Unknown Heat Source During Eve Atmospheric Insertion

[Zosma Procyon]

The latest and largest version of my lander for getting kerbals off of Eve has run into new problems I can't really figure out the cause of, much less remedy. During entry, even a very slow entry that more than circumnavigates the planet before losing enough velocity to descend in the lower atmosphere, part of my rocket is overheating to the point of exploding, causing the rocket to disintegrate. As you can see in the pictures below, there are four inflatable heat shields in front, nine in back for stability. What you can't see is that a central cluster of seven Vector engines, each of which has the widest conventional heat shield attached to it. Also the rocket is covered in parachutes and radiators. I'm not sure if radiators actually do anything during atmospheric entry, but they seem to buy me more time before the thing explodes.

Here the rocket is during early insertion. The heat shields are hot, but they don't get much hotter.

This picture shows the trouble area. As you can see in the first picture, there is no way for hot air to get through to that point. I'll explain more what you see here with the following two pictures.

The airstream shell is there to protect a vector engine, the final stage's power plant, from hot air that hot be getting through. I honestly don't now if they do that.

I have no idea where the heat causing that section of my rocket is coming from, or even if it is really heating up or hat matter. The radiators on that section have never read above 15%. All of the radiators on this rocket are attached via radial decouplers, so they can be jettisoned once the parachutes are fully inflated at about 2300 meters. When the explosions start, they seem to start around there. They might also starting n the central Vector engine cluster, but I have no way to really know. In the flight events log, it usually lists structural failures between the adapter fuel tanks above the final Vector Engine, as well as and exploding Vector engine somewhere in the rocket. FYI: I'm using vector engines because they have the highest in atmosphere thrust percentage, and can be easily clustered.

So if anyone has any ideas what is causing this problem, please say so. And does anyone know if radiators actually do anything during Eve entry? Also, if anyone has a design for a Eve liftoff vehicle that weighs less than 700 tons (my current design is over 1100) and can lift 16 or more kerbals off of Eve in one go, please share it.

[Kerbalstar]

I'm pretty sure that a structural failure is having to much weight on the connection. Try autostrutting and rigid attaching everything, and if that doesn't work manual struts. I have those problems all the time.

[Corona688]

I think I've spotted the mysterious heating source:  Your entire spaceship.  The whole thing is glowing red-hot!  I think your heat shields only protect the parts immediately behind them.

Radiators do help, but they cool things down by by making themselves hotter.  You have to keep them cool somehow or they'll just heat up until overload and stop.

Unlike radiators, thermal control systems will cool the entire ship, not just nearby parts, so you don't have to stick them in the searing re-entry heat.  If you put a bunch of thermal control systems right behind the rear heat shields, where they can actually get rid of heat, that should help.  They are fragile however and may be destroyed by aerodynamic forces.  Point them straight backwards, maybe.

Edited January 18, 2018 by Corona688

[Tyko]

It's very possible to descend too slowly. The upper atmosphere doesn't offer much drag, so cruising through it overly long will cause you to slowly build up heat without shedding much velocity. This can be demonstrated with basic capsules around Kerbin. Try setting a 50 - 60k PE and watch how much your capsule heats up compared to a 30-35k PE.

I'm not as familiar with Eve, but the same principles should apply. You could always run some test descents with really simple capsules to see what the optimum trajectory is. Just F12, place your test craft in Eve orbit and try a series of descents. Call it 'computer modeling' 

[Foxster]

In my experience, radiators or thermal thingies do little, if anything, to help with heating during re-entry. I'd say that was a dead-end you can ditch. 

The other thing is to step back and say the craft is just too darned big. Have a look at some of the other Eve craft on this site (try the challenges section) for inspiration. 

You might be having problems with the part count. The more parts the more weird stuff happens, like parts suddenly exploding for no apparent reason, usually when under thermal or physical stress. Getting rid of the rat's nest of struts would be a good start and use autostrut instead. 

Finally - I agree on the entry angle being a potential problem. There is likely a fairly narrow window that is going to work. Try everything from 10 to 60km Pe and see what works best. 

 

[FleshJeb]

Have you tried getting rid of the radiators? They're basically useless for shedding heat while moving, and they're super-draggy. The fairing (if properly constructed) should protect the Vector with no problems.

I suspect that the Vector, or a nearby part, is experiencing drag. The heat and drag occlusion model in KSP is pretty bad. Things that look like they should work, don't. Intuitive aerodynamic design in KSP is a myth. You have to look at the numbers.

