KSP Community CubeSat

[Nicholander]

I just finished revising my SketchUp design again:

Anything still wrong with it?

[henryrasia]

That is... Really cool! Deljr15, if that's a sketch I can't even think how good's your serious business!

About 3D/CAD software, I've heard of Autodesk inventor, which has a trial version, is there such a thing for solidworks? Like an education version? Another thing, sketchup doesn't seem to be geared towards mechanical engineering, though I'm not really experienced with it so I might be wrong.

That's an awesome design, deljr, though I disagree the wheel should be on the side. If it were dead center it'd have less problems stabilizing. We'd just need to figure out how to balance it with the other components. The spheres look unnecessarily large and wasteful of space ( no offense). In my mind it'd be one single tube section with the inner walls divided into sections for the moss to grow. Finally, wasn't that going to be sealed? If so we wouldn't need an opening as it'd be assembled in advance, locked into place.

Finally, about the shape of the magnetorquers, Mbobrik posted a while back some images of how they would fit inside the cube, and it's kinda like three circles inscribed at different orientations. I don't see how they could serve as heaters, though...

I'm really excited for how this is coming along! And welcome all newcomers! As always any help helps

- - - Updated - - -

That is... Really cool! Deljr15, if that's a sketch I can't even think how good's your serious business!

About 3D/CAD software, I've heard of Autodesk inventor, which has a trial version, is there such a thing for solidworks? Like an education version? Another thing, sketchup doesn't seem to be geared towards mechanical engineering, though I'm not really experienced with it so I might be wrong.

That's an awesome design, deljr, though I disagree the wheel should be on the side. If it were dead center it'd have less problems stabilizing. We'd just need to figure out how to balance it with the other components. The spheres look unnecessarily large and wasteful of space ( no offense). In my mind it'd be one single tube section with the inner walls divided into sections for the moss to grow. Finally, wasn't that going to be sealed? If so we wouldn't need an opening as it'd be assembled in advance, locked into place.

Finally, about the shape of the magnetorquers, Mbobrik posted a while back some images of how they would fit inside the cube, and it's kinda like three circles inscribed at different orientations. I don't see how they could serve as heaters, though...

I'm really excited for how this is coming along! And welcome all newcomers! As always any help helps

[MBobrik]

Lots of really pretty pictures.

Yeah, but we need moar samples. Mazon Del wrote about N x 3 samples each cca 1.5 cm x 1.5 cm x 1.5 cm, where the higher N the better.

I was thinking about somehing more compact with less unused space between the samples.

Something along the lines of

, where the sample cylinder is more or less one piece of molded acrylic, and gears and servos and support go inside the cylinder.

Edited December 15, 2014 by MBobrik

[Nicholander]

So your just gonna ignore my SketchUp thing? Okay then... Also, did you make that MBobrik, or did you find it off the internet?

Edited December 15, 2014 by Nicholander

[Mazon Del]

MBobrik: What you put there is more in line with what I was imagining for the containers. We don't actually need to have the internal walls/separators ourselves. The little capsules that the moss will be in will be their own containers. The inside of the ring only needs to have a method of securing the capsules. This might be through a screw-hole or a clip of some sort. The camera unit itself would be on the inside facing 'out' and a second one would be perpendicular to the ring facing 'through' it to give us a 'from the side' view as well. This technically means that for any given moss sample, we will either classify it as a Top, Side, or Both plant. For the plants that are growing perpendicular to gravity, we will want the top camera (and maybe every several days an image from the side). For the ones growing with/against gravity, we will want the camera from the side (and maybe occasionally from the top). For some we may specifically want both (imagine a pure sideways crossbred with a pure up).

[Mazon Del]

I just finished revising my SketchUp design again:

Anything still wrong with it?

That looks pretty good!

Only thing is that I am not 100% sure just yet the actual physical dimensions of the space for the samples. I'm waiting on some feedback from Luis. Chances are it is only going to end up 1-2 centimeters wide. I'm not 100% certain how to scale that is.

Incidentally, if the ring itself is a ring without spokes (so a ring, not a bicycle tire), then we can fit additional equipment INSIDE the ring. Depending on the form factor of our batteries or whatever, we could easily fit them in there as well.

[MBobrik]

MBobrik: What you put there is more in line with what I was imagining for the containers. We don't actually need to have the internal walls/separators ourselves. The little capsules that the moss will be in will be their own containers. The inside of the ring only needs to have a method of securing the capsules. This might be through a screw-hole or a clip of some sort. The camera unit itself would be on the inside facing 'out' and a second one would be perpendicular to the ring facing 'through' it to give us a 'from the side' view as well. This technically means that for any given moss sample, we will either classify it as a Top, Side, or Both plant. For the plants that are growing perpendicular to gravity, we will want the top camera (and maybe every several days an image from the side). For the ones growing with/against gravity, we will want the camera from the side (and maybe occasionally from the top). For some we may specifically want both (imagine a pure sideways crossbred with a pure up).

