[1.12.*] Deadly Reentry v7.9.0 The Barbie Edition, Aug 5th, 2021

[Zander]

Nathan my heat shields are overheating before reaching the end of their ablative material. They are just exploding which should not happen especially not at a 7000m/s reentry from low earth orbit. Im using the default settings for RSS with DRE

[Goozeman]

Nathan my heat shields are overheating before reaching the end of their ablative material. They are just exploding which should not happen especially not at a 7000m/s reentry from low earth orbit. Im using the default settings for RSS with DRE

What is your reentry angle? Higher reentry angles mean that your shield must withstand higher temperatures. Ablatives can only ablate so quickly, and that speed is dependent on temperature. It is possible to overwhelm an ablative (or any kind of) heat shield.

[NathanKell]

Zander: Are you using the heat shields patch from Realism Overhaul, like the OP says?

[123nick]

i was aborting from a crashing spacecraft and as my ALCOR pod was almost coming to a stop at 11m/s (far bellow the impact tolerance of said pod) it burned up on re-entry. at 64 meters from sea level. traveling at 11m/s. didnt know kerbins atmosphere was made of thermite in gas form. however, this bug was the first time experiencing it and is proably just a very rare calculation bug.

[NathanKell]

That's...very strange. Never heard of that. Let me know if it happens again.

[Goozeman]

Curious question. Is the rate at which material ablates related to velocity or temperature? I noticed on reentry with a standard Mk1-2 capsule that the heat shield was ablating rather quickly at ~350C then after a certain point it stopped ablating even though it was at ~550C.

[AbeS]

It ablates when it cools down, not when it heats up

[Ralathon]

Curious question. Is the rate at which material ablates related to velocity or temperature? I noticed on reentry with a standard Mk1-2 capsule that the heat shield was ablating rather quickly at ~350C then after a certain point it stopped ablating even though it was at ~550C.

It's both related to temperature and dynamic pressure. Dynamic pressure itself is related to both velocity and pressure. So its perfectly possible to get high ablation rates at 350 degrees as long as go fast enough while low enough.

[Goozeman]

It's both related to temperature and dynamic pressure. Dynamic pressure itself is related to both velocity and pressure. So its perfectly possible to get high ablation rates at 350 degrees as long as go fast enough while low enough.

Interesting. By my reasoning, the temperature of the heat shield is directly related to the incident heat flux (convection from the ram-heated air). The rate of heat loss by pyrolysis should be close to equal to the rate of heat input. If the shield is warming, the heat loss is slightly less than the incident flux, and visa-versa. The rate of pyrolysis is directly proportional to the heat rejected by pyrolysis. The rate of pyrolysis (ablation) should then slightly lag behind the temperature of the shield.

Obviously, there is a lower limit on the temperature at which pyrolysis occurs, but so long as you are above that, the reasoning should hold.

[NathanKell]

I'm very happy to change the ablation model. Here are the inputs:

part temperature

shockwave temperature

degrees C applied to part per second (based on density), whose formula is: (shockwaveTemp-partTemp)^temperatureExponent * density^densityExponent * heatMultiplier

Since DRE does not care about part size or mass or anything like that (at present), it deals in temperature (and temperature change), not heat flux.

So. (1) if you have a formula for heat flux, that would be great. I can calculate frontal area by checking with FAR, and also ballistic coefficient. (2) with that, I guess I can switch to a different modeling of ablation, where there's a minimum ablation temperature and after that, check temperature change between frames and modify based on your rules. How slightly is slightly? Does it vary with shield material?

[BrickedKeyboard]

What's the G-tolerance based on? I noticed around 13 Gs you start "hitting the G-limit" and you die shortly after. http://en.wikipedia.org/wiki/G-force#Human_tolerance_of_g-force Seems to suggest the range is a little broader than that.

I bring this up because we cannot trivially get capsule lift : apparently, real spacecraft use ballast to make their capsules produce lift on reentry?

[NathanKell]

NotMyRealName: The way g tolerance works is in the readme. Basically, I track cumulative G damage. Relevant section:

Kerbals have human-like G-force tolerance, which means they can survive 5Gs for about 16 minutes, 10 Gs for 1 minute, 20Gs for 3.75 seconds, and 30+ Gs for less than three quarters of a second. The tracker is reset when G load < 5. All these are tweakable in the cfg or in the ingame debug menu. It's done by tracking cumulative Gs. The formula is for each timestep, tracker = tracker + G^crewGPower * timestep (where ^ = power). G is clamped to range [0, crewGClamp]. When tracker > crewGWarn, a warning is displayed. When tracker > crewGLimit, then each frame, per part, generate a random number 0-1, and if > crewGKillChance, a kerbal dies in that part. If G < crewGMin, tracker is reset to 0.

