36-ton Eve sea level ascent vehicle

[Starlionblue]

Finally it all came together! I knew I had the Delta-V and TWR numbers required. As we all know, the biggie on Eve is drag, so I did some testing to find the most slippery configuration. I needed around 1800 liquid fuel for the first stage. I tested a few different stacks for performance:

• 2x FT-T800 plus 3x FL-T800 in parallel with Strack Tricouplers at top and bottom
• 1x C7 adapter and 1x Rockomax X200-32.
• 1x C7 adapter, 1x Rockomax X200-16, 1x Rockomax X200-8 and 1x C7 adapter (inverted).

The last configuration gave much better performance. Nice tapered shape not only at the top of the stage, but back inwards to the single Vector. That last bit made an unexpectedly massive difference.

Configuration I ended up with:

• First Stage: 1x Vector. 1x C7 adapter, 1x Rockomax X200-16, 1x Rockomax X200-8 and 1x C7 adapter (inverted).
• Second Stage: 1x Spike. 1x FL-T800, 1x FL-T100
• Third Stage: 1x Terrier. 1x FL-T200, 1x FL-T100. 

I went back and forth between the Mk1 command pod with a small nosecone and the Mk1 lander can with a fairing. The difference is marginal and both can do the job, but if you want a docking port the fairing with the lander can means it won't be sticking out in the wind.

The little six-legged contraption at the bottom serves as a stable launch pad from Eve.

Edited September 16, 2018 by Starlionblue

[Gordon Fecyk]

This craft's centre of lift is rather far ahead of the centre of mass to start. I had to add additional fins at the bottom, and that wasn't quite enough to keep things from wobbling.

Might I suggest a pair of FL-T400 tanks instead of a single FL-T800 on your second stage? Then you can set fuel flow priority on your second stage so it drains from bottom to top. This might help keep centre of mass behind centre of lift during that crucial part in the transition between Eve's lower and upper atmospheres. I might even go as far as using three X200-8 tanks to manage the first stage's centre of mass similarly, instead of one -16 and one -8 tank. Never mind that; didn't make much difference on the initial ascent.

You'll need to turn on Advanced Tweakables to enable the fuel flow priority settings.

I'm testing your design on Alien Space Programs so I can launch from Eve. I'm still finding a lot of wobble going on, and that's even with Realistic Atmospheres enabled. I've yet to try the stock atmosphere but I have some visual bugs to fix, it seems.

Edited September 16, 2018 by Gordon Fecyk

[Starlionblue]

I was fine with the wobble as long as I limit the Vector gimbal to around 65-70. With full gimbal I got overcontrol. Should have mentioned that.

Good thought on the fuel priority. Will play with that tomorrow. Thanks!

I start the turn around 10km. Once the first stage, ehm, stages, I lose quite a bit of speed but the terminal velocity goes down as well due to the change in shape so it actually works.

Side note: I also tried a four stage variant, dividing the current upper stage into two with a Spark for the fourth stage. There was no significant performance difference compared to the current design.

Edited September 16, 2018 by Starlionblue

[Gordon Fecyk]

I'm not convinced you have the delta-v for this, at least with the stock Eve atmosphere and launching from sea level. With Realistic Atmospheres it works quite well and I have 600-700 m/s left in a 60 by 60 km orbit. But with the stock atmosphere I can't even hit an AP of 75 km. I might have some pieces missing in my replication vs your original design, or you're using a much lighter craft inside.

Inside the fairing I put a Mk1 lander can, docking port, 4 x small solar panels and one small surface mount antenna. These were the Kerbal Engineer values I ended up with for each stage at each altitude:

Imgur album

 

Edited September 16, 2018 by Gordon Fecyk

[Starlionblue]

My Delta-V values seem close to yours. I have a docking port Jr. and two little solar panels. Above the fairing in the third stage is a FL-T100 tank.

I emptied the lander can of monoprop.

I did fiddle with the ascent profile quite a bit. Probably ran two dozen tests with different turn altitudes, turn profiles, Q limits. The differences can be quite significant.

