Has Kerbin's atmosphere been changed with 0.21?

[Tex_NL]

In 0.20 I could get pretty much any SSTO plane I build into orbit with relative ease. I knew the engine to intake ratio's and when to switch to rockets.

But now, in 0.21, I've got a nearly identical copy (only minor adjustments to compensate for KSPX integration, the new SAS and some CoG/CoL balancing) of one of my 0.20 planes that could reach LKO with ease can't break atmo. Pure jet engine designs that previously reached a 280x40 km suborbital flight struggle to reach anything past 40 km.

Edit: Source of the problem has been found. The issue is not with the stock atmosphere rather with the B9 intakes.

Edited July 28, 2013 by Tex_NL

[stupid_chris]

Changelog did not indicate any tweaking of the atmospheric values, so I don't believe they did, or else they just forgot to drop us a line about it. It might be the new SAS behavior that changes the way you fly and is giving you problems.

[Tex_NL]

I am pretty sure it is not the SAS. My SSTO's are all stable designs and I use MechJeb's ascent guidance (reducing the ascent profile to 10-20% works great) to get them to orbit with the same settings in 0.21 as I had in 0.20

[stupid_chris]

MechJeb is kinda wacky right now, even the newest update has quite a lot of issues.

[Tex_NL]

True. MechJeb is a bit iffy in 0.21. But that doesn't explain why a plane previously capable of reaching orbit with ease now can't break atmo. I am running out of air way before I used to.

In 0.20 most of my SSTO planes would go past 30 km before I had to throttle back to avoid flame-out. Now, same engines, same intakes, same intake to engine ratio, I start running out of air at just 20 km.

[stupid_chris]

Try piloting it manually and see if you still can't get it out. From the looks of it and because the atmosphere shouldn't have been changed in .21, it might just be MechJeb reacting weirdly to the new SAS system and no expecting it's behaviour, resulting in a efficiency loss. I could be wrong though.

[Tex_NL]

And now for the proof it is not a MechJeb issue. Even IF MechJeb caused some difference it can't be this big.

I've flown both the 0.20 original and 0.21 version of my Hammerhead SSTO with exactly the same MechJeb settings. The first two images are from 0.20. The first is from the moment the intake air starts to decrease, the second is where MechJeb started to throttle back.

And I did the same thing in 0.21. The first is again where the intake air starts to decrease, the second where it throttled back.

Someone please explain why in 0.20 I get 'events' at 30 and 40 km and in .21 the same 'events' at 10 and 20 km. A few km difference I can understand as my two versions aren't 100% the same (I had to replace the small KSPX fueltanks on the back for Firespitter droptanks) but there is a 20 km difference!

Wing surfaces, intakes, fuel tanks, engines, EVERYTHING is exactly the same in both ships! (Except the KSPX/Firespitter tanks and minor canard adjustments to balance weight change)

Edited July 28, 2013 by Tex_NL

[astropapi1]

Hmmm... I seriously don't know what's going on.

[stupid_chris]

Hmm. Confusing indeed. It could be a bug, but with a modded installation anything could be causing it.

[longhornchris]

It looks like something has been tweaked in the drag model. I know that for FAR the drag model is much more forgiving in .21 than in .20. I rebuilt my basic designs and I'm not getting to orbit with ~40% ox/fuel where in the past I was making orbit with ~20% ox/fuel. Additionally, I've got extra fuel from the atmospheric portion of the flight. I'm still hitting roughly the same speed/altitude when I switch to rockets so I think that is being driven by the engine/air intake model.

[Tex_NL]

I doubt drag could cause this major difference in altitude. Drag is a force slowing you down the faster you go, it has nothing to do with the amount of 'breathable' air.

[stupid_chris]

Someone should test with fresh stock installs of .20 and .21 to see if a stock plane gets the same drop in intake air.

