Is There a Formula For...

[Kerbal7]

Is there a formula for making rockets that don't start flipping?

The atmosphere seems awful soupy below 15Km.

I've given up making turns below 20Km or so. I have no problem getting to orbit, just curious if there is some formula I don't know.

This is a good rocket example of mine that gets to the Mun just fine but forget turning it before the boosters drop. Doesn't seem out of balance to me.

  

Edited November 24, 2017 by Kerbal7

[Ultimate Steve]

I don't think there's a straightforward formula, but you've got a really draggy payload that changes diameters (KSP makes those have a really high drag for some reason). I'd suggest putting the payload in a fairing.

And adding fins to the base. Even 4 AV-R8's will probably help a ton.

As far as flipping is concerned, you want the mass as the front and the drag at the back. Here you have the mass at the back (can't really be helped) but you have the drag at the front (can be helped).

EDIT: This was my 1500th post? Hey, I've got five dots now! Or did I get that at 1000? I don't remember...

Edited November 24, 2017 by Ultimate Steve

[Pand5461]

54 minutes ago, Kerbal7 said:

Doesn't seem out of balance to me.

Are you joking?

Stable rockets have weight on top, drag at the tail. Your design has sharp transitions 1.25->0.625m and 2.5->1.25->2.5m on the top. They create a lot of drag. And not a single fin on the bottom. With such design, the nose cap on the top and on the boosters really look like silly "look, I do care about aerodynamics" joke.

Edited November 24, 2017 by Pand5461

[Plusck]

8 hours ago, Pand5461 said:

Are you joking?

Stable rockets have weight on top, drag at the tail. Your design has sharp transitions 1.25->0.625m and 2.5->1.25->2.5m on the top. They create a lot of drag. And not a single fin on the bottom. With such design, the nose cap on the top and on the boosters really look like silly "look, I do care about aerodynamics" joke.

lol, yes.

 

I built and tried the ship. Gentle gravity turn (not enough to be really efficient) worked fine until about 500m/s but gimbal was enough to recover.

Spoiler

Still left me with 1200m/s to burn at Ap though....

Spoiler

So I tried adding more nosecones.... That didn't really work out:

Spoiler

So back to the drawing board... and this one is great. Really stable with the fuel up top. Just remember to land on a flat surface :

Spoiler

https://imgur.com/a/kXE8C

 

[Rocket In My Pocket]

Remember that a proper rocket should look like an arrow.

Weight at the tip, long aerodynamic shaft, fins at the back for stability.

[IncongruousGoat]

To add on to what people here have been saying, put a fairing around the payload and slap some fins onto the bottom of the first stage. The fairing will reduce drag at the front, and the fins will keep the center of lift back, which is what you want.

[Corona688]

It's not always possible to make it tip-heavy when 90,000 tons of engines and fuel are below the payload, so a skipper engine may not be enough control authority on a rocket that big.  Either some swivel fins or some thuds may help.

[HebaruSan]

12 hours ago, Kerbal7 said:

I've given up making turns below 20Km or so.

Are you doing the old pre-1.0 "straight up till 10 km, then turn 45 degrees" style of launch?

[Kerbal7]

3 hours ago, HebaruSan said:

Are you doing the old pre-1.0 "straight up till 10 km, then turn 45 degrees" style of launch?

I suppose.

I go straight up to about 20 km and pull a hard turn. Or when I drop the boosters. 

I make my rockets as cheap with as much Delta-V as possible. 

This is one I came up with.

Less weight and more Delta.V seem more important than drag and symmetry to me. 

Edited November 24, 2017 by Kerbal7

[HebaruSan]

21 minutes ago, Kerbal7 said:

I suppose.

I go straight up to about 20 km and pull a hard turn. Or when I drop the boosters. 

I make my rockets as cheap with as much Delta-V as possible. 

This is one I came up with.

Less weight and more Delta.V seem more important than drag and symmetry to me. 

That's part of your problem, then. Since the 1.0 release, the aerodynamics have favored more realistic gravity turns. This means starting your turn soon after leaving the pad, but continuing it very gradually to 45 degrees or so by 10 km, pointing prograde most of the way. It saves delta V by reducing your gravity losses, and reduces the chance of your rocket flipping by keeping it aligned with the aerodynamic forces. There should be various tutorials covering this on the forum and wiki.

