[It's Back!] Another Voyage Ares Mission, and A KSP'ers Crash Course in Realism Overhaul.

[MaxL_1023]

It might just be limitations of the designs you are using. It is hard to get something onto the martian surface and then back up carrying crew, especially when you need to haul the whole thing from earth. Limiting yourself to a Saturn V scale booster (single launch) makes it even harder. 

[TiktaalikDreaming]

10 hours ago, Nittany Tiger said:

I decided to move onto the MEM tests, leaving the RCS issue to later.  I actually discovered that the RCS ports on some parts weren't firing due to not being connected to any fuel sources, causing them not to function.  I was able to correct that one set of ports but not the other since it was placed on procedural fairings.  Even enabling crossfeeding and placing fuel pipes to the fairing base didn't correct the issue.  I'll look at it later.

The MEM tests have proved to be my first major roadblock in this endeavor.  On ascent, the MEM performed well, and I was able to make low Martian orbit easily.  The craft did use up a lot of RCS to steady itself during the first stage due to its offset CoM, but after dumping the tanks, it flew well.  I see this as a design issue of the MEM (not the mod, but NAR's proposed version).

On descents....I made a lot of pancakes.  This is because Tiktaalik's MEM is designed to operate differently than the one in Voyage, despite being the same craft in general.  Tiktaalik's requires a chute and ballute for a safe descent.  Voyage ditched these two parts and had the MEM just use aerodynamic braking and a powered descent phase.  Therefore, trying to do a Voyage-style descent in the base MEM results in death and destruction. 

I've made approximately 8 - 10 practice descents in total, and they've all ended in crashes.  First two were from 150km circular orbits, the next couple from 300km circular orbits.  Both times, I couldn't slow down fast enough to prevent a crash.  Furthermore, I discovered that the descent stage lacked enough fuel to make a safe descent, only containing 900m/s delta V.  My solution was to add more fuel in the descent stage to extend the usable descent burn time and hopefully have enough delta V to slow the craft down for a safe landing.

Note: The final tanks were longer than this.

I was able to squeeze in enough fuel to bring the descent stage delta V to about 1800m/s while keeping the Martian TWR at 1.0.  A few more descent tests resulted in the craft being better able to kill velocity, but I still ended up crashing due to bad piloting (practice makes perfect).  Also, the additional fuel meant additional weight, but the deorbit solid motors were still able to get the MEM into an orbit that either fell below the ground at 150 km or resulted in aerocapture at 300km.

Unfortunately, even with the additional fuel, this may still not be enough to allow for a Voyage-style ascent.  Delta V margins are still tight, and so are RCS fuel margins.  Turning the craft ate up precious RCS fuel and descent delta V, so I found that I would always run out of RCS trying to keep the craft stable during powered descent, and on the last test, even with the large gimballing range of the aerospike engine, I ended up tumbling and crashing after running out of RCS.  Every other time, I was fighting to keep the vertical velocity low enough to hope that the MEM slowed down enough from air resistance to get the surface speed below the remaining delta V.  This took practice, and I almost achieved this on my last descent attempt before the tumbling hit and I lost the lander and crew to a fiery crash.  The other times, I would end up unable to kill my descent fast enough with the engine, and I'd pancake.  This is not including my attempts at doing a skip re-entry, which uses up RCS as well as I need it to angle the MEM in the hopes of gaining a bit of lift on descent.

Therefore, I'm stuck with this issue of needing more RCS fuel for descent, which adds more weight and thus requires more descent fuel, eventually leading to an unsolvable problem thanks to the law of diminishing returns and the rules of the book.  The only possible solution to this is to use some of the ascent stage RCS on descent, but I'm not sure if there's enough margin for that.  Furthermore, I still may need more descent fuel as it is to ensure I don't run out, as I have never got the craft slowed down below 400 m/s before a crash or unrecoverable tumble (most of this velocity being horizontal).

As a result, I'm stuck wondering if it's even possible to do a Voyage-style descent in the NAR MEM.  I'm also stuck wondering what Columbia did in the book to supplement the removal of the chutes and ballutes.  I vaguely remember them removing them because they weren't going to make much of a difference, but i don't have a copy of the book with me to check what they did, or if there's any indication of how Gershon landed the Challenger.  Maybe the book took an artistic license on the whole thing, which would seem odd for a hardcore science fiction novel.  As a result, I may have to break the purity of this recreation and use chutes and ballutes for my descent, or add aerobrakes.  This is not something I want to do because I think it would ruin the challenge of doing it by the book, but if the book presents something that is impossible in the game, I may have to just ignore the book and think outside the book.

