On ablator behaviour during reentry

[Gaarst]

Playing RSS, where heatshields are absolutely necessary, I stumbled across an issue with ablator behaviour, depending on the reentry profile.

While doing returns (and reentries) from the Moon with a Mk1-2 Command Pod and a 2.5m heatshield below it I saw that ablator seems to run out unnaturally (at least for me) with some reentry profiles. Just to clear things out, I'm playing with stock aerodynamics, heatshields, thermals, and 100% thermal heating.

To illustrate my issue, I have two examples (no screenshots, but I'll do my best to describe what's happening):

1. Straight Moon return, ballistic reentry: From an apoapsis at the Moon's height, I lowered my periapsis to ~40km to aerocapture on the first pass. The craft went down fine, entering atmosphere at 11km/s and slowing down consuming its ablator as expected. At splashdown, I have 0.5 units of ablator left out of 250 initially and max-G force was at 6.5g.
   
2. "Skip" reentry: From an apoapsis at the Moon's height (again), I lowered my periapsis to ~52km to try and make a softer reentry; I also wanted to try a controlled reentry, with a controlled angle of attack to lower G-forces on the craft. The craft entered atmosphere at about 11km/s, went down to about 48km and then up again, until it exited atmosphere. During the reentry, I kept SAS on to maintain a certain angle of attack. The problem is, when I exited atmosphere, the capsule still had an orbital velocity of 7.8km/s, but consumed almost all of its ablator: <1 unit left out of 300 initially. And I still had to make a LEO reentry to get down to Earth. I didn't make it...
   

This bothers me greatly, because for me, ablator consumption is proportional to heat endured by the craft. The heat itself comes from the transformation of kinetic energy (orbital speed) to mechanical (energy) shockwave when entering atmosphere) to thermal energy (heating while reentring). Let's forget about mechanical energy and focus on kinetic and thermal: kinetic energy is proportional to the square of the speed, so the more speed you lose, the more more energy you have to dissipate through heating: KE = 1/2 * m * v²

I will ignore gravitational potential energy because both crafts have the same initally, but it does make the issue bigger (PE is higher if you're higher, so at 130km, it is higher than at 0km)

Assume m = 3000kg.

On my first example, I had an initial speed of 11km/s, and went down to 0: KE = 180 GJ. All kinetic energy was dissipated, so dissipated energy is 180 GJ.

On my second example, same initial speed: KE = 180 GJ. But I had still 7.8km/s orbital velocity when I exited atmosphere: so dissipated energy is actually 90 GJ.

I know I should account for the phase where mechanical friction slows down the craft and produces less heat, but it happens at lower speeds and altitude, so it is not really going to change the result here. Even if it is significant, it is far from half of total energy dissipated.

So far, these results are OK: I didn't go as deep in the atmosphere for the second reentry and didn't capture, so it is normal that energy dissipated is lower.

But then, why did I consume more ablator on the second reentry (~300 units) than I did on the first one (~250 units) ?

Keep in mind that in the second example, I still had to do a LEO reentry to be able to land on Earth.

Now, I see three possible solutions to this problem:

1. I s*ck at thermal physics and aerodynamics and I missed something huge when making my approximated calculations, which makes my conclusion completely wrong. If so, please help me find my (big) mistake.
   
2. Somehow, the time the heatshield is exposed to hypersonic airflow plays a role in the ablator consumption: the longer you stay exposed, the faster your ablator is burnt off. That would explain why, during my second example, in which the craft stays longer at hypersonic speeds (going and up again, without slowing down enough to change heat "regime"), more ablator is consumed while total energy dissipated is twice smaller. This or another factor which makes ablator consumption not linear.
   
3. KSP does not model aerodynamics and ablator/heating mechanics correctly. We all know this is kind of true anyway, but I wouldn't inmagine that KSP was that far from realistic/plausible results...
   

I would like to see what other people think about this, and maybe help me find a solution. If anyone would like more info, screenshots or whatever, ask and I will provide it.

Thank you for reading.

TL;DR: Too much ablator goes off for skip reentries, why ?

Here is the solution of this problem, given by NathanKell

Edited September 13, 2015 by Gaarst
Changed to "Answered" and added link to answer

[The_Rocketeer]

I don't know.

However I do know there are two kinds of heating in KSP, and ablator degradation relates to surface heating rather than 'core' heating. It could be that your first attempt is shedding more energy into 'core' heating throughout the ship rather than into surface heating. That would share the load a bit and save you some ablator.

I read about this here: http://forum.kerbalspaceprogram.com/threads/133348-Heat-shield-exploding-before-ablating

[Randazzo]

I don't know about all that, but I do know this:

Ablator consumption occurs at a rate dependent on the amount of ablator you have in total once you've reached the threshold temperature of the heatshield or other object that contains the ablator. Once you've hit that threshold, heat in excess of the max amount shed by the ablator resource has no effect on it aside from heating up the part itself.

