Eve Ascent for three Kerbals

[jost]

Hello everybody,

first post in the forum, so if I did something wrong, please drop me a note

I'm struggling to get a three-kerbal-lander to work on Eve.  I read several of the threads in the forum regarding Eve ascent. Most seem to agree, that it's best to design the lifter just for one kerman.

Although this would make things a lot easier I'm trying to do it for three kerbals, since it's part of my Eve reference mission: The idea is to send the lander first. With the next transfer window a rover with a crew of three will follow to deploy several SurfaceExperimentPacks. SEP needs a scientist and an engineer though and I need a pilot. 

At the moment after several retries, reading the forums and redesiging again I got a lander which should work.

I fail however to get it back to orbit in my tests with HyperEdit. The first stages until dropping of the bossters work fine. 
Climbing with the vector stage too. When I try to start the gravity turn with the vector at 10-15km the ship begins to tumble. When I start the gravity turn later I'm running out of fuel in the last stage. 

Screenshots
Craft file

Any hints concerning the lander design or the gravity turn?


Thanks for your time, have a nice day and regards, jost.

 

Edited April 22, 2019 by jost
Added links

[Aegolius13]

Hi, and welcome aboard!

First of all, that's a really nice build you've got going.  Looks like you've already mastered a lot of the skills needed for an Eve mission.

As far as flipping, I've had this problem too.  I found that just putting a few light fins (like the smallest triangular wing segment) did the job.  If you haven't already, you could try turning the Vector's gimbal range down.  It may be overcorrecting and contributing to instability.  The other thing that would help (for both stability and range) would be to get a nosecone on the top.  You might be able to go that, and still have an anchor point for your upper heat shield, by using a fairing with attachment nodes.  But that would add weight, which doesn't help either.

As far as getting to orbit, I think you may just not have a big enough rocket for your payload.  Maybe try using six instead of four radial boosters - I did that once, with one pair of Vectors and two pairs of Darts, and it worked pretty well.  I also imagine that bulge in the middle of your core stage is causing a lot of drag.  This is an age-old problem since building taller makes the thing harder to reenter and land with.  But you might be able to move some of that fuel to the radial boosters, by making their stacks a little higher.

 

[Foxster]

One thing that might be causing the flipping is the tip of the craft where you have a docking port. Find another way of attaching to the heatshield assembly, such as a decoupler or hardpoint. Alternatively consider not having that rear heatshield at all and put a load more airbrakes on near the top instead. 

You have an antenna and two solar panels on the side of the capsule that will be adding drag to front end. See if you can eliminate those or else move to the back of the stage. Do you absolutely need that battery on the top too?

Oh and the remains of the decouplers (the 6 dots) might add some drag - can't remember. Replace decouplers with small hardpoints anywhere you use them. 

Regarding the bulge mentioned by Aegolius13. Have a look at the newer FL-A215 and FL-A151L Adapter tanks. They have lower drag than other tanks. 

Have a check on the drag of those intakes. They used to be the best mk1 tank-toppers but they got more draggy a while ago. You might find nosecones, though heavier, are more efficient because of the lower drag.  

You don't need the Remote Guidance Unit beneath the capsule after landing and you have your crewmen onboard, so stick it on the bottom of the craft for dumping before liftoff. 

Don't forget that you are lifting here from 17m. If all else fails, find a higher spot to launch from. 

Very nice craft btw!

Edited April 22, 2019 by Foxster

[Foxster]

Oh and have you slightly messed up the fuel feed of the first two stages? Shouldn't the first stack pair feed via ducts to the second stack pair rather than the central stack? That would let you empty and dump the first stage quicker. 

[jost]

4 hours ago, Aegolius13 said:

First of all, that's a really nice build you've got going.  Looks like you've already mastered a lot of the skills needed for an Eve mission.

Thanks for the kind words, but its' less mastered but shameless copying from other Eve threads in this forum   Especially (but not only!) from @Foxster impressing low mass landers. 

