the tandem rotor configuration as closely as possible. Sure, you are unable to taxi on two wheels, you may not be able to trim rotors individually and the VRS simulation isn't the best - we are well aware of these shortcomings - but other than that it flies as expected.
You might already know the story—just days before the release of the CH47D, we got word from Asobo that SU15 would include support for tandem-rotor helicopters. Then, SU15 ended up being delayed by several weeks. We were quite comfortable with our FM, and as such decided to release as-is. To our surprise, nearly all users were very happy with the FM (CH47 is, to this very day, the best-rated helicopter on the MS Marketplace!). That being said, we continued pushing for a FM switch to the new Tandem standard.
Sim Update 15
When the SU15 Beta finally arrived in early April, one of our team members devoted almost all their time to converting the flight model. We were flying blind, with no documentation or references on how it was supposed to work, which new variables were supported, or anything else. The first few days were pretty chaotic, but we eventually managed to get a first iteration off the ground. However, it wasn’t the well-balanced flight model we were hoping for—instead, we ended up with an unstable, erratic-flying aircraft.
Creating flight models isn't easy task. It involves an incredibly long and tedious process of trial and error and guesstimating values. MSFS often requires totally wrong values to output correct behaviors. The FM research and development continued actively during the month of April, and though we managed to achieve a handful of wins, the output wasn't what we ultimately expected.
By the end of April, the state of the flight model was as follows: the helicopter could take off and fly, but pitching up or down caused unwanted yaw moments. Additionally, the aircraft didn’t roll as intended or had a severe tendency to roll to one side. From a user experience standpoint, this was completely unacceptable.
Ps. We did not get documentation on the new FM variables until early May. By then, we had already figured them out by trial and error, and helped very little.
From this point onwards, progress was very slow.
Understanding Physics
I'm a Mechanical Engineer with a long-standing passion for aviation. I've been involved in flight simulation for over 10 years and have developed a solid understanding of physical models. However, I have to admit that the tandem helicopter configuration was entirely new to me. While working on implementing this tandem flight model, I realized I had some serious misconceptions about how tandem rotor flight works. It wasn't until mid-May that I truly grasped the physics behind tandem-rotor operation.
User from the MS Forums DinoPete11 described it very well:
Tandem rotor system is much more complex than just the fact that they counter rotate.
Nose pitch in CH47D is controlled by differential collective. Nose down is achieved by front rotor collective down and rear rotor collective up. Collective up increases torque, torque up increases yaw drift. One rotor produces more yaw than the other when you push or pull cyclic.
Cyclic pitch also increases or decreases disc loading. Lateral cyclic will change disc loading equally when you’re flying perfectly into relative wind. With crosswind it will once again not be so nice and equal. In that case once again, different disc loading = different torque = different yaw.
It’s not Asobo mishap as far I have seen so far. It’s just physics.
Chinook flies in such a stable and consistent manner because its balance of weights and forces was very carefully engineered for it, and only in conjunction with complex digital AFCS / stability assist. This goes down to the point that full capacity of fuel tanks is asymmetric between left and right side. The reality is NOT tandem rotors = no yaw. Chinook does not fly straight hands off without a computer making it happen, it never did.
There lies the key - Asobo did do a great job modeling the physics of a tandem FM. However, those physics are not really representative of how the CH47D Handles in real life. The reason is that there is a Stability Augmentation System that, under normal circumstances, is constantly correcting and balancing the helicopter. No such system exists on MSFS - the tandem FM was not engineered for the CH47.
Any sort of additional stability augmentation would be upon us to develop from scratch and implement.
The Problem
Here's the problem - how to successfully complete a transition to a new, highly experimental flight model, on an aircraft which such high customer satisfaction, without affecting the user experience AND providing a better experience than before?
