It's a pity that ISEE-3 had run out of nitrogen.
Yeah. This was a big mystery to me, as I couldn't figure out how the N
2 could leak while leaving the hydrazine (N
2H
4) behind. The N
2 was in the propellant tanks, not in separate tanks, with no bladders separating it from the hydrazine. (Bladders are often used when a tank has to work in 0 G, but the spinning ISEE-3 induced fairly substantial artificial gravity in its tanks.)
I suggested the possibility that the hydrazine was simply frozen (it freezes at a relatively high +1C). They'd already thought of that, and turned on the tank heaters to ensure that the hydrazine was liquid. That didn't help. They were also careful to select only one of the two banks of propellant tanks at a time so that if one contained a leak, you wouldn't empty the other bank through it.
Finally Dennis explained the leading hypothesis to me. It was apparently known to the people who built the propellant valves that N
2 could diffuse very slowly through the seals (some form of Teflon, I think). If it was indeed molecular diffusion, this could explain why the larger N
2H
4 molecules were left behind.
I don't remember what convinced them that the hydrazine was still in the tanks; I'll have to consult my email on that.
I made an observation that I still haven't explained, though I haven't worked on it for some time. My role in the project was to demodulate and decode the telemetry (this is sort of a specialty for me). Someone else noticed a very small frequency modulation in the carrier frequency at the spacecraft spin rate. At first my only concern was to keep it from upsetting my carrier tracking loop, but then I began to wonder
why it should be present in the first place. Obviously it was caused by the S-band antenna being displaced from the spin axis, but it was supposed to be directly
on the spin axis; I estimated an offset of a few cm. But the mass properties document showed the spacecraft to be exquisitely well balanced, so why was this happening? Could something have taken a chunk out of the spacecraft in a non-vulnerable spot, perhaps during the 1985 comet encounter? Were the propellant tanks unequally filled? Or was it just the normal result of a small dynamic imbalance? I still have to crunch the data and see, but like a lot of the people on the project we were kinda burned out by the intense effort we all put in prior to the encounter.