So it doesn't matter, in terms of ablator consumption, if you spend 10 minutes at 602 degrees (or Kelvin or whatever it's supposed to be) or 10 minutes at 1500, you will lose the same amount of ablator.

Edited September 12, 2015 by Randazzo

[Gaarst]

I don't know.

However I do know there are two kinds of heating in KSP, and ablator degradation relates to surface heating rather than 'core' heating. It could be that your first attempt is shedding more energy into 'core' heating throughout the ship rather than into surface heating. That would share the load a bit and save you some ablator.

I read about this here: http://forum.kerbalspaceprogram.com/threads/133348-Heat-shield-exploding-before-ablating

I did thought about that, but it does help solve the problem: it's actually the opposite.

When you do a single-pass reentry (first attempt) you dive in the denser parts quicker, so the "skin" heats more (convection flux is applied to skin temperature) and "core" has less time to heat: conduction from skin to core takes time.

In my second attempt though, I stayed longer in the less dense parts of the atmosphere, so core had more time to heat, and skin shouldn't heat more than on the first attempt (because convection less less intense in high atmosphere). Therefore I should have consumed less ablator on the second attempt, as more heat should have been conducted to core.

- - - Updated - - -

I don't know about all that, but I do know this:

Ablator consumption occurs at a rate dependent on the amount of ablator you have in total once you've reached the threshold temperature of the heatshield or other object that contains the ablator. Once you've hit that threshold, heat in excess of the max amount shed by the ablator resource has no effect on it aside from heating up the part itself.

So it doesn't matter, in terms of ablator consumption, if you spend 10 minutes at 602 degrees (or Kelvin or whatever it's supposed to be) or 10 minutes at 1500, you will lose the same amount of ablator.

That could indeed explaine my issue.

But I think that it's counter-intuitive that the same amount of ablator is consumed regardless of temperature. Then again, I don't know much about real ablative shields, so I might be wrong.

[Randazzo]

That could indeed explaine my issue.

But I think that it's counter-intuitive that the same amount of ablator is consumed regardless of temperature. Then again, I don't know much about real ablative shields, so I might be wrong.

I've no idea how real ablative shielding works in comparison, but I imagine it might actually be less forgiving. Real ablator could be instantly vaporized resulting in instant failure.

[The_Rocketeer]

Yes I think Randazzo has the answer. Effectively, the shedding of ablator has a cooling effect on the ablator's skin. Once up to temperature, the rate of shedding is constant, but the rate of heating is proportional to velocity and air density. As long as the sum of accumulated skin heat is lower than the destruction threshold, and as long as the descent takes less time than the time taken to shed all the ablator, you're fine. However, your skip re-entry is keeping the ablator up to temperature for too long, partly because it isn't benefiting from the cooling effect of the lower atmo on the way back up (which the direct re-entry does on the way down).

For interest's sake, from Wikipedia:

The ablative heat shield functions by lifting the hot shock layer gas away from the heat shield's outer wall (creating a cooler boundary layer). The boundary layer comes from blowing of gaseous reaction products from the heat shield material and provides protection against all forms of heat flux. The overall process of reducing the heat flux experienced by the heat shield's outer wall by way of a boundary layer is called blockage. Ablation occurs at two levels in an ablative TPS: the outer surface of the TPS material chars, melts, and sublimes, while the bulk of the TPS material undergoes pyrolysis and expels product gases. The gas produced by pyrolysis is what drives blowing and causes blockage of convective and catalytic heat flux. Pyrolysis can be measured in real time using thermogravimetric analysis, so that the ablative performance can be evaluated.[17] Ablation can also provide blockage against radiative heat flux by introducing carbon into the shock layer thus making it optically opaque. Radiative heat flux blockage was the primary thermal protection mechanism of the Galileo Probe TPS material (carbon phenolic). Carbon phenolic was originally developed as a rocket nozzle throat material (used in the Space Shuttle Solid Rocket Booster) and for re-entry vehicle nose tips.

Edited September 12, 2015 by The_Rocketeer

[Vim Razz]

But I think that it's counter-intuitive that the same amount of ablator is consumed regardless of temperature. Then again, I don't know much about real ablative shields, so I might be wrong.

The confusion seems to lie with the impression that the ablator operates by transferring excess heat away from the vehicle? (Like dipping a hot pot into water, so that the steam generated -- ablating water, so to speak -- transfers heat away from the metal surface.)

It doesn't. As noted by the wiki there, it creates an insulating boundary layer of gas with (relatively) low specific heat as it vaporizes, preventing thermal conduction into the vehicle as much as possible. Until it runs out.

