This is actually a pretty good idea for warming up cold planets. Putting one in orbit around Mars, for instance, would significantly aid a terraforming project, is a smarter idea than my idea of dumping nuke waste in volcanic craters.<br /><br />Building one large enough to provide, say, 0.1 solar flux, should be sufficient to trigger a cascade CO2 outgassing sufficient to produce an earth-normal atmospheric pressure, with equatorial temps typical of temperate climates on earth.<br /><br />The object should probably orbit in a long orbit, of 20-30 days duration, like our moon. This would provide room for a martian skyhook, and provide a gravitational tide that may help restart martian tectonic processes.<br /><br />Given this, its orbit would likely be about 100,000 miles from the planet. At that distance, its emittance area would cover 3.14 x 4 x 10^10 miles^2, or just over 125.6 billion square miles. Given that the cross sectional area of Mars is (at 2,097 miles in radius) 12,560 miles^2, Mars would intercept 0.000001% of the emitted energy of such an object. In order for Mars to receive 0.1 solar flux (about 140 watts/meter^2), which is about 362,598,335 watts / mile^2, the artificial sun would need to emit 45,542,350,932,148,224,000.00 watts, or about 45 million petawatts. This power requirement demands 1,366,270,527,964 tons of uranium.<br /><br />The reaction would likely be self moderating based on the materials you put into it to start with. This is not unusual, there have been naturally occuring nuclear reactors found by geologists that burned out long ago, the most well known being one in Ghana that occured a billion years ago, which was actually four separate reactor 'cores' in different areas of a large uranium formation. These natural reactors burned for more than a million years each and had an average power rating of about 25 kw. <br /><br />The tonnage demands for such a fission star could be signficantly reduced if the fission core were surrounded by a refle