Thursday, December 27, 2012

How to revolutionize aviation industry

I have been ranting about that personal aviation is starving and struggling at best to stay alive. But there is no such reason it has to stay that way. If nothing is done, ever increasing regulation and bureaucracy and the increasing expenses will make personal aviation extinct sooner rather than later. And in my opinion, sportplanes/LSA are not the right answer. They are substitute for something else that would be really needed instead - airplanes are most useful when they are used to get somewhere - with fun and with style. Let's face it, flying traffic pattern is such hardcore fun that only dedicated pilots like me and maybe the readers of this blog will enjoy about. And to get somewhere, there is the blocker, which is weather. And I think there is a solution for that:

What is missing in general currently in current general aviation market is simply:
- affordable mass produced aircraft that are so simple to fly that much greater portion of population can do that, and with less training than currently is required for pilot's license.

Lets address this simple to fly thing: mechanically controlled aircraft can be easy to fly to an extent. For example Diamond DA40 is very easy airplane to fly. However, it still requires a lot of training to fly one. To make airplanes more accessible to people, controls should be extremely simple to learn, should not be any harder to get it than learn to ride a motorcycle, or ideally, it should be easier. So what would be the cure then?

In my opinion the answer is in fly-by-wire. No control rods, cables and the like, but instead a (doubled for redundancy) computer that controls all control surfaces and specially purpose built brushless DC servo motors which are directly connected to control surfaces. User uses a fully electronic stick that only gives input to the computer which then decides what control surfaces to move and how much to respond to the input the way user wanted. Pushrod assemblies are very complicated, have many parts, of which many are expensive. To make the plane less expensive, all expensive parts should be designed out of it. Servo motors can be made in large quantities quite inexpensive and the control system will not be super expensive either as it would use off-the-shelf hardware. Doubling that for redundancy would not be very expensive as the hardware in question is similar than what controls all quadro copters and toys like that.

Second enabler will be avionics which displays only important information to the user. E.g. user does not need to know oil pressure, oil temperature, EGT, CHT etc. values if they are all well and optimal, ECU is controlling the engine without bothering you about mixture controls or other obsolete things, and trend analysis done by the computer concludes that trends are not alarming and user does not need to worry about engine failure. Only when something would fail, or something suspicious would be found from analysed trends, the computer would inform the user about the problem. If things are ok and green, why to bother user with 100 gauge displays on a cluttered control panel. User would be left with much less workload and he or she could concentrate on looking for other traffic and flying the plane rather than trying to watch all the engine monitoring etc. values.

The system could communicate automatically with ATC and would display the pilot a very easy path to follow to make a IFR approach without needing any of the workload that is the burden for IFR-pilots. The plane would do things automatically, a bit like a copilot that would be helping you on the right seat. You would not need to worry about what is ILS alpha approach as all the necessary information would be presented you automatically and the display would show you in very clear way which way to turn your plane. You would still be able to control the plane to not lose the fun of flying, but the hard and messy parts would be handled by automation. And for pilot going to weather he or she is not able to fly through, there could be a panic button that would direct the plane to the next airport and land automatically if the pilot so wishes.

And one more enabler would be nice to have. It takes awful lot of time to pre-flight a modern high performance aircraft such as Cirrus SR22. You open the checklist and start checking this and that and that and that and there are zillion items to check before you can even start the engine. What if there would be a computer with sophisticated sensors wired everywhere in the plane and it could do the pre-flight automatically without user needing to care about it unless he or she wants. Sometimes, in case of a short trip, due to the long time spent to pre-flighting the plane, one could reach the destination by car faster by driving rather than getting to airport, pre-flighting plane, filing flight plan etc. There surely should be a faster way to do it with some technology.

It would not be the thing how fast the plane would be. That would be secondary after the plane would be fast enough. How economic the plane would be has greater importance. But how easy it is has greatest importance. Not everybody is skilled enough to ever become IFR-pilots and let's face it, single pilot IFR is a quite dangerous thing to do in hard IMC to the minimums alone without second pilot as helper. The second pilot would need to be replaced with very intelligent computer system that would take off the workload from the pilot so the pilot could concentrate on flying a perfect approach.