Thanks for your input guys! 

I tried limiting the gimbal to 10, 20 or 30 %. That didn't help with flipping, the rocket just started tumbling earlier. 
I tried adding another two booster with vector engines. This didn't helped either, it started tumbling right at the beginning. 
So I think the key is reducing drag. I think I can remove the antenna from the capsule. The relays on the bottom can should be enough for the transfer to Eve surface. The rocket itself will be piloted by Jeb, so no need for commsat connection during ascent. The battery is a different story. I play with USI life support, where you need to supply the kerbals with EC. I'm not quite sure, whether the EC in the capsule will be enough.  I will test this. 

Getting rid of the decoulpers is excellent advice, thanks! I will also try to remove the nose clip

Concerning the jr docking port with heat shield:  I tried replacing with a small nosecone, but then the ship was killed during the reentry. It seems I need two heatshields (front and top), any idea how to attach the heatshield AND the nosecone on the top?
Concerning the fuel flow: Nope, the boosters feed the central core but thanks to fuel flow priority they will do the right thing non the less  
I tried getting rid of the duct tapes but in the end it messed up my delta v displays.

So for now: Back to the drawing board! 
 

[Foxster]

Something like this would work for the heatshield. Needs tidying up a bit of course. 

I still think you might have the ducting sub-optimal. You are burning fuel from the 2nd stage from liftoff. Consider changing from this...

...to this...

...which would mean changing the symmetry of the first two stages from quad to two duals. 

[jost]

Thanks for the idea with the hard point for attaching the heat shield. I will test it later. First I will test  exchanging the intakes with nose cones. I also tried to replace the decouplers with hard points. The reults were a quick deassemblies during reentry, thus I will stick with decouplers for the airbrakes and chutes. 

If this works I will try to change the fuel lines. A question though:  You suggest to fire the second booster stage after dropping the first one or did I missunderstood you?
Because just replacing the lines without changing the engine fire actually recudes the delta v as shown in these screenshots.

Edited April 22, 2019 by jost

[Foxster]

Have a look at this craft: https://www.dropbox.com/s/al8kysx2eec844i/EveThree Fox 2.craft?dl=0

I have made the 1st and 2nd stages two separate pairs and fed the fuel from the 1st stage to the 2nd stage to the centre stack. All engines fire from launch. This increases the dV as shown by MJ from 6671 to 6693. Not a huge difference but it all helps. Plus you drop the 1st stage quicker and so reduce drag losses.

Edited April 22, 2019 by Foxster

[KosmoNot79]

the air intakes would create a lot of drag

[Foxster]

24 minutes ago, KosmoNot79 said:

the air intakes would create a lot of drag

Not as much as you'd think from their flat-top appearance. 

For use in Kerbin's thinner atmosphere they can still be the best mk1 tank-topper because of their low weight. 

[Laie]

12 hours ago, jost said:

I'm struggling to get a three-kerbal-lander to work on Eve.  I read several of the threads in the forum regarding Eve ascent. Most seem to agree, that it's best to design the lifter just for one kerman.

That's only because, technically speaking, a single person will be enough to plant flags and gather science and so on.

At first glance, your lifter should be wholly up to the task. I'd treat all four aerospikes as a single stage, drain and drop them together. In my opinion, a first stage that doesn't get you to 300m/s is not worth having under any circumstances; and if reasonably streamlined (as in your case), it's worthwhile to cling to the first-stage engines up to 500m/s or more.

13 hours ago, jost said:

When I try to start the gravity turn with the vector at 10-15km the ship begins to tumble. When I start the gravity turn later I'm running out of fuel in the last stage. 

I think that may be the main problem. You should start your gravity turn very early, practically as soon as you can. Lean it over very very gently at first, only nudging a bit above or below prograde as needed, in order to reach a certain mark --  say, 70° @ 300m/s or something. Example. I can't tell you what mark to aim for, that's different for every rocket.