The problem might be the rapid size transition from the fairing baseplate to the engine, and a bad fairing. What I would do is flip on the Thermal and Drag cheats for the pop-up menus, (Alt-F12 menu) then right-click on the engine as the craft aerobrakes. Look at the parts in the area. If anything is experiencing a lot of drag, that's the culprit.

Also, Corona, Tyko, and Foxster are probably right.

Edited January 18, 2018 by FleshJeb

[Snark]

Moving to Gameplay Questions.

On 1/17/2018 at 7:59 PM, Zosma Procyon said:

part of my rocket is overheating to the point of exploding, causing the rocket to disintegrate

Do you actually know this?  I.e. do you actually know that it's overheating, and what was overheating, and that it wasn't something else causing a problem?

What does the detailed damage report say, exactly?  (You know, the one that pops up when you hit F3.)  Could you post a screenshot of that?

Also, here's a red flag:

On 1/17/2018 at 7:59 PM, Zosma Procyon said:

even a very slow entry that more than circumnavigates the planet before losing enough velocity to descend in the lower atmosphere

...be wary!  It sounds as though you're thinking "descending slowly = safer", like dipping your toe into a hot bath-- but it doesn't work that way.  It's actually quite the opposite in many cases:  you're more likely to fry by going slowly than if you just plow straight in.

22 hours ago, Tyko said:

It's very possible to descend too slowly. The upper atmosphere doesn't offer much drag, so cruising through it overly long will cause you to slowly build up heat without shedding much velocity.

^ This.  A very-very-shallow approach gradually roasts you without slowing you much, so by the time you get down to where there's enough drag to actually start slowing you, your ship is already toasty and doesn't have much headroom to warm up further before it starts slowing down.

 

[Zosma Procyon]

7 hours ago, Snark said:

Moving to Gameplay Questions.

Do you actually know this?  I.e. do you actually know that it's overheating, and what was overheating, and that it wasn't something else causing a problem?

What does the detailed damage report say, exactly?  (You know, the one that pops up when you hit F3.)  Could you post a screenshot of that?

Also, here's a red flag:

...be wary!  It sounds as though you're thinking "descending slowly = safer", like dipping your toe into a hot bath-- but it doesn't work that way.  It's actually quite the opposite in many cases:  you're more likely to fry by going slowly than if you just plow straight in.

^ This.  A very-very-shallow approach gradually roasts you without slowing you much, so by the time you get down to where there's enough drag to actually start slowing you, your ship is already toasty and doesn't have much headroom to warm up further before it starts slowing down.

 

I'm trying to use as little Delta-V as possible to know it out of orbit.

I will clarify what I've been doing. I have been performing this tests in a Sandbox saved game, and using the Set Orbit cheat to put the test articles (my big honking rockets) into initial orbits around 106-120 km around Eve, usually in Polar Orbits alligned to pass over the highlands. I want to use as little Delta-V as possible to knock it out of orbit so I eventually have to launch less rocket from Kerbin. Eventually think I may launch it with most of the lander's tanks empty, and fuel it in Eve orbit. 

The latest tests, all spectacular failures due to random parts of the rocket flying off even though they're rigidly connected and autostructed, have started with deorbit burns of less than 100 m/s, just enough to dip the periapsis below 86km. But from the feed back i'm getting here, I'll try deorbit burns of between 400-600 m/s, and go in steeper. That will make it easier to land where I want too. Had my last few failures succeeded, they would have landed in the ocean, or on the cliffs at the outskirts of the northern polar region. Not good locations for Eve surface launch tests.

[Snark]

26 minutes ago, Zosma Procyon said:

The latest tests, all spectacular failures due to random parts of the rocket flying off even though they're rigidly connected and autostructed, have started with deorbit burns of less than 100 m/s, just enough to dip the periapsis below 86km. But from the feed back i'm getting here, I'll try deorbit burns of between 400-600 m/s, and go in steeper. That will make it easier to land where I want too. Had my last few failures succeeded, they would have landed in the ocean, or on the cliffs at the outskirts of the northern polar region. Not good locations for Eve surface launch tests.

That sounds like a plan.

Here's another potential failure mode which occurs to me.  Sorry, didn't occur to me sooner, probably because after a few years of playing KSP, "do not under any circumstances ever do a really slow reentry because I'll fry" has been ground so firmly into my KSPing instincts that I tend not to think about what happens if one does that:

Heat conducts.

If you put a hot thing in contact with a cold thing, heat will gradually leach from the hotter to the colder.  How fast it transfers depends on the parts involved, but it's a nonzero rate.

