Hyperion Project Recap

Fall 2012 – Spring 2013 — A collaboration between Arvada Associated Modelers and the University of Colorado, Aerospace Engineering Sciences Department

Joe Pirozzoli, a long-standing member of AAM, reports:

Hyperion is a research project consisting of multiple, on-going phases of development and experimentation involving many disciplines of aeronautical and aerospace engineering. The face of Hyperion is the vehicle itself which is a unique design and “proof of concept” blended body/wing conceived by the graduate students at CU, and funded in large part, by the Boeing Corporation. The Blended Wing Body Design of the Hyperion was based off of the Boeing X-48 Unmanned Aerial Vehicle. The Hyperion vehicle was constructed almost entirely from composite materials including extensive use of carbon fiber. The current design of the Hyperion aircraft was conceived in the second year of the Hyperion Project at CU, which began in Fall 2011-2012.

The current design combined the first year’s center-body, which was made partly at the University of Stuttgart, Germany and partly by CU Graduate Students (which has been previously documented), with a new set of “swept” wings. The new swept wings, unlike the wings of the first year, complied with the defined specifications of Boeing’s Blended Wing Body Design. In addition to the new design, the aircraft was now also equipped with a newly designed and custom machined set of main landing gear struts including a braking system that was incorporated for the rear, main wheels.

Hyperion was a true experiment because of all the uncertainties and questions about its design. Questions included, was it too heavy, would it be underpowered, exactly where was the center of gravity (CG), and how much control surface deflection would be required. It seemed that there were more unknowns than knowns. However, working through these unknowns with a group of extremely bright, enthusiastic, and creative individuals turned out to be half the fun and challenge. There were many frustrations along the way, but in the end, it was pure exhilaration.

Hyperion would also be a surrogate of sorts for me to collaborate with intelligent, young, prospective scientists who are very much akin to my oldest son Tony. Tony is currently enrolled at the University of Nebraska and on his way to getting a Masters degree in theoretical physics. Admittedly, I am often baffled by the work he is doing, but I never tire of trying to understand what new things science is uncovering or researching. “Hanging out” with the graduate students at CU seemed to bridge the distance from my son.

Hyperion also consisted of many other subsystems and components that were designed and being tested. Everything “Hyperion” was somewhat experimental, but also meticulously designed, tested, monitored and evaluated. I attended 2 presentations by the students to their professors and it was clear that every aspect of Hyperion was under intense scrutiny. There was very little about Hyperion that wasn’t being “graded” except for the RC systems.

Myself and Stewart Garret got involved in Hyperion back in late October 2012. Dr. Jean Koster, a professor at the University of Colorado, was looking for experienced pilots to fly their SIG Rascal 110 with “Hybrid-Electric” engines. Many others from AAM would also become involved in the project as it progressed and I am truly proud and grateful to everyone (who are too many to name) that supported and helped. AAM truly put its best foot forward and demonstrated once again, its commitment to community outreach.

AAM provided the “RC” experience and knowledge base to bring Hyperion to life. AAM’s expansive runway was also a great asset to the project. It was really the only suitable site (in the Denver area) for testing this vehicle short of a full scale runway. Due to the high cost and time involved with constructing the vehicle itself, the Hyperion team had decided not to experiment with new or untested RC systems and components. They instead used well known and fully proven RC equipment. Our job was to make sure that the RC systems and components, including the competence to fly the vehicle, were the least of their concerns. Despite all this, flying Hyperion would prove to be quite challenging and require all of my experience and skill as an RC pilot.

The success of Hyperion would lie less in the accomplishment of achieving flight, but rather more in what these students would learn about the processes of project development and management. Hyperion would also provide the valuable “hands-on” experience that fresh-faced, graduate engineers are often criticized for not having. I could have only have imagined such an opportunity as an undergraduate engineering student. Hyperion was well beyond anything one could learn in a classroom.

