VTOL fixed-wing UAV

Ali F

Member
Good evening guys,

I hope you are doing well.
I m working on a project, its a fixed-wing uav, and it will be very useful to have some advice from you regarding the design process.
The idea is to build a VTOL fixed-wing uav for mapping and environmental monitoring with:
- High wing with dihedral.
- 2m wingspan.
- Solar cells for energy.
- conventional tail.
- Autopilot.
Im asking for some advice regarding the desing process and also the material to use, and all other features.
Thank you.
 

quorneng

Master member
Ali F
A project of enormous complexity with individual features that are somewhere between difficult and impossible!
VTOL is now quite possible with multi copters but combining a fixed wing to allow conventional flight generates a whole new set of problems that require specialist solutions that are not (yet) generally available.
Solar power falls into the same category. It can be done but has many limitations.

I suggest you do some research on either multi copters or fixed wing planes to understand what is available, what their limitations are and the costs involved.
If you can't find a solar powered VTOL fixed wing plane there is almost certainly a good reason. It is technically not possible. ;)
 

Ali F

Member
Ali F
A project of enormous complexity with individual features that are somewhere between difficult and impossible!
VTOL is now quite possible with multi copters but combining a fixed wing to allow conventional flight generates a whole new set of problems that require specialist solutions that are not (yet) generally available.
Solar power falls into the same category. It can be done but has many limitations.

I suggest you do some research on either multi copters or fixed wing planes to understand what is available, what their limitations are and the costs involved.
If you can't find a solar powered VTOL fixed wing plane there is almost certainly a good reason. It is technically not possible. ;)
Thank you for all this information.
If we eliminate the solar cells + VTOL, what would be the design process of a fixed-wing uav? And what’s material to use?
 

ElectriSean

Eternal Student
Mentor
I would start with a FT Explorer. Once you get that flying well, you can add a flight controller and GPS running iNav or Arduplane. Once you get that running well would be the time to start delving into designing an airframe.
 

Ali F

Member
ElectriSean
Thank you for your reply.
But if i want to build something in my own, what is the design process?
 

quorneng

Master member
Ali F
What is the design process?
There is no simple 'one set of rules' that fits all.
Any design is always a compromise from a complex set of variables to give the best solution in the opinion of the designer to meet a particular requirement.
So the answer is
either it requires years of model design and flying experience, an understanding of the theory of aerodynamics and a working knowledge of suitable power systems.
or start with something that has already been designed by someone who has this knowledge and appears to meet your requirements.
When you have built and learnt to fly that plane well you might then be in a position to make it perform better in a way that suits you.
Which I think is exactly what EletriSean advised. ;)
 

clolsonus

Well-known member
Hi Ali, people write books on the subject so there is a tremendous amount of information and theory that could be discussed here. There are actual design processes and strategies just for designing things (anything.) There are a lot of basic formulas and relationships that can help size an airplane for what you want to do, and also estimate important things like stall speed, efficiency, range. There are tools like ecalc (web site) that can help with planning the best motors/esc/battery combinations. There is much to discuss about autopilots, control theory, sensors, algorithms, and robustness. There is plenty to learn about electronics, wiring, connectors, RF signals.

An aerospace engineer doesn't need to know everything, but they need to know some of everything. (They might think they know everything.) :) So I've done a little of this sort of thing ... enough to have quite a few strong preferences and simultaneously realize how little I know about how anything works.

I will also observe there are two main approaches taken to the overall design process. 1. you have people that know enough formulas and theory to do a design on paper and come up with a plan. 2. you have people who have been designing and building model airplanes as long as they can remember and have built up a tremendous amount of experience and intuition. Group #1 can fail by missing something obvious like making a hinge or linkage work right, or never considering where the CG should go. Group #2 is successfully building and flying great and fun designs, but it's unlikely anyone from that group will be working on the next space shuttle. The best person will have a healthy portion of both theory and experience. People that do this seriously ... no matter which group they start in, will will want to learn and absorb as much theory and subjective experience as possible.

