Jump to content

physicsless part causing drag?


EliteGuy3

Recommended Posts

So one day I was trying to get minimalist record in the K-Prize Challenge, and I wondered why the SpecialSpeckOfDust used a Z-200 battery. When testing it with a entirely clipped Z-100, the plane didn't accelerate as fast. When I replaced it with the Z-200, it flew fast enough to get to orbit and back, like the video.

Can someone explain to me why the Z-200 is better than a physicsless part in this case?

Edited by EliteGuy3
Link to comment
Share on other sites

Reading from the wiki, " The mass of the part and the drag experienced by it are added to the parent's part. " So maybe it has something to do with that....

Link to comment
Share on other sites

Hi.

 

Maybe the "Brains" can con confirm or deny what I will write here. I am not that good

When a part is "Physicless" it doesn't mean th it isn't there.

The physic isn't computed or applied where it is situated.

The mass and drag are added to the parent part.

No matter where the Z-100 is, even clipped, it contribute to drag.

 

The Z-200 is however streamlined and has less drag.

 

QED

 

ME

Link to comment
Share on other sites

50 minutes ago, Martian Emigrant said:

No matter where the Z-100 is, even clipped, it contribute to drag.

 

The Z-200 is however streamlined and has less drag.

Oh, this game....

Well thanks. The wiki too shows that physicsless parts' characteristics are adopted by their parent parts. Still don't quite understand the Z-200, maybe it's because there's a nose cone connected to it, but at least I know now about the Z-100.

Link to comment
Share on other sites

10 hours ago, EliteGuy3 said:

Oh, this game....

In this case, it's more of a nomenclature problem than anything else-- the term "physicsless" is now misleading, though it used to be accurate.

A brief history lesson:  prior to KSP 1.0, the so-called "physicsless" parts actually were physicsless.  Literally zero mass and zero drag.  After 1.0, this was fixed so that they do have mass and drag, it's just that it gets applied to the parent part (or, more accurately, the first non-physicsless ancestor part, since the parent might itself be physicsless) instead of to the part itself.

So the term "physicsless" is now unfortunate, since it's downright wrong; but it had already been around so long by the time they fixed this that the name just kinda stuck.

 

12 hours ago, EliteGuy3 said:

Can someone explain to me why the Z-200 is better than a physicsless part in this case?

It makes perfect sense, as long as you expunge that horribly misleading label "physicsless" from your mental vocabulary.  :wink:

I'll edit your quote to terminology that's more relevant to this discussion:

12 hours ago, EliteGuy3 said:

Can someone explain to me why [a stack-mounted part] is better than a [radially attached] part in this case?

It's for the same reason that any stack-mounted part does better than a radially attached one.  It's because the Z-200 is in-line with the stack and therefore doesn't add any drag to the rocket-- the craft is just a smooth cylinder, which is nice and aerodynamic.  The Z-100, on the other hand, is radially attached, meaning it's stuck to the outside of the rocket, therefore it adds drag.

12 hours ago, EliteGuy3 said:

testing it with a entirely clipped Z-100

Careful, there-- looks like you're making a mistaken assumption about how KSP aerodynamics work.  (Don't worry, you're in plenty of good company; this is an easy mistake to make.)  The key issue here is:

12 hours ago, EliteGuy3 said:

clipped

...Why did you clip it?  Let me take a guess.  "Because I don't want it to add drag, so I put it inside the part so it will be out of the airstream and won't affect my performance."  Am I close?  :)

(Or perhaps it's the variant, "Because it looks dorky so I clipped it inside to make it not visible.  But I get the side benefit that it's out of the airstream and won't add drag, right?", which amounts to the same thing in gameplay terms.)

Alas, no.  That would be a completely reasonable assumption to make, and any player could be forgiven for thinking that-- if you're in this boat, you've got plenty of company!  However, that's not how KSP aerodynamics work.

KSP's aero model is very simplistic.  It's not a simulator.  Specifically, it does no geometric modeling of occlusion, for drag purposes.  What that means is that, unlike reality, the actual location of a part doesn't matter where a part is when determining how much drag it generates.  The game uses a drastically simplified model in evaluating how much drag a part generates (do a forum search on the term "drag cube" for lots of discussion about this).  What actually matters is not where a part is, but rather, how is it attached.  The "rules" are very simple:

  • A stack-mounted part that's the same diameter as the things above/below it in the stack won't generate drag.
  • A radially-attached part will generate drag.

Thus, the radially attached Z-100 will generate drag, even if you clip it completely inside its parent part, because it's still radially attached as far as KSP is concerned.  The game doesn't "know" that it's "physically inside" the other part, when it's doing its drag calculation.  Similarly, if you put a Z-200 in-line in a 0.625m stack, it won't generate any drag at all... even if you use the editor "offset" widget to move it physically out of the center stack and off to one side.

"But that's stupid," I hear you say.  "Why would they do it like that?"

Well, not being a Squad employee, I don't have an authoritative answer, but I can make what I believe to be a pretty good guess.  I assume it boils down to one or more of the following:

  1. It's a lot easier for Squad, this way.  (Simpler to code.  Better performance.)
  2. It doesn't matter, the large majority of the time.  (Because in practice, most people don't clip their radially attached parts inside most of the time, or take their stack-mounted parts and offset them radially.  And even when people do clip radially attached parts inside, they're usually fairly lightweight anyway so the effect often isn't noticeable.)

If you'd like a more sophisticated/realistic aero model, which does take geometry into account, you can do so with mods such as FAR.  Just be aware that it'll have a pretty big effect on how your craft fly in atmosphere, so if you were to make that leap, your current ships may need some re-designing, particularly if they're heavily "atmosphere-involved" (e.g. spaceplanes).