Rascal fights in the latter months of 2012 consisted of testing two, not so unique, engines: an OS 120 4-stroke, and a Plettenberg Terminator electric out-runner (11 HP), powered by 12S lipo battery system. Eventually, the students would combine these prime movers to create an integrated IC/electric hybrid system which has yet to be fully tested and flown. The work on the Rascal would eventually lead to test flights of both a ½ scale Hyperion prototype and then the actual, full scale Hyperion.

Flight testing of the ½ scale prototypes and early taxi testing of the full scale Hyperion would prove to be extremely challenging. One of our greatest challenges was scheduling testing when the runway was clear of snow. Much of the testing was attempted from late-February through mid-April when it seemed as though a major snow storm was expected almost weekly. I would like to thank my good friend Lanny Hansen for always being ready to assist with plowing and clearing the runway using the Club’s trusty tractor.

Fight readiness and testing of Hyperion would begin in earnest in early March 2013. It began with the full integration of all the RC flight systems with the other components and subsystems including telemetry and monitoring. The students also created a flight simulator specifically for Hyperion that turned out to be quite accurate and useful. Systems integration, ground checks and flying on the simulator would lead to a series of high speed taxi tests and eventually, hopefully, a series of successful flights.

I was invited to assist with a series of flight readiness reviews and check-outs some of which took place on the CU campus. One purpose of these reviews was to assure that the RC systems were functioning and set up correctly. This included aspects of both performance and safety. These students were not well versed with transmitter programing, servo/control surface setup, ESC programming, etc. Despite their lack of experience they did an excellent job of construction and preliminary setup.

The flight readiness reviews included a couple of presentations by the students to their professor’s. These presentations shed quite a bit of light on all of the other aspects and goals of Hyperion. It was during these presentations that I learned that Hyperion’s success did not depend absolutely on a successful flight. Further, their mentors were not rushing toward that conclusion. I was also surprised at how much they valued and relied on my experience simply as an RC pilot. I felt very welcomed and appreciated during those meetings which exemplified the integrity of the project and the Hyperion program.

For anyone that was present when the Hyperion team arrived at the field, it may have appeared that things were somewhat disorganized and chaotic. Their activities were actually well practiced and choreographed. Unfortunately, there always seemed to be something that needed repair or added attention prior to each taxi/flight attempt, a situation all of us can relate to at some time when “maidening” a new aircraft. Still, the rigidness of their processes resulted in consistent and well-prepared flight readiness which gave us all the confidence that Hyperion was ready for safe testing.

There are many aspects of RC flight preparation that can make or break a successful flight, but paramount are: ample thrust and the ability to achieve proper air speed, location of the CG, and control surface deflections. All of these were in question when it came to Hyperion despite the fact that the students had calculated, simulated or experimented to find the proper values and parameters. At the end of the day, all of that is subject to the accuracy of construction. And not just for the aircraft and airfoil, but also for gaps in the control surfaces, drag of external components (such as the gear), etc.

No matter what, a light aircraft with plenty of power is the best insurance for successful flight. Unfortunately, Hyperion was neither light nor was it blessed with an overabundance of power. Hyperion weighted a robust 48 pounds. The estimated body/wing loading was a hefty 42.2 ounces (2.64#) per square foot. Not a problem if you have an equally robust power plant. Let’s face it, a lot of power can cure a lot of ills. Unfortunately, achieving maximum output and efficiency from the Terminator motor was inhibited by the use of a 20” diameter, 4-blade propeller which was necessary because of the low ground clearance from the vehicle’s tricycle landing gear. Experimentation with numerous propeller diameters and pitches provided a maximum power output of 5000 watts from the Terminator (i.e. just over 104 watts per pound). Hyperion would only have 6.7 HP for its 48 pound airframe. In comparison, my 27 pound CompArf Extra 300 is equipped with a 10 HP DA-100. Yet, the students were confident that 5000 watts was adequate, and in the end, they were quite right.