So all that said, here is some simple advice. Keep your project as simple as possible. Use off the shelf airplanes and components as much as possible. Best case scenario, if you pick a ready to fly airplane and a ready to fly flight controller (pixhawk?) you will likely spend months putting things together and figuring out how the software works. Don't pile on insurmountable design/build challenges along with just getting a basic system flying. (I saying this with the assumption you are a student juggling a full class load.) If you are a DIY'er sinking in all your evenings and weekends and daydream time into the project for the fun of it, then you might be able to get things working a lot quicker.

Pick a plane that is very easy to fly. Find something that is slow and very cheap or very quick to [re]build. Plan on having multiple mishaps. Find a good RC pilot to help you out, or plan on investing a few weeks/months and a few crashes learning basic piloting skills yourself.

Be safe. The prop(s) are like sharp knives that can spin at 1000's of RPM. Respect the danger. Remove the props when you work on or around the aircraft. I put a big slice in my couch setting up a uav in my basement. Due to a switch on my transmitter being misconfigured and then accidentally bumped ... the throttle ran up to full and shot across the room. Thankfully I got out of the way, but just about everyone has a similar story when they let their guard down or tried to cut a corner.

Now, when you find yourself at the point where you have a basic airplane flying with a basic autopilot system, you are ready to work on whatever your next priority is. Do you want to integrate a camera? Do you want to run down that path of adding vtol capability? Do you want to design a purpose built airframe? I suggest you forget the solar cells, but maybe you want to run down that path next and investigate what you can do there. It is good to establish a solid baseline of what you are confident you can fly and then add to it.

If you are doing this as a student and need to knock a home run out of the park for your first try and you only have until late April to turn in your final report, then plan carefully and realize that everything takes way longer than you think and is way harder than you'd ever imagine. If you press through, you will learn way more than you ever though possible. You'll get stuck waiting for parts to arrive, you'll break things, or plug something in backwards and let the smoke out. You'll run into something that just doesn't work and you can't figure it out. And then time runs out on the project before you can complete it and fly. So keep things as simple as possible. Get something flying as quickly as possible, even if it's just a simple RC airplane. If you've never built/flown one before, it's not hard, but it's not easy, there are plenty of places to get tripped up with just the simple things.

These projects are tremendously rewarding when they come together and fly. Everyone here knows the feeling of a successful maiden flight with a new airplane you've just spent weeks or months putting together. Good luck!
 
  • Like
Reactions: FDS

Ali F

Member
Ali F
What is the design process?
There is no simple 'one set of rules' that fits all.
Any design is always a compromise from a complex set of variables to give the best solution in the opinion of the designer to meet a particular requirement.
So the answer is
either it requires years of model design and flying experience, an understanding of the theory of aerodynamics and a working knowledge of suitable power systems.
or start with something that has already been designed by someone who has this knowledge and appears to meet your requirements.
When you have built and learnt to fly that plane well you might then be in a position to make it perform better in a way that suits you.
Which I think is exactly what EletriSean advised. ;)
Yes I understand, thank you so much for your time and advice, its very helpful.
 

Ali F

Member
Hi Ali, people write books on the subject so there is a tremendous amount of information and theory that could be discussed here. There are actual design processes and strategies just for designing things (anything.) There are a lot of basic formulas and relationships that can help size an airplane for what you want to do, and also estimate important things like stall speed, efficiency, range. There are tools like ecalc (web site) that can help with planning the best motors/esc/battery combinations. There is much to discuss about autopilots, control theory, sensors, algorithms, and robustness. There is plenty to learn about electronics, wiring, connectors, RF signals.

An aerospace engineer doesn't need to know everything, but they need to know some of everything. (They might think they know everything.) :) So I've done a little of this sort of thing ... enough to have quite a few strong preferences and simultaneously realize how little I know about how anything works.

I will also observe there are two main approaches taken to the overall design process. 1. you have people that know enough formulas and theory to do a design on paper and come up with a plan. 2. you have people who have been designing and building model airplanes as long as they can remember and have built up a tremendous amount of experience and intuition. Group #1 can fail by missing something obvious like making a hinge or linkage work right, or never considering where the CG should go. Group #2 is successfully building and flying great and fun designs, but it's unlikely anyone from that group will be working on the next space shuttle. The best person will have a healthy portion of both theory and experience. People that do this seriously ... no matter which group they start in, will will want to learn and absorb as much theory and subjective experience as possible.

