SEDSAT-2 Payload Design Notes 20070419
From SEDSWiki
SEDSAT2: Payload Study – Three proposed imaging missions
Contents |
Earth imaging
This idea has been baselined as the Primary Payload for SEDSAT-2.
This idea has been recommended by the Payload team as a possible primary or secondary payload for SEDSAT-2. Reason: Getting photos of the Earth does not seem like a very ambitious undertaking, yet produce a nice result. Pro: Modest requirements. Con: Relies on attitude control to a certain extent.
Purpose: To take visible-wavelength photos of the Earth.
Approach
The idea is to use an onboard camera to take pictures of the Earth, while in orbit. What kind of images to expect would depend on a number of factors; such as the orbit, the degree of attitude determination and control achieved, the inclination, the resolution of the imaging device, the focal length and quality of the camera optics.
Other cubesats have proven that nice pictures of Earth in the visible wavelength range can indeed be achieved. Looking at the "UXGA" sheet of the Preliminary Optics Calculations spreadsheet, one can get an estimate of what kind of area would be covered from a given altitude, and at what resolution, if the camera was aimed at nadir. The "SXGA" and "VGA" sheets give estimates based on 1.3 and 0.3 megapixel chips; it is worth noting that the resolution stays the same, but the area covered is reduced, which may be counter-intuitive.
A 1000mm focal length seems nice at first, because of the high resolution, but is probably a bit much in reality. It does not necessarily require a 1m long lens - in fact it can be a lot shorter - but the mechanical complexity increases, and factors such as light intensity, optical aberration etc become more significant. Also, while 2m resolution sounds tempting, atmospheric diffusion will reduce the practical resolution, while the small area covered will actually render the images useless for most purposes (bandwidth will prohibit covering a large area with a mosaic of high res pictures.)
Looking at nadir will give the highest on-ground resolution for any given setup, but pictures at an angle, perhaps with the horizon in view will probably be more attractive. The "Inclined" sheet takes altitude and a set of inclination angles to give an idea of what could be seen if the camera was tilted relative to nadir. The 500mm and 1000mm focal lengths have been omitted from this sheet, as they don't seem practical.
What the sheet says is basically that the angle need to be high regardless of focal length and altitute if the horizon is to be within the image. Even with a focal length of 4mm and an angle of 60 degrees, the Earth will cover 55% to 66% at the center of the image, for typical LEO altitudes, so almost any modest focal length will produce nice images. Short focal lengths can be realised using only fixed lenses, with no protruding/moving parts. Combined with a reasonable aperture this would limit the mass/volume requirements.
Conclusion
Earth imaging is a viable mission. It can be done almost regardless of focal length and camera resolution - even attitude control is not really necessary, provided spin/tumble rate is not extreme. Instead any attitude control can be exploited to decide what to take pictures of, while the optics can be engineered to specifications to get the most interesting/useful results attainable within the resource budget.
Deployment imaging
This idea has been recommended by the Payload team as a possible secondary payload for SEDSAT-2. Reason: Getting photos of the launch vehicle or other cubesats would be cool. Pro: Might be done as a "bonus" without adding much complexity to an Earth imaging payload. Con: Will require use of power before satellite status can be determined.
Purpose: To take visible-wavelength photos shortly after separation from the P-POD.
Approach
The idea is to use an onboard camera to take images shortly after the satellite has been deployed from the P-POD, hopefully getting a nice shot of the pod, launch vehicle and/or other cubesats.
While deployment of panels, antennas etc can not begin until 15 minutes after separation, the onboard systems may be powered on immediately following the release of the deployment switches, if the satellite is equipped with pre-charged batteries.
Using the spring plungers recommended by CalPoly (highlighted on this page, the satellite's relative velocity after separation will be about 3m/s, depending on friction, mass etc.
At 30 seconds after separation the distance would be roughly 0.1km, giving a resolution of 9cm if the camera is 2 megapixels, and the focal lenght is 4mm (standard focal lenght using only a primary lens). At 40mm focal length, which could be achieved without any protruding optics, the resolution at 0.1km would be 1cm (see "UXGA" page of http://protoforge.org/files/u/0k/23/sedsat2-optics.xls.) Regardless of distance, the probability of getting anything within the field of view would be roughly 1/9 and 1/900 respectively if the satellite is tumbling randomly.
Commencing too quickly after separation is not really desireable, as the tumble rate would not have to be very high to make an image blurred.
Conclusion
Successfully photographing the launch vehicle seems plausible. If the camera has a short focal length, giving it a relatively wide field of view, and a picture is taken once a second from 3 to 30 seconds after separation, the yield would in theory be three successfull photos.
Other missions
The cubesat AeroCube2 by Aerospace Corporation successfully took a picture of CP4 following launch on 2007-04-17.
Object imaging
This idea has NOT been recommended by the Payload team. Reason: Getting photos of objects in Earth orbit would be improbable, and results would most likely be useless.
Purpose: To take visible-wavelength photos of objects in Earth orbit.
Approach
The idea is to use a camera to take photos of other objects in Earth orbit. This could be space debris, satellites or the ISS. What objects could be seen would depend on factors such as their distance and velocity relative to the cubesat, their size and the resolution and focal length of the cubesat's camera.
NASA's ORDEM2000 software contains a database of orbital debris, that can be used to make predictions of the number and average relative velocity of objects of different sizes encountered during a year at given altitudes and positions. The software is meant to be used to calculate how often a satellite will be impacted, and by how large objects, but making some assumptions one can of course calculate how many objects will pass through the field of view of the camera, based on the aperture angle, distance from the camera and altitude of the camera.
For the purpose of this, it has been assumed that the camera would be spin stabilized and facing nadir, with one of CMOS chip's short edges facing the direction of travel. Looking at the "Objects" sheet of the Preliminary Optics Calculations spreadsheet available at http://protoforge.org/files/u/0k/23/sedsat2-optics.xls one can see that with a focal length of 300mm, a circular orbit at 750km altitude and a 2 megapixel imaging device, the average number of objects larger than 1m in diameter passing through the field of view would be approximately 15 per year. The average velocity of these objects, relative to the satellite, would be roughly 9.2km/s, resulting in a clearly impossible target exposure time of ~ 1/4500s.
If the debris direction of travel is evenly distributed, a guess of one in ten objects having a relative velocity that makes it possible to photograph it doesn't seem unreasonable. If that is the case, one object could in theory be photographed every eight months.
Conclusion
Given the low frequency of opportunities - which would be even lower at lower altitutes or with a shorter focal length - and the fact that at on the average fly-by distance a satellite of 2x3 meters would make a blob of 20 pixels, object imaging does not seem to be a worthwile mission.