At practically sea level on Eve

Second stage:

Third stage:

Coasting to circularisation

This orbit isn't very neat but it is clear of the atmosphere, and I have a teeny bit of Delta-V to spare.

Edited September 16, 2018 by Starlionblue

[Gordon Fecyk]

Nice job! You are certainly a better pilot than me for this. My lander can with fairing kept flipping on second stage separation, so I went with Mk1 pod plus nosecone, no fairing.

Still was short by 300 m/s. Ended up EVAing Jeb to cover the last bit, which is still OK I suppose, when you just need to get to space.

[Update] If I'm careful to keep dynamic pressure at or under 45 kPa during the first stage, I can reach orbit with just under 100 m/s to spare. I also have to start my gravity turn at 23 km, switching from SAS hold vertical to Prograde hold orbit. So the margin is really tight, but still doable. Quite the design.

Edited September 16, 2018 by Gordon Fecyk

[Starlionblue]

Not so much skill as MechJeb. 

Either way quite marginal as a design, but gets the job done. Of course, I could add a second engine and more tankage to the first stage but that bumps the weight up by a good 6-8 tons.

I did a lot of experimenting with limiting dynamic pressure. I noticed that with a slender vehicle like this the terminal velocity tends to be higher than my maximum speed throughout the ascent, so no point limiting dynamic pressure. It seems better for performance to accept higher dynamic pressure than on Kerbin, even 5 times as high.

As a general rule, wider vehicles have a lower terminal velocity. Given the soupy nature of Eve's atmosphere, making the vehicle as slender as possible with smooth, long tapers, gives disproportionately greater drag improvement compared to the same design choices on Kerbin. The thicker the atmosphere, the more frontal area and taper sharpness seem key. (I'm pretty sure the vanilla aerodynamics don't model area ruling, but Ferram does.)

Trying for a Sears-Haack body is a good starting point for Eve. On Kerbin, extreme low-drag aerodynamics are less important, so you can go for a wider design as needed structurally. Basically the thicker the atmosphere, the more slender and smoothly tapered the vehicle needs to be. On Duna, by comparison, you can get away with quite a fat "cigar" if your mission demands it, without incurring significant aerodynamic penalties.

Thanks for giving the design a shot, btw.

 

Edited September 25, 2018 by Starlionblue

[Gordon Fecyk]

I'm trying to understand something about max-Q here and how to avoid excessive stress. There's still more I can learn from this.

I have Ferram Aerospace installed; granted it's an unofficial re-compile for KSP 1.4.5, but it's still the stock parts from your original design, just using a Mk1 pod and nosecone instead of fairing and lander can, so the craft should have the same properties as on KSP 1.3.1 with an official FAR release. You brought up something I never considered in my launches before: what terminal velocity is at a given moment.

With FAR's flight details visible, I can observe the current dynamic pressure and terminal velocity. I only just noticed that terminal velocity drops sharply at a certain altitude and speed during launch, and then increases again after clearing a certain altitude.

Is it just a matter of keeping my air speed below whatever terminal velocity is? Because if I do that, I get much better performance and I can make orbit consistently. I couldn't do that at first.

Eve's one heck of a training ground for aerodynamic rocket design.

[Starlionblue]

Keeping airspeed under terminal velocity is relatively more important on Eve, because the atmosphere is so thick that drag becomes the primary constraint. On a world with a thinner atmosphere, exceeding terminal velocity does not have nearly as negative an effect.

If you flipped the rocket around, pointed the nose at the ground and let go without thrust, it would accelerate to terminal velocity, then stop accelerating. Terminal velocity is dependent on the shape of the vehicle and the density at that location. Drag is now equal to weight. You can "increase the weight" of the vehicle by adding thrust, but the drag increase is not worth it.

You can also get a terminal velocity display with Kerbal Engineer Redux. It's a selectable item in one of the configurable dashboards.

Edited September 25, 2018 by Starlionblue

[Aeroboi]

@Starlionblue Very impressive.

[Starlionblue]

Here's a link to the craft file on Dropbox, btw. Forgot I could do that! 

If you need a good low level testing spot on Eve, try lat 0, long 18 in HyperEdit.