The simulator provided a starting point for control surface deflections. The students had successfully adapted their Futaba 10-channel transmitter to an accurate model simulator of Hyperion. The program allowed us to experiment using the radio’s travel endpoints to adjust control surface deflection on the aircraft, and incorporating functions such as exponential. We adjusted the radio and surfaces such that Hyperion would fly smoothly and authoritatively. The values that were obtained from the simulator became our “low rate” starting point. Actual setup would also utilize a “high rate” for backup. The higher rates would prove to be necessary for stable flight because the simulator could not predict the loss of efficiency caused by gaps between the trailing edge of the body/wing and the control surfaces. A lesson most of learn early on in RC building and flying.

Numerous high speed taxi tests were performed to optimize both the stability and effectiveness of the tricycle landing, but also to experiment with the location of the CG. Simulations and wind-tunnel testing suggested that there may be serious issues with pitch stability if the CG was too far aft. Additionally, the location of the CG was very finite. In other words, the simulations also suggested that the CG location (for stable flight) was within a fairly narrow range.

The close proximity of the CG relative to the rear main gear also made taxiing in a crosswind quite difficult. This was exacerbated by the lack of sufficient rudder surface to induce yaw in order to assist with steering as the nose wheel was not, by itself, able to overcome a minimal crosswind. It became apparent after a short amount of high speed taxiing (and one mishap) that testing could only be continued under conditions with the wind nearly down the runway (or non-existent).

Early taxi testing without successful rotation or obviously lifting of the nose wheel also suggested that Hyperion was nose heavy. However, there was limited opportunity to shift the CG aft without adding additional weight and causing further instability on the ground. Conversely, the concern of potential pitch instability was always in the back of everyone’s mind, so the CG was moved back in small increments (i.e. 1 centimeter or less at a time).

Finally, on Saturday, April 6th, and after numerous high speed taxi events Hyperion achieved a very short flight. Hyperion successfully rotated and flew, but only briefly and without adequate pitch control. There was enough elevator to achieve rotation, but it took full deflection and when the elevator was slightly decreased it caused Hyperion to pitch immediately back downward to the runway. The resulting damage ceased the day’s testing, but also provided a great deal of knowledge and data to review and evaluate. Mainly, we were all now assured that Hyperion had enough power and lift to achieve flight. Questions about CG, elevator deflection, flap reflex, etc. were still very much up in the air (no pun intended).

There was much discussion about moving the CG further aft, but eventually that it was not implemented. Instead we focused on gaining adequate elevator authority. The elevator was modified in 3 ways: 1) positive trim up was added; 2) deflection was increased by 25%; and 3) exponential was changed from negative to positive. We concluded that the elevator needed to move quickly away from neutral and be less sensitive when the stick was at or near full deflection. This is counter to the way most flyers set up their planes, but seemed perfectly logical for Hyperion. Another modification was to increase flap reflex. Hyperion’s design was such that negative flap deflection (i.e. upward) also helped improve the elevator’s effectiveness.

After a lot of discussion and evaluation about the results on April 6th and settling on modifications, the team had to wait another nearly 3 weeks to try again. School schedules, business travel, parts availability and inclement weather postponed Hyperion for another flight attempt until Friday, April 26th. Even at that late date in April, another snow storm threatened the test flight, but snow was followed by very warm weather and the runway would be clear and dry on the day of the test. The million dollar variable would be the wind, which as everyone knows, had been very unpredictable and blustery throughout the early Spring.

Finally, the day would arrive and everything was ready except for the wind. The wind was not strong, but it was a crosswind. We did not want to risk another “taxi” mishap that would unnecessarily delay things further. However, the reality was that the school semester was coming to a close and many of the Hyperionites would be graduating and/or moving on to careers, internships or other destinies. The students were anxious to attempt a flight and had an “all or nothing” attitude. Even if it meant severe damage or destruction they were ready to risk it all in favor of a total commitment to flight.

Up until now, I had been very cautious with Hyperion. In a sense, caution had led to tentativeness and even reluctance to push for flight. Under normal circumstances, we take a new airplane to the field confident in the knowledge that a successful flight will be achieved. All of the questions and lack of success with Hyperion had created doubt and even a little fear. Success this day would depend on being fully committed to flight and to just go for it all. And fortunately the students felt the exact same way. They were willing to risk everything for one moment of success (regardless of how brief).

