In terms of architecture I did very poorly. The actual ITS is rigid-hulled, passengers travel in microgravity, life support is ISS-style, and both Mars and Earth arrival is direct aerocapture and propulsive landing. There are a lot of windows. There are no propellant depots anywhere. Oh well... I like parts of my approach better but certainly Musk's approach is less risky up front and less expensive to develop.
In terms of vehicle performance, I posted numbers for a 12m version and a 15m version. The vehicle will be 12m so I'll use those numbers. I'll compare to the reusable ship profile from Musk's talk.
In short, I didn't do very well. Details after the jump.
'Accuracy' numbers should be close to 100%. Over means I guessed high, under means I guessed low.
|dry mass, t||291.20||275.00||106%|
|dry mass, t||28.00||150.00||19%|
|launchpad mass, t||14,988.12||9,375.00||160%|
Right off the bat we can see I seriously underestimated the Isp of the Raptor engines. That of course had the follow-on effect of requiring significantly more propellant, more takeoff mass and more first-stage thrust. I was able to squeeze in a lot of extra cargo in an expendable profile but it's not an ideal design. Musk is able to put up about half that payload with all reusable components and about 40% less propellant.
He also allocates a lot of mass to the ship, which is reasonable; I wasn't sure what to use for structures, life support and thermal protection mass for a vehicle that can do multiple re-entries between maintenance periods, so I basically estimated what a simple payload fairing and bare bones upper stage hardware could do. Even his tanker design is 90 tons, where my dry mass estimate was only 28 tons; the rest of it is in extra engines, heavy-duty thermal protection and robust structure.
My architecture apparently was massively overbuilt. SpaceX will use a transport ship with a lot of extra delta-V (over 6km/s from LEO when refueled), so their trips are going to be 3-5 months instead of 6-8 months. You can get away with a lot less shielding that way, particularly if passengers are willing to take on a percentage point or two increase in their future cancer rates in exchange for cutting the trip cost in half (or better). The ships themselves will carry a strong investment in reuse, reliability and simplicity; this goes a long way towards eliminating the advantages of task-specific ships. Musk did talk about depots and other possible optimizations, but it's clear he thinks the mission should be as simple as possible in the early stage. Once the trail has been blazed then improvements can be implemented.
I'm curious to see how a cargo variant might be put together. The crew version can carry a lot of cargo (300-550 tons), but it is limited to cargoes of 12m in diameter and 12m to maybe 14m in length. The obvious payload for a cargo variant would be a massive Bigelow module, but even the BA-330 has a length of almost 14 meters and only 20-22 tons mass. It's possible Bigelow might design a module specifically to fly on a cargo IPS, but they've been burned by SpaceX in the past and seem to be serious about maintaining multiple launch options. More likely we'll see a reusable supersize payload fairing and an expendable/repurposable variant of the propulsion section of the second stage; this could take up four BA-2100 modules in one shot (around 400 tons) then stick around to provide ~200 kW of power and nearly 2000 tons of propellant capacity as a depot. Not to mention, it would have plenty of capacity to change inclinations or place those modules in lunar orbit with LEO refueling.