So all that said, here is some simple advice. Keep your project as simple as possible. Use off the shelf airplanes and components as much as possible. Best case scenario, if you pick a ready to fly airplane and a ready to fly flight controller (pixhawk?) you will likely spend months putting things together and figuring out how the software works. Don't pile on insurmountable design/build challenges along with just getting a basic system flying. (I saying this with the assumption you are a student juggling a full class load.) If you are a DIY'er sinking in all your evenings and weekends and daydream time into the project for the fun of it, then you might be able to get things working a lot quicker.

Pick a plane that is very easy to fly. Find something that is slow and very cheap or very quick to [re]build. Plan on having multiple mishaps. Find a good RC pilot to help you out, or plan on investing a few weeks/months and a few crashes learning basic piloting skills yourself.

Be safe. The prop(s) are like sharp knives that can spin at 1000's of RPM. Respect the danger. Remove the props when you work on or around the aircraft. I put a big slice in my couch setting up a uav in my basement. Due to a switch on my transmitter being misconfigured and then accidentally bumped ... the throttle ran up to full and shot across the room. Thankfully I got out of the way, but just about everyone has a similar story when they let their guard down or tried to cut a corner.

Now, when you find yourself at the point where you have a basic airplane flying with a basic autopilot system, you are ready to work on whatever your next priority is. Do you want to integrate a camera? Do you want to run down that path of adding vtol capability? Do you want to design a purpose built airframe? I suggest you forget the solar cells, but maybe you want to run down that path next and investigate what you can do there. It is good to establish a solid baseline of what you are confident you can fly and then add to it.

If you are doing this as a student and need to knock a home run out of the park for your first try and you only have until late April to turn in your final report, then plan carefully and realize that everything takes way longer than you think and is way harder than you'd ever imagine. If you press through, you will learn way more than you ever though possible. You'll get stuck waiting for parts to arrive, you'll break things, or plug something in backwards and let the smoke out. You'll run into something that just doesn't work and you can't figure it out. And then time runs out on the project before you can complete it and fly. So keep things as simple as possible. Get something flying as quickly as possible, even if it's just a simple RC airplane. If you've never built/flown one before, it's not hard, but it's not easy, there are plenty of places to get tripped up with just the simple things.

These projects are tremendously rewarding when they come together and fly. Everyone here knows the feeling of a successful maiden flight with a new airplane you've just spent weeks or months putting together. Good luck!
Hey clolsonus,
I really appreciate your detailed respone, thank you so much, its very helpful.
Just to have a better discussion, I have to say that Im an Aerospace Engineer, and as you may know, the more information you know the more you complex things, I asked for a design process just to have other ideas and processes.
Because for me, to build an airplane, you have a long list of calculations first, and you have to design a CAD model, and then do a FEA simulation to calculate the forces and moments, and also a CFD simulation to have the aerodynamic forces, you should also know which material to use and the stress strain curve and the Young Modulus...............
Because of all this huge information and knowledge I have in the field, I was blocked how I have to begin.
 

clolsonus

Well-known member
Hey clolsonus,
I really appreciate your detailed respone, thank you so much, its very helpful.
Just to have a better discussion, I have to say that Im an Aerospace Engineer, and as you may know, the more information you know the more you complex things, I asked for a design process just to have other ideas and processes.
Because for me, to build an airplane, you have a long list of calculations first, and you have to design a CAD model, and then do a FEA simulation to calculate the forces and moments, and also a CFD simulation to have the aerodynamic forces, you should also know which material to use and the stress strain curve and the Young Modulus...............
Because of all this huge information and knowledge I have in the field, I was blocked how I have to begin.

Hi Ali, I apologize for wrongly assuming. Sorry about that. I work in the aerospace engineering dept at my university, but my degree is in computer science ... so take whatever I might say with a proper grain of salt.