For the ascent profile in Mechjeb, try:

• Start turn at 10km or 362m/s
• Turn shape 63.7%

Edited September 20, 2018 by Starlionblue

[HalcyonSon]

On 9/16/2018 at 7:05 AM, Gordon Fecyk said:

I'm not convinced you have the delta-v for this, at least with the stock Eve atmosphere and launching from sea level. With Realistic Atmospheres it works quite well and I have 600-700 m/s left in a 60 by 60 km orbit. But with the stock atmosphere I can't even hit an AP of 75 km. I might have some pieces missing in my replication vs your original design, or you're using a much lighter craft inside.

Inside the fairing I put a Mk1 lander can, docking port, 4 x small solar panels and one small surface mount antenna. These were the Kerbal Engineer values I ended up with for each stage at each altitude:

Imgur album

 

Eve's atmosphere goes up to 90 km in the stock game (1.2.2).

[Starlionblue]

3 minutes ago, HalcyonSon said:

Eve's atmosphere goes up to 90 km in the stock game (1.2.2).

Indeed. This vehicle can do a stable orbit higher than 100km from an equatorial sea-level location on Eve. 

[HalcyonSon]

On 9/16/2018 at 2:19 AM, Starlionblue said:

tage, but back inwards to the single Vector. That last bit made an unexpectedly massive difference.

Configuration I ended up with:

• First Stage: 1x Vector. 1x C7 adapter, 1x Rockomax X200-16, 1x Rockomax X200-8 and 1x C7 adapter (inverted).
• Second Stage: 1x Spike. 1x FL-T800, 1x FL-T100
• Third Stage: 1x Terrier. 1x FL-T200, 1x FL-T100. 

I'll definitely give this a try.  My all - 1.25 m design is flipping and losing speed on the second stage after staging off the Vectors.  It would be massively overpowered for Kerbin and probably have enough dV for a 500 km orbit, but Eve is turning out to be a different beast entirely.

[Starlionblue]

On 9/21/2018 at 10:51 PM, HalcyonSon said:

I'll definitely give this a try.  My all - 1.25 m design is flipping and losing speed on the second stage after staging off the Vectors.  It would be massively overpowered for Kerbin and probably have enough dV for a 500 km orbit, but Eve is turning out to be a different beast entirely.

Eve is something else indeed  

All 1.25m I think is too long and slender for good control. I tried that too  

You can download the craft file from the link in my post above. The only non-stock item is the MechJeb part. 

Edited September 25, 2018 by Starlionblue

[Gordon Fecyk]

1 hour ago, HalcyonSon said:

Eve's atmosphere goes up to 90 km in the stock game (1.2.2).

I managed to get to orbit (92 by 90 km) with Starlionblue's design in the stock atmosphere after a lot of practice. So no problem there. I had to be very, very precise on my ascent and I didn't have the benefit of MechJeb.

[Starlionblue]

5 minutes ago, Gordon Fecyk said:

I managed to get to orbit (92 by 90 km) with Starlionblue's design in the stock atmosphere after a lot of practice. So no problem there. I had to be very, very precise on my ascent and I didn't have the benefit of MechJeb.

Hand cranking it on Eve ascent. Thats hardcore.

[coredumpster]

Thanks for this!

I finally landed and returned a kerbal from the surface of Eve without using any cheats.  ( I did use F5 / F9 a couple times though, as I tuned my ascent ).

I stuck this craft on the roof of a Planetary Base Systems main hub building to serve as a large / heavy stable platform for launch.

Could not pilot this by hand, especially the first stage. This is definitely a fly by wire ship for me.

Dialed the Vector down to 60% gimbal, and set MJ Ascent Guidance curve at 100%, with gravity turn starting at 2500.  Made it to a 95km x 94 km orbit with about 700 dV to spare, as I launched from Eve Highlands ( a little over 1k starting altitude ).

Rendezvoused with the Eve Cycler ship and now waiting comfortably in orbit for a return window to Kerbin!  :D

[Starlionblue]

Nicely done. And yes, I agree that it is very much FBW.

I got to thinking about the wobble on my jury-rigged launch pad. Maybe if I used some struts from the beams to the second stage to hold it steady...