This flight adventure was my attempt, using Prepar3D V4, to emulate a real-world gliding speed challenge. My P3D flight was inspired by one of Bruno Vassel’s excellent YouTube videos (see links below). In that video, Bruno broke the Utah 100 km triangle speed record of 95.2 mph set back in 2003 by flying an average 100.23 mph (162 km/h) in a Schleicher ASW 27b glider on the 5th August 2017 at Nephi, Utah (u14).
To emulate Bruno Vassel’s real-world speed challenge in a somewhat convincing way, I undertook the following actions and configuration settings:
Task replication: I replicated Bruno’s flight task by extracting the waypoint data from the IGC logfile captured on Bruno’s flight, and then converted the data to an FSX flight plan using a Python script that I had developed.
This is the resultant flight plan in Little Navmap:
Choice of glider model: The choice of glider model was easy. I used the ASW27b model from Wolfgang Piper’s collection of freeware gliders for FSX and P3D. This matches up nicely with Bruno’s glider, although there are likely some differences due to limitations in the simulator flight modelling.
Payload and ballast settings: To maximise my cruise speed performance, I set the payload and ballast in P3D to give the maximum take-off payload of 500 kg. I also set the pilot weight and ballast to bring the centre of gravity slightly forward of default to improve high speed trimming as shown in the following snapshot:
Weather settings: To approximate the weather conditions of Bruno’s flight, I initially explored the possibility of using Active Sky 2016 (AS16) to inject historical weather directly into the sim. The problem with that approach was that the AS16 add-on chose stratus clouds when the historical meta data left the lower cloud layer type unspecified and the soaring environment add-on (CumulusX!) will not produce thermals when stratus clouds are present in the lowest cloud layer.
To get around that problem, I tried using a snapshot of the historical weather in manual mode and then changed the cloud type to cumulus but then I ran into other problems. Firstly the upper level wind layers had enough wind shear to make thermalling in the simulator almost impossible. Also AS16 seemed to generate too many clouds for the specified cloud coverage, which did not look anything like the real-world conditions seen in Bruno’s video. Moreover, the extra clouds exacerbated performance issues that I have been experiencing lately in my simulator.
In the end, I decided to use the historical data from AS16 as a guide for manually creating a weather scenario using the advanced weather settings in Prepar3D. I tried to copy the AS16 reported weather conditions as closely as possible but I did modify the upper wind layers a little to reduce the wind shear to more modest levels. I also reduced the reported visibility to better match the cloud haze visible in Bruno’s video because this visibility issue had not been reported in the AS16 historical weather data. Plus I elliminated the lower level cloud layer completely, relying instead on clouds generated by the CumulusX! add-on in ‘Unblue’ mode. I added a high-level cirrus cloud layer for purely cosmetic reasons.
Soaring environment configuration: To create a convincing soaring environment (thermals) consistent with the conditions evident in Bruno’s flight video, I configured CumulusX! as shown in the following snapshot. Note also, that I was using ‘blue’ thermals (i.e. no cumulus cloud layer in P3D) and I had the ’Unblue’ option in the CumulusX! Help menu ticked. This was done, partly because my P3D installation did not seem to represent a 2/8 or less cloud coverage accurately, and also because reducing the cloud coverage helped with some performance issues that I have been experiencing lately. Unfortunately, the CumulusX! cloud textures looked awful.
Link to CumulusX! competition soaring environment add-on:
While I fell well short of the Utah 100 km speed record (only 95.88 mph as against 100.23 mph) in my attempt, I nevertheless enjoyed this speed challenge immensely. I felt pressured all the way around the set course, trying to balance minimum arrival height against the need to achieve a task average speed greater than 100 mph. I am convinced that a speed challenge, such as this, can be equally entertaining as any race against other gliders in a multiplayer event.
After lengthy delays with some technical and personal health issues, I finally…finally managed to get this latest X-Plane 11 cross-country gliding adventure recorded and edited!
I even managed, after several failed attempts, to eventually upload all two hours of it to YouTube, although I had to split the video into two parts. Such is the nature of the 2nd rate internet services that we are lumbered with in Australia.
