Why do you need concrete on the Moon?
On Earth we have about 100 kPa atmospheric pressure. On the Moon, with lower gravity and a change in gas mix, that might be reduced in a moon habitat to a pressure of 50 kPa. It would require a membrane capable of withstanding an internal pressure of 50 kPa which, for a large radius dome would require a very significant tensile strength. Concrete domes work well on Earth where there is no big pressure difference across the wall.
The biggest challenge is countering the internal air pressure. A concrete dome will not solve the problem without high tensile fibres. The problem can be countered by a layer of ballast on the top of the dome, while the lower wall would still require high tensile material.
This suggests, not a concrete dome, but a vertical axis cylindrical structure with low vertical walls and a flat roof. One possible static solution is a flat roofed area sunk into a convenient crater with the weight of ballast material on the flat film roof countering the internal pressure. The wall membrane is constrained by the rock structure, or maybe partly by the hydrostatic pressure of external material.
Ignoring the problems of construction for the moment, we need some idea of how thick the roof ballast will need to be if it is cut from the floor of the selected crater, gathered from the surface, or from a quarry nearby. On Earth there is a buoyancy due to the atmosphere, that is missing on the Moon. We can ignore that buoyancy since it is only amounts to about 1.2 kg/m3.
It is going to be simpler to design the ballast for Earth's gravity of 9.8 m⋅s-2, then convert it to the Moon's lower gravity of 1.625 m⋅s-2, which is weaker by a factor of 9.8 / 1.625 = 6.03 The ballast layer calculated for Earth will need to be multiplied by 6.03 to correct it for the Moon.
We have reduced the internal air pressure to 50 kPa which will require a force of 50k Newton per square metre.
On Earth, gravitational acceleration is 9.8 m⋅s-2. 50 kN would require 50 k / 9.8 = 5.1 tonne of ballast per square metre.
Basalt rock ballast would have a density of about 2.8 tonne/m3 on Earth, so it would require about 5.1 / 2.8 = 1.82 metre of ballast to counter the 50 kPa internal pressure. Now we convert the thickness to allow for the Moon's lower gravity. 6.03 * 1.82 m = 10.97 m, call it 11 metres.
That 11 metre layer of ballast will keep the roof membrane down. It will also act to thermally insulate the habitat from the month long cycle of lunar day/night. If the inhabitants are lucky, micrometeorites will be stopped before they perforate the inner membrane.
Will gamma rays from the Sun be a problem, as when flying at high altitude, or will the ballast fix that problem too?
Earthworks, an 11 metre thick layer will require a significant volume of cut and fill. If concrete was used instead of local rock ballast, then without tensile fibres in the concrete it would need to be a similar thickness. That is a massive amount of concrete.
So how can we support an 11 metre thick layer of ballast before inflation or in the event of a major leak. Following a small leak such as a membrane perforation, the air pressure will remain constant while the roof gradually falls towards the floor. An internal patch would be pressed by air pressure onto the membrane. If the damage was too great, ballast would flow down into the air space until the remaining membrane was pushed upwards resulting in a massive blowout. There would need to be safe areas that were protected from the low roof during difficult times.
If a tunnel was dug underground into a hill, then wall stability of the tunnel would be critical. A membrane lining would work 11 metres below the surface. Unless tunnels were cut in solid rock, deeper tunnels would need a wall under compression, probably sintered shells that lock together. Shallow tunnels would preferably be a smaller diameter, with a circumferential tensile fibre reinforced membrane to handle the hoop stress.
Whatever is done, an airlock will be needed for access between the habitat airspace and the lunar surface.
use some kind of solar powered, fancy schmansy 3D printer type robotic system to melt the lunar soil into a silica crystal casing.