So yes, there is a lot of analysis you can do on paper; it sounds like you probably know how to do more of that than I do. Some of that is useful, some of it probably isn't as useful. For example, we have people here that obsess on picking an airfoil, when in reality, just going with a clarky would get them %96.85840735 of the way there, and later if they want to get that last 3.14159265% squeezed out, that would be the time to look at better optimizing the airfoil.

Here is what I would suggest. Keep the focus on the original mission. Think carefully about the high level challenge you are trying to solve. The design choices should directly support the end goal. It's too easy to drill down to specifics and forget why you are doing things in the first place. It's too easy to decide you need certain things early in the process and they become constraints you have to struggle with from then on. Everyone does it. I do it too. So do what I say, not what I do. :) For example, does your end objective really require solar cells? Or in a moment of dreaming, did solar cells sound like a fun/good way to extend flight times so you went with that as a requirement? Maybe there are better ways to achieve the desired endurance? Likewise, 2m wingspan? The plane would ideally size itself based on how much fuel/payload you need to carry vs. required airspeed, stall speed, etc. Maybe there other constraints imposed on you, for example your boss or a customer told you it should be 2m wingspan, and then that's ok, that's what you go with and do the best you can with.

It could very well be that you just want to build a 2m wingspan vtol with solar cells, camera, t-tail. You hope the final result does some useful things or is a useful product, but you are embarking on the project to build something you've been dreaming about. That's totally fine too. I'd say most of the projects people work hard on are things they wanted to do, more than things they needed to do.

So if I have any advice at all, keep your focus on the original core task and let all these other specific details be solutions to your design goals. I know that's super generic and abstract. It's more of a philosophy than a design methodology. It's one of those things that is easier said than done.

And then I still stand by my other advice. Get something flying early early early. Pick off the shelf components and cobble them together. Get something flying that is 80% of what you want (or maybe just 50% but it's flying.) Once you have a prototype working, then you can make better choices about where to invest your limited development time and effort in order to make your design stand out. You may want to design an airplane from scratch, but it's probably a terrible business strategy. From personal experience it is far too easy to spend 5 years developing something in a cave, and then another 5 years after that because it's not yet perfect, and you could be 10 years down the road having done a lot of great work and an awesome product, but the world has already moved on to other things. At the end of the day, if you manage to scratch out an existence along the way, do good work, learn a ton, then it's probably more than worth it, even if the product itself isn't a home run. You'll be right there with the rest of us also trying to scratch out an existence one way or another.

Those are my thoughts for whatever they are worth. :)

Best regards,

Curt.
 

quorneng

Master member
Hi Ali
I think clolsonus has said all that is required but do try to avoid over thinking things.
Unlike full size designing a model plane is a one man task. To achieve a result is an meaningful time scale you have to cut corners which in effect means you will simply have to copy what others have done or maybe use what is commercially available. This involves doing research (the internet?) to find out what has been achieved and what is available.

It is worth remembering that at small sizes scale effects have a huge impact on the strength of structure and that air itself has rather different characteristics. Models tend to operate in conditions (very low Reynolds numbers) that are outside much of the full size testing so there tends to be more of 'it works' rather than trying to mathematically check if it is the best possible result.

If you sketch out what you think might be what you are looking for I am sure there will no shortage of advice on sites like this.
 

evranch

Well-known member
Just some more info on VTOL if you are still interested in that route.

Ardupilot supports quadplanes, which are a standard fixed wing with a main propulsion motor and 4 motors to lift and balance it for VTOL. However there are issues that come with a quadplane that as clolsonus says may not help your core mission. Those motors and props will be large to lift a heavy, long endurance airframe.

- extra complexity
- extra weight
- extra drag
- extra cost
- a surprising amount of fuel burn for the vertical portion of the mission

They look really cool but there is a reason most people doing fixed wing mapping have settled for a hand or catapult launchable platform. A hand launch gets you into the same terrain as VTOL, with much less complexity. Put some skid plates on the bottom and you can bring it down almost anywhere but a boulder field.

