Curiosity Style Egg Drop Lander

[K^2]

You are still looking at less than a second of operation. In that time frame, you need to do a couple of dozen of corrections. You need both precision and fast response there.

In terms of servo mechanics, they are PWM controlled. Older style servos, the cheap ones, consist essentially of a capacitor to turn PWM signal into average reference voltage, a potentiometer hooked up to servo arm to turn arm's position into another voltage, and a differential amplifier hooked up to these two voltages to drive the electric motor. There are all sorts of delays there. It will take a good fraction of a second just for the reference voltage on the cap to adjust to the new PWM control signal.

There are much better servos out there. With stepper motors and digital controls. These can have absurdly short response time and more precision than you'd need. They tend not to find as much use in hobbies, though, because of higher cost. So you might have trouble finding some that would work for you.

In principle, you might be better off just going with steppers and setting up your own controls. After all, it's the nozzle position that you are interested in. Not the servo position.

P.S. All of that said, this might, indeed, be a better option in terms of delays. At least, it shifts mechanical problems of valve control to digital problem of nozzle control. Maybe I'm just better with the later, but I think I'd also go with that. Keep in mind, however, that you are going from a semi-stable problem to a totally unstable one. So you'll have to introduce stability in your control algorithm. You'll probably end up needing better response than PID, but understanding how PID works would be a fine first step.

Edited July 11, 2014 by K^2

[cantab]

I'll just give a few ideas for simple approaches. Even if you want to ultimately have something sophisticated, simpler prototypes could be useful.

For a non-throttling "suicide burn" design, trigger the rockets with a static line. Runs from the drop point and when it snaps taut it yanks on the valve or switch.

To release the egg, use a mechanical hook or catch of some sort that automatically lets go when the egg touches down. The release of the weight will then send your skycrane flying away, add some ballast to the egg if need be.

For the more control-oriented approach, using valves or gimbals, make the rockets water jets fed by an umbilical hose. You'll have a constant ship mass, unlimited flying time, and an easy emergency shutoff.

[christheman200]

K^2, I've witnessed the extreme accuracy of stepper motors in the past. They seem like a better option.

If the rocket is designed to be aerodynamically stable, it should be statistically more stable, because any variation in the thrust of each engine in a 4 thruster system would make the rocket unstable, whereas in a single engine scenario, there is an overall smaller chance of this being as significant.

Cantab, I would like to use your umbilical hose approach, but it would make the craft unstable from any interaction of the mass of the hose with the lander.

I'm continuing research regarding the performance of the rocket.

[K^2]

Hobby rockets achieve stability due to the fins. You will have aerodynamics working against you. The whole thing is going to be unstable, both statically and dynamically. You'll have to achieve active stability with control software. That's the only way.

[magnemoe]

Hobby rockets achieve stability due to the fins. You will have aerodynamics working against you. The whole thing is going to be unstable, both statically and dynamically. You'll have to achieve active stability with control software. That's the only way.

If the drag point is high up it will be stable, so some sort of air brakes high up will keep the top up while having most mass down.

Could it use an laser to measure distance to ground? know you can get pretty cheap ones for construction, this trigger the rocket, it the rocket trust is consist you could just set this up for a soft landing using trial and error.

[K^2]

If the drag point is high up it will be stable

Sure. If you plan to bury the probe in the ground. If you are planning to slow it down, then you can't count on drag as you touch down. Not to mention that it's a big "if" to begin with. Sure, one could always build a lawn dart. But seeing how the system has to be capable of maintaining stability at zero velocity, it's just extra weight and complexity.

[magnemoe]

Sure. If you plan to bury the probe in the ground. If you are planning to slow it down, then you can't count on drag as you touch down. Not to mention that it's a big "if" to begin with. Sure, one could always build a lawn dart. But seeing how the system has to be capable of maintaining stability at zero velocity, it's just extra weight and complexity.

You are right I was thinking of stability during the drop phase, and this will need multiple engines, an system with only one engine could probably be make stable enough without active control.

However something who hit the ground in huge speed and disintegrating while the engines spray water everywhere would be very kerbal.