[Kerbal7]

15 hours ago, Pand5461 said:

Are you joking?

I will be glad to see a picture of your rocket that cost LESS than 35,000, and can put three kerbals into orbit, with MORE than 4,283 Delta-V left in the tanks.  

[Zhetaan]

2 hours ago, Kerbal7 said:

Less weight and more Delta.V seem more important than drag and symmetry to me.

If you do nothing else, try putting some winglets at the bottom in 4-symmetry (angled at 45 degrees to avoid clipping your boosters), get rid of the front nosecone, and put the command pod on top of the crew cabin (with a parachute on top).  Then see whether you can get the same rocket into orbit for nearly the same price, but note how much more delta-V you have when you get there.

The pancake fuel tank in 2.5m has the same amount of fuel as an FL-T800, but since you need a wide stance and low centre-of-mass for landing on the Mun, the pancake tank is the better way to go.

Anyway, you did ask for the formula, and that's it.  Concentrate the mass forward as much as you can, and concentrate the drag aft as much as you can.  The imbalance is not in the rocket's shape per se but rather in the aerodynamic forces that act on it.  As you go from subsonic to supersonic speed, the drag on the nose begins to act strangely and will try to flip you.  If you have mass at the front, that gives those forces less lever arm to act.  If you have draggy bits at the rear, that gives your rocket more lever arm to counteract those forces.

Other things to try include setting fuel flow so your tanks drain lowest first going up, and at the top--especially where you have the pancake tank with the landing legs--you should consider a fairing or a Poodle engine, but one major diameter change won't knock you out of the sky.

[Kerbal7]

1 hour ago, Zhetaan said:

If you do nothing else, try putting some winglets at the bottom in 4-symmetry (angled at 45 degrees to avoid clipping your boosters), get rid of the front nosecone, and put the command pod on top of the crew cabin (with a parachute on top).  Then see whether you can get the same rocket into orbit for nearly the same price, but note how much more delta-V you have when you get there.

 

I have added 4 steerable winglets to the bottom.

If I put the command pod on top of the crew cabin, the payload becomes uncontrollable and flips on reentry. Parachutes down. Parachutes don't open, death, destruction, hair, teeth, and eyeballs, are all that's left of the passengers. The crew cabin on top of the command pod is self-righting on reentry. Although you wouldn't think so. I've tested both configurations a lot.   

With this rocket I blast off. When I drop the boosters, I start my turn. Around 15 Km or so. It's easy to fly, IF, you don't try to turn before you drop the boosters. If I try to turn before I drop the boosters, 15-20 KM, it'll spin out.

I was just wondering if there was a cheap and easy way to turn before around 20 Km or so. 

I always play career mode. So it's all about getting the most Delta-V for the least amount of money. 

[Gordon Fecyk]

I don't know about a formula, but it is possible to get your ship to orbit.

A little bit of redesign helped. I put your command pod at the top and passenger module under that, then added a 1.25 to 2.5 m adapter to reduce the drag a little bit. I also angled the rocket slightly eastward because you had launch clamps and thought, why not? 

If you have advanced tweakables turned on, you can increase the priority of the bottom liquid fuel tank. This will help keep the centre of mass ahead of the centre of drag a little. I also reduced the thrust of the Kickback SRBs to 75% of normal, so the Skipper's gimbal had a better chance of holding course. Jebediah here did the rest, using prograde hold.

I didn't know what was in that service bay. I added a small reaction wheel and put the batteries inside the bay.

53 minutes ago, Kerbal7 said:

If I put the command pod on top of the crew cabin, the payload becomes uncontrollable and flips on reentry

Back when I was playing with Alien Space Programs, I had a return craft that has two passenger cabins that would also flip. But it was just solid enough that it didn't come apart. Having a small reaction wheel in the service bay helped get through the worst of the re-entry, so when it flipped it was already slow enough to stay together. The passengers probably lost their lunch inside, though.

Otherwise, you want the heaviest bit of the return craft close to that service bay.

Edited November 25, 2017 by Gordon Fecyk
About that topside command pod...

[Zhetaan]

30 minutes ago, Kerbal7 said:

If I put the command pod on top of the crew cabin, the payload becomes uncontrollable and flips on reentry.

With this rocket I blast off. When I drop the boosters, I start my turn. Around 15 Km or so. It's easy to fly, IF, you don't try to turn before you drop the boosters. If I try to turn before I drop the boosters, 15-20 KM, it'll spin out.