Apart from having some more fun with another virtual gliding adventure in X-Plane 11, my task was intended to be an experiment. My question I wanted to be answered by this task was this:
Can this simulator successfully and convincingly implement a cross-country gliding task of about two hours duration, which is based solely on ‘blue’ (cloudless) thermals?
To answer that, here are the links to the videos on my YouTube channel, and my conclusions and X-Plane 11 settings can be found below.
This was overall another enjoyable flight adventure using X-Plane 11, and that was in spite of some personal pain discomfort at times during the flight. X-Plane continues to surprise and please me with what it can provide to a virtual gliding enthusiast like myself, and it does this regardless of some pretty significant limitations in the modelling of a realistic soaring environment. For example, there are no cloud based thermals in X-Plane. Every glider soaring session in X-Plane 11 is effectively what glider pilots call a “blue” thermal day. Imagine what a great virtual gliding sim this would be if X-Plane had competition soaring environment functionality similar to that provided by the CumulusX! add-on for FSX. Blue thermal days are fine occasionally, and they do occur sometimes in the real-world, but thermals with cumulus clouds on top make for a more tactical cross-country challenge or glider race. Completing a cross-country task without a land-out is less of a lottery with cloud based thermals.
THERMAL LIFT MODELLING
One good thing I can say about the X-Plane 11 thermals is that do appear to be modelled reasonably realistically in terms of their lateral structure. The thermals exhibit a noticeably irregular lift profile when circling in them, making centring on the core quite challenging. I liked this and believe it is much better than the thermals in FSX and Prepar3D, which are too circular in lateral profile. With the weather settings that I used, the X-Plane thermals seemed lumpy and bumpy – just like the real-world thermals as I remember them. Also, the lift at the tops of the thermals tapers off reasonably convincingly in X-Plane, but I am still uncertain about whether there is any modelling of the overall vertical lift profile for X-Plane thermals.
There is still some more testing to be done. Some questions about thermal modelling remain in doubt or are still unanswered by just one test task. These include:
Do the thermals exhibit lift weakening close to ground level?
Do the thermals widen and lift become stronger in the upper levels?
Are the life cycles of thermals modelled in any way?
One tick in the box for X-Plane 11 that I can report is that an examination of my flight tracks when circling in thermals indicates that X-Plane seems to be making some attempt to model either thermal lean or drift of thermals in response to the wind.
EFFECT OF GROUND FEATURES & LAND CLASSES
On the other hand, I have not seen any clear evidence yet that X-Plane 11 thermals might be linked to ground features, nor have I seen confirmation yet that the thermal distribution might be linked to land classes such as large water bodies, forests, built up areas etc. The land class question might be resolved definitively when I fly future tasks in more diverse areas than was the case at Cunderdin.
One current disappointment is that the X-Plane 11 cloud shadows exhibit a very annoying and distracting flicker on my system, which I cannot seem to get rid of. To get around this problem for this task, I eliminated the lower cloud covering, leaving only the high cirrus layer. The cloud shadows from the cirrus layer still flicker but are less noticeable due to the lighter shading from these high level and thinner clouds. Since recording this flight, I have found a plugin that allows me to make all cloud shadows totally transparent, effectively hiding cloud shadows, so that this problem does not irritate me. Nevertheless, I would like to have visible cloud shadows that do not flicker – so I am still looking for a better solution to this problem.
THE FLIGHT PLAN
The flight plan is a 160 kilometre triangle based on a real-world task at Cunderdin in Western Australia. The original task was flown by Luke Dodd in an Astir CS sailplane in 2000. For my own convenience, I chose not to convert the task turn point coordinates, and instead, I modified the turn points slightly, placing them over the grain elevator structures at Kellerberrin and Wyalkatchem. Also I placed the start and finish lines over the glider clubhouse at Cunderdin airfield with an arbitrary maximum start height of 6,000 feet MSL and a minimum safe finish height of 500 feet AGL. My chosen start and finish line height limits are typical of the limits set for UK Virtual Gliding Association tasks.
The Plan-G flight planner shows that I should be able to complete the task in under two hours if I can achieve a task average speed of 83 kph or better. As the 2-part video above shows, I completed the task faster than that and was able to land and park the glider before the sun set.
