The Falcon 9 booster tank is made of aluminum lithium alloy. Welding it without forming hydrogen bubbles is tricky. SpaceX uses friction stir welding throughout.

I had to crop the welding jig from the photo because of ITAR restrictions. Let me just say that it is very clever, and the challenges are like building a ship in a bottle.

I have also collected some interesting details on the Falcon 1 flights so far:

Flight 1: The booster engine cutoff early, and the rocket went only one mile up. The video of what happened next has not been seen outside SpaceX. Without stage separation, the whole rocket tumbled back to crash 300 ft. from the launch pad, causing a gigantic fireball. There are no people on the atoll during launches, but there were plenty of cameras catching the action, and the smoke that enveloped the entire launch area. The remains of the rocket were collected from the reef at low tide.

Flight 2: Programming mistake in the upper stage led to improper feedback control loops and it wobbled out of control.

Flight 3: While the lack of stage separation has been discussed, it is a bit more interesting. The upper stage ignited inside of the interstage coupler... and that destroyed the fragile upper stage bell (which is as thin as a Coke can).

Flight 4: Perfect flight, but ironically, the first without a real payload. The Malaysians sat this one out, and will fly their satellite in Flight 5. The dummy payload in this flight was machined aluminum with a realistic mass profile and was named the RATsat after the initials of the employees who cranked it out over a weekend.

After entering orbit, they restarted the sustainer motor, a capability test that was not pre-announced.

[Update: This post and text were before I had any involvement with SpaceX. And the subsequent Flight 5 was a total success, with accurate satellite deployment]

Flight 4 is memorialized in the Space X Music Video.