[John FX]

A fixed amount of Dv released in a fixed amount of time will slow a craft from a fixed speed to another fixed speed. As has been said, have a piece of string that activates some dumb boosters that slow the craft to a set speed (ideally zero or close to) and have some springy legs able to absorb the force of a small fall (from just above zero to your margin of error)

An SSTL if you wish.

i made something similar in KSP. It was a pod, a chute and three seperatrons. I set the chute to slow me down to about 80m/s (and make sure I was pointed down) then the three seperatrons fired just before impact and slowed me down to the exact speed that would break all my landing struts but not explode anything. (about 38m/s would you believe)

I had to use KOS to trigger the seperatrons at 45-48m which meant either an altitude accuracy of 3m or a timing accuracy of 0.03 seconds or so...

I went for the dumb altitude trigger and it worked quite well.

[christheman200]

The lander may not be as unstable as you think at low velocities. The drag should still push in the same direction, and once it reaches a velocity of zero there will be a drag force of zero.

I like the simplicity of a dumb landing system, but that's not what I'm looking to get out of this project.

The official DragonEgg lander blog will be put up very soon. The idea of a kickstarter campaign to raise funds for this project is creeping into my mind.

[mythic_fci]

Can the lander be attached to a hose or pipe to feed it, or must it be a freeflying craft? If it must be freeflying, you may want to chain a large bottle of pressurized water (or two) to a controllable valve (Use something like those on lawn hoses, with a remote-controlled motor to adjust the throttle. Also attach servos on the valve for control If you can attach a limp hose (ex. No force applied), then use the same throttle valve mentioned above, but with a hose attached instead of bottles of pressurized water.

Edited July 15, 2014 by FCISuperGuy

[christheman200]

I have started an official twitter account!

You can follow me at @LansdaleEng, or via https://twitter.com/LansdaleEng

I will be posting daily updates on the lander, as well as engineering related subjects. At the moment, my father and I are working to put up a custom domain, and set up a blog sorts.

Plans for today include working on a simulator, and making some graphics.

[christheman200]

FCI, it is not feasible to run the craft off of an external source. The lines leading into the craft would destabilize it.

I have already constructed two large water chambers, with two more on the way.

[mythic_fci]

FCI, it is not feasible to run the craft off of an external source. The lines leading into the craft would destabilize it.

I have already constructed two large water chambers, with two more on the way.

Maybe I'm a little over-thinking this, but I've made a similar thing using an external computer program to stabilize the ship even with a moving CoM.

Nice to see the water chambers are up and running. Maybe you can use my suggested controllable valve...?

[christheman200]

There might be a few too many variables with a moving centre of mass to get a properly stabilized craft. As well, most of the weight belongs to the water held in the bottles, so there wouldn't be much mass, making the craft even more vulnerable to effects from the hose.

If you have some free time today (a couple of minutes), would you mind finding a lightweight valve that you think would work with your method? I've been a bit stumped in what to use.

What type of fitting should I use to let the nozzle gimbal?

I've also put out a request for a graphic logo over in the space lounge: http://forum.kerbalspaceprogram.com/threads/86074-In-Need-of-Some-Graphics

If someone wouldn't mind making a logo for the project, we'd be looking much more professional!

[christheman200]

I've now modelled a gimballing nozzle, to be 3D printed in ~1 month.

By fiddling with an online simulator, I have found that my craft will have around 60-75 m/s of delta-v. I'm debugging my own simulator at the moment.

As well, my website is very much a work in progress.

[christheman200]

Getting back to this thread. The project is far from dead!

I'll be travelling from LA through San Francisco starting Thursday for a week or so. If any of you can manage to get me a tour of Caltech, JPL, SpaceX, etc., send me a message. That would be a really big deal for me! I only have about a 50 percent chance of getting tours of the above at the moment. As such, there won't be much construction and testing going on for a while. To make good use of this time I've created 6 pressure vessels, which will have cured by the time I get back.

What's happened in the time since my last post? I have pressure tested a tank to 100psi through launching a water rocket, and even after many massive impacts the bottle held strong. I haven't created a coupler to chain together multiple tanks yet. I've also assembled 6 more tanks, which will cure while I'm in California. Recently I created a gimballing nozzle from Styrofoam, albeit without a nozzle, yet. I am going to figure out how to easily make a flexible nozzle over the next week. Last night I hooked up a servo to my arduino, and I hope to control it with a wii nunchuck, acting as a crude gyroscope, which will be taken from accelerometer data.