I was just wondering if there was a cheap and easy way to turn before around 20 Km or so. 

I always play career mode. So it's all about getting the most Delta-V for the least amount of money. 

Right; I always put a heat shield on the bottom of my pod-cabin stack.  That makes it (just) self-stable but you need the heat shield to push the centre of mass to where it's needed to get results from that particular arrangement.

The only thing I can tell you, then, is that you've found an edge case and that you emphatically should not expect similar results with different rockets.  It's true that the atmosphere is substantially thicker below 15 km, but it's also true that prior to version 1.0, it was almost semisolid--it made more sense to burn straight up for 10 km and then turn hard over because it was more efficient to get out of the soup and then try to fly.  Post 1.0, the atmosphere is supposed to model reality somewhat more accurately; it is by no means perfect, but it is supposed to be good enough that the losses due to gravity from burning straight up outweigh--completely--the losses due to drag of remaining lower in the atmosphere for a few more tens of seconds.

Therefore, it seems to me that you've got an unusually draggy rocket (that much is obvious from the design) but you also have it coupled with enough raw power that it can essentially brute-force its way to orbit.  What is your pad thrust-to-weight ratio?

Anyway, I don't think I can do anything more with this specific rocket without either making it unflyable or else making it a completely different rocket, but I do have something for you to take a look at that will fit well with your design approach:  Cheap and Cheerful Rockets.  The thread is a bit old (some of the formatting was lost in the last forum update), but the principles are post-1.0 and therefore mostly should still work.  The important thing about the philosophy is that it focuses more on getting the most payload fraction to orbit for the least amount of money, which is in many cases more important than getting the most delta-V into space.

[Kerbal7]

3 hours ago, Zhetaan said:

Therefore, it seems to me that you've got an unusually draggy rocket (that much is obvious from the design) but you also have it coupled with enough raw power that it can essentially brute-force its way to orbit.  What is your pad thrust-to-weight ratio?

...but I do have something for you to take a look at that will fit well with your design approach:  Cheap and Cheerful Rockets.  

Thanks a lot for the link! I'll check it out. 

You can download the rocket here. It's all stock parts except for MechJeb and Kerbal Engineer. The .craft file   https://nofile.io/f/jsvO492HdHM/MOON+SHOT.craft   Put it in orbit and back to Kerbin.

This is the rockets TWR, etc.

[Plusck]

6 hours ago, Kerbal7 said:

I have added 4 steerable winglets to the bottom.

If I put the command pod on top of the crew cabin, the payload becomes uncontrollable and flips on reentry. Parachutes down. Parachutes don't open, death, destruction, hair, teeth, and eyeballs, are all that's left of the passengers. The crew cabin on top of the command pod is self-righting on reentry. Although you wouldn't think so. I've tested both configurations a lot.   

With this rocket I blast off. When I drop the boosters, I start my turn. Around 15 Km or so. It's easy to fly, IF, you don't try to turn before you drop the boosters. If I try to turn before I drop the boosters, 15-20 KM, it'll spin out.

I was just wondering if there was a cheap and easy way to turn before around 20 Km or so. 

I always play career mode. So it's all about getting the most Delta-V for the least amount of money. 

I know my post was very much made in jest, but it was intended to be serious too. My gravity turn in that album was intended to show how it can work fine - even with your original rocket format - as long as you take it easy, turn immediately off the launchpad and follow prograde without trying to turn too much in those first seconds. I got to orbit first try with only one hairy moment when it looked like it might flip, but didn't. It wasn't perfect (still 1200 m/s to add at Ap) but it was certainly better than flying straight up to 15km.

However, if you add just a couple of those short 1.25-2.5 adapters, it makes a huge difference to the drag. I added three (one for each jump from one form to the next) but I think it would have been fine not using the second one (under the lander's fuel tank).

See here:

Spoiler

No need for advanced nose cones really. I don't think they help much in this case and they are expensive and heavy.

Instead of those hydraulic manifold thingies, I used the smallest radial connectors up top and a strut at the bottom of the boosters.

And I added elevons but really I don't think it was necessary and that added another couple of thousand funds. Total: 34k

And the whole process of getting to orbit is pictured here:

https://imgur.com/a/4Zgp1

That leaves over 2k dv still in the upper stage. Steady as a rock even at 500m/s at 10 km altitude.

