Beneficiation

An electrostatic ilmenite separator could process 15,000  tons of regolith per year working 10 hours a day, 150 days a year, and processing 10  tons of regolith per hour.  It would amass about 0.6 tons per ton-hr. of regolith, or in this case about six  tons.  It would use 1.2 kW/ton of input mass or about 12  kilowatts of electricity.  The electrostatic separator could output 100 kg. of ilmenite for every ton of input mass, or about one  ton per hour  in this case (1).

Thus, enough ilmenite could be extracted every hour to make 320  kilograms of titanium. That's 480 tons of titanium every year from just six  tons of beneficiation equipment upported to the Moon.

Extraction

The ilmenite must be reduced with hydrogen gas to yield water, titanium dioxide and iron.  The water is electrolyzed to recover hydrogen and gain oxygen.  Hydrogen leakage must be replenished by volatiles mining or ice mining in shadowed polar craters, if there really is ice in those craters and not just trapped hydrogen from the solar wind. 

A fluidized bed or numerous fluidized beds for the hydrogen reduction could be made on the Moon from blister steel made on the Moon.  Stainless steel would be needed for these so that it doesn't rust.  Until we develope methods of extracting chromite, probably by electrostatic separation, from mare soil, or find deep layers of chromite in central crater peak upthrusts, we must upport chromium for limited amounts of stainless steel for space applications like this. 

These fluidized beds would be rectangular because construction and welding will be easier that way. We will not have to roll steel into curved plates.  We will simply pour out molten steel plates in shallow compacted regolith sand molds outvac and cover the steel with slag to prevent evaporation into the vacuum.  Then weld them up and add webs in the corners for strength against internal pressure.  We could make all pipes from blister steel and electrolysis units too on the Moon, though we might need to upport platinum electrodes for the water electrolysis units. 

Making electrostatic separation units on the Moon should be possible also, as these are not too complex for a clever mechanical engineer with some steel, some tools and some electricity to work with. 

The output from the hydrogen reduction fluidized beds ( H2 redux units) will be TiO2 and iron particles fused together.  These will be ground, iron magnetically removed, and possibly treated with CO gas to form iron carbonyls to get the TiO2 purified.  The TiO2 will them be reduced to titanium and oxygen in FFC cells upported to the Moon along with calcium chloride flux.  Eventually we will have plenty of calcium and some chlorine from volcanic glass beads to make enough flux for the FFC cells.  We might even mine enough carbon to make electrodes for FFC cells made on the Moon. 
Titanium
1) H. H. Koelle  "Pilot Production at the Moonbase 2015"  pg. 11
<http://www.highfrontier.org/Archive/Jt/Koelle%20PILOT%20PRODUCTION%20at%20the%20
MOONBASE%202015.pdf>
David A. Dietzler, 2007
Titanium, Not So Simple

Do not fall into the trap of thinking that titanium is easy to produce and can replace steel on the Moon.  While titanium might have special uses and can be formed into many small complex parts by 3D stereolithography also called 3D sintering, it is much harder to work with than steel.  Alloys of titanium need aluminum and vanadium, but vanadium is extremely rare on the Moon.  Titanium is harder to weld and it has a high melting point-1800 C. so it will be difficult to find mold materials.  It cannot be heat treated to soften it and make it easier to work as steel can and it cannot be heat treated to harden it as steel can.  One of the beauties of steel is that you can cast a part, anneal it to make it soft and workable, then harden it to make the part strong by proper heat treatment.  You can also case harden steel, therefore you can make a gear or cam out of soft steel that has some flexibility for stress tolerance and case harden surface to prevent wear.  We can't do any of this with titanium.

To make titanium we need: 1) Regolith mining tractors  2) electrostatic ilmenite separators and power supplies for those  3) A fluidized bed with H2O condensors and electrolysis cells to recover H2 from H2O formed as well as O2 storage systems consisting of compressors and space radiators and LOX tanks 4)  hydrogen gas and H2 storage systems, probably compressed H2 storage tanks and compressors  5) Ways to separate iron from TiO2 from the fluidized bed.  Grinders, magnetic separators, acid leaching or carbonyl iron extractors and carbon monoxide gas  6) FFC cells that are fairly heavy to deoxidize TiO2.  These require calcium chloride and chlorine is rare on the Moon though calcium is abundant.  They also have large carbon electrodes.  7) Power supplies for the fluidized beds and FFC cells  8)  To make electrostatic separators, fluidized beds and associated pipe, tanks, etc., devices for separating iron from TiO2 and FFC cells on the Moon we will need steel so steel comes first and to make steel we basically need lots of cast basalt and fused silica "lego blocks" although that is a simplication.  See:
Blister Steel   and  Blister Steel 2

LEFT: An FFC cell
As with steel, we will need robots to handle the masses of ilmenite, TiO2+iron and sponge titanium output from the FFC cells.  Titanium is not magnetic like iron and steel so we won't have the convenience of a robot crane with an electromagnet to handle titanium etc.  We will need robots with gripping mechanical hands and buckets of TiO2+iron.  We will also need work chambers with concrete floors as we will for iron and steel.  Electron bean stereolithography generates lots of X-rays and apparatus on Earth is mostly radiation shielding by mass.  On the Moon we will set up 3D sintering machines with walls of cast basalt and regolith surrounding them for radiation protection.  These machines will work well with free vacuum.