So let's consider what happens when you try a really slow I-want-to-fry-myself atmospheric entry such as you've been trying, using the big 10m heat shield.

That heat shield has no ablator, so it has no particular ability to eliminate heat.  It simply has a really high max temperature it can stand (much higher than most other parts), and I'm guessing a relatively low thermal conductivity (so it doesn't immediately explode the thing that it's connected to).  So let's compare what happens on a "normal" reentry, and compare that with the slow bake you've been doing:

• Normal reentry:  Heat shield rapidly gets really hot, e.g. over 3K degrees.  The thing next to it starts to warm up.  However, because you're diving deep into the atmosphere and the big heat shield is super draggy, the craft slows down pretty rapidly, and fairly soon, it's slow enough that it's done with the "hot" part of reentry.  The rest of the ship doesn't blow up or anything, because the "bake time" is only a minute or two and finishes quickly enough that there's not enough time to transfer a lot of heat from the heat shield to the rest of the craft.
• Slow-bake reentry:   The front end of the craft experiences blowtorch heat, and stays that way for a really really long time.  So the heat shield gets up over 3K degrees... and sits there.  And heat starts to leach.  And the rest of the craft starts to warm up, because there's plenty of time for the heat to gradually percolate through the ship as it tries to equalize the temperature.  And so the heat shield stays at its constant hot temperature, while the rest of the craft gradually gets warmer ... and warmer ... and warmer.  It's not going to stop until either you slow down enough to not be hot anymore, or else the entire ship is the same temperature as the heat shield.  But since you're on a slow-bake reentry, the ship takes way too long to slow down, and the temperature starts to rise and rise until the first part gets over its tolerance (probably at 2K degrees), then that part explodes and the ship quickly goes kablooie.

Basically, once you start a reentry sequence, a clock starts ticking, and you need to decelerate below charbroil speed before the clock runs down.  How fast the clock runs down depends on the thermal properties of the parts involved and also on the ballistic coefficient of your ship, but given typical KSP designs, a slow-bake reentry will often take more time than you have available.

[Zosma Procyon]

55 minutes ago, Snark said:

That sounds like a plan.

Here's another potential failure mode which occurs to me.  Sorry, didn't occur to me sooner, probably because after a few years of playing KSP, "do not under any circumstances ever do a really slow reentry because I'll fry" has been ground so firmly into my KSPing instincts that I tend not to think about what happens if one does that:

Heat conducts.

If you put a hot thing in contact with a cold thing, heat will gradually leach from the hotter to the colder.  How fast it transfers depends on the parts involved, but it's a nonzero rate.

So let's consider what happens when you try a really slow I-want-to-fry-myself atmospheric entry such as you've been trying, using the big 10m heat shield.

That heat shield has no ablator, so it has no particular ability to eliminate heat.  It simply has a really high max temperature it can stand (much higher than most other parts), and I'm guessing a relatively low thermal conductivity (so it doesn't immediately explode the thing that it's connected to).  So let's compare what happens on a "normal" reentry, and compare that with the slow bake you've been doing:

• Normal reentry:  Heat shield rapidly gets really hot, e.g. over 3K degrees.  The thing next to it starts to warm up.  However, because you're diving deep into the atmosphere and the big heat shield is super draggy, the craft slows down pretty rapidly, and fairly soon, it's slow enough that it's done with the "hot" part of reentry.  The rest of the ship doesn't blow up or anything, because the "bake time" is only a minute or two and finishes quickly enough that there's not enough time to transfer a lot of heat from the heat shield to the rest of the craft.
• Slow-bake reentry:   The front end of the craft experiences blowtorch heat, and stays that way for a really really long time.  So the heat shield gets up over 3K degrees... and sits there.  And heat starts to leach.  And the rest of the craft starts to warm up, because there's plenty of time for the heat to gradually percolate through the ship as it tries to equalize the temperature.  And so the heat shield stays at its constant hot temperature, while the rest of the craft gradually gets warmer ... and warmer ... and warmer.  It's not going to stop until either you slow down enough to not be hot anymore, or else the entire ship is the same temperature as the heat shield.  But since you're on a slow-bake reentry, the ship takes way too long to slow down, and the temperature starts to rise and rise until the first part gets over its tolerance (probably at 2K degrees), then that part explodes and the ship quickly goes kablooie.

Basically, once you start a reentry sequence, a clock starts ticking, and you need to decelerate below charbroil speed before the clock runs down.  How fast the clock runs down depends on the thermal properties of the parts involved and also on the ballistic coefficient of your ship, but given typical KSP designs, a slow-bake reentry will often take more time than you have available.

