why is this rocket unstable?

[king of nowhere]

once more, i find myself with a rocket that looks very good, should have no problem whatsoever, and yet will capsize immediately at start.

this thing won't fly straight. it's turning to the side already at 50 m/s. to compensate that, i tried to turn on the rhino. well, this rocket either crashes anyway, or will require a lot of steering to try and keep it pointed the right way.

it clearly has too much drag in front. but how? what is there to possibly make to much drag? i have a nose cone and fuel tanks. 4 small solar panels, whose effect should be negligible.

I tried to shift the large tank from the back to the front in an attempt to move forward the center of mass, and it helped somewhat. as soon as i shut down the rhino, the ship started rotating.

I just want to understand which of those seemingly normal parts is actually much more draggy than would be apparent from its design

[RealKerbal3x]

@king of nowhere Without trying your craft myself, I'd say at least part of the problem is lack of control authority. As far as I can tell, the only part on your craft providing control is the tiny reaction wheel in the probe core at the top - you're lifting off on SRBs that don't have gimbal control, and the fins you're using are fixed and can't contribute to control. Since you don't want to use the Rhino when lifting off, I'd recommend adding some larger moving fins.

[Snark]

Your CoM is too far from the front of the ship; it's basically in the rear.  And since you're taking off on the SRBs alone, you don't have any significant engine gimbal to give you any active steering.

Suggestion:  Move the SRBs down.  I mean, way down-- mount the radial decouplers as low as you possibly can on the central stack, then mount the SRBs as low as you can on the decouplers.  And put stabilizing fins on the bottom of the SRBs.

Also:  Are those SRBs flexing at all?  Are you using autostruts (or regular struts) to stabilize them and hold them rigidly in place?

 

[Superfluous J]

You don't have the gui on so I'll add: is whatever your control part is, pointed "up"?

[king of nowhere]

20 minutes ago, RealKerbal3x said:

@king of nowhere Without trying your craft myself, I'd say at least part of the problem is lack of control authority. As far as I can tell, the only part on your craft providing control is the tiny reaction wheel in the probe core at the top - you're lifting off on SRBs that don't have gimbal control, and the fins you're using are fixed and can't contribute to control. Since you don't want to use the Rhino when lifting off, I'd recommend adding some larger moving fins.

I am using the rhino to provide control. and i have 2 more reaction wheels inside the cargo bay.

17 minutes ago, Snark said:

Your CoM is too far from the front of the ship; it's basically in the rear.  And since you're taking off on the SRBs alone, you don't have any significant engine gimbal to give you any active steering.

Suggestion:  Move the SRBs down.  I mean, way down-- mount the radial decouplers as low as you possibly can on the central stack, then mount the SRBs as low as you can on the decouplers.  And put stabilizing fins on the bottom of the SRBs.

Also:  Are those SRBs flexing at all?  Are you using autostruts (or regular struts) to stabilize them and hold them rigidly in place?

 

yes, they are using regular struts, no problems there

2 minutes ago, Superfluous J said:

You don't have the gui on so I'll add: is whatever your control part is, pointed "up"?

yes, it's one of the first things i did check

 

Edited January 23, 2021 by king of nowhere

[bewing]

As RealKerbal3x said, you don't have any control authority. Most rockets do not go straight all by themselves. You need something on it that can be steered. Either gimballing engines, or aerodynamic control surfaces. Maybe you don't need much, but you certainly need something.

Reaction wheels are minimally effective. In general, try one -- if it doesn't work, then you need to use something more.

So, possibility #1 is to turn on the rhino, but keep it throttled down to less than 10% thrust. Or, start it out at 100%, and then throttle it way down after you make your gravity turn.

Possibility #2 is to stick on some steerable fins. And then you need to stick them as far as possible from your CoM. The argument is tedious, but control surfaces with 100% of their surface being steerable are best. Elevons or tailfins are preferable. Attach them to your nosecone at the very front of your craft. If you don't want to carry that mass all the way to orbit, put a decoupler on the nosecone. (And then remove those "fins" from the back end -- they aren't doing you any good, because they are too close to your CoM.)

 

Edited January 23, 2021 by bewing

[lemon cup]

What’s the idea behind the shuttle-style cargo bay as your payload option of choice? Rule of cool? Is the launcher reusable in some way? Both are acceptable answers, but if you are in fact looking for the most effective launcher the best option would be to ditch the idea and go with a traditional booster at the bottom and fairing at the top. The payload bay + whatever you have inside is serving to make the middle of your rocket less dense, and is definitely contributing to the bad drag characteristics.

If you are determined to use the shuttle bay though, I’d recommend adding more fuel tanks above the bay, maybe partially defueling the bottom tank that the Rhino is attached to and setting the top tanks’ priorities to “1”. (So that they drain first and don’t starve the Rhino mid-flight)

Also consider replacing the SRBs with liquid fuel boosters, with engines that gimbal (!) and lighting the Rhino from the get-go with its thrust turned down as @bewing suggested.