What's happening is that though your Kerbals can survive 13Gs for some time (~20 seconds probably), they've been at at lower G level for a while, so their cumulative damage has built up. Now, you subject them to 13Gs for more than a few seconds or so, and they conk out, because they've already suffered damage at lower G levels.

We can get capsule lift: use CoMOffset = 0, 0, X in the part.cfg, where X is some small number like 0.06 for the Mk1-2 pod (play around and see what you get). Presto, ballast.

For the 4m-size one in RO, I'm trying CoMOffset = 0, 0, -0.192

[BrickedKeyboard]

Yes, but the reason real life spacecraft use ballast is apparently you jettison it later before deploying chutes. Just what I read, don't shoot the messenger.

I hope your buddy Medieval Nerd can go and modify the capsule lifts in his realism mod.

[FlowerChild]

Nathan, if you will forgive a moment of self-promotion, I've finally publicly released the mod I've been talking about so much here if you'd like to check it out:

http://forum.kerbalspaceprogram.com/threads/61632-Plugin-0-22-Better-Than-Starting-Manned-Career-mode-redefined?p=837988#post837988

Please let me know if the way in which I reference Deadly Reentry is cool, as I definitely don't want to cause any offense with any of this. Once again, mucho thanks once again for all the help you've provided in making this happen

[NBen]

I'm getting an issue in deadly reentry where the 1.25m heat shield is not working.

I have the DeadlyReentryCont_v4 folder. I extracted it and put the DeadlyReentry folder into the rootfolder/Kerbal Space Program/GameData folder like you are supposed to.

In Career Mode, in the science center, I click on the starter tech tree (the one that you start with), and I notice that it isnt ''researched'' yet. I click to research it and the game asks me to complete the research.

I go to the Vehicle Construction Building and its greyed out, saying it needs to be researched.

I go back to the science center, and it's unresearched again.

I think it has something to do with the parts.cfg for the 1.25m shield (since the .625m shield is unlocked and works properly) but I don't know where I need to look to correct it in the cfg file to properly implement it into the tech tree.

[drtedastro]

It ablates when it cools down, not when it heats up

Abes@, I 'think' that what you are referring to is when the heat shield is like the tiles on the shuttle. They get rid of heat as they cool down, but they do not 'ablate' in the sense that the shields on the Mercury, Gemini and Apollo capsules did. Those blunt heat shield lost heat through the ablative process as they lost "ablated' mass during re-entry and that carried the heat away. The part that was melting broke off and the heat it contained went with it...

Two different methods, two different systems. The method on the shuttle was to be reusable, albeit with some tile repair. The other shields were one use only....

Just one of many refferences... http://www.space.com/22046-nasa-orion-spacecraft-heat-shield.html

[NathanKell]

BrickedKeyboard: I don't recall many cases of ballast being jettisoned; IIRC in some capsules it was moveable, so that you could control the reentry angle. That said, if you want that, use a pod with a separate heatshield and offset the heatshield's CoM.

FlowerChild: as I said there, awesome! The way you're referencing Deadly Reentry is absolutely fine by me.

NBen: weird. I know people have had similar issues with mechjeb. I think it's just KSP being funky with adding a part after a career game has started? Also, in the latest DRE, I think I changed the part's technode, so you can't use it as a decoupler early (since the Mk1 pod already has a shield, you don't need a regular 1.25m one until later).

[Goozeman]

I'm very happy to change the ablation model. Here are the inputs:

part temperature

shockwave temperature

degrees C applied to part per second (based on density), whose formula is: (shockwaveTemp-partTemp)^temperatureExponent * density^densityExponent * heatMultiplier

Since DRE does not care about part size or mass or anything like that (at present), it deals in temperature (and temperature change), not heat flux.

So. (1) if you have a formula for heat flux, that would be great. I can calculate frontal area by checking with FAR, and also ballistic coefficient. (2) with that, I guess I can switch to a different modeling of ablation, where there's a minimum ablation temperature and after that, check temperature change between frames and modify based on your rules. How slightly is slightly? Does it vary with shield material?