Since the plane would be fly-by-wire, it could be made more stable and the ride could be augmented to be less bumpy by automation. So that the plane would be always rock solid, in perfect trim etc. in all flight conditions that would make flying the IFR-approaches very easy and safe. This would enable really this pre-requisite of getting somewhere. Who would want to drive 1000 km with a car slowly averaging 80 km/h at best if there was a way to do it with airplane with no more expense, as easily as with the car, and much more conveniently and certainly with a lot more style. That would make airplanes desirable again to general public, that they currently are not. Only hardcore pilots (like myself) have a desire for the current selection of airplanes.

None of these things I presented here are technically impossible to do and neither they would or should cost much. Everything presented above is doable with even a less sophisticated hardware than a typical smartphone that costs no more than 400 euros. The authorities with ancient rules, regulations and bureaucracy is a thing that makes everything expensive, and I could imagine that some certification process could kill this idea for starters even before it began and this kind of installed system would cost due to certification costs, millions per one single unit. But lets assume that this would not happen and things would go the right way for a change.

To get somewhere, regulation should be largely reinvented in order to not kill personal aviation. Well it is already killed by certification requirements and regulations, it is more like a zombie now, but I would be really happy the GA to be reanimated and to be a real business again, and this time reinvented to be better than ever before. This is a market that would exist if it was not overregulated and that would employ a lot of people and would revolutionize how people travel from point A to B - instead of sitting in Boeing or Airbus and licking their knees in tourist class, they could enjoy similar freedom as driving a family car and getting to the destination with their own schedule without queues, waiting, walking kilometers in huge terminals of huge airports, and without associated anxiety etc. Instead of a large jumbo-jet now and then, there would be demand for hundreds of thousands to millions of small inexpensive and easy to fly aircraft that would be designed for serious transportation and not to be toys like LSAs. Current airplane market is more like a joke. To make it a real market, it would need to be completely reinvented and revolutionized, otherwise it is a dead end that leads to nowhere else than the gradually reduced amount of pilots and the remaining pilots facing more and more ridiculous expenses which will further reduce their interest and numbers which is no good for the individuals nor businesses.

Thursday, October 25, 2012

(Wing) mold making for rapid prototyping

I am progressing forwards from concentrating on aerodynamics to also fabrication and optimizing the fabrication process. Been doing hand layups for a quite long time now, but I need to start doing shape accurate parts. Testing aerodynamics requires very high accuracy.

I have been doing several molds lately. One was almost successful, but it was a lot of work and it still had non-sharp edges.

So what I am trying to do is a wing mold with CNC fabrication. I have a CNC mill that can carve blue styrofoam, wood, MDF, Corecell etc. and is large enough for producing a half (lower or upper) of a wing section (with the limitation of the length, have to glue a big mold from ~1.2 meter pieces together).

I would need a rapid prototyping technique to produce shape accurate molds with glossy or at least almost glossy surface. So that manual work would be minimal. This is especially pronounced in case of making wing molds.

The problem I am facing is this:
- if I carve the plug to blue styrofoam, and then paint it and polish it, there is the downside:
- the styrofoam can be only painted with a paint that does not have solvent in it, and the only paint that will fill the surface is solvent free epoxy primer. The problem with that is that it also makes sharp edges round, and especially in a small scale (RC scale) the roundness becomes way too big to be acceptable. The edges where the lower and upper half meet, should be also absolutely accurate and sharp. Doesn't happen with this technique.

Has any reader used molding epoxy? I saw some picture of a mold being filled with a molding epoxy and then milled with CNC again to shiny surface directly (?). Would that be viable option for my use? As it would be for rapid prototyping and for fabricating many wings (and not just one pair), it should be somewhat reasonably cost effective. Making the mold from wood is not completely inexpensive either - requires a very thick perfect wood block (or MDF block). I could not afford consider replacing the styrofoam with a huge solid mold plastic block (that is used in industry for prototyping shapes with CNC), because the same volume is much more expensive, would be possibly fine for a CNC model of a small device, but for making a mold for large wing the cost hikes out of the roof very quickly. Styrofoam is cheap and very easy for the machine to carve, but that's the best part of it, otherwise it is really poor material.

Any first hand experiences on this?

Tuesday, October 23, 2012

Propeller placement article (external link)

There are many considerations where to put a propellers in a small aircraft or RPV. The common place to put them is at the nose. The biggest reason and driver for this placement is that it is advantageous for CG location. However, from aerodynamic standpoint that is not very optimal. There was discussion at HBA Forums about propeller placement and this document was linked (it studies difference of prop placed in pusher and tractor configuration):

Optimal place is behind the wing, a bit above the wing centerline (only small part of the prop circle goes below the wing). This placement has the typical CG challenges with it. And it will require either pylon on the wing, or a pylon on the fuselage (assuming a single fuselage). There is then the question about the effect of the body to the prop located near the fuselage behind the trailing edge of the wing. There might be unfavorable flow due to the effects of the wing-fuselage joint that this study did not take in account.

