KSP 1.0.2 re-entry behaviour suddenly changed and MK16 parachute explodes on landing

[RidiculousRichard]

KSP: 1.0.2.842 32 bit on windows 7 HP 64 bit

build id = 00842

2015.05.01 at 21:02:12 CEST

Branch: master

Problem: Consistent increased re-entry effects, mk16 parachute explodes on landing and pilot orientation hold suffers high frequency oscillations and moves away from the selected orientation

Mods:

Precision node

Kerbal alarm clock

Reproduction steps from save file:

1) Jettison upper stage

2) Use Jeb's piloting skills to hold prograde or retrograde; restore orientation hold before entering the atmosphere

3) Allow the command pod to perform re-entry (it is already on a re-entry course)

4) Deploy parachute ~2km above terrain

Log file and save file obtainable at http://s000.tinyupload.com/?file_id=71057878838376790157

Reproduction effects without save file:

1) Build a rocket with a mk1 capsule and mk16 parachute re-entry vehicle

2) Launch into orbit and get into a 80km x 30km re-entry sub-orbital trajectory

3) Use piloted prograde or retrograde hold to see bizzare behaviour

4) Watch re-entry using SAS in orientation mode

5) Deploy parachute ~2km above terrain

Update: The re-entry and parachute behaviour is occurring for all games (new and existing) on my KSP 1.0.2 install

Hello,

Within the last week a strange thing started to consistently happen while playing my first KSP game (Career mode Medium difficulty) running v1.0.2.842

I installed Precision node and Kerbal alarm clock fairly quickly after getting the game and they are the only mods I have ever used. The game has been functioning normally with those mods for about two weeks with Kerbin moon return missions

* At the current time putting a single mk1 capsule with a mk16 parachute into a 80km x 30km sub orbital trajectory will cause re-entry effects to start between 65km and 68km at ~2.3km/s. Previously these effects only started at ~36km

* At the current time the mk16 parachute will explode due to overheating on touchdown despite functioning normally during the final descent. Previously this did not occur.

* At the current time using the prograde or retrograde hold provided by a pilot results in the craft pointing vector oscillating rapidly and steady drift away from the appropriate orientation. I have never really used this much before but don't remember noticing this obvious problem; Basic SAS works fine

I did have a self inflicted physics oops as discussed in http://forum.kerbalspaceprogram.com/threads/123867-What-are-the-default-physics-settings but that was rectified by following the suggestion. The physics.cfg is normal (I have checked that the file contents have not changed since)

I have not upgraded to 1.0.4 (title screen still shows 1.0.2.842) and have not upgraded the two mods to avoid incompatabilities. I obtained KSP via steam and have copied the entire directory off to another location to stop it being messed with and I start KSP via a link to the copied KSP.exe rather than via Steam (32 bit on Windows 7 home premium 64 bit). I have not knowingly experienced any of the bugs that the 1.0.4 change log suggests are fixed in that revision. I have not experienced any of the self help issues

I am concerned that my returning 1st interplanetary mission is going to run into some kind of difficulty (unusual re-entry effects and things exploding are not a good sign when your rapidly approaching kerbals are depending on a predictable aero braking manoeuvre).

Is there a known recommendation to restore the 'normal' previous behaviour (or was that itself the bug)?

Is my only hope to abandon my KSP 1.0.2 game and start a brand new game from the beginning with 1.0.4? (loosing all game progress is not a desirable outcome)

I could obtain and post some screen shots showing the new behaviour if that is likely to assist.

Reporting information and save game at the top of this post

For those that fixate on such things I am deliberately still playing KSP 1.0.2 with my sole existing game to avoid any possibility of the changes from breaking / penalising my in-progress game as a result of the game rule changes. The next game played would be a hard difficulty career game and would be played with the latest release at that time once I've finished the current difficulty; I'm am not a fan of rapid feature updates (as opposed to rapid bug fixes and occasional feature expansion updates) and consequently I don't buy pre-release games but each to their own.

Thank you for any suggestions you might have,

the lives of several kerbals are on the line (and the accident investigation boards are getting tired of "oops" as the cause of death),

Richard

Edited June 29, 2015 by RidiculousRichard
Trying again with a new file host

[RidiculousRichard]

An update for anyone who comes across this:

1) It is not clear if there is any heating rate dependence on the altitude at which re-entry effects are shown (36km vs 86km) but I believe that there is although it is not too severe and should be a cause of concern (I can't get a 36km re-entry effects to make an absolutely unbiased comparison)

2) So far the only component that explodes due to overheating as soon as a re-entry vessel touches down is the Mk16 parachute (but I'm only in the early career game); you are simply denied the recovery refund on the parachute

3) Pilot prograde / retrograde can be erratic; don't leave a vessel unattended in these modes. I've not had any problems with using basic stability which so far has always been fine.

4) Always keep a pristine zipped up copy of any KSP version; I'm not that trusting in it's consistency over time

Please do comment if you have any reason to suspect any of these observations may be untrue

Edited June 28, 2015 by RidiculousRichard

[Brigadier]

Well, as far as the re-entry heating is concerned, since I understand the heating algorithms have been significantly changed with the 1.0.x releases, try using a heat shield on the re-entry vehicle. The parachute is now burning up if you release it too soon (this happened to me until I understood the more realistic physics being applied to re-entry), so wait until the heating effects are gone before deploying. Right click on the parachute before deploying and look for the recommendation on chute deployment in the part window (I've seen Unsafe, Risky and Ok flags so far).

I've also noticed that the SAS modes are much more sensitive. I tend to leave SAS ON and manually adjust the vehicle orientation on descent or use RETROGRADE but sometimes my capsule flips, even with a nose cone chute. It also means watching your battery power carefully.

[RidiculousRichard]

I've only played the demo and 1.0.2 so I'm not sure what the changes were that caused a fuss.

For light-weight landers using a capsule base has been very reliable. Batteries on the side of a vessel are most likely to overheat.

The mk16 problem above only happens at the moment of touch-down; they function properly during the terminal descent. As you say, opening them in a fireball is not a good idea.

Regards,

Richard

[Guest]

After jumping through the captcha hoops to download the log and save file, it said requested download does not exist. Parachutes exploding after landing is not normal. In fact, I've never seen it happen. I was downloading the file mainly to see if I could reproduce it, just because this sounded fun. So, without the craft file or the log, I really can't do anything other than speculate. Do you have any rtg's on the craft? Any fuel cells? It sounds like something is heating up the parachute. Re-entry effects at 60km+ is not default. Maybe one of the settings is cranked up above normal?

[RidiculousRichard]

I'll check the download and fix it tonight

From memory all that I have on such pods are:

Mk 1 capsule

Clipped Octobody thing (for rescue / kerbal football)

Small battery

Mk16 parachute

I may have:

2 basic solar cells

2 radial parachutes (paranoia backup)

[RidiculousRichard]

Download link updated to http://s000.tinyupload.com/?file_id=71057878838376790157

I've concluded that filedropper.com is a scam to try and get you to pay for their 'PRO' version (I should have twigged when pasting the link; it could never be as nice as that. They put some effort in mind as the download works when you try it shortly after uploading)