[Tex_NL]

I think you're right about that Chris. To exclude as many variables as possible I am currently re-downloading 0.20.2 and 0.21.1.

I'll build and fly a simple stock plane in 0.20 first, then transfer it to .021. to fly it again for comparison. After that I'll build an identical craft from scratch in 0.21 to confirm findings.

Any suggestions to improve the planned process are welcome.

Edited July 28, 2013 by Tex_NL

[tavert]

Seems exactly the same to me if you use stock intakes and engines - I'm getting the same performance on several craft/ascent profile combinations that would be extremely sensitive to any change in intake air mechanics. The only things that changed with the part stats are the radial intakes are now 10x lighter (so same mass as the ram intakes, less intake area, but more intake air capacity), the basic jet engine generates half as much heat, spools up slightly faster and down 2x slower, and there's the new jet engine visual effects.

The intake air mechanics seem to work fundamentally the same between 0.20 and 0.21, I suspect it's the B9 intakes and/or engines that have been tweaked between the two versions. Some of the B9 engines had been missing things like "useVelocityCurve = True" in 0.20, I would not be at all surprised if the B9 parts went through a bugfixing / rebalancing between 0.20 and 0.21. And I have no idea whether anything in FAR changed, haven't played with it yet. But from your screenshots looks like you weren't either.

Edited July 28, 2013 by tavert

[Tex_NL]

As far as I know there has been NO update for B9. I used exactly the same in both 0.20 and 0.21

Far is not of any issue. It's of course possible could cause some effects be but in my case it isn't. I have never used FAR in any installation.

[numerobis]

I'm busy today, but I'll inspect tomorrow whether the intake algorithms changed.

One thing: can you try setting your physics delta t to 0.03 in both versions? That affects how much intake air you produce per second when the intakes are full. It shouldn't explain the change at high altitude, just the change in when you see the number start to drop.

[Tex_NL]

OK, I just finished a stock intake test.

Full throttle take off and wheels up ASAP. Pulled the nose up to 30 degrees and left it there for the entire test. Three ram intakes for one turbojet engine.

Here are the results:

0.20

Intake air reduction at ~11,5 km

Full throttle flame-out at ~25 km

0.21

Intake air reduction at ~11,5 km

Full throttle flame-out at ~25 km

Conclusion: No significant difference. The cause of my original findings are probably rooted in the B9 mod.

Edit: Bac9 just released a new B9 version. Changelog does mention changes to engines so this might be solved. I'll keep you posted.

Edited July 28, 2013 by Tex_NL

[Tex_NL]

The new B9 3.3 did not solve the issue.

In the mean time I tested the efficiency of other B9 intakes to compare them with stock. All B9 intakes pretty much all suffer from the same. The RBM, RNM and Sabre intakes are supposed to be superior but now are on par with the stock circular intake.

[bac9]

The new B9 3.3 did not solve the issue.

In the mean time I tested the efficiency of other B9 intakes to compare them with stock. All B9 intakes pretty much all suffer from the same. The RBM, RNM and Sabre intakes are supposed to be superior but now are on par with the stock circular intake.

They were never supposed to be superior, that's incorrect.

Here is a quote from Taverius regarding the balance of all parts in B9 in comparison with stock parts.

[Tex_NL]

They were never supposed to be superior, that's incorrect.

Here is a quote from Taverius regarding the balance of all parts in B9 in comparison with stock parts.

http://i.imgur.com/rFgkP3B.jpg

I have no reason to doubt your statement but do you have an explanation in the performance difference between 0.20 and 0.21? If they aren't supposed to be even slightly superior have I been designing my planes on a bug/error?

[Char0093]

I have no reason to doubt your statement but do you have an explanation in the performance difference between 0.20 and 0.21? If they aren't supposed to be even slightly superior have I been designing my planes on a bug/error?