According to the article, it was possible to increase quite significantly the Clmax of the wing with the rear placement of the propeller due to the suction effect to the wing. This leads to interesting thought about a line-thruster - multiple small electric motors turning multiple relatively small props behind the trailing edge of the wing, providing suction to the whole wing surface, or at least large part of it. Interesting question then would be that would a varying thrust angle be beneficial, should the pylons be actually mounted on the flaps? Downside of this is that this may lead to flap mechanism that is not very lightweight as the flaps have to take all the torque and push from the motors. Normal flap mechanisms would not like that.

Any comments on this?

Saturday, August 18, 2012

High Altitude UAV Launch for Microsatellite Launcher Rocket

Rockets lifting off from the ground are inefficient and require enormous amount of fuel and especially oxidizer despite there is plenty of oxygen in the low atmosphere. Rocket engine is very inefficient too, and to save fuel, the best way is to limit the time the rocket engine is needed => more reasonable amount of fuel needs to be carried and subsequently smaller rocket is needed.

To launch microsatellites, a similar approach than Burt Rutan's WhiteKnightTwo (or Stratolauncher) is the best way to go. The difference for a microsatellite rocket is that the payload is much more modest and a much smaller and much less expensive airplane can be used. The airplane can be either UAV or optionally piloted aircraft. If it is just pure UAV, it can be made smaller and less expensive.

The principle of this airplane would be series hybrid propulsion. There would be a small but powerful modified series turbo liquid cooled motorcycle engine turning a generator (which is also a brushless DC electric motor, but bigger than the propulsion units). The actual propulsion would be provided with several smaller electric motors which would turn relatively modest size propellers (vs. the need for prop size in the high altitude). The large number of the propulsion units would compensate for the diameter. Very large diameter in the prop would cause severe geometric problems for the aircraft (landing gear, wing placement etc. as evidenced by Grob's high altitude aircraft), but instead lowering the disc loading of the prop by increasing the number of the props, is much less expensive and much less complicated way to do it, especially since the construction with electric motors approaches trivial.

The aerodynamics of the aircraft would be based on a high aspect ratio long wing and conventional tail placed aft of a ideal laminar pod-shaped fuselage (for UAV pod-shape would be even easier because there is no need for seating people inside and no bumps / deviations of the ideal shape are needed). The airfoil designed for the purpose would be obviously a low Re airfoil with maximized L/D at moderately high Cl (unlike the cruise-oriented airfoils presented earlier on this site).

The airplane would lift the rocket with the hybrid electric propulsion to 50000-80000 ft like e.g. the Stratolauncher does, but with the difference that the aircraft would be very inexpensive. Even 80000 feet should be possible with hybrid propulsion since when the dual turbo can not provide enough power anymore (it would cut out pretty much after 60000 according to some studies done for HALE-UAVs), it can continue climb purely by using batteries. There is a limit though how much batteries can be carried along with the plane to not lose too much of the payload (the rocket). This is a compromise that needs to be optimized carefully to maximize benefits.

The airplane would be obviously fully reusable and good for a long long time. Different kind of payloads could be accommodated, depending on the need. E.g. rocket carrying multiple microsatellites or a rocket carrying one payload or a equipment that would study higher atmosphere (without rocket) or possibly a pressurized pod housing a human (as alternative to the rocket). It would be highly interesting concept to design & implement. It requires some budget, but can be done with quite modest budget not commonly considered in the aircraft/spacecraft circles.

Marching order would be this:
First study items for this would be:
1. Determine how large rocket would be needed for reaching orbit from case 1) 50000 ft and case 2) 80000 ft. Someone familiar with rocket equations could contribute if you would be interested. Please note comments are unfortunately moderated due to spam-comment problem, and may not appear immediately (only when I have time to go to press approve).
2. Then the size of the aircraft that will lift the rocket can be determined by requirements set by 1).
3. Determine the propulsion system needed for lifting the 2) to case 1) 50000 ft and case 2) 80000 ft.

Conservative assumption would be that the aircraft would reach 50000 ft and horizontal speed would be negligible at launch altitude, and the size of the rocket would need to be determined based on that altitude (and initial speed). Then if the aircraft could be made to reach 80000 ft, it would be all plus, bigger payload could be accommodated to the rocket.

