Rover Overstressed when opening parachutes

[EnderPig]

So I'm quite new to Kerbal Space Program and I have been following Scott Manley's career mode tutorials. I have learned a lot. I am at episode 24 now where you try to send a rover to Duna. I have done everything except the landing. And I just can't do it. When I enter the atmosphere, then try to slow down and open the parachutes, my rover suddenly explodes and gets flung out of the craft. A message pops up saying: "[Rover Max Model S2]: Overstressed". I have no clue why this happened. I tried to follow the tutorial the best I can. Is it the parachutes or is something wrong with the rover? I can't figure it out. I had some trouble with the separator on top of the rover while building the spacecraft because I didn't understand what Scott wanted me to do with the offset. The only thing I changed on the rover was that I added a few batteries under the control. Does anyone have any idea why this happened? Also the only mod I use is Mechjeb 2 (For docking). I did comment on the video as well but I don't think anyone is going to see it. I am using the newest version of KSP (Steam doesn't want me to update it or anything right now)

Video of the explosion: https://www.youtube.com/watch?v=jigbC7BYZ8w

Tutorials I used to make the spacecraft and the rover: https://www.youtube.com/watch?v=LzQ4t7gjKcs

https://www.youtube.com/watch?v=Qb6CVX6QwLA

 

https://www.youtube.com/watch?v=Om1VOBKSRWA

http://imgur.com/XoBQD52 Here the Rover is still in it's right position.

http://imgur.com/0zy8x6r Here I have opened the parachutes, the wheels are gone and the top of the rover is stuck on the top. (Later on somehow the separator and the top part of the rover fells of the craft too)

http://imgur.com/yU7ZxAr Here is a F3.

What should I do to fix this? What do I need to change to make this work?

 

 

Edited October 3, 2016 by EnderPig
Added Pictures

[Plusck]

I'm not sure that that offset thing is really a good idea in this case.

However, the problem is that without offseting the separator, both up from the rover and down from the roof of the service bay, the rover is going to clip into the service bay and this is guaranteed to cause problems.

The correlative problem is that by offsetting the separator like that, you are weakening the joint between the separator and the rover.

If you have an "overstressed" message for a rover wheel (yes, it is one of the wheels that is overstressed), it's certainly because it's hitting something inside the bay. You should open up the service bay to look at what's happening on the way down, but I would expect that the rover is swivelling from side to side in there.

You should also take a close look at F3 after an accident like that, so see what collided ito what.

Edited October 3, 2016 by Plusck

[EnderPig]

I added pictures of the F3 and what happens inside the service bay. What should I do in this situation and what is the way to fix this so I can safely land?

[John Cochran]

First question I'd ask is what version of KSP are you using and what version was Scott Manley using for the tutorial? The wheels in 1.1.x are rather ... fragile compared to earlier versions of KSP. Assuming my old eyes are up to the task, I believe Scott Manley's tutorial is using version 1.0.0 where the wheels aren't nearly so fragile.

[EnderPig]

4 minutes ago, John Cochran said:

First question I'd ask is what version of KSP are you using and what version was Scott Manley using for the tutorial? The wheels in 1.1.x are rather ... fragile compared to earlier versions of KSP. Assuming my old eyes are up to the task, I believe Scott Manley's tutorial is using version 1.0.0 where the wheels aren't nearly so fragile.

I am using the newest version of KSP on Steam. I think Scott Manley used 1.0.0.

Edited October 3, 2016 by EnderPig

[Foxster]

Do you have to get this particular craft down or are you OK to make changes in the VAB and try again?

[Palaceviking]

How about a very high aerobrake to reduce speed?

Also how many parachutes have you got, dont open them all at once, stagger the deployment.

Or switch to 1.2- the new wheels are brill!

[Plusck]

41 minutes ago, EnderPig said:

I added pictures of the F3 and what happens inside the service bay. What should I do in this situation and what is the way to fix this so I can safely land?

I strongly suspect you can't do anything at all, other than reloading a quicksave and hoping that the game blesses you with a more forgiving Kraken.

Then again, someone may come along with a second opinion and it'll all work out ....

 

Incidentally, the F3 info I was talking about is at the beginning of the log, that has been scrolled past in your screenie    Also your pics are so dark I can hardly see a thing

However, it looks like the clearance between the service bay and wheels is just too tight. They are probably clipping into one another at times, and this is causing a Kraken-like effect on the wheels. The stack separator is probably too weak (due to offset) to hold the rover properly. The problem could well have started when you lifted off from Kerbin: if you have a savegame from that time take a long hard look inside the service bay in orbit to see if any wheels look like they're "stuck" in the service bay floor.

