KSP Community CubeSat

[Nicholander]

Also, how will we get the CubeSat to start spinning once it's in LEO?

[K^2]

Also, how will we get the CubeSat to start spinning once it's in LEO?

Should be possible with magnetotorquer. That's one of the cases where I could use axis tumbling to actually make the sat do what I want. But I'll need to do some sims to make sure.

Basically, I want a principal axis running along the length of the sat, and magnitotorquer across. Torquing it towards parallel and anti-parallel with Earth's field will cause sat to spin along longitudinal axis. Axis tumbling will take any spin around mag field lines into longitudinal or radial spin, and I can adjust these with the torquer using a modified PID.

Alignment with the sun can be done using a second magnetotorquer. Once the sat is spinning, it will work like a gyro. So a longitudinal torquer will cause rotation about the field lines, allowing the sat to orient itself towards the Sun.

So the entirety of attitude control consists of a pair of coils with ferrite cores and amps to drive them.

But again, I'll check that it all works in simulation first.

Edited July 18, 2014 by K^2

[dharak1]

Earlier you said we may need to balance the stuff inside to prevent it from wobbling. What level of balance would we need? Would the weight at each end depend on the chips on the boards or could we be a dozen or so grams off?

[christok]

What I mostly need to know from someone who's bio savvy, is what's the minimal diversity needed for a closed system to survive for up to half a year? What I need in there to get the ball rolling? And what temperature ranges should be maintained? Ideally the temperature range should be as high as possible. That will probably affect the choice of the species.

When you lock microorganisms in a closed space, there is an initial extinction event as the local ecosystem goes towards a new equilibrium (the process is called "succession") but then the microbial density often goes up*. In general, the simpler organisms tend to survive more while the more complex organisms are likely to die off during succession. Over time the culture becomes senescent and mostly dies off, a process which unfortunately can take vastly different amounts of time for seemingly identical cultures**. However, the culture will survive essentially indefinitely if the entire range of metabolic features is present from the start***.

A quick trip to your local pond can get you everything you need, although you want to be very careful in handling the resulting culture. A culture of potentially pathogenic organisms is far more dangerous than the few you had contact with when collecting the samples. This may make an ISS deployment problematic. Which is sad, because an ISS-launched cubesat can sometimes be recovered for ground-based study.

In a monoculture, the process of growth and die-off is divided into five phases, known as the "lag phase" (initial adaptation in which bacterial numbers are stable), "exponential growth phase", "stationary phase", "death phase" (in which bacterial numbers decrease exponentially) and "survival phase" (the long tail of the death phase, theoretically never reaching zero and not mentioned separately by all sources)*.

If we think our sat could survive for months, it might be good to go with some variation of a standard ecology experiment called the "Winogradsky column". (It is a diverse ecology in mud and nutrients, with aerobic and anaerobic microorganisms, some autotrophic and others heterotrophic, and possibly some plants.) The gravity gradient will be extreme with such a small centrifuge so I'm not sure how accurate it would be but any data is better than what we have now. In particular, I am concerned about the fact that plants would be at the top where the gravity is lower and that we might not get a long enough column for the anaerobic layer to survive. I'm not sure if any culture could survive very long without that anaerobic layer.

* M.V. Kilgore, Jr., A.T. Mikell, Jr. & D.L. Pierson, Microbiological issues of space life support systems. (In S.E. Churchill, ed. Fundamentals of space life sciences. Vol 2. 1997. Krieger. ISBN 0-89464-051-8.) pp. 289, 297

**R.B. Thompson & B.F. Thompson, All lab, no lecture: An illustrated guide to home biology experiments 2012. O'Reilly. ISBN 978-1-449-39659-6. p 88.

***M. Nelson, Bioregenerative life support for space habitation and extended planetary missions. (In S.E. Churchill, ed., Fundamentals of space life sciences. Vol 2. 1997. Krieger. ISBN 0-89464-051-8.) p. 318.

[Mazon Del]

Later today (after lunch) I am going to send an email to Mary Musgrave about our interest in the project to see what information or aid she might be willing/able to provide.