At your TWR, it will be about 35 seconds until you reach 300m/s. That is actually a lot of time, though it may seem much shorter as you try to control your rocket. Anyway, try to get it on the right track early in the ascent, then toggle SAS to prograde or just take the hands off. If the rocket is aerodynamically stable, hands off tends to work better than SAS -- and IMO it's totally worthwhile to add a few tailfins if needed. The 10kg "Basic Fins" already go a long way, and will withstand entry if protected by a heatshield.

[Aeroboi]

17 hours ago, jost said:

I read several of the threads in the forum regarding Eve ascent. Most seem to agree, that it's best to design the lifter just for one kerman.

I would suggest this since 1 kerbal launch vehicles often need less less parts or stage complexity so for a very early learner who practices Eve launches it is advised to try this.
Honestly, it doesn't matter much if you really want to try. 3 kerbals is one extra pod or a larger one meaning a different launch vehicle, either 2.5m sized or 1.25m sized with asparagus like yours is but with extra engines, also like your's is. If your willing to attempt 3 kerbal launch vehicles like your doing for a first time eve launch vehicle then it involves somewhat more engineering compared to a 1 kerbal launch vehicle.

Idea...

In the VAB take off the whole lander part including the stages during the part of the accident.
Activate CoM and CoL markers and activate RCS build Aid mod to view wet/dry com markers (yellow and red CoM orbs)

Is the red orb below the Center of lift? If so you need to find a way to lower the center of lift during that stage, for instance use small elevons at the bottom. Either way have some fins near the bottom even if the CoL is close to the CoM. Even if the dry CoM is close to the CoL the dry mass of the engines plus the CoL location could swing the vessel around. Remember that a rocket without fins also has a center of lift due to body lift even on Mk1 fuselages. For that matter the height of the asparagus tanks themselves could alter the aerodynamic difference.

Also, struts also induce drag albeit little. I see you have quite a few on top. Remember that the parent part creates the drag, that is the part where the strut part is first connected to. Is it the Mk1 capsule, if not try to attach them from the heatshield to the pod. I believe that isn't possible directly so place something between the heatshield. Also know you can activate "Autostruts" in the main menu "general" settings that can strut a part by right clicking it and then use the autostrut function in the context menu.
 

Edited April 22, 2019 by Aeroboi

[jost]

Again thanks a lot for your ideas. 

Sadly none of them gave the intended results. I tried to screenshot every case: 

https://imgur.com/a/rDp1YwH 


More and more I think @Foxster is right that I need to choose a higher start position. On the other hand in my last tries I already selected positions on midlands or highlands biome. 

I also think, that @Laie is right, that I should start the gravity turn earlier with the original craft. However I have no idea, when this earlier should be. 
Everytime I tried a gravity turn with the Vector stage I ended up with a flipping rocket, loosing velocity. I tried at 10km and 20km .
Adding a fin (see screenshots) didn't work since some of the fins disapeared during reentry ): 

Any ideas?

Regards, Johannes.

[Ratwerke_Actual]

20 hours ago, jost said:

Hello everybody,

first post in the forum, so if I did something wrong, please drop me a note

I'm struggling to get a three-kerbal-lander to work on Eve.  I read several of the threads in the forum regarding Eve ascent. Most seem to agree, that it's best to design the lifter just for one kerman.

Although this would make things a lot easier I'm trying to do it for three kerbals, since it's part of my Eve reference mission: The idea is to send the lander first. With the next transfer window a rover with a crew of three will follow to deploy several SurfaceExperimentPacks. SEP needs a scientist and an engineer though and I need a pilot. 

At the moment after several retries, reading the forums and redesiging again I got a lander which should work.

I fail however to get it back to orbit in my tests with HyperEdit. The first stages until dropping of the bossters work fine. 
Climbing with the vector stage too. When I try to start the gravity turn with the vector at 10-15km the ship begins to tumble. When I start the gravity turn later I'm running out of fuel in the last stage. 