In general, I don't agree with SM's suggestion to offset the stack separator. IMHO it's just asking for problems. Parts that separate cleanly away from other parts should (IME) never be offset since they snap to their proper place in the editor and offsetting them away from that ideal position is just risky. You can offset parts on the "outside" edge of decouplers all you want (the bits that pull away from the craft and remain attached to the debris), but not to the inside edge, and a separator has two such edges. Add an octagonal cubic strut to the top of the separator and offset that into the roof of the service bay until alignment is perfect, by all means, but remember to make sure you have headroom to get the rover out. And be absolutely sure to reduce the separation force to zero or you will kill the rover wheels.

The only real solution (unless, as I said, someone else comes along with a better explanation and solution, which is always possible) is to give up and redesign/redo the mission, or cheat.

To cheat:
1- switch off damage via alt-F12 until you're on the surface, see if you can extract the rover at all (not necessarily possible, mind you), or
2- rebuild that part of the ship with a stronger attachment (me, I'd add a strut from the bay to at least one bit of the rover) and better wheel clearance, and hyperedit it into Duna orbit.

Edited October 3, 2016 by Plusck

[EnderPig]

30 minutes ago, Foxster said:

Do you have to get this particular craft down or are you OK to make changes in the VAB and try again?

I am OK with that, I just need to know how to make it work then (pictures would be nice too)

15 minutes ago, Plusck said:

I strongly suspect you can't do anything at all, other than reloading a quicksave and hoping that the game blesses you with a more forgiving Kraken.

Then again, someone may come along with a second opinion and it'll all work out ....

 

Incidentally, the F3 info I was talking about is at the beginning of the log, that has been scrolled past in your screenie    Also your pics are so dark I can hardly see a thing

However, it looks like the clearance between the service bay and wheels is just too tight. They are probably clipping into one another at times, and this is causing a Kraken-like effect on the wheels. The stack separator is probably too weak (due to offset) to hold the rover properly. The problem could well have started when you lifted off from Kerbin: if you have a savegame from that time take a long hard look inside the service bay in orbit to see if any wheels look like they're "stuck" in the service bay floor.

In general, I don't agree with SM's suggestion to offset the stack separator. IMHO it's just asking for problems. Parts that separate cleanly away from other parts should (IME) never be offset since they snap to their proper place in the editor and offsetting them away from that ideal position is just risky. You can offset parts on the "outside" edge of decouplers all you want (the bits that pull away from the craft and remain attached to the debris), but not to the inside edge, and a separator has two such edges. Add an octagonal cubic strut to the top of the separator and offset that into the roof of the service bay until alignment is perfect, by all means, but remember to make sure you have headroom to get the rover out. And be absolutely sure to reduce the separation force to zero or you will kill the rover wheels.

The only real solution (unless, as I said, someone else comes along with a better explanation and solution, which is always possible) is to give up and redesign/redo the mission, or cheat.

To cheat:
1- switch off damage via alt-F12 until you're on the surface, see if you can extract the rover at all (not necessarily possible, mind you), or
2- rebuild that part of the ship with a stronger attachment (me, I'd add a strut from the bay to at least one bit of the rover) and better wheel clearance, and hyperedit it into Duna orbit.

How would I get the rover moving if it has struts on though? Probably going to edit it into orbit.

Edited October 3, 2016 by EnderPig

[Plusck]

12 minutes ago, EnderPig said:

I am OK with that, I just need to know how to make it work then (pictures would be nice too)

Post the craft file (if you can) and I'll take a look, if you want.

12 minutes ago, EnderPig said:

How would I get the rover moving if it has struts on though?

Struts aren't "real" parts. They can't hold something together that isn't held together via the vessel's tree structure. The struts will break automatically when the "real" connection between the parts (i.e. the stack separator) is broken.

 

edit: just went back and took a better squint at the pics. So yes, it's clipped into the roof rather than the floor. Odd. I still suspect that a wheel-clipping-into-floor kraken made it do that, but I have no idea how

Edited October 3, 2016 by Plusck

[EnderPig]

Before I launched, I made sure that the wheels at least aren't out of the bay or in the bottom part.

