It seems my KSP game may be experience more drag than others

[Tiefman]

Hello everyone first post here, I'm pretty desperate for help. I made a basic orbit rocket with ~3680dV, plenty to orbit. Shared the craft with some friends, and was suprised to see they had way more leftover dV than I did. At first stage seperation, they had almost 150 more m/s of speed than I did. We tested further by sending the rockets STRAIGHT UP WITH NO INPUT and once again my rocket fell ~100 m/s shorter speed wise. I have absolutely no clue why this may be. I have tried looking for help on the reddit, the discord, but no luck. No one seems to understand what is happening. Ive reinstalled KSP, launched with no mods, deleted Physics.cfg, all that. Still no luck.

Here is a video of speed loss in ascent profile: https://www.youtube.com/watch?v=AnYEx61OOMI
Here is a video of speed loss going STRAIGHT UP: soon

[Tyko]

Can you post your craft file? That way others can try

 

And welcome to the forums..  

Edited February 21, 2019 by Tyko

[MechBFP]

Did you make sure to actually delete all the mods? Your GameData folder has nothing in it at all but folders with “Squad” in the name?

[OHara]

With the first video up so far, it looks like a difference in drag, that begins when you cross the speed of sound.  If you can upload your craft to kerbalx.com or dropbox and post a link here, I'm sure we will figure it out.

There is one easy mistake -- connecting the tank below a cargo bay to some contents of the cargo bay rather than the cargo bay itself -- that happens to be treated rather unkindly by KSP's simplified model for aerodynamics, and gives this symptom.

[Vanamonde]

Hmm. I don't know what's going on here but it sounds like a modding issue? At any rate, thread moved to modded support. 

[soulsource]

1 hour ago, Vanamonde said:

Hmm. I don't know what's going on here but it sounds like a modding issue? At any rate, thread moved to modded support. 

I agree. For instance, FAR on default settings, has significantly lower drag than stock.

[OHara]

On 2/21/2019 at 9:06 PM, Vanamonde said:

sounds like a modding issue?

This turned out to be not a mod problem after all, but a configuration problem with freshly-installed KSP.  (Maybe not worth moving again, though, as google can help others find the solution here.)

The solution was found via the reddit thread and is now reported on Squad's bug-tracker :

Delete the PartDatabase.cfg and have KSP forced to generate it when it is started.

[Laie]

FWIW, I'm experiencing the same right now. Been working on Eve lifters:

1. rollout, hyperedit,
2. launch, saveload, launch, saveload...
3. eventually saveload to before rollout and back to VAB for tweaking.

Then there was a sudden drop in craft performance.

I'm positive that it just happened between two rollouts. I didn't even restart KSP, much less managed any mods. At first I was suspecting my small change to the craft to have an outsized effect... it took me a while to figure out that all craft are underperforming. Vessels that almost-but-not-quite managed to make orbit now stall as I cross the sound barrier.

Deleting the partdatabase had no effect.

Doing a diff with the partsdatabase from my 1.6.0 install which I haven't touched in a while, I only see differences where parts are in one install but not the other. Parts that are present in both installs (like, stock parts) have the same drag cubes.

I checked physics.config if I accidentally messed up something, touched a slider in the debug menu perhaps, but can't find a thing. File is identical to the physics.cfg of my ksp1.6.0 install, or 1.4.5 for that matter.

Settings config also shows only one small difference, SHOW_WHATSNEW_DIALOG_VersionsShown --  I refuse to believe that this is in any way related to the problem at hand.

Edited February 23, 2019 by Laie

[Vanamonde]

13 hours ago, OHara said:

This turned out to be not a mod problem after all, but a configuration problem with freshly-installed KSP. 

Off to Gameplay Questions, then, where others with similar problems are more likely to see it. (Sorry for bouncing the thread around, but we try to keep things organized.) 

[Laie]

@Vanamonde: I'm afraid this is something very very weird.

1. I've re-installed KSP, copied over the save with craft files. Still much drag.
2. several iterations of reloading KSP, deleting PartDatabase, reloading again: no effect.
3. starting a whole new save in the whole new (stock+MH) install, re-creating a simple Eve lifter and moving it to Eve, it suffers from high drag. (Eve lifter? Drag? Yes, a lot more than what I'm used to)

I don't have the slightest.

[Vanamonde]

Then the thread is staying put for the moment anyway. 

[OHara]

2 hours ago, Laie said:

sudden drop in craft performance.     I'm positive that it just happened between two rollouts [rolling the craft out of the assembly building].

That is a little different to the trouble the OP had. They had a persistent differences between their install and their friend's install.

You do have the Terrier on your Eve craft, and that was the bug causing the OP difficulty.  The new variant-model Terrier makes a more drag in a rocket stack than the old (thread link).  Of the new variants, the 'shroud' variant is least bad if you have a game-generated PartDatabase.cfg, the 'bare' variant is least bad if you have PartDatabase.cfg  from the KSP store zip file.

Maybe you replaced the old Terrier with the new between your two trials?   Otherwise, if you start a thread with a kerbalx link to the craft file, we can compare experiments.

[Laie]

30 minutes ago, OHara said:

Maybe you replaced the old Terrier with the new between your two trials?   Otherwise, if you start a thread with a kerbalx link to the craft file, we can compare experiments.

Yes, I did. Performance comparison between the old craft and the MH version: no comparison.

On 1/14/2019 at 2:20 AM, OHara said:

most of the Making History engines, the Engine Plates, and Structural Tubes cause extra drag when they are inserted into a stack

Ah-hah! Check, and check.

So it was a small change to the craft having an outsized effect, after all. I even suspected the plate at first...

The next vessel I tried was an Eve lifter that was almost successful -- it had a Vector for the second stage, mounted in the stack to a thrust plate. I figured that if the problem was thrust plates, it should work better if I replaced that assembly with a Mastodon. Then it performed worse. Much, much worse.

I'd never have guessed that an inline, same-diameter engine would create more trouble than a thrust plate assembly. Still: my fault for tinkering with the craft in any way. It was no longer the same vessel that almost worked before. Yet I was totally sure that it couldn't possibly perform worse, and never stopped to check the original. I've done so just now and, lo and behold, it works as it used to.

I tried more things that eventually convinced me that my install is broken, but in hindsight, all of it can be explained by by what OHara said over in the other thread. @OHara Thank you so much.

[Foxster]

13 hours ago, Laie said:

FWIW, I'm experiencing the same right now. Been working on Eve lifters:

1. rollout, hyperedit,
2. launch, saveload, launch, saveload...

You have to be careful with HyperEdit, it can produce some weirdness. A step you didn't mention I find is important...

1. Rollout

2. Use stock alt-F12 to set orbit around Eve

3. Use HE to land

If you go straight from the VAB launchpad to landing on Eve (or elsewhere) then you get weirdness.