1.875m Parts

[Guest]

Cost is literally just a variable right now, so implementation for a mod depends entirely on when that variable is read since you'll want to update that cost as the procedural part changes. If the cost is updated regularly (vessel simulation or the like, this also unsurprisingly lays some groundwork for delta-V calculations) there is no problem. If the cost is updated when a new part is added or removed, you only run into the problem right before you launch directly from the VAB/SPH since I would assume the vessel cost would be calculated on load.

In practice, cost will likely be updated continuously in a simple vessel-sim like manner since the user can tweak things like fuel. This means procedural mod authors will have no problems updating that single variable and having the correct cost applied. The formula a particular author uses will, of course, be up to them to decide.

[Eric S]

Maybe a surface area calculation for the tank itself?

Depends on whether you want to balance against stock or reality, and there are a few other factors in the latter that might affect the scaling as well.

Stock scales fuel capacity, mass, and cost linearly based on volume. The really small parts don't fit on the scale, and the SLS parts have their own scale, however.

Reality isn't so clear cut. If you're wanting to scale the cost/mass directly linearly to the surface area, then you're assuming that the tank has the same thickness at any scale. At some point on the high end, the same thickness won't even be able to support its own weight. If you double the dimensions of the tank, you've increased the mass that the tank can hold by 8 times, but haven't noticeably increased the tanks ability to constrain the fuel within. That's not going to end well if anything goes wrong. The more mass per surface area, the less lateral force it would take to rupture the tank.

Things don't even have to go wrong, really, because a large tank is probably going to have more mass on top of it than a small tank. The slightest sheer force would make the tank collapse for a sufficiently large tank.

This is similar to the arguments against giant ants, or giants in general. Double the height and maintain the same proportions, and you've increased the mass by 8 times, but only increased the creature's ability to survive the stress of the mass by 4 times.

I like Kasuha's approach, and if we were aiming for that level of accuracy, that would be a good start. All things said, I'm fine with the linear scaling that stock KSP tanks use, reality would be far too complex to try to model here.