Additionally, I know that for $7K we can get 21+ weeks of work out of a group of roughly 3 students with additional faculty advisor (professor) from Worcester Polytechnic Institute. They have a display showing a science package they managed to get up on the space shuttle once, and they are quite close with NASA as a result of running various NASA tech challenges, so they are not strangers to this.

I might hit up a professor or two that I know that could be interested in this.

Edited July 18, 2014 by Mazon Del

[Rakaydos]

Ss far as getting a biomass through flicht and null g... what about an expandable habitat, which compressed id completely filled with water, but once in orbitvand spinning it is inflated eith compresded air to provide a mixed environment.

Thoughts?

[Creature]

* M.V. Kilgore, Jr., A.T. Mikell, Jr. & D.L. Pierson, Microbiological issues of space life support systems. (In S.E. Churchill, ed. Fundamentals of space life sciences. Vol 2. 1997. Krieger. ISBN 0-89464-051-8.) pp. 289, 297

**R.B. Thompson & B.F. Thompson, All lab, no lecture: An illustrated guide to home biology experiments 2012. O'Reilly. ISBN 978-1-449-39659-6. p 88.

***M. Nelson, Bioregenerative life support for space habitation and extended planetary missions. (In S.E. Churchill, ed., Fundamentals of space life sciences. Vol 2. 1997. Krieger. ISBN 0-89464-051-8.) p. 318.

I'll have to look if I can find that Churchill's book. It's a bit old but shouldn't be an issue. Thanks for bringing it up

I think the study should be tied to the current scientific field in some way and not just throw something random together. There are ways to do even molecular level assays with very light equipment using normal lateral flow strips and microfluidistic valves. It's just a matter if suitable assays for a meaningful target are available (commercially or via university) since we can't develop them on our own. You do need a camera to read the result though but it's required for any growth studies too. There are even quantitative assays that utilize cellphone cameras but I'm not sure if any are commercially available yet.

[Mazon Del]

The expandable environment route is a risky one. We cannot just have a plastic bag inflate and call it good. First it needs to be strong enough to take the pressure of the air (I think the air pressure at sea level is 14.7-ish PSI). Then it needs to be a material that will not degrade in the high UV environment or outgas into the enclosed space. Finally, if we want to have the plant chamber exposed to the sun, it is very difficult and expensive to get any sort of flexible window to do this. If we have a rigid container, we can simply have one of the walls be a plate of an appropriate material with the proper coatings to do the job while the method of attachment is very simple.

As far as weight balancing goes, I am a fan of having the probe be bilaterally symmetrical. Think an O'Neill cylinder http://en.wikipedia.org/wiki/File:Spacecolony3edit.jpeg except there are two 'land' pieces and two windows.

One way to help ensure the soil stays where it is (if we go the soil route instead of something like aqua/aero-ponics) is to get a mesh of some sort to place over the soil. If designed properly, it could be possible for the mesh to hold the soil in place, but when the plants begin to grow they poke through. There are meshes used for this purpose on recently sculpted land masses, though admittedly they are not going to be dealing with 5Gs and vibrations.

I am not super crazy about the tethered cubesat approach (where the cubesat separates and spins, connected by a tether). It will help with the 'gravitational' gradient definitely, but I am worried about the complexity required. If anything I think this approach ends up being even more dependent on proper balancing than a 'simple' axis rotation.

Though it would reduce the size of any given section as well as increase the mechanical complexity somewhat drastically, one possibility that occured to me is that we could actually break out the habitat into sections that deploy axially. I am attempting to think of a good way to explain it, but imagine if we had a 3U cubesat. The middle segment is command and control, the two ends each contain 4 habitats. Once in orbit, the system deploys the habitats where they hinge out unfolding from the center until the probe assumes a sort of barbell shape. Once it begins to spin, you now have 8 habitats all with the same gravity, each with its own plant payload (greenbeans in one, bamboo in another, potatoes, etc). Of course each habitat would likely be shaped like a triangular prism to fit this design. An additional downside would be that our sensor system would likely be reduced because we would be required to give each habitat its own sensor system. It would be way too difficult to have a probe that could move from one habitat to the next (a moving mass in a small rotating system would be a bad idea, plus the mechanical complexity). As far as another advantage, if we did it this way we likely wouldn't care too much about the orientation of the probe with relation to its spin. At least 4 of the habitats are likely to receive light.