My advice for a UAV design would be to get your wing loading as low as possible, allowing payload flexibility and hand launching and landing at minimal speeds. I'm working on a delta wing design right now that has been doing everything I want from it, with a very simple build, easy hand launch and broad flight envelope.
 

Ali F

Member
Hi Ali, I apologize for wrongly assuming. Sorry about that. I work in the aerospace engineering dept at my university, but my degree is in computer science ... so take whatever I might say with a proper grain of salt.

So yes, there is a lot of analysis you can do on paper; it sounds like you probably know how to do more of that than I do. Some of that is useful, some of it probably isn't as useful. For example, we have people here that obsess on picking an airfoil, when in reality, just going with a clarky would get them %96.85840735 of the way there, and later if they want to get that last 3.14159265% squeezed out, that would be the time to look at better optimizing the airfoil.

Here is what I would suggest. Keep the focus on the original mission. Think carefully about the high level challenge you are trying to solve. The design choices should directly support the end goal. It's too easy to drill down to specifics and forget why you are doing things in the first place. It's too easy to decide you need certain things early in the process and they become constraints you have to struggle with from then on. Everyone does it. I do it too. So do what I say, not what I do. :) For example, does your end objective really require solar cells? Or in a moment of dreaming, did solar cells sound like a fun/good way to extend flight times so you went with that as a requirement? Maybe there are better ways to achieve the desired endurance? Likewise, 2m wingspan? The plane would ideally size itself based on how much fuel/payload you need to carry vs. required airspeed, stall speed, etc. Maybe there other constraints imposed on you, for example your boss or a customer told you it should be 2m wingspan, and then that's ok, that's what you go with and do the best you can with.

It could very well be that you just want to build a 2m wingspan vtol with solar cells, camera, t-tail. You hope the final result does some useful things or is a useful product, but you are embarking on the project to build something you've been dreaming about. That's totally fine too. I'd say most of the projects people work hard on are things they wanted to do, more than things they needed to do.

So if I have any advice at all, keep your focus on the original core task and let all these other specific details be solutions to your design goals. I know that's super generic and abstract. It's more of a philosophy than a design methodology. It's one of those things that is easier said than done.

And then I still stand by my other advice. Get something flying early early early. Pick off the shelf components and cobble them together. Get something flying that is 80% of what you want (or maybe just 50% but it's flying.) Once you have a prototype working, then you can make better choices about where to invest your limited development time and effort in order to make your design stand out. You may want to design an airplane from scratch, but it's probably a terrible business strategy. From personal experience it is far too easy to spend 5 years developing something in a cave, and then another 5 years after that because it's not yet perfect, and you could be 10 years down the road having done a lot of great work and an awesome product, but the world has already moved on to other things. At the end of the day, if you manage to scratch out an existence along the way, do good work, learn a ton, then it's probably more than worth it, even if the product itself isn't a home run. You'll be right there with the rest of us also trying to scratch out an existence one way or another.

Those are my thoughts for whatever they are worth. :)

Best regards,

Curt.
Thank you so much for your effort, I really appreciate that.
So, from all what you said, the steps are:
- Get something that fly properly as soon as possible.
- Pick an existance design.
- wingspan is related to how much lift you need.

My project is for mapping and environmental monitoring, so i choosed the fixed-wing for the long endurance they offer, also if I can get a high Aspect Ratio wing ( like a glider wing ) it could be perfect, because in this case you dont need to much power.
So the components that I need are:
- Camera.
- Pitot-Tube.
- Altimeter + Barometer + Thermometer.
- Autopilot.
- Sensors (gas + obstacles).

The VTOL is not only for the take-off and landing, but its also provide hovering, and this is important to take pictures or measurements in a specific position.
 

Ali F

Member
Just some more info on VTOL if you are still interested in that route.

Ardupilot supports quadplanes, which are a standard fixed wing with a main propulsion motor and 4 motors to lift and balance it for VTOL. However there are issues that come with a quadplane that as clolsonus says may not help your core mission. Those motors and props will be large to lift a heavy, long endurance airframe.