Well, let's turn it around. Why do we need the capsules, when the moss can be directly in the cylinder sections themselves, and orientation of the moss from/to/perpendicular to gravity achieved simply by positioning and orienting the moss inside the compartments ? Because we won't have two cameras. The problem is depth of field of the camera. To have depth of field 1.5 cm long, the camera lens must be at least 15 cm from the sample. So the camera must be inside the ring, and the image bounced once through a mirror on the opposite side of the sat. Otherwise the camera has to sit on the other side itself, which will give us cca 8 cm camera distance and only 5 mm of depth of field, which means the sample would have to be only 5mm thick, otherwise only first 5mm of it would be in focus and the rest would be blurred.

[Nicholander]

Also, if you're wondering, the pressurized moss area is 9.4 Cm in diameter and 3.5 Cm tall.

[Nicholander]

...the camera lens must be at least 15 cm from the sample.

Wait, 15 Cm? The CubeSat is only 10 Cm, so how the heck are you going to do that? Unless you meant 1.5 Cm.

[MBobrik]

So your just gonna ignore my SketchUp thing? Okay then... Also, did you make that MBobrik, or did you find it off the internet?

Why ignore ? it is one of many iterations we will go through. And yes. I made it in SketchUp, learning how to use it in the process. It took me almost 2 hours, mostly learning the GUI and controls, but the second time I believe I could do it in a few minutes.

[MBobrik]

Wait, 15 Cm? The CubeSat is only 10 Cm, so how the heck are you going to do that? Unless you meant 1.5 Cm.

That's why I said we have to bounce the image through a mirror on the opposite side of the sat. So the path of the light will be from the sample, to the mirror on the other side, that will be like 7.5 cm, and then back to the camera which will be located inside the inner empty space of the ring, this way doubling the distance.

[deljr15]

MBobrik: What you put there is more in line with what I was imagining for the containers. We don't actually need to have the internal walls/separators ourselves. The little capsules that the moss will be in will be their own containers. The inside of the ring only needs to have a method of securing the capsules. This might be through a screw-hole or a clip of some sort. The camera unit itself would be on the inside facing 'out' and a second one would be perpendicular to the ring facing 'through' it to give us a 'from the side' view as well. This technically means that for any given moss sample, we will either classify it as a Top, Side, or Both plant. For the plants that are growing perpendicular to gravity, we will want the top camera (and maybe every several days an image from the side). For the ones growing with/against gravity, we will want the camera from the side (and maybe occasionally from the top). For some we may specifically want both (imagine a pure sideways crossbred with a pure up).

If we are going to put capsules inside of a ring like that it will in effect create the same wasted space of my design.

On the topic of capsules. If you can give me some basic shape/dimensions. I will revamp that sketch.

If the acrylic can withstand atmospheric pressure that may be a better option.

Edit:If the acrylic can withstand atmospheric pressure without being too heavy. That may be a better option. Also got to think about mounting and a drive system.

Edited December 16, 2014 by deljr15

[deljr15]

I really like that design! Simple, light, and probably doable without lubricants, which is kind of important.

Its probably even possible to extend the part of the axle that goes toward the wall such that you can fit a heater and the batteries behind the dishes. Again, that way you'd help balancing and heat up both things that need heating the most with a single heater.

What I mean with balancing: If I remember correctly the center of mass must be within 3cm of the center of volume. By putting the battery outside helps counter-weighting (is that a word?) the electronics that might be on the other opposing wall.

On the topic of balance we will need to have all the parts to even begin to attempt. I agree that we will probably need more mass behind it.

That is... Really cool! Deljr15, if that's a sketch I can't even think how good's your serious business!

Thanks

About 3D/CAD software, I've heard of Autodesk inventor, which has a trial version, is there such a thing for solidworks? Like an education version? Another thing, sketchup doesn't seem to be geared towards mechanical engineering, though I'm not really experienced with it so I might be wrong.

Yes there is 29 day trial is available on their website. Also they have a free program called "eDrawing Viewer" that can open and play with solid files.

That's an awesome design, deljr, though I disagree the wheel should be on the side. If it were dead center it'd have less problems stabilizing. We'd just need to figure out how to balance it with the other components. The spheres look unnecessarily large and wasteful of space ( no offense). In my mind it'd be one single tube section with the inner walls divided into sections for the moss to grow. Finally, wasn't that going to be sealed? If so we wouldn't need an opening as it'd be assembled in advance, locked into place.

