• 0
eightiesboi

Need some input for DRE reentry and flipping

Question

Hey all,

I am having problems making a return vehicle that won't flip on re-entry. The closest I've gotten is 28,000 m from a 200,000 m circular orbit. I have a couple of parameters that are important for my return vehicle, and I am hoping I can get some suggestions. I *have* read the entries on flipping rockets (or not flipping them, as the case may be), but I've had little luck getting the vehicle to not flip. The only mods I am using are DRE, RCS Build Aid, SETI Probe Parts, and Station Parts Expansion Redux and their appropriate dependencies (TweakScale, etc). You can find the .craft file here: https://1drv.ms/f/s!AoyHZiRU1jT-yfUB-64QbAPkq90q-w

The vehicle is basically a mk 1 capsule on top of a mk 1 crew compartment on top of a PPD Hitchhiker on top of a 2.75 m (tweaked) heatshield. (Note: ignore the decoupler, fuel tank, and engine. They are only present for testing and are not part of the actual vehicle.). The CoL seems to be behind the CoM on entry. I have tried moving the  CoL forward and back, using fins. I have added additional weight directly under the shield (my DCoM isn't that different than my CoM, thanks to the relatively heavy Hitchhiker). I have added an additional (small) reaction wheel (which is what got me down below 30,000 m). In my actual career, I *don't* have access to bigger reaction wheels. 

Steps I take. From a circular 200km orbit, deorbit to a 30,000m PE. Decouple test engine and begin descent. Hold retrograde with SAS until explosion.

I have tried RCS on and off, precision and full power modes. I want to avoid rolling the vessel for spin stabilization for roleplay reasons. I do not want to move o a lower orbit first or skip in the atmosphere and lower the AP.

Suggestions?

Kerbal%20Space%20Program%202_16_2018%201

 

 

Edited by eightiesboi
Clarification regarding engine

Share this post


Link to post
Share on other sites

19 answers to this question

Recommended Posts

  • 0

Well, for starters, I'd look into getting both the RCS Build Aid and Correct CoL plugins. They're absolutely invaluable for missions like this. It'll tell you exactly how your CoM and CoL are shifting around as you descend and decelerate. I strongly suspect that the loss of ablative from your heatshield is pushing your CoM closer to the capsule, and once it gets high enough that it passes your CoL, you flip end for end and *poof*, your ship burns up. You need to look into ways of pulling the CoL "aft" during reentry. Airbrakes would be my recommendation, especially the Ablative Airbrakes from the mod of the same name.

Second, you're coming in too steep. I'd try making the reentry from 100km, or even 75 km. Steep reentries are tricky under the best of circumstances, let alone with a craft whose center of mass will shift significantly during reentry. May I ask why it's imperative you make a direct descent instead of going to a lower orbit first?

Third, that lander design... *shudders* No offense, but that is a design that would perform about as well as it looks... Hint: Not well. You'll want a much more streamlined design. That will help with the CoL/CoM issues, too.

Share this post


Link to post
Share on other sites
  • 0

Leaving the engine and tank attached until you are nearly landed may help. The engine is quite heavy compared to the rest, and that mass may be enough to force it to not flip.

(Maybe even land on them, and don't decouple them at all. Save yourself some money that way. -- And get rid of the heatshield, of course.)

Edited by bewing

Share this post


Link to post
Share on other sites
  • 0

Too much drag in the bottom part of the craft, specially the flat heatshield and the transition between the Hitchhiker and the Cabin. (open nodes cause huge amounts of drag in stock aerodynamics).

I 'd ditch the Hitchhiker, use a bunch of Cabins in radial symmetry around the central core and protect the whole thing with a non-staged 2.5m fairing. (leaving a hatch out if that is necessary)

 

BTW, CoL indicator don't tell you the whole story, you need to also consider the not-lift aerodynamic forces and (in the case of stock indicator) lift from non-wing parts.

Share this post


Link to post
Share on other sites
  • 0
16 hours ago, MaverickSawyer said:

Well, for starters, I'd look into getting both the RCS Build Aid and Correct CoL plugins. They're absolutely invaluable for missions like this. It'll tell you exactly how your CoM and CoL are shifting around as you descend and decelerate. I strongly suspect that the loss of ablative from your heatshield is pushing your CoM closer to the capsule, and once it gets high enough that it passes your CoL, you flip end for end and *poof*, your ship burns up. You need to look into ways of pulling the CoL "aft" during reentry. Airbrakes would be my recommendation, especially the Ablative Airbrakes from the mod of the same name.