- extra complexity
- extra weight
- extra drag
- extra cost
- a surprising amount of fuel burn for the vertical portion of the mission

They look really cool but there is a reason most people doing fixed wing mapping have settled for a hand or catapult launchable platform. A hand launch gets you into the same terrain as VTOL, with much less complexity. Put some skid plates on the bottom and you can bring it down almost anywhere but a boulder field.

My advice for a UAV design would be to get your wing loading as low as possible, allowing payload flexibility and hand launching and landing at minimal speeds. I'm working on a delta wing design right now that has been doing everything I want from it, with a very simple build, easy hand launch and broad flight envelope.
Thank you for you response evranch.
The VTOL is not only for the take-off and landing, but its also provide hovering, and this is important to take pictures or measurements in a specific position.
 

clolsonus

Well-known member
Thank you so much for your effort, I really appreciate that.
So, from all what you said, the steps are:
- Get something that fly properly as soon as possible.
- Pick an existing design.
- wingspan is related to how much lift you need.

Again, this is all from my own tiny slice of personal experience and perspective, I would never claim this is the only way or even the best way. There is almost always many good ways to accomplish a task.

So I think yes, if you can get something relatively simple in the air, that will give you a baseline of experience and something to work forward from. It's good to make design decisions to improve your work, but it is better to make a change for a specific reason rather than sitting in front of a blank piece of paper trying to decide everything at once. If you pick an existing design (or existing ARF/RTF) that will just help you get flying sooner. Yes, wing span (multiplied by wing chord) == wing area is the biggest factor in deciding how much payload you can carry.

My project is for mapping and environmental monitoring, so i choosed the fixed-wing for the long endurance they offer, also if I can get a high Aspect Ratio wing ( like a glider wing ) it could be perfect, because in this case you don't need to much power.

One thing I would caution you on is over optimizing speed, endurance, and payload early in the project. It is easy for this to result in an airplane that is extremely difficult (scary, sketchy) to launch and land. RC airplanes are usually built super light, super over powered, often have big fat wings and the purpose to have large margins. The pilot is on the ground not onboard the airplane so your response is slower and usually less precise ... you need bigger margins in the design to fly safely.

The other reason is that if your design is too hard to hand launch, now you need to design a launcher. A launch system will be trickier than you first imagine. No one want to go home with a shredded airplane, but this happens far too often if the complexity goes up and the wing loading goes up.

In my humble opinion, the best thing you can do for endurance is to keep a big wing, but make everything as light as possible. Assuming you are cruising around at the most efficient speed for the airplane, then the amount of power required to sustain flight is most proportional to aircraft weight. If you want to stay in the air a long time, fly slow, keep the throttle low, keep your design very light weight.

You may find that if your purpose is taking pictures, you may want to slow down anyway. If you fly fast, you may have issues with blurring and smearing of the imagery, and possibly not being able to trigger the shutter fast enough to keep up with the amount of overlap you want.

Personally, I would go with something that is 20% slower or 20% less endurance if it's easy to launch and land and fly. I've done both ... we had a custom designed flying wing UAV that had great specs on paper, but was really difficult to launch and safely. We had several mishaps and set backs due to how difficult the aircraft was to fly. In the end I achieved better results with an old skywalker (with a T-tail !) than we had with our custom designed, composite, professional aircraft. And the skywalker tracks like an ocean crab with ADD. More recently I have been flying with an X-UAV talon. This flies a lot nicer, but is bigger and and higher wing loading so it's more tricky to launch and has a lot faster airspeed on landing which makes landings a little trickier too. These are boring designs that have been around for years, but you can still buy them because they work pretty well for a lot of purposes.

So the components that I need are:
- Camera.
- Pitot-Tube.
- Altimeter + Barometer + Thermometer.
- Autopilot.
- Sensors (gas + obstacles).

The VTOL is not only for the take-off and landing, but its also provide hovering, and this is important to take pictures or measurements in a specific position.