The configuration for the aircraft could be quite simple:
- 2 small pods with two tails (or interconnected tail) like Burt Rutan's high altitude lifters leaving the center section for the payload
- one pod could house the piston engine + fuel
- another pod could house the batteries
- four wheel landing gear would be located to both pods, similarly than White Knight

Monday, June 4, 2012

Pipistrel Panthera videos

I found some interesting videos from Youtube describing the new Pipistrel Panthera four seat aircraft that has incredible performance and economy compared to competition in the similar class:

John McGinnis' Synergy Aircraft got funded in Kickstarter

Roughly 4 hours ago, John McGinnis succeeded to raise the goal for funding his Synergy aircraft on Kickstarter. The funding exceeded the goal by couple of tens of thousands. I think this is a remarkable achievement and it proves that also aircraft projects can be crowd-sourced. Congratulations to John and good luck with his project. Now it is his turn to finish the plane and get it on the air. This is now on critical path since as the plane is funded, the reputation of crowd-funding aircraft projects in the future will have this as a case example, so he better succeed. Looking forward to seeing Synergy to fly to Oshkosh! Please make it happen!

Thursday, May 24, 2012

John McGinnis has Synergy Aircraft as Kickstarter project

I met John McGinnis in Oshkosh and he is a humble homebuilder guy with a dream and passion bigger and stronger than most of the aircraft fellas and he has unorthodox design that will be one of the coolest looking homebuilt aircraft when finished & successfully in the air. It is a kind of like flying wing but elevons placed in a box tail in the wing tips -  a configuration which is supposed to reduce drag. This is combined with very aerodynamic pod shape fuselage and laminar flow airfoil wings. And on top of that the plane will run on Deltahawk diesel engine and will work in places where AVGAS is no longer available and can just fill the tanks with JET-A, even in places like Kuujjuaq or Pangnirtung if he chooses to go for that kind of adventure. You can help this project getting this bird to the air by backing John's Kickstarter project:

Saturday, March 10, 2012

Idea: Helicopter with the blades extendeding from the outer edge of a lens shaped lifting body

I was today walking with a dog and it occurred to me that with a fly by wire system a helicopter which would have blades outside of a circular lifting body and that would have a pusher prop (cruise prop) on top of the said flat lifting body in a pivoting pylon with thrust vectoring, could possibly fly stable and avoid the not that good things of a helicopter and be more efficient VTOL craft than a conventional helicopter. The bearing on the outer ring would possibly though be some sort of a challenge. On a big one, the outer ring could be not connected to the inner ring at all, but it could actually use maglev technique to levitate in a hole without physical contact with the bearing. This could come with too heavy weight penalties, so not sure if it would be feasible. So this is just a out of the box thinking idea because I have always felt that there is something wrong with helicopters, I have never really liked how they look like and how they function, there must be a better way to do that.

Cyclic system could be hidden inside the center section so that it would not cause drag like on a helicopter where there are lots of mechanical parts directly in the airflow causing drag. In cruise the plane would be a gyrocopter rather than a helicopter because it would have positive angle of attack unlike helicopter which cruises at negative angle of attack (and only blades have positive angle of attack).

I would like to try this as RC-model, but the challenge would be to fabricate such a large and yet lightweight bearing that would be size of the craft itself. The outer shell with the bearing would have a gear tooth that would be driven with a pinion gear that would be connected to the engine. This would also function as gear reduction drive. In addition to the ring drive to takeoff and landing, there would need to be a generator for driving the pylon mounted thrust vectoring prop that would be run with an electric motor.

The shape of the lifting body would be a suboptimal airfoil because it would need to be symmetrical and have both sharp leading edge and trailing edge but. This kind of craft would be naturally unstable but it could be artificially stabilized with software.

Thursday, March 1, 2012

Cross sections in Rhino from a loft

I wrote a long time ago about how to do cross sections in Rhino. I mentioned that it needs to be Object intersection. If you do Boolean Intersection, this technique is not going to work. I found that the earlier instruction might have been a bit confusing because there are multiple ways to do intersections in Rhino.

The original article is here: Please read it.

The function what you have to use is Intersect (to do Object Intersection).
If you type it to the command area, type Intersect, not Intersection or Booleanintersection.
You know that you have chosen wrong intersection method if you have to select two set of objects. Multiselect everything (your cutters and the lofted surface) at the same time and do Intersect. It works.