Edit: and to answer one of your questions, the tapered tanks at the top are more draggy than they look, more-so than a payload fairing or nose cone by mass ratio. Second worst offenders being the size 3 cargo bays!

Edited January 23, 2021 by lemon cup

[king of nowhere]

37 minutes ago, bewing said:

So, possibility #1 is to turn on the rhino, but keep it throttled down to less than 10% thrust. Or, start it out at 100%, and then throttle it way down after you make your gravity turn.

 

i already did, at 30% thrust. as i said, it helps - and it's the reason i could use the rocket in the first place - but it's still unstable.

17 minutes ago, lemon cup said:

What’s the idea behind the shuttle-style cargo bay as your payload option of choice? Rule of cool? Is the launcher reusable in some way? Both are acceptable answers, but if you are in fact looking for the most effective launcher the best option would be to ditch the idea and go with a traditional booster at the bottom and fairing at the top. The payload bay + whatever you have inside is serving to make the middle of your rocket less dense, and is definitely contributing to the bad drag characteristics.

 

i would have liked a fairing myself. the cargo bay is because i need to deorbit a vehicle, so i made 2 things with a claw on one hand an an inflatable thermal shield on the other. and the claw is not a valid attachment point for the rest of the rocket. so i stuck a separator on the shield, and stuck those two probes on the cargo bay on opposite ends. it's just a convenient way to save hassle on a launch. the whole vehicle is a one-shot i'm not going to use again, so i also spent minimal time on it.

as a matter of fact, i already launched it (with some difficulty) and successfully performed its mission (again, with some difficulty, mostly because i forgot to put batteries on the probe, so it was a race against time to stick them in place before they run out.)

the thing is, i wanted to figure out what was the problem with the rocket to learn something. i've flown much more unstable rockets much more easily, and i was trying to make sense of stuff. Low baricenter and small fins are good explanation, but i'm not sure they justify it in full. I was half expecting to be told that the Mk3 cargo bay or some of the adapter has horrible aerodinamic properties...

[lemon cup]

40 minutes ago, king of nowhere said:

so i made 2 things with a claw on one hand an an inflatable thermal shield on the other. and the claw is not a valid attachment point for the rest of the rocket.

Oh! Do you know how to use the interstage node feature on stock payload fairings? It allows you to stack two or more craft that would not otherwise be stackable (like your claw contraptions) inside a single payload fairing. 

Right click the fairing in the VAB and turn “Interstage Nodes” on. 

For a more detailed tutorial check out my answer in this thread:

 

[king of nowhere]

56 minutes ago, lemon cup said:

Oh! Do you know how to use the interstage node feature on stock payload fairings? It allows you to stack two or more craft that would not otherwise be stackable (like your claw contraptions) inside a single payload fairing. 

Right click the fairing in the VAB and turn “Interstage Nodes” on. 

For a more detailed tutorial check out my answer in this thread:

 

neat! every time i tried to launch multiple probes with a mk3 cargo bay i ended up losing more time trying to deal with the idiosyncrasies than i would save in multiple launches. skipping that part is very useful.

[Snark]

3 hours ago, king of nowhere said:

I am using the rhino to provide control.

The problem is that your CoM is so low that the Rhino can't help much.  The control authority of a gimbaling engine is directly proportional to how far behind the CoM it is, so it has a decent lever arm to work with.  In your case, the Rhino is so close to the CoM that its gimbaling doesn't help you all that much.

3 hours ago, king of nowhere said:

and i have 2 more reaction wheels inside the cargo bay.

Probably nearly irrelevant, especially on a craft that size.  The issue is that as a rocket goes faster and faster, the aerodynamic forces get huge during the "max Q" part of ascent-- easily enough to totally swamp the (relatively) tiny torque that reaction wheels can provide.

If a craft is aerodynamically unstable, then reaction wheels probably won't help it much.  The way to fix it is to make it more aerodynamically stable by moving the CoM forward, and to provide control authority in the form of fins and gimbaled engines that are as far behind the CoM as possible.

4 hours ago, king of nowhere said:

I tried to shift the large tank from the back to the front in an attempt to move forward the center of mass, and it helped somewhat.

This is not a bad idea.  One question:  Did you tinker with the fuel priorities of your tanks so that they drain from the bottom up?

I ask because, by default, they'll drain all together in parallel.  Whereas if you change it so that the bottom one drains first, then that will help shift the CoM forward more.

[king of nowhere]

1 hour ago, Snark said:

This is not a bad idea.  One question:  Did you tinker with the fuel priorities of your tanks so that they drain from the bottom up?