Honestly (and again this is just from my tiny sliver of personal experience) I would say this: If hovering is a hard requirement and you need a camera to take pictures, just go to the dark side and buy a dji drone (probably a mavic 2 pro, but maybe a phantom 4 pro v2 since those were just re-released.) You sacrifice flight time and range, but you will be flying the same weekend the system arrives, and it will be hard to beat the results without a *lot* of your own effort. You can get more batteries and do multi-battery missions if you need to cover a bigger area. The cost will look daunting, but if you add up everything you'll spend building a system yourself from scratch, plus cost yourself at $5/hour (or even $1/hour) for all the time you put in, plus assign some cost to the delay in getting your first system up and flying ... you will probably be way better off spending $2000-2500 on a nice off the shelf system.

The main reason to not just buy a DJI is if you really really really want the experience of building your own system from scratch. That's a totally valid reason, (I've done that myself, even designed my own flight controller from scratch) but just enter the process realistically. You don't build everything yourself from scratch to save time or money or get a better result in the end. You would do it to learn a ****LOT**** and for the experience and satisfaction.

Best regards,

Curt.
 

evranch

Well-known member
Thank you for you response evranch.
The VTOL is not only for the take-off and landing, but its also provide hovering, and this is important to take pictures or measurements in a specific position.

You don't really need the pitot tube unless you are flying at the edge of your envelope. Pitot tubes are notorious for calibration issues causing crashes and aren't really needed with modern airspeed estimation algorithms. Your other avionics should be included in your autopilot package as long as you don't mess up and buy something without a barometer that was meant for a race quad (oops)

For your usage you may want to reconsider the fixed-wing portion of your UAV, depending on how far you have to fly between measurements.
Note that taking pictures in specific locations is a solved problem for fixed wing. You can do centimeter-accurate georeferencing with an RTK GPS onboard, measurements from the IMU and hot-shoe feedback from the camera as you snap pictures while flying at relatively high speeds. Keeping your airspeed up is critical to getting good endurance.

In fact Reach can do this out of the box: https://store.emlid.com/product-category/rtk-receiver/reach/

However, if you need dwell time for gas sensing, this is a totally different usage case. Switching in and out of quadplane for repeated hovers will burn a lot of fuel for acceleration and deceleration. You need to transition from above Vs to hover, then accelerate back to Vs for every point. You may not end up using the wings all that much unless your points are a mile apart. At that point, consider whether the bulk of your time will be spent in hover or in transport, and that a quad burns pretty much the same fuel in hover or in transport. You may just want to go with a large multicopter.

On the other side, consider the location accuracy required. What resolution do you need gas samples at? Do you need to hover, or could you sense gas while flying a 10m radius circle? My big wing can fly this circle all day drawing 5 amps, while a much smaller quad will hog down 10-20 amps easily to hover.
 

Ali F

Member
Again, this is all from my own tiny slice of personal experience and perspective, I would never claim this is the only way or even the best way. There is almost always many good ways to accomplish a task.

So I think yes, if you can get something relatively simple in the air, that will give you a baseline of experience and something to work forward from. It's good to make design decisions to improve your work, but it is better to make a change for a specific reason rather than sitting in front of a blank piece of paper trying to decide everything at once. If you pick an existing design (or existing ARF/RTF) that will just help you get flying sooner. Yes, wing span (multiplied by wing chord) == wing area is the biggest factor in deciding how much payload you can carry.



One thing I would caution you on is over optimizing speed, endurance, and payload early in the project. It is easy for this to result in an airplane that is extremely difficult (scary, sketchy) to launch and land. RC airplanes are usually built super light, super over powered, often have big fat wings and the purpose to have large margins. The pilot is on the ground not onboard the airplane so your response is slower and usually less precise ... you need bigger margins in the design to fly safely.

The other reason is that if your design is too hard to hand launch, now you need to design a launcher. A launch system will be trickier than you first imagine. No one want to go home with a shredded airplane, but this happens far too often if the complexity goes up and the wing loading goes up.

In my humble opinion, the best thing you can do for endurance is to keep a big wing, but make everything as light as possible. Assuming you are cruising around at the most efficient speed for the airplane, then the amount of power required to sustain flight is most proportional to aircraft weight. If you want to stay in the air a long time, fly slow, keep the throttle low, keep your design very light weight.