I'm not sure what you mean by dead center. It is currently dead center about the axis of rotation.

[deljr15]

The problem is depth of field of the camera. To have depth of field 1.5 cm long, the camera lens must be at least 15 cm from the sample. So the camera must be inside the ring, and the image bounced once through a mirror on the opposite side of the sat. Otherwise the camera has to sit on the other side itself, which will give us cca 8 cm camera distance and only 5 mm of depth of field, which means the sample would have to be only 5mm thick, otherwise only first 5mm of it would be in focus and the rest would be blurred.

How about a lens fixed to the camera?

That's why I said we have to bounce the image through a mirror on the opposite side of the sat. So the path of the light will be from the sample, to the mirror on the other side, that will be like 7.5 cm, and then back to the camera which will be located inside the inner empty space of the ring, this way doubling the distance.

I don't see mirrors working in a low budget sat. The alignment is too mission critical and far too easy to get knocked out. Also having vision tunnels in the sat will waist far too much space.

I can't seem to find the quote where someone mentioned putting the camera inside the ring. On that topic whats wrong with one of the corners looking in? (Other then DOF issues MBobrik brought up)

[Mazon Del]

Well, let's turn it around. Why do we need the capsules, when the moss can be directly in the cylinder sections themselves, and orientation of the moss from/to/perpendicular to gravity achieved simply by positioning and orienting the moss inside the compartments ? Because we won't have two cameras. The problem is depth of field of the camera. To have depth of field 1.5 cm long, the camera lens must be at least 15 cm from the sample. So the camera must be inside the ring, and the image bounced once through a mirror on the opposite side of the sat. Otherwise the camera has to sit on the other side itself, which will give us cca 8 cm camera distance and only 5 mm of depth of field, which means the sample would have to be only 5mm thick, otherwise only first 5mm of it would be in focus and the rest would be blurred.

The capsules are a relatively standardized system to grow small amounts of moss inside, plus they can incorporate additional lenses to allow for the user to be able to use the naked eye instead of alternate lens systems. Basically the cover (injection molded plastic I think) is shaped to have a lens in it that provides some magnification. I would need to ask Luis about how easy it is to specify the magnification level with this. Alternatively, if we decide we do not need them, we can do a lens-less design. In the early days of this thread we found some links to a few electronic microscopes that were very small form factor devices that seemed to be able to look at objects immediately in front of them.

Ideally we want to reduce the optics requirements as much as possible.

Incidentally, having only 5mm for depth of field is pretty much perfect. The starting moss is likely to be approximately a thin paste on the bottom of the capsule (nutrient gel, plus the starting moss material). It will grow into the 5mm capsule over time. Probably about the 2-ish weeks of the experiment. More than enough time for us to get the information we are trying to find.

As far as why we want to have the mosses oriented all in the same direction despite the different growth directions, it reduces the variables under test and we don't need to care about the possibility that over the period of the experiment the moss could slip 'down' its container.

I would say it is far too early to declare with certainty that we won't have two cameras.

deljr15: Right now I do not actually have the specs for the capsules because I was told they can be made to order to our needs. So far it seems that our needs are driven by our rate of image downloading. No reason to cram in a bunch of moss if by the end we will only have been able to take the pictures required for a third of what we put in. Right now I am waiting for Luis to respond with how often a given moss plant needs to be imaged for useful data.

I am getting the sense (but Luis has not stated so in these literal terms) that technically a single image at the conclusion of the 0.1G level and another at the 0.3G level should be sufficient for a given plant (if it has just enough resolution to tell the difference between the rather large sized cells of the moss, then he can sort of back-calculate the information. But this is just a tentative opinion of mine based on some information he had said the first time we met. It could turn out that we need images every several hours for this to work. I don't know just yet.

Speaking of our data rates, have we managed any updated estimates on what sort of data speeds we might hope to get with the various bits of hardware under consideration?

When I get home, I'll toss in a solidworks mockup of what I have in mind.

[deljr15]

So what kind of capsule are we talking about?

Petri dish

test tube

smaller sphere

[Mazon Del]

If I remember right, think a tiny petri dish with a cover that has a dome on it. (It was several months ago that I saw it.)

[deljr15]

So tomorrow morning I will draw up a petri dish with a OD of 1.5cm and a inside space 5mm tall

I can't seem to find the quote where someone mentioned putting the camera inside the ring. On that topic whats wrong with one of the corners looking in? (Other then DOF issues MBobrik brought up)

Well I can answer myself on this one... that would put the camera underneath the sample. So I need to think about getting that camera inside.

