random_people
Member
Long time, hardcore lurker here. Finally have the confidence and time put into a design to share and try and get some feedback. This is a long one, please bear with me, want to share where this project came from and where we are attempting to go with it.
Quick intro: Currently Mechanical engineer in my senior year studying in NYC. Always had an interest in Aerospace: started in rubber band models 13 years ago, then played with model airplanes and rockets, combined them, VTOLs, UAVS, and now finally my first full scale aircraft with some quadrotors on the side.
Currently working on wing structure design for TerraSoar, a grant funded art project for a roadable electric aircraft that will become part of additional art pieces. The present configuration was chosen as a good middle ground to serve as the base for a velomobile - canard pusher aircraft that will be built in Brooklyn, New York. It was initially designed as an electric powered ultralight, but we found out on our visit to Air Venture this year that the empty weight limit currently includes the batteries. summer of redesign later, TerraSoar is now an aircraft for the Light Sport Experimental Classification with a weight of approximately 750 lbs empty including batteries. With a 30 kW motor for air propulsion and 2x 2kW hub motors in the rear wheels for ground propulsion, TerraSoar will be partially road-able but with limited performance. Keep in mind this aircraft/car is part art, sometimes at a cost of its performance. The following are renders of the aircraft in a previous iteration in both flight and a possible land configuration. The wings are a bit bulky and might need to be completely removed for land travel.
I have some renders of the current design in flight configuration with wingtips posted below.
THEN: Performance estimates for TerraSoar were generated with a modified RONCZ worksheet in EXCEL while XFLR5 was used to more accurately model the effect of the canard-wing interference on values of wing lift coefficient CL. Given the Eppler 1230 airfoil and the Roncz R1145MS airfoil, suitable incidences and positions for the wing and canard were found that provided the desired lateral stability . However there were limitations. However Fuselage panel analysis in XFLR5 is unsatisfactory for modeling a full aircraft because it lacks viscous.
In ordered to gain a more accurate and complete verification of the design we are in he process of analyzing the complete aircraft in CFD. Now, I do understand the more or less universal policy against CFD, or Colorful Fancy Diagrams, mainly that it should be avoided unless used very, very carefully. This is something that my partner and I are being very wary of and diligent about. We have spent the last month comparing experimental results for finite wings in wind tunnels with those of identically modeled XFLR5 and CFD values. Varying turbulence models, mesh size, and boundary layer treatments were explored in CFD. we have finally found a combination that produces consistent results in the same range of Re and AoA we expect to see on the full scale aircraft.
NOW: The part of this airplane that is officially being analyzed for this project is the wing loads and necessary structure. Based on performance value estimates like stall speed and maximum speed given our design, we have established our flight envelope that we will be using to define our load cases for the wing and canard structures design. Given a 1.5 Ultimate F.S. and additional requirements made in ASTM-F2245, design loads were selected as N = +4.5 and -2.5 at and above maneuvering speed. We understand the we need to define loading cases at the outer limits of the aircrafts flight envelope, however we had some questions regarding the treatment of AoA in these analyses.
QUESTIONS:
1. In establishing the static load case at Vs, or stall speed do we simply need to: model the flow around the aircraft at the necessary AoA and lowest speed, then take those aerodynamic (pressure) forces found and multiply them by N to get the design load, then multiply them again by the ultimate factor of safety to then finally get the load to design the structure to support? What AoA do we use for negative loading?
2. Along the same lines as question 1., when analysizing for the static Vne case, do you simply set flow velocity at Vne and the AoA at 0, take those aerodynamic forces found on the flight surfaces, multiply them by N and again by the Ultimate F.S. for the final load? Would we need to also test at additional angles of attack?
3. In regards to the additional required load cases, we will be considering all of those at the boundaries of the flight envelope, as well as wing torsion due to control deflection, gust loadings at Vc and Vd, and asymmetrical loadings as per ASTM-F2245. Are there any other ones we need to run for a light sport aircraft that flies relatively slow? Flutter maybe?
4. When considering gusts at Va and Vd, we couldn't exactly find out how to handle basic analysis with the advanced and time consuming dynamic analysis aside. It seemed as though a gust could be simply treated at an incremental raising and lowering of angle of attack relative to the wing. Is this satisfactory for estimating the gusts on an aircraft such as this which will cruise at about 60 knots?
This site has been one of the most immense helps when it came to gaining any sort of footing on aircraft design, and I look forward to doing my part here when I can. Please let me know what you think about the questions as well as the design in general.