[christok]

I'll have to look if I can find that Churchill's book. It's a bit old but shouldn't be an issue. Thanks for bringing it up .

Microcosm has it for $100 (http://www.astrobooks.com). I can assure you they have everything we need (since I bought so much of it already ) but you can always check if Amazon et al. have a better deal.

Speaking of which, we need to really start digging into the literature. I'm currently checking C.D. Brown's Elements of spacecraft design (2002. AIAA. ISBN 1-56347-524-3) to see what sorts of margins we will need in our initial design.

- The closest to our category of spacecraft of a completely new design needs to count on a 50% mass increase from the initial bidding stage. (p. 26)

- Our design should also not exceed 50% of the throughput capacity of the computer we choose. (p. 37)

- Our thermal design needs the ability to handle 35C higher than the expected temperature maximum we calculate and 35C lower than the minimum we calculate (10C on either side for the "flight allowable" temperature, another 10C for our qualification and an additional 15C for the difference between what we design for and what we qualify the satellite at). We should calculate the thermal requirements on a per-component basis rather than a craft average. (p. 37) I am particularly worried about thermal design. As this satellite will carry living organisms with definite temperature requirements, without the battery capacity for any significant heating on the night side and with a very high surface:volume ratio, we will need to take care.

- 40% margin on battery capacity. (p. 38)

- 100% force/torque margin on the worst-case performance of mission-critical deployables. (p. 38)

These numbers are of course not based on cubesats so we may have to fudge a little here and there but it's good to learn from what has already been done. Looks like we have our work cut out for us.

Deployables are of course always a problem on spacecraft, especially cheap ones, but I was thinking something regarding our solar panels. We can significantly increase our power budget if the panels point towards the sun. Perhaps we can initially have panels on four sides of the cube, facing outwards so as to provide a smaller amount of power as long as any of them face the sun. When we are ready to spin up, we rotate the craft so that one of the other two sides faces the sun. On that side, the panels are hinged (without any motors) and the opposite side is attached with a once-off release mechanism. We then release the panels and spin the satellite up, allowing centrifugal effects to extend the panels. I'm sure someone in the industry has already thought of (and used) this, does anyone have any info?

Edit: Mazon Del, I was paying so little attention I didn't even notice that's exactly what you proposed for the experiment. Not intended as an insult. I think it is better suited to the panels than the habitat but feel free to post counter-arguments.

Edited July 18, 2014 by christok

[Mazon Del]

Edit: Mazon Del, I was paying so little attention...

Unending fury upon you! Hehe.

Those percentages sound pretty good. Honestly it shouldn't be too difficult to find a processor that far exceeds our computational requirements while being electrically affordable.

I do agree though that thermal management is a bit of an issue. I can't help but wonder if maybe something to do to help with this, is assuming we went with the O'Neill Cylinder approach, that the hull of the craft has some insulation on it, and when we are heading into darkness, the solar panels close up again. These would of course be a sandwhich of solar panel, hull material, and insulation. As far as mechanism design is concerned, if we had the open/close systems be a wormgear driven linear actuator they would be slow but have massive torque behind them to ensure mechanical power.

As far as what parts move around, I don't care too much at this point, I'm just throwing up ideas as I have them.

I attempted to send an email to Musgrave, but unfortunately the only email I can find for her is a uconn.edu one, and it sends my emails back as undeliverable. Either because I cannot pass through, or because that email is incorrect. >.<

Also, I was slightly wrong on the cost at WPI. It averages $7,500 per student. But one student is still committed to 21+ weeks of work (estimate about 30-50 hours per week depending on how into it the student is) plus a faculty advisor.

[K^2]

We then release the panels and spin the satellite up, allowing centrifugal effects to extend the panels.

There are a couple of minor problems with that. First, it requires spin to be maintained. That might be fine, as attitude control I have in mind sort of assumes it as well. The other part is that during deployment, that will change moment of inertia tensor rather suddenly. That can be trouble.

These might be lesser problems than trying to articulate the panels, however. And by now, I'm also thinking that the craft should be a sunflower, with side panels deploying to face back, so that the craft can be oriented to expose both the eco-chamber and the solar panels to the Sun.