Edited October 3, 2016 by EnderPig

[Kelderek]

Before you go completely back to the drawing board, you can try to adjust your parachute settings.  First, take one of your parachutes (use a drogue chute first if you have one, otherwise go with your largest one) and have it deploy at the maximum altitude on its slider.  Then have each additional parachute deploy in 100-200m increments lower altitude (depending on how many you have total).  Then as you descend, target the first parachute and watch the part that shows whether it is safe to deploy and pop it immediately once you see it is safe.  Your parachutes will semi-deploy all at once (you haven't changed staging), and then they will fully deploy in a staggered sequence.

By starting at a higher altitude you reduce the strain on the vessel because they deploy in thinner atmosphere and by staggering the chutes it breaks up the total deceleration into smaller events.

It may not work based on your design, but it's worth trying before you go back to the VAB.

[EnderPig]

15 minutes ago, Kelderek said:

Before you go completely back to the drawing board, you can try to adjust your parachute settings.  First, take one of your parachutes (use a drogue chute first if you have one, otherwise go with your largest one) and have it deploy at the maximum altitude on its slider.  Then have each additional parachute deploy in 100-200m increments lower altitude (depending on how many you have total).  Then as you descend, target the first parachute and watch the part that shows whether it is safe to deploy and pop it immediately once you see it is safe.  Your parachutes will semi-deploy all at once (you haven't changed staging), and then they will fully deploy in a staggered sequence.

By starting at a higher altitude you reduce the strain on the vessel because they deploy in thinner atmosphere and by staggering the chutes it breaks up the total deceleration into smaller events.

It may not work based on your design, but it's worth trying before you go back to the VAB.

Sadly that didn't seem to work with me. As soon as I deploy the parachutes the rover just goes crazy.

[EnderPig]

42 minutes ago, Plusck said:

Post the craft file (if you can) and I'll take a look, if you want.

Struts aren't "real" parts. They can't hold something together that isn't held together via the vessel's tree structure. The struts will break automatically when the "real" connection between the parts (i.e. the stack separator) is broken.

 

edit: just went back and took a better squint at the pics. So yes, it's clipped into the roof rather than the floor. Odd. I still suspect that a wheel-clipping-into-floor kraken made it do that, but I have no idea how

Here is the craft: http://www.mediafire.com/file/zdxe1zm5hoodq2f/Ender_Duna_Explorer.zip

[EnderPig]

Okay. So I manually opened the parachutes one by one and landed on a hill, then rolled to the bottom and I got back up again. So I have succesfully landed now.

[Plusck]

8 minutes ago, EnderPig said:

Okay. So I manually opened the parachutes one by one and landed on a hill, then rolled to the bottom and I got back up again. So I have succesfully landed now.

Wow - cool.  So I'm apparently far too pessimistic about the chances of success in some cases.

I took a look at the craft file, and made the sort of changes I suggested earlier. Basically: clipped a few batteries into the body of the rover to improve headroom, replaced separator with decoupler and zero ejection force (frankly, I don't understand why SM uses a separator - it is heavier and leaves more bits lying around). Added a docking port to "control from" and a strut to hold it tight inside the service bay.

A bit pointless since it wasn't necessary, but it's too late now so here it is

[Kelderek]

1 hour ago, EnderPig said:

Sadly that didn't seem to work with me. As soon as I deploy the parachutes the rover just goes crazy.

So are you saying that when you staged the parachutes and they semi-deployed that was enough to cause the damage? I had assumed your problem was with full deployment, not semi-deployment.  In that case, I can see why manually popping the parachutes one by one instead of through staging worked better.

[Plusck]

8 minutes ago, Kelderek said:

So are you saying that when you staged the parachutes and they semi-deployed that was enough to cause the damage? I had assumed your problem was with full deployment, not semi-deployment.  In that case, I can see why manually popping the parachutes one by one instead of through staging worked better.

That's what I'm confused about.

I was going to suggest staging the parachutes successively in my first post, but I watched the video again and noted that the rover exploded when the parachutes were armed but hadn't even started to semi-deploy. So I deleted that part of the post and didn't think any more about it.

That's what's really bugging me about the fact that EnderPig managed to land by doing exactly what you suggested. Judging by the video it shouldn't work, and I don't understand why.

[moogoob]

3 hours ago, Plusck said:

Post the craft file (if you can) and I'll take a look, if you want.

Struts aren't "real" parts. They can't hold something together that isn't held together via the vessel's tree structure. The struts will break automatically when the "real" connection between the parts (i.e. the stack separator) is broken.

This is something that I forget I had to learn back in the day. You expect struts= hold things together against all odds. In reality they're good for that, but best for shoring up parts that are supposed to decouple. You can imagine it like a one-way collet - holds together when its supposed to, detaches when the parts they're attached to separate.