As I'm sure you're aware, reentry heating is complex, so this is the best that I can come up with for a simple heating/ablation model.

The shock temperature you have already, and is found simply from the normal or oblique shock relations. There are two primary modes of heat flux into the shield, convection and radiation. Convection dominates at lower re-entry speeds, and radiation dominates at higher speeds. I'll just deal with convection for the time being.

For convective transfer:

q'' = h*(T-T_inf)

where q'' is the heat flux per unit area, h is heat transfer coefficient, T is the temperature of the object, and T_inf is the applied temperature.

Relations for h can be found from the Nusselt number correlations. For laminar flow (valid at high altitudes):

Nu = C*Re^.5*Pr^(1/3)

Nu = h*L/k

Making some simplifications, one can arrive at

h = C'*sqrt(rho)

where C' is a tweakable constant, and rho is atmo density.

For the ablation of the heat shield, the Arrhenius equation should suffice. It is:

dm/dt = B*m*exp(-T_0/T)

where B and T_0 are tweakable constants. One caveat is that the rate of ablation never really hits zero (it gets close) with this model, so a lower cut-off would be beneficial. Or don't and call it vacuum ablation :-P.

Finally then, the rate of heating/heat rejection is

m*c*dT/dt = -dm/dt*L -epsilon*A*sigma*T^4-h*A*(T-T_shock)

where m is the mass of the shield, c is the heat capacity of the shield, L is the heat of pyrolyis of the shield, epsilon is the emissivity of the shield, sigma is the Stefan-Boltzman constant, and A is the area of the shield. d()/dt is the time rate of change of that variable.

Do note that T appears without a finite difference in the rate of ablation and radiation terms, so it must be given in absolute temperature, i.e. either Kelvin or degrees Rankine.

With regard to the tweakables:

C' is simply a function of the flow. Shield properties do not come into play.

B and T_0 are properties of the shield, and must be sorted by trial and error with possibly some help from literature. c, L and epsilon are properties of the shield that could be estimated or obtained from literature about the material.

Finally, radiative or thermal soak type shields can also use this model simply by dropping the ablation term.

Edited December 15, 2013 by Goozeman

[drtedastro]

For the 'thermal soak' type shields, with the ablation term dropped, will there need to be a 'max K' or max temperature that they can absorb / soak up to before they fail?? Will this need to be another term / parameter for the final amount of heat / energy they can take before breaking down and failing?

[Goozeman]

For the 'thermal soak' type shields, with the ablation term dropped, will there need to be a 'max K' or max temperature that they can absorb / soak up to before they fail?? Will this need to be another term / parameter for the final amount of heat / energy they can take before breaking down and failing?

Setting the ablation term with a ridiculously high B and T_0 could have the same effect as a catastrophic failure at a prescribed temperature. A very low B and T_0 would ablate slowly, but also start at relatively low temperatures.

[NBen]

NBen: weird. I know people have had similar issues with mechjeb. I think it's just KSP being funky with adding a part after a career game has started? Also, in the latest DRE, I think I changed the part's technode, so you can't use it as a decoupler early (since the Mk1 pod already has a shield, you don't need a regular 1.25m one until later).

No, there is something wrong there, because I looked at the cfg file for the 1.25m heatshield and its starting technode is listed as "techRequired = start"

so that means it should be available from the get go.

Another thing I've noticed, just to see if there were issues in sandbox mode, but the 1.25m heatshield is listed in the structure tab *twice*, so I'm not sure if there was a copy/paste thing going on, or if there is duplicate code hiding somewhere that's causing a conflict.

here is the code for the config file in question in its entirety. Not sure if you've updated the file in the interim, but I downloaded the DeadlyReentry v4 folder from the dropbox site on Dec 13, 2013.

PART
{
// --- general parameters ---
	name = 1.25_Heatshield
	module = Part
	author = Bobcat

// --- asset parameters ---
	mesh = model.mu
	scale = 1.0
	rescaleFactor = 1.25
	CoMOffset = 0, -0.2, 0

// --- node definitions ---
// definition format is Position X, Position Y, Position Z, Up X, Up Y, Up Z
	node_stack_top = 0.0, 0.06196643, 0.0, 0.0, 1.0, 0.0, 1
	node_stack_bottom = 0.0, -0.01, 0.0, 0.0, 1.0, 0.0, 1

// --- FX definitions ---
	fx_gasBurst_white = 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, decouple
	sound_decoupler_fire = decouple

// --- editor parameters ---
	cost = 200
	category = Structural
	subcategory = 0
	TechRequired = start
	entryCost = 0
	title = 1.25m Heatshield
	manufacturer = Bobcat Ind.
	description = Sturdy thermal shield to keep the fiery death on the outside of the pod. Make sure the shield points to the travel direction while reentering, or the pod may still get heated up. Recommended for any pods reentering atmospheres.