Here's a video of the last launch of the day:

The plan is as follows while on vacation:

Figure out how to get the nozzle to gimbal

Find an easy method of connecting pressure tanks

Learn some control logic

I have no knowledge of control logic, so this is where I will need to spend some time learning in the future. If you have a favourite tutorial on control logic, feel free to reply with a link.

[Javster]

Personally I'd go for a capsule with a parachute and maybe some kind of landing rocket

[DasValdez]

The water in your tank may slosh unpredictably, throwing off your COM. In the tank, you're going to need room for air to compress which leaves room for slosh. Getting the control and gimballed nozzle to work without this variable will be hard enough... you may first need to test drops with just the chute and tank to determine how it plays out.

Baffles are the first thought, but I'm not sure how you'd install them in a home built reinforced bottle. Stuff a loofah in there, maybe? Maybe a separate tank up top for air, feeding into the water tank, and let gravity do the rest? Again, a chore to set up safely on a home built bottle.

Good luck!

[Westi29]

Just have a giant parachute that will basically slow you down to 1m/s or less, then when near the ground, activate the two water rockets and use a kinex winch or something similar to lower the egg to the ground and detach.

I would solve all of the problems with a big parachute, and make the water jet sky crane more for show than anything else.

[Nuke]

i saw a two servo solid rocket gimbal design the other day which looks promising:

http://hackaday.io/project/1162-Guided-Rocket

[christheman200]

Slosh may be a large issue, but I feel that it is unlikely to cause much grief. Your post made me realize that the air may leak out if g forces start going upwards. Under a parachute this should not happen.

I haven't seen that gimballed rocket before. It looks very interesting! I might just steal some of that code.

[christheman200]

Just quick update to say that everything is going well.

The school year has started up, and I should have 2 or 3 classes to work on this project for the rest of the year. We have a 3D printer, which I am going to be setting up over the next week.

I now have an 8 foot 4 inch tall water rocket in my house. I'm developing the recovery system for a general unguided flight test, which is quite close to being done. It will be based off of Arduino/Freeduino boards, and since I only have an Uno at the moment I may have to wait for my Pro Mini Freeduinos to come in before I launch. Losing a 40 dollar Arduino isn't my idea of fun, even though the whole rocket has cost much more than that so far. The head of our First Robotics team is working on a table which will balance a ball by shifting the table upon which it sits, and this happens to be pretty similar to my gimballing engine. Hopefully her work can help me out with my project.

California was awesome! I passed by SpaceX, Tesla Motors, Space Shuttle Endeavour, the Orion Capsule, got past the front gates of JPL, and toured Caltech. For those of you who know him, I met Veritasium after my campus tour. There were also numerous giant telescopes.

[Payload]

^Yes, a common misconception with water rockets is that they are more efficient with convergent-divergent nozzles.

I've managed to push a pop bottle to a pressure of 120 psi before, albeit I reinforced the bottle with plastic around the outside. However in retrospect it was a very bad method of reinforcement, and I expect that the bottle used could have withstood the pressure.

If you want to hold 120 psi with a small metal tank, it can be incredibly thin. The problem with this is that it is very hard to form the metal this thin. Realistically, it is possible to make your pressure tank as thick as the plastic in a pop bottle, which will make it too heavy.

One of the most important things about making metal tanks is that they are much, much more dangerous than plastic bottles. If the tank explodes, you'll have deadly shrapnel going everywhere, and cut metal is very sharp.

Yes, chutes are crucial.

I'm on our school's FRC team. I can't wait for school to start up again.

They sell your small lightweight pressure vessels in stores all over the world. They come in two different kinds. Some of them hold Nitrous Oxide for whipped cream dispensers. Some of them hold Co2 for airguns. They look like this.

The plus side is they are safe and pre-filled.

[dharak1]

Not trying to derail the thread or anything but what would happen if I used my spare whip-its inside of a pellet gun? Do they have different pressures? Does the density effect how the gun fires? I suppose these are kind of related to how he chooses his gas.