Alternative deployment can be done with a pair of servos at the corners. Each one can extend two panels adjacent to that corner. The 10g micro-servos will be able to do the job just fine.

[Mazon Del]

It is certainly possible to have a mechanical system in place such that when the panels are fully deployed they are locked in place.

However I believe the idea of actuated panels is better, mostly because it allows us to play tricks like closing the panels for heat conservation during night time (clearly we'd have at least one panel on the outside, just for safety in case the batteries run dry while on the dark side). And frankly, I never trust purely mechanical systems in a robotic system.

What could be interesting if we do go with the bilateral oneill cylinder and we have something like bean sprouts (something vinelike that tries to grab onto nearby stuff) we could set up a few guidewires to control the development of the plants and see how well they work given different 'gravity' geometries. An example being that if we had a guide wire cross straight through the middle and one of the sprouts grew along the length. After a certain point halfway up the plant, the 'gravity' vector changes so the top of the plant suddenly believes it is upside-down.

What I'd expect for plants near the edges of the growth trays is that they would start to grow upwards, but the 'taller' the plant gets the more it is pulled to the side of the area (towards the windows). As a result of the tip being closer to what it thinks is horizontal (remember, all outward directions are down in the spin gravity), it compensates and tries to grow towards the center of the cylinder, repeat. This actually would be pretty useful data to have if true. It could show just how the plant growth occures while in high gradient areas. Perhaps this could be utilized somehow for more efficient crop growth in space? Or maybe it is something that extra care is needed to avoid.

[christok]

There are a couple of minor problems with that. First, it requires spin to be maintained. That might be fine, as attitude control I have in mind sort of assumes it as well. The other part is that during deployment, that will change moment of inertia tensor rather suddenly. That can be trouble.

I assumed we maintain spin based on the experiment we have been discussing. I had assumed the craft would spin up very slowly, although I may be guessing wrong about how much torque you want from those magnetos.

These might be lesser problems than trying to articulate the panels, however.

I agree. We'll need hinges that won't cold-weld, for a start.

And by now, I'm also thinking that the craft should be a sunflower, with side panels deploying to face back, so that the craft can be oriented to expose both the eco-chamber and the solar panels to the Sun.

Like Mazon Del, I still feel we should have some solar panels pointing out in the packed configuration. That would work if we had a second joint, but it does of course add complexity if we move them with servos. I am also not comfortable with servos repeatedly deploying and re-deploying in flight; that's bound to fail over a multi-month timeframe. (Possibly necessary depending on our thermal conditions.)

I am also afraid of overheating in the sunflower configuration. If more of the craft is in the solar panels' shadows, that should improve our ability to heat-sink the experiment.

Alternative deployment can be done with a pair of servos at the corners. Each one can extend two panels adjacent to that corner. The 10g micro-servos will be able to do the job just fine.

However I believe the idea of actuated panels is better, mostly because it allows us to play tricks like closing the panels for heat conservation during night time (clearly we'd have at least one panel on the outside, just for safety in case the batteries run dry while on the dark side). And frankly, I never trust purely mechanical systems in a robotic system.

I want to avoid garden-variety motors as much as possible. They're highly prone to failure in spaceflight conditions and there are all sorts of issues with things like lubrication, cold welding, etc. (Non-spaceflight) COTS motors are out of the question. Anything with a decent probability of working (especially repeatedly) is expensive to buy and we have too high a chance of mission failure if we try and design it ourselves. If it is problematic to deploy the panels with spin, I have two alternative ideas. 1: We release them with (very weak) springs before spinning up. 2: We add a few small coils of wire and deploy with electromagnetic repulsion. The nice thing about number two is that we can also retract the panels easily for night-time insulation, although we will need a mechanism to keep them in position without draining power.

What could be interesting if we do go with the bilateral oneill cylinder...

I'd definitely want to put the biosphere in a (flattened) cylinder. I'm thinking that this will encourage currents of air and water to compensate for our very small biosphere by distributing nutrients and heat as well as help compensate for asymmetries in mass distribution about the axis of rotation.