The Hyperion was readied for flight, minor issues were resolved and modifications were implemented. The vehicle was ready, but the wind was proving to be a pest. A fairly steady NE cross wind prevailed for most of the morning. Everyone was anxious, but I wanted the least of my concerns to be the wind. Finally, the wind began to wane and change direction. As it switched from NE to SE it also diminished. Additionally, there were brief periods when the velocity was very low. We positioned ourselves at the West end of the runway and waited for a lull in the wind. When the moment came, the goal would be to achieve a maximum throttle taxi as quickly as possible.

Stewart was at my side and prepared to make dual rate switches and trim adjustments as needed. We waited for approximately 10 minutes, which seemed like a lifetime, but finally a window of opportunity presented itself. I advanced the throttle and was able to taxi straight down the center line and achieved full throttle as I passed the first set of pilot stations. I waited a few moments longer then progressively pulled back on the right stick. As Hyperion passed the last pilot station it rotated and began to fly. It climbed out fairly aggressive, but not too steeply and accelerated once it was in the air. All the doubts about thrust and weight were erased.

I continued to climb at full throttle then began a fairly aggressive left turn. The first thing I noticed was a distinct lack of rudder authority. Thus, I was forced to bank very steeply in order to complete the turn without ending up over Leyden road. On the first downwind (westerly) leg I began experimenting with part throttle and control surface effectiveness. Reducing the throttle would be necessary for landing and to conserve the batteries which were not expected to last more than 4 minutes. The ailerons were trimmed, but sluggish, so I switched to high rate (+33%).

Well past the field I began another less aggressive left turn to the south. At reduced throttle and with a lack of rudder authority this took me well south of the parking lot and eventually back toward the east. It would fly at ½ to ¾ throttle, but I began to lose pitch control. It was on the upwind leg that pitch control completely evaporated, Hyperion was extremely nose-heavy. Hyperion was heading downward at approximately 10 degrees with the elevator at full deflection. We acted quickly. I advanced the throttle and switched the elevator to high rate. At the same time, Stewart increased the flap reflex. Pitch control returned before making a other wide-sweeping left turn that would put Hyperion back north of the runway and again downwind. By now it was readily apparent that landing was not going be an easy affair.

I also knew that I may not have time for several landing attempts so I committed to attempt a landing on the next upwind pass. I flew well downwind and made another sweeping left turn to a final approach. As I throttled back progressively, I also let Hyperion pitch forward at a fairly steep attitude and glide path. Just off the end of the runway and at approximately 50-100 feet high I again lost pitch control. Additionally, I could not keep Hyperion from drifting to the north side of the runway. Without pitch, I didn’t dare bank and the rudders would not hold it on course. Within 150 feet of the runway I knew I had 2 choices: throttle up and try to regain pitch control, or simply let Hyperion come to earth. I knew Hyperion was on a heading that would put it on the north apron so I elected to let it crash land there.

Before I took off, I had already considered the very real possibility (based on everything we had learned up to that point) that I may not have full pitch control, particularly at slower speeds. Under that scenario I would have to fly at a very high rate of speed and essentially fly it into the runway in order to land. An extremely risky and dangerous proposition with a 48 pound beast that had very little rudder authority. No, I had already contemplated the results of such a botched maneuver including it “cart-wheeling” down the runway or skipping off the concrete and landing in the pits or careening full speed into the fencing. The choice had already been made and I had the presence of mind in that last 150 feet to simply let Hyperion come to rest well west of the pit area and on the reasonably soft north apron.

Surprisingly Hyperion’s damage was fairly minimal. Hyperion can certainly be repaired and flown again although I am not aware of its future plans. The project will certainly continue and I sincerely hope that AAM continues to be an integral part of it. I know one thing for sure, despite all the commitment, work, frustration, stress, and uncertainty… I had a blast and would do it again in a heartbeat!!!

Posted in CU AES, Hyperion.