Link to comment
Share on other sites

54 minutes ago, Snark said:
  • A stack-mounted part that's the same diameter as the things above/below it in the stack won't generate drag.

Err, zero drag, really?

Less drag than, say, if it was attached to a part of disparate size but surely not no drag?

Link to comment
Share on other sites

6 minutes ago, Foxster said:

Err, zero drag, really?

Less drag than, say, if it was attached to a part of disparate size but surely not no drag?

AFAIK, yes.  Though to be honest, now that I think about it, it's been so long now that I don't know whether the reason I think that is "because I heard it from a knowledgeable source" or "because I just assumed that."  :wink:

Certainly if it's not zero, it's small enough that it might as well be-- at least as far as designing your ship is concerned.

If someone has an authoritative answer to "is it actually zero or just very tiny", however, I'd love to hear it so the matter's settled.

Link to comment
Share on other sites

I'll have a look in a bit (just getting into Prey atm) but I'm pretty sure that all parts contribute some significant amount of drag, even when attached in the middle of same-sized parts. With the amount of drag increasing sharply as the parts increase in size. 

Though, as I said, it is quite a bit less than odd-sized parts joined together. 

Link to comment
Share on other sites

20 minutes ago, fourfa said:

Yeah, stack attached parts make drag - just not very much.

And also importantly, it is highly dependent on angle-of-attack.  Stack mounted fuel tanks (and other parts) generate plenty of drag when they are sideways to the airstream (relative to their attach node placements).


The drag cube system is fairly complex, taking into account if parts are attached to attach nodes to determine occlusion for a specific 'face' of the cube.  If something is attached to a node, it will be occluded (partially or fully) by the part on the node (depending on the size of the drag cube for the part on that node), and generate less drag (down to zero drag at 0' AOA and full occlusion) -for that face- (generally a top or bottom face).  Which faces are used for calculate depend on the angle-of-attack (actually I think they are all used, but the contributing factor/scalar for each is AOA dependent, and interpolated across faces where more than one is non-occlued and exposed to the airstream).

Now, the drag-cube system also simulates parasitic / skin drag on the 'side' faces of the drag cube even when at zero angle of attack.  So even a part that has fully occluding parts on its top and bottom nodes and at zero AOA it will still produce some 'skin' drag.

When at an angle-of-attack inbetween direct-on and side-on, it interpolates the drag between that produced by the 'occluded' faces and the non-occluded side faces, using I believe the cosine of the angle relative to the faces.  So if the top of a fuel tank is occluded by another part, but it is at a 45' angle of attack to the airstream, it would put out roughly 70% of the drag (cosine) compared to if it were completely non-occluded.

Edit:  Honestly, I'm unsure if the cosine use is correct or not, it may be more linear than that....

Edited by Shadowmage
Link to comment
Share on other sites

Wow, never thought this forum would get the attention of @Snark.

6 hours ago, Snark said:

The game uses a drastically simplified model in evaluating how much drag a part generates (do a forum search on the term "drag cube" for lots of discussion about this).  What actually matters is not where a part is, but rather, how is it attached.  The "rules" are very simple:

  • A stack-mounted part that's the same diameter as the things above/below it in the stack won't generate drag.
  • A radially-attached part will generate drag.

 

Ok,

So in summary, in-line parts can avoid generating (lots of) drag, while it doesn't matter where you place a physicsless radial part because the army monkey. I apologize for my discombobulated confusion over this topic. However, I'm glad it put light on physicsless parts, which I sure needed to learn about.

Link to comment
Share on other sites

Just for completeness, here's the reason some parts are "physicless"...

There are some parts you just want to stick on the side of a craft singly. Stuff like ladders and science experiments. If they followed the usual rules then your craft would always be going off-course due to drag and/or mass imbalance. 

So, some parts that are like this add their drag and mass to the part they are mounted to, so avoiding them causing these problems. 

Edited by Foxster
Link to comment
Share on other sites

12 minutes ago, EliteGuy3 said:

it doesn't matter where you place a physicsless radial part

FTFY.  It doesn't matter where you place any radial part; it's not just physicsless ones.  They generate drag regardless of physical location.

An interesting side note:  KSP doesn't model occlusion for drag purposes, but it does model it for reentry heat.  (For example, heat shields cast a thermal "shadow" behind them during reentry.  That's why the 10m inflatable shield will typically protect your whole ship, regardless of how things are attached in stacks or radially or whatever.  Anything that fits behind the shield is safe.)

So, clipping parts won't save drag, but it might save their getting burned off.  An example of exploiting this is putting airbrakes on the back end of a ship so that the heat shield on the front end stays pointed forward.  Normally you can't use airbrakes very well for reentry, since they have a really low heat tolerance (just 1200 K) and get burned off.  But if you've got a big heat shield, the airbrakes will do just fine, as long as they fit inside the heat shield's "shadow".  IRL that wouldn't work, but it works in KSP because the heat shield blocks the heat without blocking the drag.

Link to comment
Share on other sites

5 hours ago, Snark said:

It doesn't matter where you place any radial part; it's not just physicsless ones.

As I said,

6 hours ago, EliteGuy3 said:

I apologize for my discombobulated confusion over this topic.

Again, thanks for the advice. Clipping radial parts only help for re-entry heating, and applies to all radial parts. See, this how I live up to my motto, "Discombobulated Engineer."

Link to comment
Share on other sites

  • 8 months later...
This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...