I'd like one as well,Mostly because I was curious and decided to test this for myself.As I'm a tad pressed for time,I don not have Screenshots,but I ran a plane I like to use in .20. A Concept I thought up years ago,the SR-71 HAIF (High Altitude Intercept Fighter). Admittedly,it's FAR's SR-71 Blackbird with a bit of touching up with the B9 pack with Lazor and Pirated weapons.

Changed in .20 on the FAR SR-71,are the intakes to the shock cones of B9, and 4 Diverterless intakes.Removed the adapter in favor of the B9 S2 Reaction wheel and 2 person cockpit along with Nosecone. Also added a pair of Strucual Wings and with small control surfaces to even up the CoL a bit and give her some more Lift at higher Altitudes. Along with Lazor System Stack and 4x Wasp Missiles on the Under belly. Mech Jeb only for Intake Air Readouts,as the Nose cone Avionics worked fine enough.

She'll do Mach 5 at about 26 km,all day everyday. Maximum Altitude hit was 35 km before there was no air left. Engines can run to 24 km at maximum throttle.

now..Here's where it gets interesting.Remove the Lazor Stack and the 4x Wasp Missiles in .21..EVERYTHING ELSE the same....

Need to throttle back at 17 km to keep the engines running. and Highest I can get is 24 km at Mach 3.38,and even then I dont have enough thrust to keep her Level so she falls out of the sky.

Something here is obviously different..Not with the Design,but with the parts or the Atmosphere.Yet,looking over the Change logs,there should be nothing affecting it this much.I've ALSO noticed even with STOCK parts,if you are close to 10 kg/s of air to your minimum Air needed,you'll Flame out.This is Different from .20,where I could go BELOW the Minimum Air Needed by at least 2 kg/s before it Flamed out.

Back to the HAIF,I did another Test Flight...Watching at when I start to lose how much Air she takes in,Both .20 and .21 Set on the runway with 337 Kg/s of Air feeding in....

.20 Starts losing Air at 11 km.

.21 Starts losing Air at 6.5 km.

Mods I am Running are as Followed:

Mech Jeb

Kethane

Procedural Fairings

Kerbal Alarm Clock

Kerbal Live Feed

Romfarer lazor System

FAR's .Craft of the SR-71B (FAR IS NOT INSTALLED)

[Tex_NL]

I am both disappointed and relieved I am not the only one running into this problem.

If it was just me it could have been something wrong with my install (or even with my hardware) but now that there is at least one more it more or less proofs it is something related to the mod itself.

[katateochi]

Is this solved? The thread is marked answered but I don't really understand what the answer is (which might just be me being thick!)

I did this test on the craft I developed in 20.2 with an older version of B9 (R3c);

I ran the same craft in 21.1, first with the new version of B9 (R3-3) and then with the previous version (R3c).

With the current version of B9 (R3-3) in 21.1 - 26km was the max alt I could reach and the top speed was 1138ms.

With the previous version of B9(R3c ) in 21.1 - 55km was the alt I flamed out at with a top speed of 2169 (and my AP was at 70km).

I had a look in the .cfg for the RBM intakes in the different versions and there is a difference, I don't understand what the specific implication of the values are but I can see that this must be what's responsible for the change;

So is the answer that the current version is how these intakes are going to remain, or are they likely to return to their former glory?

[a.g.]

â–¼ R3.1

* Intake stats recalculated using a different formula (factoring the root of mass / 0.01 for unitScalar, and factoring (unitScalar+sqrt(coeff))/(unitscalar+1) into the speed). Nothing changed gameplay-wise, but that will fix improper flow reporting in KSP context windows. Nope, RBM intakes never were 10 times more powerful than others, KSP lied to you. It will continue to lie about air speed though.

From B9 release notes, with the totally incorrect part crossed out.

[Tex_NL]

Is this solved? The thread is marked answered but I don't really understand what the answer is (which might just be me being thick!)

...

The original problem has been answered. Kerbins atmosphere has not been changed. But there is still something weird going on with the B9 mod.