Screenshots
Craft file

Any hints concerning the lander design or the gravity turn?


Thanks for your time, have a nice day and regards, jost.

 

https://youtu.be/w5YgAFebkpA

You are welcome to any ideas from my rig.... I had higher standards, e.g. 3kerbals, Hab unit, and Kerbin return vehicle, all launched from the beach.

This thing did a 30ton payload to orbit launch, so it's overkill for your mission.

Sadly the only vid I kept was the bad gravity turn version. But one run got to orbit on 5100m/s  Δv.

Get some research done on using crossfeed and fuel prority settings to stage without fuel lines.

Edited April 22, 2019 by Ratwerke_Actual
spelling

[Laie]

1 hour ago, jost said:

I also think, that @Laie is right, that I should start the gravity turn earlier with the original craft. However I have no idea, when this earlier should be.
Adding a fin (see screenshots) didn't work since some of the fins disapeared during reentry ):

Zero heatshield at the bottom? Then it's going to be tough. Though I wonder how the lander can survives in that case. Anyway, the proper place for tailfins would be as low as possible, practically touching on the Vector. In that position they may actually be shielded by the lander can.

As for initiating the turn. You have Mechjeb, why not use SmartASS? I've had great success with that thing.

1. dial in 90/90 for pitch and heading, leave "roll" unchecked. Don't activate it just yet.
2. turn on SAS in stability assist, take off
3. as soon as you're off the ground, hit "execute" on Smart ASS.
   If your rocket starts flipping then and there, it may be necessary to tweak gimbals or PID settings or both. More below.
4. do a slow clicky-clicky to gradually tweak the pitch. Perhaps one degree at 50m/s, another at 80m/s, and then one more click for every 10m/s of airspeed gained. That means you start clicking faster as acceleration picks up.
5. set yourself a target pitch-by-airspeed you want to reach. If you find that you're coming out too high or too shallow, click faster or slower. Ideally you should never deviate from prograde by very much, but well.... ideals. To be pursued but never attained.
6. as you cross your mark, hit the SRF+ button and then just follow prograde.
7. you can still adjust your course by using the +-1 buttons when going prograde, but try to avoid it. Or at least to get it over with early. Once your going at two or three times the speed of sound, you really want to point your rocket straight into the wind, and not deviate by one bit until you're above 50km.

As for tweaking gimbals and stuff: First of all, make sure that you have roll disabled on the Vector. Next, reduce gimbal range. Some 20% gimbal ought to be fully enough for your vessel. If you still flip, open "Attitude Adjustment" and try the "hybrid" controller.

ETA: really, disable roll on the engine. Looking at your screenshots, I have an inkling that this is the cause of your problems. Either that, or the boosters are so wobbly that they induce torque, in which case you need moar struts (auto- or otherwise).

The whole stack from the Terrier down should have Autostruts on, and set to grandparent. Aerospikes should also be set to grandparent, and nosecones/intakes to root part.

Edited April 22, 2019 by Laie

[Foxster]

Try it with the mod Gravity Turn Continued to automate the flight. If you can make orbit with some dV to spare using it then you'll know the craft is capable of making orbit.  

Once you have a good craft that can make orbit "in the simulator" then you can fly it yourself manually. 

[jost]

WOW! Gravity Turn Continued makes a BIG difference. I used it with my original design after removing the solar cells and , disabling roll of the vector. Limiting gimbal to 20 led to nice explosion.
So I changed gimbal to 100%, kept roll disabled  activated Gravity Turn Continued with destination height of 90km and finally orbit:
Orbit with Gravity Turn

Since I'm using MechJeb anyway I'm quite satisfied with it, whether I use MechJeb for Autopilot or something doesn't matter to me
Designing rockets is fun, flying them not so much, at least for me
Tomorrow I will try whether it will still work with added battery and solar cells. While life support isn't actually an issue (the kerbals can survive up to 15 days without EC), I prefer to have a minimal control for docking later. 
And although I don't like flying rockets at least one time I want to do it for my own

Thanks to @Foxster and @Laie for their helpful hints!

