VSP is incredibly effective at putting together advanced aircraft concepts that can ask fundamental “What if?” questions – and help to quickly get answers whether it’s worth continuing a new line of thought. A recent conceptual design study that we conducted at NASA Langley asked the question “Is it possible to reduce the operating costs of an advanced General Aviation (GA) aircraft by a factor of 10?” Asking such a question requires establishing a good baseline model (such as the Cirrus SR-20/22), and then putting together a good conceptual geometric representation of the proposed solution to compare to that State-of-the-Art (SOA) reference. So I thought you’d like see one of the VSP model concepts we came up with to meet this question, and a few of the key results. So why ask this question?
The average price of 100LL right now is $6.10 per gallon, and a typical GA aircraft gets 12 mpg; this means fuel cost is about half the total operating cost and this problem is only going to get worse. 100LL is going to have to go away, it is the major source of lead pollution. It’s not a reasonable fuel to be using going forward, and specialty replacement fuels that are manufactured just for the GA market are likely to be even more expensive once the EPA and FAA finally get rid of 100LL. The Single Engine Piston market is going to die a slow death unless we can provide new technology solutions that are environmentally friendly and economically sustainable.
Using electric propulsion is an incredible new degree of freedom for aircraft designers to radically alter how an aircraft performs, especially in terms of efficiency and operating costs. But right now electric propulsion is fundamentally limited by electric energy storage such as batteries, and for this reason the skeptics refuse to consider it a viable solution. At around 200 Watt hours/kg, they are correct that the specific energy is about 60 times better for AvGas at 34,700 Whrs/gallon. But you’ve got to look beyond this simple comparison. Even a modern GA aircraft engine like the IO-550, the thermal efficiency is 28% and the specific power is about .5 hp/lb; versus the electric motors that we have available today at 93% with 3 hp/lb (including controller, with the motors we’re developing now are 96% efficient at 5 hp/lb). So that factor of 60 is immediately decreased by a factor of 3.3 based on the efficiency (energy required) alone. If you do a detailed accounting of the SR-22 propulsion system, you can save 498 lbs by replacing the reciprocating engine with an electric motor; which can go straight into added weight for batteries. Next consider the real range required for a typical GA flight, where a 200 to 300 mile range would be sufficient. Last, flying around with an Lift to Drag ratio of 12 just isn’t going to cut it anymore (yes, that’s what the SR-22 gets at its cruise speed of ~200 mph, but a much more respectable L/D of 18 at 150 mph. Achieving higher L/D’s is clearly possible as we go to motor glider concepts, and electric propulsion dramatically cuts down on installation losses such as cooling drag (6 to 9% on the SR-22), scrubbing drag (about 2%), and propeller blockage which drops installed propeller efficiency (which can be alleviated by compact electric, lightweight motors that can be positioned just about anywhere on the airframe). Put this together with advanced batteries, like Envia, that have already demonstrated 400 Whr/kg in the lab, and very reasonable electric aircraft solutions can be achieved.
Baseline vs Advanced Concept: We did our comparisons to the SR-20/22 because it is a good representation of the SOA in performance and operating costs, we started our advanced concept development based on a concept that offered a much higher efficiency starting point. I selected the SWIFT (Sweep Wing with Inboard Flap Trimming) concept developed by Steve Morris and Ilan Kroo (Stanford) as my concept starting point. It is an extremely efficient (L/D > 26) and lightweight (420 lb) motor glider solution that Aeriane fabricates in France. But it’s only a single person, low speed motor glider that required substantial changes to turn into a 4 place. We wanted to make an electric SWIFT concept into as efficient of a GA aircraft as possible, and maintain that high L/D ratio (but with a substantially larger fuselage) at a much higher speed (150 mph), so we incorporated several advanced technologies we have been developing – such as a Boundary Layer Ingestion Inlet for fuselage drag reduction, and wingtip vortex propeller interaction for induced drag reduction.
Attachments: I’ve attached a few slides that show the approach, and assumptions that went into this simplified analysis. I’ve also placed the resulting concept geometry into the VSP Hanger. VSP permitted us to do a good internal layout to insure a fair cabin volume comparison – this was important since the bulky reciprocating engine was removed and permitted a fuselage wetted area reduction. We chose back to back seating because flying wing concepts have a very limited pitch trim capability, so we needed to keep the CG excursion as small as possible. But as you’ll see if you look at the slides, we needed even more pitch trim capability so we adapted the SWIFT to a C-wing configuration (which also helped get us to a higher CLmax, which also is fundamentally limited with flying wings). Only tail volume coefficients were used for tail sizing, and no control analysis was performed – and with flying wings, this is an important simplification that needs to be further considered. But the results indicated that as long as ranges of about 200 miles are acceptable (including a 30 min reserve), that weights comparable to other 4 place aircraft can be achieved. Our analysis including directly exporting VSP wetted areas (from the tsv drag buildup file format) into an Excel spreadsheet that lets us quickly and easily understand the parasite drag. Coupled to a vortex lattice analysis (such as Vorlax or AVL), we were able to have a good understanding of the induced drag as well. We then used empirical methods to account for the change in fuselage drag from the BLI propulsor, and the induced drag from the wingtip wake vortex propeller. Vortex lattice codes do a great job of capturing the non-planar induced drag benefit of the C-wing layout, so there wasn’t much uncertainty from this unconventional configuration. We didn’t do a structural analysis, but those are the next steps we’ll be taking as we go into further depth towards understanding this concept, and others.
So what was the resulting conclusion? You can look at the attached slides to see more details (or read the attached AIAA paper that explains why we’re looking at such concepts). But relating to the original question, the total energy required for the SR-20 (the slower, lower power version of the SR-22) is about 382 kW hrs. The total energy required for the electric Swift is 65 kW hrs. Using the average electric utility rate of $.115 /kW hr and AvGas cost of $6.10 /gal, the resulting difference in energy cost to fly 200 miles is $67.40 for the SR-20, and $6.14 for the eSWIFT – an 11 times difference in energy cost. Certainly a more complete life cycle analysis is required to amortize the batteries over their typical 2000 cycle life and take into account the difference in maintenance costs of essentially care-free electric motors versus reciprocating engines. But even at current electric motor and battery costs, the propulsion system is less expensive than using existing aircraft engine solutions. Even ridiculously expensive batteries appear to be a good deal compared to a Teledyne Continental IO-550N costing $60,000 for 310 hp. So it does appear possible to achieve an order of magnitude reduction in operating costs, and this isn’t even taking into account electricity rates as low as $.06 /kWhr for commercial users at off peak hours if batteries were to be charged at night. Personally, I believe there are exciting opportunities for dramatic improvements with electric aircraft over the next 10 years – and it’s great to have tools like VSP to help understand these differences. Now if someone would just give me a vortex lattice method that’s coupled to a propeller lifting line or actuator disk method I’d be in even better shape to analyze such concepts as the eSWIFT.