What I'd expect for plants near the edges of the growth trays is that they would start to grow upwards, but the 'taller' the plant gets the more it is pulled to the side of the area (towards the windows). As a result of the tip being closer to what it thinks is horizontal (remember, all outward directions are down in the spin gravity), it compensates and tries to grow towards the center of the cylinder, repeat. This actually would be pretty useful data to have if true. It could show just how the plant growth occures while in high gradient areas. Perhaps this could be utilized somehow for more efficient crop growth in space? Or maybe it is something that extra care is needed to avoid.

Plants also grow towards the light. I think we should be more interested in how the organisms respond health-wise to partial gravity than the directions they grow in.

[K^2]

Mazon Del, we are talking about total of less than 400ml of volume. What you suggest sounds a bit too complex for that.

If there is budget for an inflatable, that would be something to try. But that presents all sorts of new issues and new ways for the whole thing to fail.

christok, good arguments. If centrifugal deployment is really that much more reliable, I can work around it in the spin-up procedure. It will be a very gentle spin-up either way, since magnetotorquers will take time to get the thing up to speed.

[indroth]

Base budget can be designed around an off-the-shelf MPU. Like I said, I'll look into some software error-correction options. Error-correcting memory is awesome, but it's pricey, and rarely includes anything more fancy than a parity bit. I can build an architecture with a co-processor whose sole job is managing memory in Z_p using polynomial error correction. But that will slow things way down if each read/write will have to go through a separate processor.

Perhaps a Byzantine Fault Tolerant system with multiple off the shelf microprocessors will suffice? This has the added advantage of providing failure rate and failure mode data for off the shelf electronics in LEO. At the same time, it does add size, mass, heat, power drain and software complexity to the controller, so it may not be worth the trouble.

[Mazon Del]

Oh yes, when I say O'Neill Cylinder for this, what I have in mind is that two opposing sides of the cubesat hold growth trays and the other two opposing sides are windows, all sides would of course be flat rectangles/squares.

As far as health vs direction, they can both be done at the same time. Really they both WILL be done simply because we are not going to be controlling every plant's growth individually so if we have a camera module we will be able to make the sort of inferences that I was discussing. With the addition of a single wire (if we have the right plants) we suddenly get to observe another effect to. I doubt the addition of a single wire will break the mass budget on us if we wanted to add it.

As far as motors and an open/close mechanism, if we can find a motor rated for the conditions we will be dealing with and metals that are acceptable it isnt that difficult to design a reliable mechanical system for opening and closing the sunflower's 'petals'.

As you bring up air currents, I actually propose we have a small fan for air circulation. One of the facts I got from Musgrave's documentation was that plants suffer from a problem NASA realized humans have. If a human is strapped down and sleeping (basically not moving) he will eventually suffocate on his own CO2 because it isn't mixing with the surrounding air. For plants it is the same thing, except they would be unable to reach CO2 because of all the Oxygen. This actually brings up another issue she spoke about. Depending on our plant choice, we would need some method of converting the oxygen to CO2. Otherwise our plants will kill themselves by removing what they need to breath entirely from the environment. One solution provided was to have a local storage of CO2 that you can use to keep the plants 'fed'. Unfortunately I am uncertain what view NASA and the launcher people have on the idea of compressed gas on board the cubesats, even in small amounts.

On another note, one of my specialties is robotic arms. It would be quite possible in the Earth-like environment of the interior to have a very small robotic arm to allow us to poke at the soil or plants or move our camera around. Given that this space is not subject to vacuum or freezing temperatures, standard COTS micro-mini servos should suffice.

Incidentally, what would be a fascinating thing to try to do if we had the space and mass budget, would be to try and design a mini-return module (difficult yes, but maybe not impossible for a small enough container). If we rigged it right, we could have the on board cameras constantly piping visual data to a small SD card (or something similar) in the module. If recovered we could add info to the discussion about what it looks like inside a craft as it re-enters. If we had the mini-arm, we could even save a sample of plant matter. Though of course, the main question is if we would be allowed to try rather than if we could make it happen.

[Mazon Del]

Mazon Del, we are talking about total of less than 400ml of volume. What you suggest sounds a bit too complex for that.

Thus the disclaimer about the complexity. I can make it happen, I've actually done something similar for a class project once. But I would be hesitant to do it on a spacecraft, simply because of the propensity for something to go wrong beyond our ability to fix.