The current version I'm working on has 14 inflatable heat shields (5 fore and 9 aft), around 100 parachutes, dozens of panel radiators, dozens of thermal control systems placed wherever they might be shielded from the entry air stream, 8 air breaks, 40 heavy duty landing legs, and a partridge in a pair tree. I want it to at least get to the point where it is descending slowly under all of those chutes so I can test the radiator dumping stage.

[foamyesque]

I think you're probably overbuilding, @Zosma Procyon, and it's likely the root cause of your issues. You don't need all that to land on Eve at all. The radiators can shield parts behind them from thermal flux (like anything else) and can do so a bit more effectively than most because of their high temperature tolerances, but they don't actually act as active radiators during entry. They can be deleted entirely. Likewise you're running an awful lot of struts, which will ruin your aerodynamics on the way up unless you get rid of them.

Me, I don't like parachutes; I actually think I've said this in another question of yours :D. They produce a jolt to the ship that can cause structural failures and you can't pick your landing spot. I'm big on winged landers, which if you fly a cobra entry on Eve will bring you to a stop really quickly; an 100x80 orbit will become an 80x0 trajectory by the time you reach 80km. Your huge inflatable heatshield drag should be doing the same...

 

Are you landing fully fuelled? Because that makes things more difficult; your ballistic coefficient is much larger and it takes a lot more braking to execute a good entry or landing, and the stresses on the entire ship become harder to handle. Particularly with large Eve missions, ISRU is an alternative to consider; you can use your lander's tanks to get it to Eve, which saves enormously on launch, then land it empty/nearly empty, ISRU your tanks back up, and then leave.

 

As far as returning six Kerbs go, I give you this:

 

(Ship file: https://www.dropbox.com/s/hh7h2jsqfn7isvr/Eve - 6man.craft?dl=0)

It needs some work to make it a fully mission capable plane, but the ascent is easily doable; it'll take six Kerbs to Eve escape from 800m.

Edited January 20, 2018 by foamyesque

[Zosma Procyon]

Well I almost just had a successful test, but struct a kraken. My main chutes were deployed and an fully opening, then for no reason it briefly accelerated to a sixth the speed of light and flew apart. This has actually happened a few times. 

I'm going to open a thread to report this kraken.

[foamyesque]

Excessive struts can do weird things where they keep bouncing tension around your craft until a strut snaps, and then you get bits flying everywhere, occasionally (as here) at physically implausible velocities.

[Zosma Procyon]

51 minutes ago, foamyesque said:

Excessive struts can do weird things where they keep bouncing tension around your craft until a strut snaps, and then you get bits flying everywhere, occasionally (as here) at physically implausible velocities.

I'm just going to start over from scratch again.

[Cpt Kerbalkrunch]

@Zosma Procyon, not sure if you're actually looking for advice or just using these threads as sounding boards in a way, but there comes a time when you have to stop and realize that you've stopped trying to design a ship to complete a mission and are now trying to design a mission to fit the ship you've built. I don't really believe any ship is necessarily too big. You can make an Eve lander as large as you want to. And I always manually strut everything; so I don't see a problem with those either. However, there is definitely such a thing as having so much more than you need that it's become detrimental. And I'm not speaking theoretically about what should work, but from experience about what does work. Even the slow descent works just fine for me. The reason these things work is that all the parts of my ship are shielded properly. It usually takes quite a bit of testing to get it right; but it always works. And I have never used a single radiator or thermal control. Those are just completely unnecessary. And not unnecessary like people will say about extra Kerbals. You want to bring extra Kerbals, and that's fine. I do it, too. But unless you just want to bring radiators for some reason, ditch 'em. And all those 'chutes. I'm not even sure what to say. I've told you on another thread that my 500 ton lander uses 15 'chutes. Six drogues and 9 regular. I specifically build each lander with descent engines that first lower my orbit, and then burn again just before touchdown for a smooth landing. I think your design has gotten a bit out of hand. I would recommend you takes step back and look at what you're doing and ask yourself if you can do better. There's no shame in failure in this game. It is not the same as defeat. You are not giving up, just changing tactics. Take all that you've learned and build your ship again from scratch. That sounds like a huge undertaking, but we all do it all the time. And for me, everytime I do, things always work out better. Anyway, just my recommendation. Hope all goes well.

Crap. Kinda got ninja'd by the OP himself. 

Edited January 20, 2018 by Cpt Kerbalkrunch