Second, you're coming in too steep. I'd try making the reentry from 100km, or even 75 km. Steep reentries are tricky under the best of circumstances, let alone with a craft whose center of mass will shift significantly during reentry. May I ask why it's imperative you make a direct descent instead of going to a lower orbit first?

Third, that lander design... *shudders* No offense, but that is a design that would perform about as well as it looks... Hint: Not well. You'll want a much more streamlined design. That will help with the CoL/CoM issues, too.

@MaverickSawyer, @bewing, and @Spricigo, thank you for your suggestions. 

I use RCS Build Aid (it's right in my OP), and I don't think Correct CoL doesn't really help me in this instance. I have tested the idea of loss of ablative, but that isn't it (tested by cheating in more ablative and by adding weight greater than the original loss). The problem I have here is that while KSP shows CoL in the VAB, it doesn't show drag, which is what I need. I can move CoL all the way to the capsule, but that hasn't helped at all. Because I have added more weight to just under the heatshield, I don't think shifting CoM is a problem (and indeed, RCS Build Aid shows my CoM and dCoM to be nearly identical, even if I remove the ablator).

I don't have access to airbrakes yet. 

I am using DRE, and steeper reentries (with less linger time in the upper atmosphere) are usually suggested. I have plenty (even when not cheating for test purposes) of ablative left (tested by disabling blowy-uppy due to heat), so reentry heat isn't a problem. Also, 200km is still low Kerbin Orbit. The reason for the direct descent is this will eventually serve as an escape capsule for my station, and it is powered by SRBs that have a specific DV setup to get it home. I don't normally have problems with a 30km descent from LKO with any vessel other than this one (including some monstrosities that would scare the dead); what are the tricky circumstances you mention?

And the actual design in my fully modded (i.e. non-test) environment is more streamlined with an adaptor going from the Hitchhiker to the Crew Tube, and some bells and whistles. I simplified it (and the mods I am using) for testing and for posting here, so that anyone who wanted to try it themselves could download the .craft file and not have to install a ton of mods. The fully streamlined and much prettier escape vehicle explodes too. 

6 hours ago, bewing said:

Leaving the engine and tank attached until you are nearly landed may help. The engine is quite heavy compared to the rest, and that mass may be enough to force it to not flip.

(Maybe even land on them, and don't decouple them at all. Save yourself some money that way. -- And get rid of the heatshield, of course.)

The engine isn't part of the actual vehicle. I edited the OP to clarify this. Also, I have really bad luck getting engines down without having them blow up. I don't know if it is easier in Stock (I assume so), but nearly every time I have tried to deorbit a service module with DRE my engine goes bye-bye. Lol!

 

4 hours ago, Spricigo said:

Too much drag in the bottom part of the craft, specially the flat heatshield and the transition between the Hitchhiker and the Cabin. (open nodes cause huge amounts of drag in stock aerodynamics).

I 'd ditch the Hitchhiker, use a bunch of Cabins in radial symmetry around the central core and protect the whole thing with a non-staged 2.5m fairing. (leaving a hatch out if that is necessary)

 

BTW, CoL indicator don't tell you the whole story, you need to also consider the not-lift aerodynamic forces and (in the case of stock indicator) lift from non-wing parts.

In the actual vessel I have an adaptor between the Hitchhiker and the Crew Cabin, but that one flips too. I hadn't thought of the cabins around a core, but I use CLS and that may not work for me unless I get really creative. I may try that though. And yes, CoL indicator not indicating drag is definitely one of my problems!

 

Honestly, I think what is frustrating me the most on this is that my module is front-heavy (relative to the retrograde direction of travel, as it should be), has it's CoM in front of its CoL (again, relative to the retrograde direction of travel), and has sufficient control authority (additional SAS in one test, RCS) that it shouldn't be flipping (I don't think). There is no reason that I can see why the Mk 1 capsule, which (with the monoprop tanks considered) weighs 1.5t, should flip in front of the Hitchhiker and heatshield, which combined weigh between 4.4t (with full ablator) and 3.6t (if the ablator was completely used up) or 4.1t (which is about where the ablator is when the flip occurs). The crew cabin weighs 1t, so I understand why it would be the fulcrum point, but a lighter object on the end of a cylinder shouldn't flip in front of a heavier object at the other end of a cylinder in these circumstances (1900m/s+ in the direction of the heavier end). Lawn darts don't work that way. And remember, I have tried adding more weight to the retrograde front side, with no helpful results.