We certainly don't want to push to production any solution that would result in a worst user experience in any way. No iteration of the flight model to this date has been anywhere close to "Acceptable" by itself, and the idea of pairing it with SAS (Stability augmentation system) introduced several problems:
- Any SAS System that we experimented with would be based off the AFCS Autopilot System on the CH47D, which essentially works by sending commands to cyclic and collective. This is fine on AFCS as changes are rarely violent and sudden, however, SAS requires constant input which would result on a very annoying " visual stuttering" of the Cyclic and Collective on the virtual cockpit
- SAS Requires deadzones on the controls, which is automatically a barrier to entry to the product. Our AFCS does require deadzones as well, but we can get away with it as an "optional feature that you may use on demand" - migrating FMs would force all users to comply with the requirements to accomodate for the SAS module.
- We have reasons to believe that FMs on Xbox present slightly different performance characteristics, likely due to a slower sampling and calculation rate. These systems are based on very finely tuned PID controllers. These controllers would likely require fine-tuning for Xbox, which is practically impossible to do due to the fact that Xbox is as closed-off platform and uploading flies for testing takes several days.
- The difficulty of developing a SAS system from scratch is very high and beyond the scope for a product of this category.
- Zero helicopters with SAS have made it to Xbox. Not necessarily a blocker, but not a good sign.
- Branching off PC and Xbox versions is NOT an option - we have made a commitment to offer the exact same product and characteristics on both PC and Xbox, not to mention the difficulties of maintaining different versions of the same product. It is just unfeasible to do.
It isn't looking great.
The Solution
The only viable solution at this point is to keep the current flight model as it is and educate the public on the reasoning behind this decision. No iteration of the tandem flight model by itself is good enough, or accurate enough to make it into production at this stage. Any change will ultimately affect the user experience.
Not long after we came to this realization, we noticed a modder on the MSFS Forums offering an alternative: a free mod that not only replaced the current flight model with a proper Tandem Rotor system but also added the necessary SAS module to make it flyable. While it inevitably suffers from the issues we mentioned earlier (stuttering, dead zones, and Xbox incompatibility), it does provide PC users with the option to install, activate, and uninstall the module as they see wish.
Such an experimental and limited module cannot be included in the production version. Freeware modules like this tend to iterate and mature quickly, but our complex infrastructure of distributors would slow down the process of releasing new versions of the mod. Additionally, flight models impact too many areas of the product to simply offer them as an "on-off option" on a tablet, meaning we would essentially need to distribute two versions of the product—bringing us back to the challenge of ongoing maintenance. Once again, quoting the developer of this mod (DinoPete11):
A freeware modification means we can freely experiment until it’s really, really good, without being responsible for people’s purchase money. There may be some regressions along the way but we’ll get there.
Hence, our best shot was supporting the freeware modder and provide any help necessary to push the mod out.
The Mod
DinoPete11's mod is available for Download here (PC ONLY): https://drive.google.com/drive/folders/17UlWASO-fWBU86cp_GMDarar-9y1Ztfg
This is a work in progress project. The features may not be compatible with all platforms, and may not be to the liking of all users, so it is not an official update - it’s a modification developed in coordination with Miltech developers, but not by Miltech developers themselves.
The goal is to try to overcome some of the limitations that prevented the addon from transitioning to native tandem rotor flight model.
Feedback is very welcome. See docx file in documentation for keybinds for custom AFCS stability assist system. It is necessary to approximate chinook flying experience with native CFD tandem rotor instabilities.
Copy the folder and its subdirectories to community folder, at the same level as your Miltech CH-47 addon. Do not overwrite anything in the core CH-47 addon, this is a separate addon package that modifies the behavior of the helicopter. Put it in community and your CH47 will behave differently.
KNOWN ISSUES:
- Blade slap sound is tied to certain elements of miltech rotor configuration. Switching to tandem rotor breaks them, bladeslap sound will not be audible. Fix is being worked on
- Available power might still be insufficient for MTOW, power tuning is still WIP
The Future
Our efforts for an official, production transition to a tandem flight model have stopped for all the reasons listed above. We will see how/if FS2024 changes the situation, or how the new sim may affect the current flight model. It's too early to comment on this.
Regardless, two things are for sure:
- We will continue collaborating with the modding community to enhance the current solution for PC
- CH47D will be available on FS2024, for no upgrade fee, within the first few days after the simulator has released.
Linked from: