radial attachments doing weird things with my SSTO

[fommil]

I have an SSTO with a Mk3 profile and I have an EP-37 engine plate at the back (which is just slightly larger than the Mk3 to be annoying) with 5 RAPIER engines (4 radial + 1 center).

It is a decent SSTO, but I wanted to improve it, so I swapped the engine plate for 4 radial attachment points (BZ-52) to get the same 4+1 profile. According to the wiki, these points have no drag but using the debugging menu tells a different story. The drag from the radial attachments and the fuselage is quite significant, but still comparable to what it was before. However, my SSTO is not an SSTO anymore. It fails to ever generate enough thrust to get past mach 1.

What's going on? This is very confusing. This doesn't seem to be purely drag related. Is it possible that one of my engines is getting clipped and not contributing the thrust that it says it is? Wouldn't it be heating up the surrounding engines in that case?

I ended up using a smaller engine plate and displacing the engines, which is lighter than before and less drag, I think (I'm eyeballing it, so I'm not able to tell the difference when it's marginal). But this is still weird, and I'd like to know why I can't use the radial attachments instead of an engine plate (I have no need for the decoupler or the extra mass!)

Before anybody flags it up: the Mk3 engine mount is useless here because it doesn't let me change the number of engines. I can put it between the fuselage and engine plate, to get less drag, but the extra mass cost is crazy. Would be better spending that mass budget on another engine!

Edited November 2, 2022 by fommil

[Lt_Duckweed]

To see the exact amount of drag each part is producing, and the total drag the crat is producing overall, you can enable a couple of options in the debug menu.

Alt-F12 -> Physics -> Aero -> Check "Show Aero GUI" and "Aero Debug Data in Part Action Menus"

Drag in KSP is calculated based on the surface area and shape of parts.... kind of.

When you node attach parts, the game reduces the surface area of the face the node matches with but the surface area of the corresponding face of the attached part (so node attaching a 1.25m fuel tank on the front of a 2.5m tank reduces the front face area of the 2.5m tanks by 25%, and completely eliminates the rear face area of the 1.25m tank)

If a part isn't node attached it is considered fully exposed to the airstream, unless it is in a fairing/bay.  This is what is happening with the radial attachment points.  This means they are creating a lot more drag than the engine plate would (since the engine plate is using node attachment)

For more info on how the drag calcs work, check out this video: 

 

[Gargamel]

Moved to gameplay questions 

[fommil]

1 hour ago, Lt_Duckweed said:

To see the exact amount of drag each part is producing, and the total drag the crat is producing overall, you can enable a couple of options in the debug menu.

Yup, I'm using that. The total drag at some fixed speed and altitude doesn't seem to differ too much between the version with the engine plate and the version with the radial attachments. That's what's so weird about this! It's possible I'm measuring wrong so maybe I should do this properly and actually write stuff down on some paper. I was especially confused to see that the radial mounts contributed any drag at all, because the wiki docs claim that they do not have any physics, I guess that's just out of date.

1 hour ago, Lt_Duckweed said:

If a part isn't node attached it is considered fully exposed to the airstream, unless it is in a fairing/bay.  This is what is happening with the radial attachment points.  This means they are creating a lot more drag than the engine plate would (since the engine plate is using node attachment)

OK, that makes a lot of sense. Engine plate it is! And even better if I use one that matches the size exactly. That means using the Mk3 Mount or a Mk3 convertor to 3.5m and then the 3.5m engine plate. The extra parts contribute more drag but the total must still be less than what it was.