I'm open to suggestions.

The RCS is overpowered, so eats propellant faster than needed.  I have been shunting the thrust limiter to around 10%.  It then lasts heaps longer, and doesn't tend to do the vibrating over-correction thing.

The chutes don't yet have working RO+RSS config.  So, they just act like standard chutes, meaning they won't do much to stop you from pancaking.  They might help a teeny bit though.  The issue with trying it using just blunt body aero plus rockets is there's not enough rocket in that descent stage, because the aero braking is next to useless.  You've added a chunk of fuel, but a TWR of 1.0 won't do a lot of slowing you down.  So, you'll end up needing more of it.  You might be better off with a set of retro solid boosters with higher thrust, to do an initial deltaV.  Maybe attached to the disposable side walls.  Or hack a higher thrust into the descent engine.

 

Or find some other way of increasing the drag.  Can't see too many options without either chutes or some kind of aerobrake

[Nittany Tiger]

3 hours ago, MaxL_1023 said:

It might just be limitations of the designs you are using. It is hard to get something onto the martian surface and then back up carrying crew, especially when you need to haul the whole thing from earth. Limiting yourself to a Saturn V scale booster (single launch) makes it even harder. 

These are all tests for the big mission.  I'm practicing and perfecting each part of the Ares mission before I do it all in one go.  For the MEM tests, I've just been using Hyperedit to put the MEM in orbit around Mars.

The Saturn VB is not taking the crew + MEM to Mars.  The Saturn VB is just lofting half of the Ares ship to orbit where it will meet with a propulsion stack, dock with that, and then do the Mars mission.  The rest of the stack, according to the book, took nine launches of the Saturn VB to assemble and fuel.  I could do the nine assembly flights, but I feel it's not that important to the main mission recreation.  Furthermore, the limitations of the game would make it hard to assemble the propulsion stack properly.  I could use docking ports and Kerbal Inventory System, but in my mind, it wouldn't look right, unless I turned the S-II interstage into a docking port and decoupler.  I guess assembling the propulsion stack could be something for later.  As far as there being limitations with the Saturn VB, I haven't found there to be any, especially after rebuilding the Ares Mission Module.  That module masses about the same as the Skylab, and the Saturn VB can loft that + MEM + Apollo CSM + an Orbital Maneuvering Module into orbit with fuel to spare.

I'm thinking of illustrating the entire mission before I perform it, maybe to clear up what I'm doing since I know not everyone has read Voyage.  Plus, I can make some fun drawings.

22 minutes ago, TiktaalikDreaming said:

The RCS is overpowered, so eats propellant faster than needed.  I have been shunting the thrust limiter to around 10%.  It then lasts heaps longer, and doesn't tend to do the vibrating over-correction thing.

The chutes don't yet have working RO+RSS config.  So, they just act like standard chutes, meaning they won't do much to stop you from pancaking.  They might help a teeny bit though.  The issue with trying it using just blunt body aero plus rockets is there's not enough rocket in that descent stage, because the aero braking is next to useless.  You've added a chunk of fuel, but a TWR of 1.0 won't do a lot of slowing you down.  So, you'll end up needing more of it.  You might be better off with a set of retro solid boosters with higher thrust, to do an initial deltaV.  Maybe attached to the disposable side walls.  Or hack a higher thrust into the descent engine.

 

Or find some other way of increasing the drag.  Can't see too many options without either chutes or some kind of aerobrake

I'll drop it to 10%.  It defaults to 60%.  I very much notice SAS and MechJeb's Smart ASS overcorrecting and thus gulping RCS fuel, so I tend to use it sparingly on descents.  I actually was pulsing the RCS initially until the lower atmosphere, where I would purposefully burn it off to drop the mass of the MEM.

Since you mentioned aerobraking to be useless, I've started to wonder if FAR is just underdoing the drag of the MEM.  Maybe the part drag needs to be increased.  I can't be sure with theoretical calculations.  I say this because on descent, I barely experience any heating of the MEM during descent.  Some ablator gets burnt off, but not much, and there's no re-entry flames.  I'm thinking maybe the drag needs to be increased.  As I mentioned before, I get about 1 g of deceleration from aerobraking at the lower 20km of the atmosphere without engine assistance.  Without cracking open my fluid mechanics textbooks and doing the math, I can't say if that's accurate.  What I can do is post this screenshot of some numbers to give you an idea of what's going on, and I can take more.