Edited by eightiesboi
Typo and slight rant

Share this post


Link to post
Share on other sites
  • 0

https://imgur.com/a/Lbimt

I have tried your design ( i have used the normal rcs and the 2.5 meters heatshield), and i didn't find any real big problem, even diving from almost 300kms. Heat wise, not considering the heat shield, the Hitchhiker container never reached 65% of his termal capacity (1300c) way far from the 85% soft failure point of DRE. The only problem with your design is that while going at "slow" speed (under 1500m/s) it creates some aerodynamic torque, but if you put some dumb fins, everything is solved.

  • Like 1

Share this post


Link to post
Share on other sites
  • 0
20 minutes ago, Flavio hc16 said:

https://imgur.com/a/Lbimt

I have tried your design ( i have used the normal rcs and the 2.5 meters heatshield), and i didn't find any real big problem, even diving from almost 300kms. Heat wise, not considering the heat shield, the Hitchhiker container never reached 65% of his termal capacity (1300c) way far from the 85% soft failure point of DRE. The only problem with your design is that while going at "slow" speed (under 1500m/s) it creates some aerodynamic torque, but if you put some dumb fins, everything is solved.

Interesting. I never make it down to 1500m/s, so I'm not worried about the aerodynamic torque (yet), but I'm curious why it didn't flip for you. Did you have RCS on? If so, note that the regular RCS thrusters are more powerful than the SETI ones. I don't have access to the regular ports yet. OTOH, if you had RCS off and still didn't flip, then there's something different between our two installs. 

Would you mind uploading your .craft file? Also, you mentioned you are running heavily modded... Are you using FAR

Edited by eightiesboi

Share this post


Link to post
Share on other sites
  • 0
41 minutes ago, eightiesboi said:

Interesting. I never make it down to 1500m/s, so I'm not worried about the aerodynamic torque (yet), but I'm curious why it didn't flip for you. Did you have RCS on? If so, note that the regular RCS thrusters are more powerful than the SETI ones. I don't have access to the regular ports yet. OTOH, if you had RCS off and still didn't flip, then there's something different between our two installs. 

Would you mind uploading your .craft file? Also, you mentioned you are running heavily modded... Are you using FAR

I put my rcs off for the entire run ( as you can see from the info pane on the top right of the screen, i still have 140/140 monoprop) and i'm not running FAR. This is the craft file https://db.tt/C5X7xM8wpL ( to go to the 1st version you just have to delete the fins)

Share this post


Link to post
Share on other sites
  • 0
38 minutes ago, Flavio hc16 said:

I put my rcs off for the entire run ( as you can see from the info pane on the top right of the screen, i still have 140/140 monoprop) and i'm not running FAR. This is the craft file https://db.tt/C5X7xM8wpL ( to go to the 1st version you just have to delete the fins)

Okay, just gave it a go and ker-boom. I made a video, link below. I am curious--are you using SSPX(r) (Stockalike Station Parts Expansion Redux) Hitchhiker or the standard Hitchhiker? If you don't have that mod installed, I will try again with the stock version of the Hitchhiker. (Note that I deleted your fins, but as you can see, I had exposed skin on the Hitchhiker within the first 10k or so.

Video here: https://1drv.ms/v/s!AoyHZiRU1jT-yfUJ02ayt_wrVPdc-Q

Share this post


Link to post
Share on other sites
  • 0
2 minutes ago, eightiesboi said:

Okay, just gave it a go and ker-boom. I made a video, link below. I am curious--are you using SSPX(r) (Stockalike Station Parts Expansion Redux) Hitchhiker or the standard Hitchhiker? If you don't have that mod installed, I will try again with the stock version of the Hitchhiker. (Note that I deleted your fins, but as you can see, I had exposed skin on the Hitchhiker within the first 10k or so.

Video here: https://1drv.ms/v/s!AoyHZiRU1jT-yfUJ02ayt_wrVPdc-Q

i'm using the sspx container, and i have seen your video and i have no idea why your heatshield is conducting the heat so well, when mine doesn't break a sweat...

Share this post


Link to post
Share on other sites
  • 0
17 minutes ago, Flavio hc16 said:

i'm using the sspx container, and i have seen your video and i have no idea why your heatshield is conducting the heat so well, when mine doesn't break a sweat...

One thing I noticed is you are using the Stock heatshield. But if we are both using DRE, that shouldn't make a difference. Also, it's not so much that it is conducting heat as it is there is exposed skin on the Hitchhiker (you can see it in the dialog boxes).