Edited November 25, 2017 by Plusck

[Kerbal7]

27 minutes ago, Plusck said:

I know my post was very much made in jest, but it was intended to be serious too. My gravity turn in that album was intended to show how it can work fine - even with your original rocket format - as long as you take it easy, turn immediately off the launchpad and follow prograde without trying to turn too much in those first seconds. I got to orbit first try with only one hairy moment when it looked like it might flip, but didn't. It wasn't perfect (still 1200 m/s to add at Ap) but it was certainly better than flying straight up to 15km.

However, if you add just a couple of those short 1.25-2.5 adapters, it makes a huge difference to the drag. I added three (one for each jump from one form to the next) but I think it would have been fine not using the second one (under the lander's fuel tank).

See here:

  Reveal hidden contents

No need for advanced nose cones really. I don't think they help much in this case and they are expensive and heavy.

Instead of those hydraulic manifold thingies, I used the smallest radial connectors up top and a strut at the bottom of the boosters.

And I added elevons but really I don't think it was necessary and that added another couple of thousand funds. Total: 34k

And the whole process of getting to orbit is pictured here:

https://imgur.com/a/4Zgp1

That leaves over 2k dv still in the upper stage. Steady as a rock even at 500m/s at 10 km altitude.

Thanks a lot for all the advice!

I will test the adapters and see the performance difference it makes.  

[Spricigo]

7 hours ago, Kerbal7 said:

I always play career mode. So it's all about getting the most Delta-V for the least amount of money. 

So you are failing twice.

1st by overengineering the rocket, with much more DeltaV than necessary for the mission.  You carry fuel you don't need, consequently you use more fuel.

2nd not following a good trajectory you pile up a lot of gravity and cosine losses. 

You can do your mission for a lot less, but you really need to rethink your strategy.

 

[Kerbal7]

21 minutes ago, Spricigo said:

So you are failing twice.

1st by overengineering the rocket, with much more DeltaV than necessary for the mission.  You carry fuel you don't need, consequently you use more fuel.

2nd not following a good trajectory you pile up a lot of gravity and cosine losses. 

You can do your mission for a lot less, but you really need to rethink your strategy.

 

1st. I haven't stated the mission. How can you tell me the rocket has too much Delta-V for the mission when you don't know what the mission is? How much Delta-V is needed for a mission is relevant to my piloting skills any-ways. Maybe I'm a really bad pilot and need 5,000 Delta-V  to get into orbit. Do you know? How would you know?

2nd. How much is "a lot?" How much unnecessary loss am I losing from gravity and cosine? Have you calculated this number? What is it?     

Edited November 25, 2017 by Kerbal7

[Spricigo]

2 hours ago, Kerbal7 said:

1st. I haven't stated the mission. How can you tell me the rocket has too much Delta-V for the mission when you don't know what the mission is?

You named the rocket "MOON SHOT" and said it "get to the mun just fine".  But above everything else you said, and I quoted, you try to get the most Delta-V  So, you stated objective is to have much more deltaV then necessary for the obvious mission of landing on the Mun and return.

2 hours ago, Kerbal7 said:

How much Delta-V is needed for a mission is relevant to my piloting skills any-ways. Maybe I'm a really bad pilot and need 5,000 Delta-V  to get into orbit. Do you know? How would you know?

Your posts don't mention any difficulty doing the required manuever (except the gravity turn) required for a Munshot. Thus I have no motive to assume otherwise. 

2 hours ago, Kerbal7 said:

2nd. How much is "a lot?" How much unnecessary loss am I losing from gravity and cosine? Have you calculated this number? What is it?     

Judging by the screenshots posted you used about 3.9km/s to put it in orbit. I suppose you could reduce it to 3.3km/s without much redesign. 

[Plusck]

1 hour ago, Kerbal7 said:

1st. I haven't stated the mission. How can you tell me the rocket has too much Delta-V for the mission when you don't know what the mission is? How much Delta-V is needed for a mission is relevant to my piloting skills any-ways. Maybe I'm a really bad pilot and need 5,000 Delta-V  to get into orbit. Do you know? How would you know?

2nd. How much is "a lot?" How much unnecessary loss am I losing from gravity and cosine? Have you calculated this number? What is it?     

@Spricigo is right, even though he may have come across as a bit abrupt.

