They didn't think it could be done.

Unbeknownst to me, when I started on this program at Lockheed, the Program Manager thought the mechanical requirements were too great. He was partially correct, in that it couldn't have been done with the software and process that were in use at the time. Thankfully I was able to get approval to update both for the program - not that the other divisions like it at first.... Prior to this program there was no electonic concurrent engineering.

Pictured above is the completed "Doghouse" - the antenna payload without the reflectors. Please compare it to yesterdays electronic picture. The major difference is that the feeds are stowed. The installation of the two big feeds happens in two major steps. First we install them to the structure in the deployed position aligned to a few thousandths of an inch. Deployed because that is how they are used and the most critical position. A slight miss and in stead of covering China it is pointing in the Pacific. Next we stow them up and "hope" they fit to the mounting structure - the tolerance for fitting was tiny.

This is when the concurrent engineering comes into play. In the past when components were installed it was often found that some other program had changed design or location and there was now an interference. Something would have to be modified and possibly retested - never mind the documentation.

I was working closely with Frank (arms wide open above) when we first stowed the feeds. They were much more complicated, heavier and larger than any other feed structures. Usually the antenna is a 12 pound composite reflector. These were over 10 feet tall and 150 pounds. As the feed completely stowed - it happens very slowly - we looked at each other, and Frank had a look of amazement in his eyes. "I've never seen anything fit together like that in my 15 years here" was what he said. That is what CAD is most suited for.