Okay, out of curiosity, I ran the test again without SSPX(r) and without DRE. I had no problems. @Flavio hc16, are you certain you have both of those installed, latest version? I noticed you don't have the DRE icon on screen. If you don't have it installed, and you want to try again, would you mind installing it and giving my craft a try? If you do, would you mind if we shared some logs and config files, so I can figure out what is different between our installs?

 

Edited by eightiesboi
Ran second test without DRE and SSPX

Share this post


Link to post
Share on other sites
  • 0
6 hours ago, eightiesboi said:

There is no reason that I can see why the Mk 1 capsule, which (with the monoprop tanks considered) weighs 1.5t, should flip in front of the Hitchhiker and heatshield, which combined weigh between 4.4t (with full ablator) and 3.6t (if the ablator was completely used up) or 4.1t (which is about where the ablator is when the flip occurs). The crew cabin weighs 1t, so I understand why it would be the fulcrum point, but a lighter object on the end of a cylinder shouldn't flip in front of a heavier object at the other end of a cylinder in these circumstances (1900m/s+ in the direction of the heavier end). 

I see a misconception here. 

The fact that "part in front" is heavier than "part in rear" is not enough to ensure the craft will not flip. 

Fist because "portion of the craft in front of CoM" have exactly the same weight than "portion of the craft behind of CoM". The CoM is the pivot point.

Second because how much drag a parr is causing have nothing to do with its mass.

 

 

 

Share this post


Link to post
Share on other sites
  • 0
28 minutes ago, Spricigo said:

I see a misconception here. 

The fact that "part in front" is heavier than "part in rear" is not enough to ensure the craft will not flip. 

Fist because "portion of the craft in front of CoM" have exactly the same weight than "portion of the craft behind of CoM". The CoM is the pivot point.

Second because how much drag a parr is causing have nothing to do with its mass.

 

 

 

I don't think there's a misconception. I understand that the CoM is the pivot point. The CoM at its furthest back is right about where the crew cabin meets the Hitchhiker. That's why I referenced it. The majority of mass is concentrated at the retrograde front of the vessel. And yes, I agree with you that drag is unrelated to mass, but I think we all agree that an object with its mass concentrated in its heavy blunt end and a relatively light tapered end will travel blunt end first during reentry (or it should). 

Share this post


Link to post
Share on other sites
  • 0

All of you that made suggestions, thank you. I ran another series of tests, this time substituting the stock Squad heat shield for the DRE one (which is what I used on the original vessel). When I used the Squad heat shield, on the same vessel I started with, I did not flip. When I used the DRE heat shield, I did. Since the shields *should* perform identically, I will report this as a possible bug in the mod. 

Again, thank you all.

Share this post


Link to post
Share on other sites
  • 0
9 hours ago, eightiesboi said:

.but I think we all agree that an object with its mass concentrated in its heavy blunt end and a relatively light tapered end will travel blunt end first during reentry (or it should)

No, we don't agree with this. And not even is important if we agree with this or not because we need to accept the fact regardless of if it meet or expectations.

The drag is concentrated further ahead of where the mass is concentrated. This cause your craft to flip. In other word, the mass of your ship is not concentrated in the blunt end.

Share this post


Link to post
Share on other sites
  • 0
3 hours ago, Spricigo said:

No, we don't agree with this. And not even is important if we agree with this or not because we need to accept the fact regardless of if it meet or expectations.

The drag is concentrated further ahead of where the mass is concentrated. This cause your craft to flip. In other word, the mass of your ship is not concentrated in the blunt end.

I appreciate your input. We disagree. 

Share this post


Link to post
Share on other sites
  • 0

It's less about mass than it is about density and aerodynamics.  Just eyeballing it, the big one's 4 times the crew cabin's volume, probably 8+ times the capsule's.

Share this post


Link to post
Share on other sites
  • 0
14 hours ago, eightiesboi said:

I appreciate your input. We disagree. 

Didn't matter if you agree with me. The issue is that the game don't agree with you. 

And, unlike me, the game will not give up if you refuse to acknowledge how aerodynamics works.

 

 

Share this post


Link to post
Share on other sites
  • 0
5 minutes ago, Spricigo said:

Didn't matter if you agree with me. The issue is that the game don't agree with you. 

And, unlike me, the game will not give up if you refuse to acknowledge how aerodynamics works.

 

 

Actually, as I posted a day ago, changing the heatshield to stock solved my problem, which may have been tweakscale-related. The game agrees with me and my understanding of aerodynamics just fine now. My test vessel is stable through descent without any flipping, even if I nerf the reaction wheel authority. Again, thank you and everyone else who responded. The last few exchanges give me something to consider if a future vessel proves to be unstable. 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now