I ask because, by default, they'll drain all together in parallel.  Whereas if you change it so that the bottom one drains first, then that will help shift the CoM forward more.

no. in the past i tried a couple times to set up fuel priorities, but it never worked correctly. since then, i don't try anymore; i never needed it

yes, the low torque from the engine having too little leverage was probably the main reason. after i shifted the lower tank to the top, it became usable.

[vv3k70r]

On 1/23/2021 at 6:31 PM, king of nowhere said:

this thing won't fly straight

Some control surface?

Wings do not help on themselvs - they need some moving pars on the edge.

 

 

 

[king of nowhere]

4 hours ago, vv3k70r said:

Some control surface?

Wings do not help on themselvs - they need some moving pars on the edge.

 

 

 

most rockets fly well without control surfaces.

exquisitely crafter rockets fly well without control, gimbaling, or sas. though i don't put that kind of effort for a single use rocket i'm never going to use again.

[vv3k70r]

50 minutes ago, king of nowhere said:

most rockets fly well without control surfaces.

Because they are designed corectly for specific circumstances - like planed flightpath throu atmosphere. They are not stable in all cirmustances - only in those planned and bit around. Your fins give stability if rocket go fast and have COM on top. See proportion of drag for those fins during reentry (use F12).

This one You show is not stable by first look. This are imaginable proportion (we call this "beauty" - it depend completly on proportion). You will come to this after more rockets and be able to send them without any tests. All physics is just about proportion.

If it looks good it fly well.

[king of nowhere]

5 minutes ago, vv3k70r said:

Because they are designed corectly for specific circumstances - like planed flightpath throu atmosphere. They are not stable in all cirmustances - only in those planned and bit around. Your fins give stability if rocket go fast and have COM on top. See proportion of drag for those fins during reentry (use F12).

This one You show is not stable by first look. This are imaginable proportion (we call this "beauty" - it depend completly on proportion). You will come to this after more rockets and be able to send them without any tests. All physics is just about proportion.

If it looks good it fly well.

more rockets? i've played 1500 hours on this game. I've launched hundreds of rockets. and most of them fly straight at the first try, unless i'm intentionally trying to do something crazy.

and that rocket in the picture, it looks good to me. it looks much better than other things i've launched that flew true anyway. good proportion. aerodinamic shape. a lot of my rockets didn't have either.

I mean, look at this other thing

To send this thing in orbit, i just strapped 4 boosters underneath it. No control besides the reaction wheels. a big draggy thing on top. Does that thing balanced on 4 sticks look like a good rocket? But it flies smoothly, with a perfect gravity turn. And I made it 6 months ago, when i didn't have half the experience i have today.

And on the other side, i had a rocket with a perfect rocket-like shape, good aerodinamics, just an apparently minor problem of center of mass being low, and that couldn't get to orbit.

when this kind of #### happens, i don't want to just fix the rocket. I want to figure out why a rocket that looked good, wasn't.

[vv3k70r]

2 hours ago, king of nowhere said:

more rockets? i've played 1500 hours on this game.

It dosent play any roll. Lerning curve depend on Your g-factor. Number of iteration required for any folowing sigma is square root of previous one.

2 hours ago, king of nowhere said:

And on the other side, i had a rocket with a perfect rocket-like shape, good aerodinamics, just an apparently minor problem of center of mass being low, and that couldn't get to orbit.

As You see this problem is not minor one. It is critical.

2 hours ago, king of nowhere said:

when this kind of #### happens, i don't want to just fix the rocket. I want to figure out why a rocket that looked good, wasn't.

People here told You why. It is obvious.

You have geometrical proportion of COM to TWR to drag to prograde. Drag is airspeed depended, TWR is mass dependent from moving COM.  You get one of those outside this area - it is why forklifts, cranes, cars, trucks, aeroplanes fall. Same rule for rockets. This area made of proportions is world You live in. If You left this area You are not in the world where You are alive.

You have a range of values that must be kept in narrow corners of this area. You pushed one outside the borders and all others comply. You ask for disaster and You get one.

Get back to living area with those values - COM up, drag down and more.

2 hours ago, king of nowhere said:

To send this thing in orbit, i just strapped 4 boosters underneath it. No control besides the reaction wheels. a big draggy thing on top.

Yep - so this work, lot of drag, high TWR and You get stability.

Rocket is not a gentle glider, it is brute force to orbit.

Edited January 25, 2021 by vv3k70r

[Incarnation of Chaos]

On 1/23/2021 at 5:55 PM, king of nowhere said:

no. in the past i tried a couple times to set up fuel priorities, but it never worked correctly. since then, i don't try anymore; i never needed it

yes, the low torque from the engine having too little leverage was probably the main reason. after i shifted the lower tank to the top, it became usable.

I wrote this long, expository post....

Then looked a page back and realized you did exactly what i was telling you to do in it. Putting the lower tank on top shifted the COG upwards, which also had the effect of giving your engine gimbal a larger lever arm to work with. And since now it didn't want to tip nearly as badly, the gimbal wasn't needed as much either.