Thanks guys!
Quick intro: Currently Mechanical engineer in my senior year studying in NYC. Always had an interest in Aerospace: started in rubber band models 13 years ago, then played with model airplanes and rockets, combined them, VTOLs, UAVS, and now finally my first full scale aircraft with some quadrotors on the side.
Currently working on wing structure design for TerraSoar, a grant funded art project for a roadable electric aircraft that will become part of additional art pieces. The present configuration was chosen as a good middle ground to serve as the base for a velomobile - canard pusher aircraft that will be built in Brooklyn, New York. It was initially designed as an electric powered ultralight, but we found out on our visit to Air Venture this year that the empty weight limit currently includes the batteries. summer of redesign later, TerraSoar is now an aircraft for the Light Sport Experimental Classification with a weight of approximately 750 lbs empty including batteries. With a 30 kW motor for air propulsion and 2x 2kW hub motors in the rear wheels for ground propulsion, TerraSoar will be partially road-able but with limited performance. Keep in mind this aircraft/car is part art, sometimes at a cost of its performance. The following are renders of the aircraft in a previous iteration in both flight and a possible land configuration. The wings are a bit bulky and might need to be completely removed for land travel.
I have some renders of the current design in flight configuration with wingtips posted below.
THEN: Performance estimates for TerraSoar were generated with a modified RONCZ worksheet in EXCEL while XFLR5 was used to more accurately model the effect of the canard-wing interference on values of wing lift coefficient CL. Given the Eppler 1230 airfoil and the Roncz R1145MS airfoil, suitable incidences and positions for the wing and canard were found that provided the desired lateral stability . However there were limitations. However Fuselage panel analysis in XFLR5 is unsatisfactory for modeling a full aircraft because it lacks viscous.
In ordered to gain a more accurate and complete verification of the design we are in he process of analyzing the complete aircraft in CFD. Now, I do understand the more or less universal policy against CFD, or Colorful Fancy Diagrams, mainly that it should be avoided unless used very, very carefully. This is something that my partner and I are being very wary of and diligent about. We have spent the last month comparing experimental results for finite wings in wind tunnels with those of identically modeled XFLR5 and CFD values. Varying turbulence models, mesh size, and boundary layer treatments were explored in CFD. we have finally found a combination that produces consistent results in the same range of Re and AoA we expect to see on the full scale aircraft.
NOW: The part of this airplane that is officially being analyzed for this project is the wing loads and necessary structure. Based on performance value estimates like stall speed and maximum speed given our design, we have established our flight envelope that we will be using to define our load cases for the wing and canard structures design. Given a 1.5 Ultimate F.S. and additional requirements made in ASTM-F2245, design loads were selected as N = +4.5 and -2.5 at and above maneuvering speed. We understand the we need to define loading cases at the outer limits of the aircrafts flight envelope, however we had some questions regarding the treatment of AoA in these analyses.
QUESTIONS:
1. In establishing the static load case at Vs, or stall speed do we simply need to: model the flow around the aircraft at the necessary AoA and lowest speed, then take those aerodynamic (pressure) forces found and multiply them by N to get the design load, then multiply them again by the ultimate factor of safety to then finally get the load to design the structure to support? What AoA do we use for negative loading?
2. Along the same lines as question 1., when analysizing for the static Vne case, do you simply set flow velocity at Vne and the AoA at 0, take those aerodynamic forces found on the flight surfaces, multiply them by N and again by the Ultimate F.S. for the final load? Would we need to also test at additional angles of attack?
3. In regards to the additional required load cases, we will be considering all of those at the boundaries of the flight envelope, as well as wing torsion due to control deflection, gust loadings at Vc and Vd, and asymmetrical loadings as per ASTM-F2245. Are there any other ones we need to run for a light sport aircraft that flies relatively slow? Flutter maybe?
4. When considering gusts at Va and Vd, we couldn't exactly find out how to handle basic analysis with the advanced and time consuming dynamic analysis aside. It seemed as though a gust could be simply treated at an incremental raising and lowering of angle of attack relative to the wing. Is this satisfactory for estimating the gusts on an aircraft such as this which will cruise at about 60 knots?
This site has been one of the most immense helps when it came to gaining any sort of footing on aircraft design, and I look forward to doing my part here when I can. Please let me know what you think about the questions as well as the design in general.
Thanks guys!
