Thrust vectoring is one way to control aerial vehicles. It’s become more popular as technology advances, finding applications on fifth-generation fighter aircraft, as well as long being used in space programmes the world over.[RCLifeOn] decided to try and bring the technology to a prop-powered RC aircraft, in an unconventional way.
After attending a lecture on compliant mechanisms and their potential use in space vehicles for thrust vectoring control, [RCLifeOn] decided to try applying the concept himself. His test mechanism is a fixed-wing with a single-piece motor mount that has enough flex in the right places to allow the motor (and propeller) to be moved in two axes, achieving thrust vectoring control.
After printing a compliant motor mount in a variety of materials, one was selected for having the right balance of strength and flexibility. The vectoring mechanism was fitted to a basic flying wing RC aircraft, and taken to the field for testing. Unfortunately, success was not the order of the day. While the mechanism was able to flex successfully and vector the motor in bench testing, it was unable to hold up to the stresses of powered flight. The compliant mechanism failed and the plane nosedived to the ground.
[RCLifeOn] suspects that the basic concept is a difficult proposition to engineer properly, as adding strength would tend to add weight which would make flight more difficult. Regardless, we’d love to see further development of the idea. It’s not the first time we’ve seen his 3D-printed flight experiments, either. Video after the break.
19 thoughts on “Thrust Vectoring With Compliant Mechanisms Is Hard”
I love the sense of humor this guy has and the effort he took to post even one of his failed experiments. Keep up the good work and I would love to see updates about this project :-D
RCLifeOn: wishing you much success in your ongoing experimentation!
Y’know, space vehicles have a smaller and much more well-behaved set of stresses in flight. The bench test doesn’t have any aerodynamic stresses due to the movement of the vehicle through the air, nor does space, but the plane flying around does. That may be the root of the issue.
When the compliant stuff breaks but the non-compliant materials work … sounds like FAA needs to upgrade their compliance guidelines
What’s the advantage of a compliant mount over something like a more traditional gimbal?
I really don’t get it either, I guess it would be cool to replace a bunch of bushings, fasteners, and 3d printed parts with one 3d printed part.
On the other hand you are trading an assortment of simple parts with one very complex one.
For atmospheric applications, the only upside would be decreased manufacturing time and simpler assembly, but honestly it might just make things harder to service as you would need to replace an entire mechanism rather than just a bolt or bearing. All in all, down here on earth you are right, compliant mechanism have very limited use cases. For space however they are absolutely brilliant.
One of the biggest issues for moving assemblies in space is lubrication. Normal greases and lubricants will boil off and/or freeze solid, so quite a lot of research has been done into things like dry lubricants, non-lubricated joints, and even entirely non-contacting joints (eg magnetic bearings). As well, all of those systems add weight and increase complexity/part count, which is dicey for an application where servicing is basically impossible (exception being the ISS which is one of the few instances in the space industry where they need as many things to be serviceable as possible)
Complaint mechanisms solve all those issues very elegantly. First, they have no moving surfaces like rolling or sliding joints, so lubrication becomes totally irrelevant. Second, they use a single part, so you don’t have to worry nearly as much about things like vibrating your fasteners loose during launch, forgetting to insert pin number 284 out of 400 during assembly, galvanic corrosion between parts made of different metals, etc. Third, since all motion is achieved through elastic deformation, if some kind of glitch or error occurs that causes the spacecraft to lose track of the mechanism’s state/position, so long as the mechanism is still intact you can easily reacquire its position by letting the mechanism passively relax back to its original state. Normal joints can only do this with the assistance of spring loading. Fourth, instead of paying for hundreds of aerospace grade components to be manufactured, you just have to pay for one (metal 3D printing is the only method I’ve heard used for space related complaint mechanisms I believe, but I have heard of non-space related ones made using CNC and even chemical etching for micron scale mechanisms).
There are likely even more benefits to using these in space that we haven’t even realized yet, but as I hope I have illustrated the benefits that compliant mechanisms provide to space applications are quite numerous.
