What's the best way to clip an RTG?

[KingDominoIII]

I was looking to replace my solar panels with RTGs on an SSTO to provide more consistent power and save drag. To clip it into the fuselage and eliminate its drag, should I surface mount and then offset it, or put it between two parts and offset them back together?

[Regor]

11 hours ago, KingDominoIII said:

I was looking to replace my solar panels with RTGs on an SSTO to provide more consistent power and save drag. To clip it into the fuselage and eliminate its drag, should I surface mount and then offset it, or put it between two parts and offset them back together?

Forget this and look at @fourfa’s response.

Edited April 8, 2020 by mabdi36
I am Stupid

[fourfa]

Neither. Clipping parts visually out of the way does not ever eliminate drag.  As for part 2 of your question, yes open nodes cause increased drag, but mis-matched nodes are just as bad or worse. For instance, if your plane is Mk1 part -> RTG -> Mk1 part, you’ll have huge drag at both the 1.25m-to-0.625m and 0.625m-to-1.25m nodes (the RTG is 0.625m or Mk0 as the game sees it - look at the diameter of the green node before attaching, these tell you what matches)

The correct answer is service bay, cargo bay, or fairing.  That is the only way to eliminate drag entirely for your various service parts. All my SSTOs have one, and I test carefully that the parts inside are actually shielded from drag (it’s quite easy to have a part that looks like it should be shielded actually be getting drag). Try alt-F12 -> Physics -> aero -> show aero data in right-click menu (something like that), then right-click the part in question.

Aero mods like FAR change everything but since you didn’t mention it and are asking here instead of the FAR thread I’m assuming you’re playing stock

Edited April 8, 2020 by fourfa

[KingDominoIII]

On 4/7/2020 at 8:19 PM, fourfa said:

Neither. Clipping parts visually out of the way does not ever eliminate drag.  As for part 2 of your question, yes open nodes cause increased drag, but mis-matched nodes are just as bad or worse. For instance, if your plane is Mk1 part -> RTG -> Mk1 part, you’ll have huge drag at both the 1.25m-to-0.625m and 0.625m-to-1.25m nodes (the RTG is 0.625m or Mk0 as the game sees it - look at the diameter of the green node before attaching, these tell you what matches)

The correct answer is service bay, cargo bay, or fairing.  That is the only way to eliminate drag entirely for your various service parts. All my SSTOs have one, and I test carefully that the parts inside are actually shielded from drag (it’s quite easy to have a part that looks like it should be shielded actually be getting drag). Try alt-F12 -> Physics -> aero -> show aero data in right-click menu (something like that), then right-click the part in question.

Aero mods like FAR change everything but since you didn’t mention it and are asking here instead of the FAR thread I’m assuming you’re playing stock

What if I clip it into a .625 segment?

[fourfa]

4 hours ago, KingDominoIII said:

What if I clip it into a .625 segment?

Should be OK?  Though there are some parts with weird drag no matter what (for instance the tri-structural adapter last I checked) so I don't want to swear.  I'll check it out later since I know it can be kinda tricky to parse the data in the AeroGUI

[fourfa]

@KingDominoIII Kinda surprising results.  The RTG seems to have strange drag.  When node-attached inline, the part behind it gets very high drag, almost as if it still has an open node (my sounding rocket reached orbit with 628m/s of delta-V remaining).  But the drag penalty of just surface attaching it isn't bad - in fact it's better than the inline test (flying the same fixed trajectory, reached orbit with 683m/s).  Worst of all is surface-attaching the RTG, then putting Small Nose Cones on its open front and rear nodes (reached orbit with 288m/s).  I also tried the Very Small Nosecone from Breaking Ground (445m/s remaining), and the Tiny Nosecone (680m/s).

What this tells me is that inherently the part doesn't have high drag at all, the extra mass penalty of nosecones don't offset whatever bare drag penalty it has.  So I revise my original advice - you should surface mount and offset it.  It won't be shielded from drag, but it hardly matters, and it probably will be less draggy than the solar panels you want to get rid of.

[Lt_Duckweed]

@fourfa the reason you are seeing weird results is drag is not calculated off of nodes, but off of the surface area of a parts drag cube.  When you node attach parts, the surface area of both parts that is in contact is subtracted from the drag cube.  The important thing here is that the surface area does not necessarily match the node size.

In the case of the rtg, it's node size is .625m, but the surface area associated with the node is less than a .625m cylinder, so when you attach it to a .625m fuel tank, the rtg end is covered completely, but the fuel tank end is NOT fully covered.

An opposite example is the 1.25m reaction wheel.  It has a 1.25m node, but it has a surface area that is larger than a 1.25m cylinder, so when placed inline with 1.25m fuel tanks, the fuel tank ends will be fully covered, but the larger surface area of the reaction wheel will NOT be fully covered.

Edited April 14, 2020 by Lt_Duckweed