Flight Plan (click to enlarge)
AIRFIELD LAUNCH POINT LOCATION FOR THIS TASK
For this task, I setup for an aerotow launch from Cunderdin, Western Australia on runway 32 at 15:50 local time on a sunny autumn (fall) afternoon. My intention was to release at 3,000 feet AGL, climb to the maximum start height of 6,000 feet before crossing the start line, hopefully by about 16:00. This would give me about two and quarter hours to complete the task before the sun sets. I felt that this would give some incentive to keep pushing harder, in effect a race against the sun, rather than having me taking it too cautiously and cruising around the course at the best L/D speed.
In a recent post I discussed my first cross-country glider adventure using the Schleicher ASK 21 glider model in X-Plane 11. That flight, which was published on my YouTube channel, was principally a ridge soaring task set at Zell am See in the Austrian Alps. This post details and links to another YouTube flight, which is a repeat of that task, but in the FSX Steam Edition simulator.
The YouTube video of this flight is best viewed in HD full screen mode if your Internet service supports it.
My intention in repeating the same task in these two flight simulators was to compare the virtual ridge soaring experiences that could be delivered by both products. To give some validity to this comparison between X-Plane 11 and FSX, I used: the same cross-country task course, the same time and date, near identical weather conditions, and identical glider model types, although the glider models were made by different developers. The only common glider model type available in both of my sims at the time of these comparison flights was the ASK 21 2-seater training glider. This meant using an add-on glider model in FSX. Leaving aside the motorized (self launching) versions, there were two add-on glider models of the ASK 21 available to me in FSX. For this flight I chose to use the Aerosoft ASK 21 rather than the ASK 21 available from Wolfgang Piper’s collection of freeware glider models because I am most familiar with that model from Aerosoft.
To be fair to X-Plane, I have no doubt that if I had compared both sims without any add-ons or plugins, then X-Plane would be streets ahead of FSX in any virtual gliding comparison. However, I am mostly interested in comparing the best possible virtual gliding experiences that these two flight simulators can provide me, including any available add-ons or plugins that I currently use to improve the overall experience.
Bear in mind also that I intend this to be a two part comparison. These first two flights provide a comparison based on a cross-country task in an alpine location where ridge or slope lift is the predominant energy source for the glider. In the second part of this comparison between X-Plane 11 and FSX, I intend to create a cross-country task in a flatland location where only thermal lift is available.
There is a discussion of the flight plan for this task in the YouTube video linked above.
Configuring the Soaring Environment in FSX
Obviously, time and date are the easiest components of the soaring environment to match up in the two simulators. Thus, in FSX, I set these to: 14 July 2016 at 14:00.
Matching FSX weather settings with those used in X-Plane turned out to be more problematic. Initially, I intended to use default FSX manual weather settings in conjunction with the CumulusX! add-on to re-create the soaring environment as near to that used in my X-Plane 11 flight as I could manage. However, I ran into a commonly occurring bug in FSX whereby multiple wind layer settings can be corrupted by the addition of turbulence and/or wind gusts in some layers. This led to failure of wind generation in FSX and thus no ridge lift in my first failed attempt. Moreover, turbulence and wind gusts in default FSX weather seem to be unrealistically abrupt and make the variometer very difficult to read because it does not have a gust filter. To work around these issues, I opted to use the Active Sky 2016 add-on in custom manual weather mode to re-create the weather conditions. Even with the less abrupt wind variation generated by AS2016, the wild excursions of the Total Energy variometer needle on the ASK 21 proved to be difficult to read at times. I seem to recall that the X-Plane ASK 21 had a similar issue with its variometer.
Following are some screenshots of my Active Sky 2016 weather settings.
NOTE: I started by selecting the default Realism Mode in AS2016. I then disabled any AS2016 wind updrafts or downdrafts so that CumulusX! could do its thing without unnecessary interference or augmentation to lift production. I also set wake turbulence to zero because that effect seems too violent and the motion unrealistic for a glider following a tow plane.
This was principally a ridge soaring task, however, I felt that it was both reasonable and realistic to provide some thermal lift, which I hoped might assist with ridge transitions.