// attachment rules: stack, srfAttach, allowStack, allowSrfAttach, allowCollision
	attachRules = 1,0,1,1,0

// --- standard part parameters ---
	mass = 0.2
	dragModelType = default
	maximum_drag = 0.2
	minimum_drag = 0.2
	angularDrag = 2
	crashTolerance = 9
	breakingForce = 630
	breakingTorque = 630

	maxTemp = 1800

	fuelCrossFeed = False

	MODULE
	{
		name = ModuleHeatShield
		direction = 0, -1, 0 // bottom of pod
		reflective = 0.05 // 5% of heat is ignored at correct angle
		ablative = AblativeShielding
		loss
		{ // loss is based on the shockwave temperature (also based on density)
			key = 650 0 // start ablating at 650 degrees C
			key = 1000 64 // peak ablation at 1000 degrees C
			key = 3000 80 // max ablation at 3000 degrees C
		}
		dissipation
		{ // dissipation is based on the part's current temperature
				key = 300 0 // begin ablating at 300 degrees C
				key = 500 360 // maximum dissipation at 500 degrees C
		}
	}
	RESOURCE
	{
		name = AblativeShielding
		amount = 250
		maxAmount = 250
	}

	MODULE
	{
	    name = ModuleDecouple
	    ejectionForce = 40
	    //explosiveNodeID = top
		isOmniDecoupler = true
	    staged = true
	}
}

[NathanKell]

Goozeman--thanks so much!

When I get some time after .23, in combination with hooking this into FAR's atmospheric density system (and using code Ferram is helpfully giving me for determining shockwave temperature in non-Earth atmospheres) I'll try to implement that model for ablation.

Also, a PSA. As part of that effort (and necessary for it) I plan to make DRE require FAR. It will make my coding life much easier, and also means fewer environments for me to try to balance for and support.

Speak now if you Have Issues with this plan.

(FlowerChild, I know right now you don't support FAR, but you were looking into it with KIDS. How much of an issue would it be for you if DRE required FAR?)

[NathanKell]

NBen: I download DRE v4. I start a new career. 1.25m heatshield shows up, and there's no duplication (There is a similar-looking 0.625m heatshield in the parts list, however).

All I can say is, if you have duplicate shields, you have an install issue; and if it's not unlocking for you, you've been hit by the KSP issue where adding mod parts to an install, and resuming a career that has already unlocked the node they're in, can sometimes lead to issues even if you go back and unlock the part. That's a KSP thing, don't think I can do anything about it on my end...

[drtedastro]

Goozeman--thanks so much!

When I get some time after .23, in combination with hooking this into FAR's atmospheric density system (and using code Ferram is helpfully giving me for determining shockwave temperature in non-Earth atmospheres) I'll try to implement that model for ablation.

Also, a PSA. As part of that effort (and necessary for it) I plan to make DRE require FAR. It will make my coding life much easier, and also means fewer environments for me to try to balance for and support.

Speak now if you Have Issues with this plan.

(FlowerChild, I know right now you don't support FAR, but you were looking into it with KIDS. How much of an issue would it be for you if DRE required FAR?)

Will it be 'possible' that we can turn off FAR for landing or have some method of it working with MechJeb for Landings and such. Any method so that FAR and MJ play well together would be great. does not have to be perfect, but just so they will be 'kind' to each other.....

Then all would be perfect, I love DRE, FAR and MJ... Thanks for asking and thanks for all of your work, to all.....

[NBen]

NBen: I download DRE v4. I start a new career. 1.25m heatshield shows up, and there's no duplication (There is a similar-looking 0.625m heatshield in the parts list, however).

All I can say is, if you have duplicate shields, you have an install issue; and if it's not unlocking for you, you've been hit by the KSP issue where adding mod parts to an install, and resuming a career that has already unlocked the node they're in, can sometimes lead to issues even if you go back and unlock the part. That's a KSP thing, don't think I can do anything about it on my end...

alright, I'll look into it on my end.