In theory, no/fewer bearing surfaces, no fatigue wear. Flexures/compliant mechanisms last forever when well designed. This seems to be poorly designed with respect to the stresses. But hey, it’s a learning experience.
You can machine them out of monolithic material giving greater control over crystal alignment to stresses, they can be made very tiny, out of materials that normally couldn’t be used, eg glass or silicon.
Agree, at first glance, that should have been machinable from a block of material. Printing it in the orientation that he did is likely a huge problem. Also, in CAD, for this application – I don’t think FEA is optional. 3D printing just makes this harder.
That said – a few properly designed pieces that assemble (as opposed to a monolithic part) could very well work. I don’t think the project is doomed, just that particular approach.
The little I know about compliant mechanisms comes from robotics. The use there is to allow for safer interaction with people and the world but they carry a whole host of challenges because of their springiness.
They also are not necessarily constructed from a single monolithic piece. All that is really required is something elastic or springy between the actuator and the part being actuated.
I’m really struggling to understand what problem in rocket thrust vectoring is better solved with a compliant mechanism.
As for manufacturing something by cutting it out of a single piece, I’m reminded of gear-head arguments over forged vs billet parts. Forged parts are generally considered stronger. I also see potential issues with the different processes the complaint parts of a mechanism would need to undergo vs the rigid parts for something that is a single piece.
> Flexures/compliant mechanisms last forever when well designed.
I’m not buying this. I cannot imagine a living hinge made of a sliver of cheap plastic outliving a hinge made of steel with ball bearings.
“When well designed” is the key phrase.
With a good understanding of the relevant material properties (yield strength, fatigue limit, etc) and the stresses involved during use, it is absolutely possible to design a flexure (or living hinge, for that matter) that can withstand millions+ of cycles under load.
Not to say that 3D printed beams are the best/cheapest/easiest material for flexure design, but it is definitely possible.
Something with a pretty limited range of motion using a self lubricated bushing has a pretty long lifetime too.
But maybe the clue as to why use a compliant mechanism has to deal with friction?
There will still be losses incurred in deforming the compliant material though.
Matt, check out plastic vs. elastic deformation. With the base of this project being an aircraft, I would think the weight and cost of the metal would void any benefit.
This is where the low cost per print is allowing people to get goofy and say “Why Not?”
That’s a different point than they were making though- they were merely replying to the question of the advantages a compliant mechanism over more traditional gimbal- you’re comparing steel parts to plastic parts…
And even if you compare steel parts to plastic parts there are cases where a plastic flexure/compliant version could greatly outlive the steel version. For example, when immersed in water…
For similar materials the flexures can vastly outlive the bearing based designs. It’s all about making sure the mechanism never gets distorted enough to have material fatigue.
Look into the Clock of the Long Now- designed to last at least 10,000 years. It has many parts that are flexures precisely because they outlive -significantly- the bearing based designs. That said those parts are made from high quality materials to begin with- like wire EDM’d steel.
people seem to like the idea of single piece prints over subassemblies. but i find its almost always better to use subassemblies and appropriate hardware to join them. also if something breaks reprint that part and salvage the rest of your part. if that part fails repeatedly, then you know its a weak point and can redesign it for strength. it also means you can use a smaller, cheaper printer.
Hmmm, 2 degree of freedom flex joint that resists torque. Sounds like a rag joint to me. Fix one end to the airframe and the motor to the other end. Then use fishing line tied to an arm extension on the servos to gimbal it. Polypropylene or polyethylene from food packaging make inexpensive and long service life flexures.
i think id go with a delta platform driven by 2 servos with a fixed linkage on the 3rd corner. would have been more robust and more compact.
Wow, that doesn’t look easy.
What about having two control systems. Traditional and the new one. The traditional one would help you get it flying and trimmed. Then disengage it and try and control with the new system… A thought…
Please be kind and respectful to help make the comments section excellent. (Comment Policy)