It is difficult match thermal conditions between X-Plane and FSX because FSX has the benefit of the CumulusX! add-on with comprehensive user control and documentation, whereas X-Plane provides minimal support for random thermal generation with only a percentage coverage slider and a thermal height slider for user control. I chose to use the default settings for CumulusX!. The default settings in CumulusX! provided a smaller thermal coverage than my chosen setting in X-Plane, but I wanted to provide similar overall thermalling advantage to the task bearing in mind that CumulusX! provides cloud based thermals, which are easier to intercept, whereas X-Plane thermals are invisible “blue” thermals.
Following is a screenshot of my CumulusX! soaring environment settings.
NOTE: The Min and Max Lift Ceiling settings in CumulusX! do not apply for this task because CumulusX! gives preference to any cumulus cloud base setting in FSX. This was set by AS2016 to 10,153 feet.
Results and Conclusions
Ridge lift conditions were remarkably similar between these two flight simulators for this task. This was a surprise to me. I was expecting FSX with the CumulusX! add-on to provide a better ridge soaring experience.
The level of difficulty in completing the task was also much the same in both sims, although I achieved a better average task time in X-Plane. I believe that the faster task time in X-Plane was probably more the result of my mistakes and tactical choices than to any differences in the way these simulators render the task. Also, due to my inexperience with X-Plane, I may have set the thermal conditions in X-Plane a little more advantageous than was needed in my effort to offset any anticipated disadvantage due it being limited to blue thermals only.
Both simulators had convincing flight models for me, although fluidity and dynamic movements seemed noticeably better in X-Plane using X-Camera and HeadShake plugins. I was using Chaseplane and A2A Accufeel add-ons with FSX, but I could not achieve the same level of immersion using the settings that tried for this task.
Sound effects seemed slightly better to me in FSX. X-Plane lacked any sound effects for the spoiler operation. The spoilers should make noticeable clunks when opening and closing and they should create extra drag and rumble noise effects. Modelling of aircraft vibration with the spoilers deployed was absent from both sims. The tyre squeal on a grass landing in X-Plane was awful and unrealistic.
The aerotow launch was definitely more realistic in FSX compared to X-Plane. In X-Plane the tow plane veered off the runway centreline as soon as the ground roll commenced, which I assume was a lack of control over the wind vane effect on the tug aircraft due to the cross-wind component in this task. Also the X-Plane ground roll seemed far too short. Moreover, by default, X-Plane commences the launch automatically as soon as any glider flight loads. It would be much better if there was an option to have the launch manually started after instrument and control checks.
In regards to the alpine scenery around Zell am See, I believe that the X-Plane default scenery and mesh had the edge over FSX with Orbx global textures and Pilot’s Ultimate Mesh. This was mainly due to the apparent stretching of textures in FSX on steep mountain slopes. Also the scenery coloration in X-Plane seemed more convincing to me. River boundaries looked slightly less artificial in X-Plane, but roads seemed to be more unrealistically placed over the terrain mesh than in FSX, although both had problems. Each of the sims had a scenery anomaly. In X-Plane it appeared on the second leg of the task and it looked like a tile seam mismatch. In FSX it appeared on the last leg of the task and it looked like a data issue in the terrain mesh.
The lighting and shadows are noticeably better in X-Plane. There was, however, some annoying flutter in cloud shadows evident in X-Plane. On the other hand, FSX does not even have cloud shadows. CumulusX! has an option for cloud shadows for CumulusX! generated clouds but this option has issues, particularly in alpine areas, and I rarely, if ever, use it.
It would be unfair to compare airport scenery between the sims for this task because X-Plane was using default, but I inadvertently chose a location for this task that included an add-on airport scenery from the Orbx European freeware airport packs.
In my opinion the issues mentioned and any that I may have overlooked are not major. Overall, I am mightily impressed with both simulators and their ability to provide a convincing and entertaining virtual ridge soaring task with the Schleicher ASK 21 model glider. In the case of X-Plane, this is quite surprising to me given that X-Plane does not have a sophisticated soaring environment compared to FSX with the CumulusX! add-on. It will be interesting to see whether both simulators will compare as equally when tested with a flatlands thermalling task. So that is my next intended YouTube project.