Oh boy... if only I had a dollar every time I hit a site that I frequent that someone referred to this old article...
Prime example:
refer to this Japanator article and my comments in it.To sum up what I said there and add on to it:
"According
to the US Government a B2 Stealth bomber costs 2 BILLION dollars to
build each currently, so this rooms plenty of room for labor and extra
parts. Superconductive motors, sure, but gas turbine engines for 52
million each? and what kind of computer costs 1.55 million?? The super
computers on the F35 aren't even half that much.
There is a serious flaw though with their method in using
electricity to power it from gas turbines and using electric
superconductors. HEAT.
According to my calculations, for a gundam to move at a rate of
60ft/s (assuming it's 60ft tall and roughly 20 tons of weight- 60ft/s
then would be a brisk walk for it), it would take approximately
331811.8 Joules of force. Convert that to Watts by dividing by 1 second
(since that’s how much power it would take to move 60ft/s for one
second), and you get somewhere around 3 MEGAwatts. Even if you divide
that up among several turbines, that level of electricity would fry any
size wire you could make, or at least require it to be frozen to
superconductive levels that'd be nearly impossible in such an
environment.
Instead, a better method would be to use individual fuel cell
systems installed at each joint or at least on a smaller individual system setup instead of one big power system, so that if one goes down all the
others can still function, and in addition to taking a severe load off
of any power system, you wouldn't need the massive would be cables that
would otherwise be necessary, and leave more room for ammo and guns.
(As you may or may not be able to tell, I research and calculate
this kind of stuff for fun- oh, and that stunt in the 5th episode of
gundam 00 would've taken well over 300MW of power for Kyrios to push a
block roughly 100 times the gundams size- current day full size 4th
generation nuclear reactors can barely output that much)"
After researching further since then, I believe a solid oxide fuel cell system would give the optimal power, durability, and efficiency that we would need for such a project.
"I can honestly say, this is the goal that brought me into engineering, so I have done alot of work on this sort of thing. As for physics computations and the type of stress on the tendon sheets for balancing, assuming they use the same ratio of strength to height that is in human beings, there's no reason that aluminum alloys with over 40 times the strength load to weight ratio wouldn't be able to handle it. As for balancing, there are already several examples of dynamically balancing robots on two legs out there (Dynamically balancing means it can walk in a similar way real people do by reacting to shifts in weight and forces - NOT pre-programmed leg movements like say the Honda Asimo). The entire concept is very possible, the issue is it's not very practical, at least for most uses anyway. Something along the lines of an Arm Slave would be alot more practical for military use, but Gundams just look so much cooler.
What bugs me about their calculations is they're thinking from a government price index perspective. You don't need the best most advanced computer in the world, you just need one that's plenty fast enough and extremely durable. Heck, for the kind of computations something that you'd need for something like this, a cluster of 4 quad-core processor computers would be OVERKILL. Just build those inside individually reinforced cases and strengthen the parts for movement and shaking and you've got a great computer that'd cost less than a fraction of a Blue gene IBM machine, not to mention being a heck of a lot more reliable if it takes a hit (in a cluster if one dies you can just switch to using the others). You could say similar things about the other stuff on the list too."
Fair warning, I wrote this over a year ago.