You may find that if your purpose is taking pictures, you may want to slow down anyway. If you fly fast, you may have issues with blurring and smearing of the imagery, and possibly not being able to trigger the shutter fast enough to keep up with the amount of overlap you want.

Personally, I would go with something that is 20% slower or 20% less endurance if it's easy to launch and land and fly. I've done both ... we had a custom designed flying wing UAV that had great specs on paper, but was really difficult to launch and safely. We had several mishaps and set backs due to how difficult the aircraft was to fly. In the end I achieved better results with an old skywalker (with a T-tail !) than we had with our custom designed, composite, professional aircraft. And the skywalker tracks like an ocean crab with ADD. More recently I have been flying with an X-UAV talon. This flies a lot nicer, but is bigger and and higher wing loading so it's more tricky to launch and has a lot faster airspeed on landing which makes landings a little trickier too. These are boring designs that have been around for years, but you can still buy them because they work pretty well for a lot of purposes.



Honestly (and again this is just from my tiny sliver of personal experience) I would say this: If hovering is a hard requirement and you need a camera to take pictures, just go to the dark side and buy a dji drone (probably a mavic 2 pro, but maybe a phantom 4 pro v2 since those were just re-released.) You sacrifice flight time and range, but you will be flying the same weekend the system arrives, and it will be hard to beat the results without a *lot* of your own effort. You can get more batteries and do multi-battery missions if you need to cover a bigger area. The cost will look daunting, but if you add up everything you'll spend building a system yourself from scratch, plus cost yourself at $5/hour (or even $1/hour) for all the time you put in, plus assign some cost to the delay in getting your first system up and flying ... you will probably be way better off spending $2000-2500 on a nice off the shelf system.

The main reason to not just buy a DJI is if you really really really want the experience of building your own system from scratch. That's a totally valid reason, (I've done that myself, even designed my own flight controller from scratch) but just enter the process realistically. You don't build everything yourself from scratch to save time or money or get a better result in the end. You would do it to learn a ****LOT**** and for the experience and satisfaction.

Best regards,

Curt.
Yes I see my friend.
So for my first uav, I will eliminate the VTOL + Hovering possibilities for a fixed-wing uav, and work on a hand launching one and FPV.
What do you mean with the big wing? The wingspan or the area?
I would say that I prefer the SkyWalker uav's for all what they can offer, a SailPlane is also great in terms of endurance and power efficiency.
So now, I can say that all I need is:
- FPV.
- Altimeter + Barometer + Thermometer.
- Autopilot.
- Sensors (gas + obstacles).
- Low speed uav.
- High wing.
 

evranch

Well-known member
Big would refer to area, as the goal is low wing loading. Excessive wingspans tend to get floppy on RC planes.

The Skywalker is a classic and proven platform that will get you in the air right away, taking data and learning how you want to develop your system. Good choice.
 

Ali F

Member
Big would refer to area, as the goal is low wing loading. Excessive wingspans tend to get floppy on RC planes.

The Skywalker is a classic and proven platform that will get you in the air right away, taking data and learning how you want to develop your system. Good choice.

This is what you meant by the Skywalker, right? Because there is also a Flying wing called the Skywalker.
1582630449898.png


So now, if I can pick a RTF Skywalker, with an FPV and an Autopilot (pixhawk + mission planner) will be great, and after successful flights, I can add sensors and all the measurement instruments, that will be good?
 

quorneng

Master member
Ali F
Don't assume a glider is the most efficient configuration. Efficiency depends on what you are trying to achieve.
A modern high performance glider is designed to do one thing really well - to fly as far as possible for a given loss in height. To do this it has to sacrifice strength so it has a very limited payload capacity and restricted manoeuvrability.
Broader shorter wings may be a bit less aerodynamically efficient but structurally are much better so for the same weight as a glider wing it can carry a 'proper' payload and with a bigger area means it flies slower which for a power plane means less power is required.

If you look at a Skywalker in this light it is actually more 'efficient' at what it does than if it had a high performance glider configuration.
There is no simple 'one solution fits all' but just finding the best compromise to met requirements if indeed there is one. ;)