[Idrryl]

This sounds absolutely amazing, a space game community doing actual science in space? I'm all for it

[Mr_Brain]

This seems absolutely awesome. I have read the entire thread, but voted for "mission to Phobos" in the poll before doing so – which was probably a little too early... Well, I guess it still works as an Ultimate Mission. If this project gets All The Funding. A lot of All The Funding.

Anyway, unfortunately, as nothing more than a teenager who's interested in space, I probably won't be able to contribute much except for some money. However, if/when this becomes a Kickstarter campaign at some point, I'll try to collect some additional donations at my school (just assuming the campaign doesn't end before my school vacation does) – I know a science teacher who, as far as I know, loves astronomy, I am quite confident he'd do his best to support this project.

On a more unrelated note, the students exiting the Austrian equivalent of a high school have to write a (please excuse this probably bad translation) "pre-scientific paper" from next school year onwards, which is basically something that remotely resembles a scientific paper, but on a much lower level. As I'm still looking for a topic (and I'll have to write it in one to one and a half years, so that's enough time for this to progress quite a bit, I hope), I might use this project as a topic, if it's allowed (by both you and the education system).

[Guest Fyre Flare]

Kickstarter start NOW!

[Rakaydos]

This seems absolutely awesome. I have read the entire thread, but voted for "mission to Phobos" in the poll before doing so – which was probably a little too early... Well, I guess it still works as an Ultimate Mission. If this project gets All The Funding. A lot of All The Funding.

I'm still working on the Phobos mission, in another topic, but my main limiting factor is hard data.

Even so, it's looking like Phobos wont be the first mission, even with *ALL THE FUNDING*. The first thing needed is to perfect the GTO->Lunar->Solar-> Earth maneuver chain, while demonstrating at least a year's endurance without the computer going bonkers from space radiation, and practice the mars aerocapture here at earth where we have better command and control.

Once that's down and we can relaiably acchieve it, then we can use all that, then use earth to reach for mars, aerocapture and aerobreak down to phobos, and attempt a landing. It's not a science mission, it's an engineering challange- to boldly go, so to speak.

[Nicholander]

I think that we shouldn't mess with deployable solar panels, i think static ones will provide enough electricity.

Also, Fyre Flare, hold your horses! We couldn't even put up a funding goal right now because we don't know how much this will cost!

[K^2]

I think that we shouldn't mess with deployable solar panels, i think static ones will provide enough electricity.

Sat needs to expose eco-chamber to the Sun, which means that side panels will be poorly illuminated. We could get more power out of 4 0.5U panels that deploy to face the Sun, then from 4 full 1U panels fixed on the four sides. In fact, with deployment, it might be enough to just have two 0.5U panels.

Seeing how solar panels are the most expensive part of the basic setup, I think it's worth extra complexity to try and keep the cost down.

[Nicholander]

Okay... I see. Also, K^2, how's that "shopping list" you said you would make earlier to show how much our Base Goal CubeSat would cost?

[christok]

Light-hearted and yet so true:

(The topic is chapter 13.1: Moving parts.)

"You really don't want to do that," the systems engineer explains, trying to remain rational.

Then anger.

"I thought you said you wanted a reliable, cheap satellite..." The designer imagines a black smudge on an otherwise perfect flight success record.

Then comes recrimination.

"I should have stuck with a simple career, brain surgery or President of the United States."

Finally, acceptance.

"This is going to cost you big bucks."

...

Nasties like flexibility, and poor alignment, leading to increased friction, jamming, and galling. It means force and restraint, momentum and potential deformation. Thermal inhomogeneity leading to dimensional instability. Vacuum welding and the problems associated with non-outgassing lubrication. Fine structures that can resonate and break. These are the lions and tigers and bears of spacecraft mechanisms.

-- R. Fleeter, The logic of microspace. 2000. Microcosm. ISBN 1-881883-11-6. p. 120.

I'm still cool with a hinged solar panel deployment by centrifugal effects. Especially if the hinge is somthing simple along the lines of a length of fabric or flexible plastic rather than cold-weld prone metal hinge. (We'd need to choose a good material that won't become brittle and break after launch but that should be doable.)