[Hagen von Tronje]

I'm 99% sure that Randazzo's answer is pretty much the whole of it, at least in-game. Realistic or not, ablator lifespan is pretty much a function of how long the heat shield spends exposed to skin temperatures above 600K. Longer reentries will indeed consume more ablator even if they are gentler reentries, so in cases where such things matter, an "optimal" reentry will presumably want the shortest reentry trajectory possible without exceeding the skin heat threshold of any parts not fully occluded during reentry.

[The_Rocketeer]

Made me think of this. I guess the effects are comparable, tho in this case the medium (hot water) represents the ablator, and the ball represents the atmosphere:

Edited September 12, 2015 by The_Rocketeer

[NathanKell]

Sorry, Randazzo, that's not at all correct.

Ablator loss in units = lossConst * e^(lossExp / part.skinTemperature) * ablatorResource.amount, clamped to 0 if part.skinTemperature is less than ablationThreshold.

The cooling flux is then just the loss in units * density * resource.hsp * pyrolysisLoss

What's actually going on here is an interesting fact about reentry, which is that the convective transfer coefficient is a function of the square root of density, while drag is a (linear) function of density. That means that, until you hit turbulent flow*, a steeper reentry has a higher peak flux but a lower heat load. This doesn't mean a lot at KSP scale, but in RSS it means that if your shield can take a high instantaneous flux, and your payload can take the Gs, it's better to go steep.

*When you hit turbulent flow, the convective coefficient spikes up. The switch from laminar to turbulent is a function of (density * velocity) so if you're still going fast down low (which you will be, the steeper your reentry) you can hit it. That never ends well (and it's what leads to the general rule that the heat load of a reentry is similar across all entry angles).

For example, Mercury had a very steep ballistic reentry (the reentry orbit was around 320km x -200km, after 170m/s retrofire from a 150x250km original orbit) to minimize heat loading on the capsule.

EDIT: You can grab AeroGUI, it has a convective flux counter (hit alt-I, then click thermal, and reset the counter just prior to atmospheric interface, then compare total loads from different reentries). http://forum.kerbalspaceprogram.com/threads/117141-1-0-AeroGUI-v1-0-28-Apr

[Gaarst]

Thank you all for your answers !

I think I got what NathanKell means, and I had no idea about the actual cause of this issue ^^

I'll try the AeroGUI thing to try to learn to play with this mechanic to get the best reentries.

Marking the thread as "Answered".

[cantab]

As far as KSP goes, ablator will deplete any time the heatshield is hot enough. (It will deplete if you fly too close to the Sun; a Moho mission in stock is liable to destroy ablator). It's possible that on the skip the heatshield was slow shedding that excess heat and so still ran down the ablator.

[Randazzo]

Sorry, Randazzo, that's not at all correct.

Ablator loss in units = lossConst * e^(lossExp / part.skinTemperature) * ablatorResource.amount, clamped to 0 if part.skinTemperature is less than ablationThreshold.

The cooling flux is then just the loss in units * density * resource.hsp * pyrolysisLoss

What's actually going on here is an interesting fact about reentry, which is that the convective transfer coefficient is a function of the square root of density, while drag is a (linear) function of density. That means that, until you hit turbulent flow*, a steeper reentry has a higher peak flux but a lower heat load. This doesn't mean a lot at KSP scale, but in RSS it means that if your shield can take a high instantaneous flux, and your payload can take the Gs, it's better to go steep.

*When you hit turbulent flow, the convective coefficient spikes up. The switch from laminar to turbulent is a function of (density * velocity) so if you're still going fast down low (which you will be, the steeper your reentry) you can hit it. That never ends well (and it's what leads to the general rule that the heat load of a reentry is similar across all entry angles).

For example, Mercury had a very steep ballistic reentry (the reentry orbit was around 320km x -200km, after 170m/s retrofire from a 150x250km original orbit) to minimize heat loading on the capsule.

EDIT: You can grab AeroGUI, it has a convective flux counter (hit alt-I, then click thermal, and reset the counter just prior to atmospheric interface, then compare total loads from different reentries). http://forum.kerbalspaceprogram.com/threads/117141-1-0-AeroGUI-v1-0-28-Apr

Never noticed any observable difference in stock KSP, time spent over the ablator threshold is time spent over the ablator threshold. It only ever seems to change as the ablator gets depleted, which makes sense with that formula there.

Just tried my first reentry in RSS. Didn't go well.

[NathanKell]

It's because as you go lower, the flux increases and the temperature increases, so the loss rate would be higher...but it's also multiplied by the remaining quantity of ablator, so the more you lose the slower you lose. Those balance.