Conversation
i dunno if this fixes it yet but im eep so here it is
Fixed it!
dont need this anymore
|
Ignore all the readme commits, im new to this |
|
This underestimates the dry mass - I point to a NASA sizing paper for more information. If you turn this into the volume of a sphere based on the radius of the tank r and a cylinder of height h, where h is the total height of the tank - 2 * r and then multiply that by the density, you'll be a lot closer to how mass works IRL. |
| double radius = Math.Pow(OldVolume * 0.001 * 0.75f / Math.PI, 1f / 3); | ||
| // calculates volume of sphere slightly smaller 4mm in radius than the previous volume | ||
| //then converts from m^3 to L | ||
| double dV = 1.33 * Math.PI * Math.Pow((radius - 0.004), 3f) * 1000; |
There was a problem hiding this comment.
Why are you converting to litres? Density is in kg/m³, not kg/L.
There was a problem hiding this comment.
Liters because the value from oldVolume is the tank volume in liters, converting to m^3 allows me to find the radius, then I create a smaller volume using a slightly smaller radius, which gets converted back to L so that the difference can be found between the two volumes
There was a problem hiding this comment.
I agree the units I used aren't standard, this is just how I elected to do it on my first pass
| additional mass savings as they progress, as if they are using better alloys | ||
| */ | ||
| double density_2195A = 0.00271; | ||
| //whyyyyyy does ksp use base units of tons bro (its not that bad) |
There was a problem hiding this comment.
This refers back to my above comment - aluminium density in tons/m³ is 2.685 tons/m³, which would be a lot nicer to read than converting from m³ to L and then using this small number.
There was a problem hiding this comment.
Ill take this into consideration when I work on it some more, this is effectively just tons/L, which is super unconventional but let's me multiply it directly with the new wall volume calculated before
|
I compared a tank with similar dimensions to the SLWT of the space shuttle and the Centaur III upper stage. Results were as follows: Centaur III approx tank mass: 2247(Total Mass)-190 (RL-10C) = 2057 kg New Method: Centaur III approx tank mass: 2247(Total Mass)-190 (RL-10C) = 2057 kg Considering this information, I believe the method I present is accurate enough and models real-life fuel tanks quite well in terms of mass. On the other hand, the current method seems catastrophic in the way it calculates dry mass. You may disagree with my method, but the numbers speak for themselves. EDIT: SOURCES: |
|
Attempted to reproduce your findings using the numbers given. These screenshots should be accurate for the current RP-1 version (v4.2.0.0). Centaur III stage (Modern Balloon Tank, filled with the RL10A-4-2N propellant mix.) The tank dry mass ends up at 1.12t without MLI, 1.29t with 100 MLI. I couldn't exactly reproduce the shape of the SLWT, so I just made sure the bottom diameter and length matched, and tweaked the top diameter until the total wet mass roughly matched the wet mass of the SLWT in the source given. This is definitely not perfectly accurate. Dry mass reads 44.1t without MLI, 45.5t with 100 MLI. |
|
Thank you @periodically-makes-puns for the extra info! |
|
Two discoveries:
New method (No Configs):Centaur 3 mockup tank: SLWT: 
Old method (No Configs):Centaur 3: SLWT: Conclusion:The old method actually seems to under-report slightly for the Centaur 3, then proceeds to over-report for the SLWT. The new method isn't completely accurate, but it eliminates the issue of massive discrepancies in large tanks compared to IRL. |
3 participants
Current method causes mass to increase with the cube (volume) of the tank, resulting in very high and unrealistic dry masses for tanks.
Proposed method calculates the wall volume of a hypothetical spherical tank, then multiplies the density of aluminum to calculate a more accurate mass profile.