With 2000 m/s still left on your upper stage in LKO, and 2400 m/s in your lander, your ship is able to land and return from the Mun, Minmus, Ike, Gilly. Going to Dres and landing would be challenging. Going to the Jool system would be doable, but you'd be limited to visiting Bop or Poll.

So for a Mun or Minmus mission, there is definitely a lot you could pare down.

 

Still, one thing I don't understand is why you're having trouble with re-entry with the mk1 pod on top. I assume you have the HECS probe core availabe, so use it! If you come in from the Mun or Minmus with approximately 30-32km Pe, and have "retrograde hold" set (on surface mode), you should have absolutely no problems:

Spoiler

 

If you revisit that album:  https://imgur.com/a/4Zgp1

I did a redesign, cutting out the upper stage altogether, putting a Poodle on the lander and putting the Mk1 command pod on top. It still has enough dv to go to the Mun and back, plus it has no aerodynamic issues, plus it costs about 5k funds less.

But still, there is more you could change. Those SRBs are too big and expensive for what they do. Smaller SRBs, with fuel on them to feed the main stack, and less fuel on the main stack, will all cost you less and give you more dv. If you activate the advanced tweakables you can micromanage fuel use across decouplers and make sure you rocket is never carrying too much fuel or empty fuel tanks for no reason.

[Kerbal7]

1 hour ago, Spricigo said:

You named the rocket "MOON SHOT" and said it "get to the mun just fine".  But above everything else you said, and I quoted, you try to get the most Delta-V  So, you stated objective is to have much more deltaV then necessary for the obvious mission of landing on the Mun and return.

Wrong. 

Nowhere have I stated my mission objective with this rocket is to go to the Mun, or Minmus, or a rendezvous, or anything else. And for a very good reason, that isn't the objective of this rocket.

My Mun rocket is much smaller than this rocket. This rocket is simply a test rocket to see how cheap I can put 3 Kerbals into orbit with the most Delta-V left over. That's it. It's a design exploratory rocket. What works, what doesn't.

It might be possible to shave off 600km/s to get it into orbit. We'll see as we make changes. 

1 hour ago, Plusck said:

@Spricigo is right, even though he may have come across as a bit abrupt.

With 2000 m/s still left on your upper stage in LKO, and 2400 m/s in your lander, your ship is able to land and return from the Mun, Minmus, Ike, Gilly. Going to Dres and landing would be challenging. Going to the Jool system would be doable, but you'd be limited to visiting Bop or Poll.

So for a Mun or Minmus mission, there is definitely a lot you could pare down.

 

Still, one thing I don't understand is why you're having trouble with re-entry with the mk1 pod on top. I assume you have the HECS probe core availabe, so use it! If you come in from the Mun or Minmus with approximately 30-32km Pe, and have "retrograde hold" set (on surface mode), you should have absolutely no problems:

This is not a Mun rocket. It's simply a test rocket.

I am glad to know this spacecraft can make it to Duna because that's my next goal. My first trip to Duna!

If you put the command pod on top of the crew cabin it will flip uncontrollably on reentry. I've tested it. 

Edited November 25, 2017 by Kerbal7

[Plusck]

10 minutes ago, Kerbal7 said:

If you put the command pod on top of the crew cabin it will flip uncontrollably on reentry. I've tested it. 

No, it doesn't. I've done this a lot and I tested it just now (results in the album I linked to).

All you need is retrograde hold and a balanced craft, and a Pe at the right height for wherever you are coming from (generally about 32 km always works).

I'll put up the craft file for the one in that album later.

 

If you look again at your pic of your craft with the KER details visible, "TORQUE" is listed as being 0.67 for your final stage. That means it's unbalanced. It should be 0.00. That could well be what is ruining your re-entry stability.

[Kerbal7]

Just now, Plusck said:

No, it doesn't. I've done this a lot and I tested it just now (results in the album I linked to).

All you need is retrograde hold and a balanced craft, and a Pe at the right height for wherever you are coming from (generally about 32 km always works).

I'll put up the craft file for the one in that album later.

 

If you look again at your pic of your craft with the KER details visible, "TORQUE" is listed as being 0.67 for your final stage. That means it's unbalanced. It should be 0.00. That could well be what is ruining your re-entry stability.

Ahhh, !

The probe core on top rights the spacecraft on reentry? Thanks! I'll check that out!