[StrandedonEarth]

Well, the first thing to know about Eve is...

Spoiler

Personally, I only know one way to get three kerbals off Eve...

Spoiler

 

 

More seriously, the only advice I have that hasn't been mentioned already is the gravity turn. I thought that on Eve, it was better to go straight up for a little while, to get out of the thickest part of soup as quickly as possible. This will lessen aero forces making it less likely to flip, as well as reducing drag from trying to go fast through the soup. I can't remember how high I went before starting the turn last time I did an Eve ascent...

Edited April 22, 2019 by StrandedonEarth

[KosmoNot79]

13 hours ago, Foxster said:

Have a look at this craft: https://www.dropbox.com/s/al8kysx2eec844i/EveThree Fox 2.craft?dl=0

I have made the 1st and 2nd stages two separate pairs and fed the fuel from the 1st stage to the 2nd stage to the centre stack. All engines fire from launch. This increases the dV as shown by MJ from 6671 to 6693. Not a huge difference but it all helps. Plus you drop the 1st stage quicker and so reduce drag losses.

I heard that they're drag inducing unless the intakes are closed

[Foxster]

4 hours ago, KosmoNot79 said:

I heard that they're drag inducing unless the intakes are closed

Nah. That was the case several years ago but not anymore. 

[Aeroboi]

16 hours ago, jost said:

Since I'm using MechJeb anyway I'm quite satisfied with it, whether I use MechJeb for Autopilot or something doesn't matter to me
Designing rockets is fun, flying them not so much, at least for me
Tomorrow I will try whether it will still work with added battery and solar cells. While life support isn't actually an issue (the kerbals can survive up to 15 days without EC), I prefer to have a minimal control for docking later. 
And although I don't like flying rockets at least one time I want to do it for my own

While I would recommend Mechjeb to begin with for ease of game play I do have some stock tips.
Always have SAS stability hold "ON"
Always try to tap W,A,S,D, instead of trying to hold it. That means try to hold W for 0.5 seconds, release W then press again for 0.5 seconds and keep tapping to slowly pitch over. Otherwise immediate control inputs can cause a rocket to flip or pitch wildly causing aero drag and Dv losses.
Whenever I create a Eve rocket I keep trying to aim for a specific speed at a specific altitude and then hold prograde from there on.
If it happens I arch to high I have to aim for a slower speed a a lower altitude. For instance activating prograde hold at 16km with 600m/s @ 30 degree pitch gets me close to 90km Apoapsis near orbital velocity when nearing 60km altitude on a specific rocket. Your rocket has a best launch profile also and the launch profile differs per each design, try to find yours, write it up somewhere so you know how to launch the rocket the same way.

[jost]

Hello everybody,

I was busy with reallife  so no time for kerbal. However I again revised the craft and again managed to ascent with the help of Gravity Turn Continued and MechJeb
Main change: I added a fairing because otherwise my solar cells were toasted during ascent and replaced the probe core with a science storage unit. I also added a nose cone. 
Since this led to more mass I replaced the FL-100 with a toridorial fuel tank. 
At the moment I'm to lazy to try to do it manually, maybe I will do this on my real Eve mission without HE aid

Again many thanks for your hints folk however I now have a new dilemma: The forum expect that I mark one answer as "best answer". However I can't decide since I ended with using advice from any posts. So: What should I do?

If anybody of you is interested here is another imgurl album:
https://imgur.com/a/jvo0E9z 
And the craft, if some want to try the craft: https://drive.google.com/open?id=1JXJ6260lyFaQIDTQWLuK3DT8zpER7zmq 
Please note: I never managed to launch from sea level, midlands or highlands should do. 

Again thanks for every hint folks, I will now beginn the real thing aka building the interplanetary fleet for the big Kerbal Eve expedition

Regards, jost.