Re: Can failed NASA missions be launched again?

Douglas,

There is usually a significant difference between what we call flight hardware (FH) and engineering hardware (EH). Since the space radiation environment is a bit hard on active devices, the parts procured for long duration, geosynchronous or deep space missions are radiation-hardened versions of what is normally found for terrestrial applications. As you might expect, these devices cost significantly more than commercial off the shelf (COTS) parts. This cost difference can be as much as 2 orders of magnitude per device. We typically refer to COTS as "crap" off the shelf, in case you were wondering. COTS devices cost a significant amount to qualify for space and often fail, thus we prefer rad hard (RH) where possible.

For particularly difficult units, an EH version will be made to test out the circuits very rigorously prior to flight and to identify hazardous/mission threatening conditions that need to be avoided to ensure mission success and be available should something go wrong with the FH version. The EH version will typically use COTS parts that have the exact same electrical characteristics that the RH version at a small fraction of the cost. Only very rarely is there a flight spare FH version, usually for reasons of cost. A neat example is the engineering unit for the Hubble wide field planetary camera 2 (WFPC 2). During operations, the shutter assembly sent back an error command that made us think the shutter blades had jammed together (two moving blades opened or closed one at a time to provide exposure uniformity). Careful inspection showed that the encoders had a slot that was degrading the light signal to the photodiode sensor. A quick test revealed that the degradation of the light looked much worse with the slot than with no slot as per the qualification testing. A program patch was sent to Hubble and WFPC 2 was back on line in less than a week. This was only possible because we had the engineering unit to learn from. (That was really hard and really fun to figure out.)

However, there have been occasions where a spare unit that was a full flight spare has been used for a different mission. A fine example of an attempt to do this was TRIANA (for those that follow this business, the derisive nickname for this project was GOREsat). It had the spacecraft bus from a canceled mission, the slight spare sun sensor from another mission, a camera that needed a little modification to go from EH to FH version, and on and on. Politics killed this particular bird as we found out that you can only go so fast when building a spacecraft. TRIANA was never launched and is in an inert environment storage container two buildings from where I am currently sitting. I fully expect that it will either be flown as is later or cannibalized for parts at some point in the future.

As for costs, that is of course an extremely variable thing. If there were full flight spares of each instrument, then only the spacecraft bus needs to be built. Launch costs are fairly static as NASA tends to write launch contracts that are worded something like: we want 10 of company X's Type Y rockets over a Z year period and we'll let you know when we want them. This saves us a huge quantity of money and generally does a great job of keeping launch costs under control.

In general, for science missions, unless there was a Congressional mandate for a particular mission, if the launch vehicle fails, that's it for the project. The exception being satellites that are part of an ongoing series like the TDRS birds (our data relay satellites), NOAA weather satellites, and the occasional earth observer. The main reason for not automatically doing the bird over is that so much changes between the start of a project and its launch, that the bird can be significantly improved before the next launch so you end up redesigning the instruments anyway.

The one thing we did to help us control costs and try to give easier access to space is create certain classes of science missions SMEX (small explorer) and MIDEX (mid-size explorer) that have a nearly uniform spacecraft bus on to which you attach your instruments. The idea being that the solar cells are more or less uniform, the power distribution is the same, the guidance is the same, the communications are the same and the only things that change are what's glued to the outside of the bus. This also saved us a huge quantity of money and provided much greater access to space for science investigators than NASA had previously. Cassini is the last of our "all the eggs in one basket" type missions that spent on the order of $1 billion to get everything out of the mission possible with the absolute minimum of risk. By comparison, the SMEX and MIDEX missions are often well under $100 million each. The trade off was the acceptance of risk of mission failure. (Better, faster, cheaper usually meant that you could have any two at the expense of the third. The odd thing is that for the most part, this worked well.)

To answer your last question, no one would fund a project that planned for failure the first time. We plan to succeed every time, so we only bid for one launch.

Thanks for a fun question.

Scott Kniffin
Radiation Effects and Analysis Group
NASA GSFC