Bug? Warped through Jool's atmosphere

[Loren Pechtel]

My objective was to barely skim through the top of Jool's atmosphere to fulfill a "fly on Jool" term in a contract.

I hit it with periapsis at 199.9km at I think 10x, the contract didn't complete and I didn't dewarp.  I saw a 199 number on the altitude, though.

I came back around at 199.5km and 5x, still no dewarp but when I manually dewarped the contract completed.

Also, it appears the rally contracts have a bug.  It says "newly launched vessel" but it doesn't seem to care.  Vessels that meet the requirements can complete them even if they were already in flight when the contract was accepted.  It's only looking at what the vessel does after acceptance.  (Said Jool-skimming vessel also managed to complete a contract that said "suborbital flight on Jool" on the orbit that it failed to find the atmosphere.  Of course the vessel meant to complete that contract is already launched, hopefully something else will come along for it to do.)

[bewing]

The point of timewarp is to save time by skipping things. I don't think any contract testing is done at all under timewarp.

[Loren Pechtel]

1 hour ago, bewing said:

The point of timewarp is to save time by skipping things. I don't think any contract testing is done at all under timewarp.

That's not the important point.  The important point is I was in the outer edge of Jool's atmosphere but still in timewarp.  It didn't force a dewarp like it should have.

[egoego]

It shouldn't happen with a 10x time warp, but when time warping on rails, the orbit simulation becomes less accurate (time wise), which means that only some points on the orbit are calculated. With high time warp values (1000x) it can happen more easily that an important event like entering an atmosphere can be missed by the lower resolution used for rails calculation. (I see that happen a lot when I drop down from Minmus to do a aerobreaking but with 10000x timewarp i skip right through.)  

Maybe your PE was so high, that your path through the atmosphere was too short and the 10x timewarp was enough so the simulation missed it. Depending on where/when you entered timewarp you could have hit the atmosphere or not.

Or you don't have the orbital drift compensation enabled and the PE drifted out of the atmosphere.

Edited March 29, 2017 by egoego

[Loren Pechtel]

5 hours ago, egoego said:

It shouldn't happen with a 10x time warp, but when time warping on rails, the orbit simulation becomes less accurate (time wise), which means that only some points on the orbit are calculated. With high time warp values (1000x) it can happen more easily that an important event like entering an atmosphere can be missed by the lower resolution used for rails calculation. (I see that happen a lot when I drop down from Minmus to do a aerobreaking but with 10000x timewarp i skip right through.)  

Maybe your PE was so high, that your path through the atmosphere was too short and the 10x timewarp was enough so the simulation missed it. Depending on where/when you entered timewarp you could have hit the atmosphere or not.

Or you don't have the orbital drift compensation enabled and the PE drifted out of the atmosphere.

Yeah, I've seen the bit of blowing right through an aerobrake in very high warp--it usually steps down as it approaches but it can do the whole flyby in one tick and not notice the problem.  That's not what happened here, though, I was actually flying the rocket and not at high warp.  On the second pass I had enough time in atmosphere to see that it didn't automatically dewarp and I hit it manually.

Experiment:  I cheated a simplistic rocket into Jool orbit and tried lowering the periapsis.

199.8--I made the passage through the atmosphere in warp.  Plenty of time there, it didn't notice.

I tried to do 199.7 but the burn was too small, I had to do 199.6 to get MechJeb to light the engine.

Dewarped just above 199.7km.  I was in a low orbit, it had already been in atmosphere for a while.

The dewarp is triggering too low.