[MrOnak]

Woah, lots of new posts. Just a quick note before I run off to work, since someone mentioned acryllic: You want to be careful with outgassing, especially since you have a camera lens right in front of it which may get coated with the outgassed stuff.

I didn't have proper time to find plastics in them but both ESA and NASA have extensive outgassing databases.

Once a suitable container is found I'd say

1. get one or a few
   
2. put moss inside
   
3. let it grow at in the fridge for a week or so
   
4. then one day put it in dry ice for 45 minutes, then heat it up with a 100 Watt bulb for another 45, rinse repeat for a day
   
5. bring it back to a lab and see if it survived that
   

If not - time to figure out insulation.

[Newt]

Many posts indeed. Does putting the camera on the ring directly really make sense? The issue of how exactly to arrange the camera system seems to have been discussed at length already, and it seems that that really does not make much sense at all. The ring should rotate so the camera can view the samples one by one, and directly into their viewers.

There was also mention of using several cameras, rather than one. As I saw this, it could be a backup measure--not a means of getting more data-- as if we have one camera only, and that gets damage or fails to connect, or whatever else, the mission becomes, well, pointless.

Down link speed is rather important too, but more for the question of image size and number. This seems to vary quite a good deal by the radio we get (i.e. how much money we spend on a transmitter). Do we really need to talk back up to the satellite very much? It does not seem like we need more than maybe to ping it it get it to send down the images. If that is the case we could probably set up a much simpler comm system than otherwise. Tomorrow I plan to read some more into exactly how fast we may be able to gather data from the sat, and how to talk to it generally, so I may be able to talk more about this then (if I have access to the forum).

Thanks for the 3D work, that really helps. I will be sure to get that viewer so I can look at the models myself more closely.

[MBobrik]

@camera.

There is a mathematical formula that connects camera params like chip size and resolution, lens focal distance and aperture with depth of field and distance of the viewed object from the camera lens, while the focal fistance and aperture themselves are constrained by required field of view, chip light sensitivity and wavelength of the light itself. And for rasp pi camera module the lens parameters are as fixed as chip properties because the lens is directly in the package with the chip. So technically, the minimum required distance between the sample and camera lens is a fixed function of the required depth of field.

If we need 5 mm depth of field, the camera lens has to be at least 80 mm from the sample.

If we need 15 mm depth of field, the camera lens has to be at least 150 mm from the sample.

Adding microscope optics to the rasp pi cam , or even using a handheld microscope instead, would not help because microscopes have depths of field fractions of a milimeter at most.

@using mirror

My idea was a fixed mirror glued directly to the sat frame so that it won't lose alignment unless the sat itself is bent out of its shape. But in this case we are screwed anyway, aren't we ? It would of course require free space in-between, but that would be only marginally bigger than what will be needed w/o mirror, because to get at least the 5 mm DOF, the camera has to be at least 80 mm from the sample, which for all practical purposes means on the other side of the sat as the sample.

@data rate constrains

A few pages back, I computed a rough estimate - we can transfer cca 50 images per one communication window, which occurs ~ 2x per day per ground tracking station. So if we want to take picture each hour, and have some redundancy, we need one ground tracking station per 3 samples. Of course we can increase this by increasing the picture taking interval or/and decreasing image resolution. Or by adding moar ground tracking stations, which, given they are just ham radio with a good antenna, should not be all that difficult.

Edited December 16, 2014 by MBobrik

[Newt]

Mirrors and other optic systems also make me nervous. A single mirror, is not too much, but we should bear in mind that this thing is going to be horribly shaken and bounced around before the experiment can start. There is a possibility that the mirrors will be knocked out of alignment then, and we need to be sure that cannot happen if we go that route. This could even happen with the glue, if we do not fix the mirror down well enough.

I will look for that data constraint page, thanks MBobrik.

[deljr15]

Well here is a drawing for a sample container. Mazon Del can you give your input.

If this container is along the lines of what your thinking I will revamp my design at lunch.

Edited December 16, 2014 by deljr15
Updated Link

[MBobrik]

@Mazon Del

I would like to reiterate my question. Can we completely do away with the sample containers and plant the moss directly into the compartments of the sample holding cylinder ?

And I add another question. Are/can the samples be sealed inside their compartments or containers, or do they need to share (a presumably much larger) volume of common atmosphere ?

[deljr15]

@Mazon Del

I would like to reiterate my question. Can we completely do away with the sample containers and plant the moss directly into the compartments of the sample holding cylinder ?

And I add another question. Are/can the samples be sealed inside their compartments or containers, or do they need to share (a presumably much larger) volume of common atmosphere ?

The only problem I see with doing away with sample containers is sealing the larger container.