For the record, this is the MEM with the added descent fuel.

Sounds like what you suggested is the best idea.  Either up the thrust, add disposable retros, or add aerobrakes.  Could also hack the drag of the MEM, but that might break reality, and I'd rather not cheat physics.  The only way to check if the drag is too low is to figure out how much drag the craft should be experiencing on descent in the real Martian atmosphere.

[Dafni]

Sorry, cant help you as such, just root through my memory for what was in the book exactly.

IIRC they dont mention much about fuel consumption on descent and ascent of the MEM. They just describe a fully powered landing, where Gershon is able to fly across the surface for a bit to select a landing site, running low on fuel, a lot like Armstrong and Aldrin in the Apollo11 LEM.

The ascent is not even described in the book, it end with York stepping out on to the surface and walking around for a bit. Previously they only describe the ascent in hardware terms: the center stack separates from the landing frame, using the same aerospike engine as it did for the descent... no mention of tanks etc, IIRC

I might pick up the book again over the weekend and re-read the landing sequence if you want.

Edited September 15, 2016 by Dafni

[TiktaalikDreaming]

4 hours ago, Nittany Tiger said:

I'll drop it to 10%.  It defaults to 60%.  I very much notice SAS and MechJeb's Smart ASS overcorrecting and thus gulping RCS fuel, so I tend to use it sparingly on descents.  I actually was pulsing the RCS initially until the lower atmosphere, where I would purposefully burn it off to drop the mass of the MEM.

I drop it to 9-10% and pulse it.  Hopefully in 1.2 there won't be such a need to pulse, but it may be a while before RO etc is updated for 1.2.  On ascent, I have RCS on very briefly to start, but then reserve it for docking, the gimbal on the ascent motor should do most of the flight control.

Quote

Since you mentioned aerobraking to be useless, I've started to wonder if FAR is just underdoing the drag of the MEM.

Well,, it's not useless, it's just no-where near sufficient.  You'd usually still be going at around 1km/s when you hit the ground, and there's not 1000m/s of delta V to spare.

As to whether that's wrong, I doubt it.  A brief read on the various mars rover landings suggests that Mars is a particularly nasty mix of atmosphere enough to be an issue, but not enough to be useful.  And the current procedure seems to be enormous drag chutes followed by some sort of rocket assist for the final landing.  Or giant bouncy ball.  But you can't really put astronauts in one of those bouncy ball things.  Or an ascent stage for that matter.  The fact that getting a Mars landing in RO+RSS is really hard, seems accurate rather than troubling.

Quote

Sounds like what you suggested is the best idea.  Either up the thrust, add disposable retros, or add aerobrakes.  Could also hack the drag of the MEM, but that might break reality, and I'd rather not cheat physics.  The only way to check if the drag is too low is to figure out how much drag the craft should be experiencing on descent in the real Martian atmosphere.

I should mention I haven't read whatever book is being discussed, and so this MEM is not based on that, but the NAR studies.  So, the engines, the fuel supply, all that is working on the assumption of those chutes helping.  So, to use these parts, without chutes, you need to find at least a few hundred m/s in delta V of deceleration.  Based on http://www.projectrho.com/public_html/rocket/realdesigns2.php#rockwellmem, you'd need to find about 660m/s (dif between MACH 3.5 and 1.5).  That said, that descent plan does seem to assume more aero deceleration before the chutes than what is experienced, and much higher up too.  (I'm assuming an Earth surface level MACH reference, rather than local speed of sound)

[Nittany Tiger]

6 hours ago, Dafni said:

Sorry, cant help you as such, just root through my memory for what was in the book exactly.

I could buy the book myself, or take you up on that offer.   I actually want to go through and recreate other missions from the book like Moonlab and Apollo N.

3 hours ago, TiktaalikDreaming said:

I drop it to 9-10% and pulse it.  Hopefully in 1.2 there won't be such a need to pulse, but it may be a while before RO etc is updated for 1.2.  On ascent, I have RCS on very briefly to start, but then reserve it for docking, the gimbal on the ascent motor should do most of the flight control.

I'll do that.

3 hours ago, TiktaalikDreaming said:

Well,, it's not useless, it's just no-where near sufficient.  You'd usually still be going at around 1km/s when you hit the ground, and there's not 1000m/s of delta V to spare.

As to whether that's wrong, I doubt it.  A brief read on the various mars rover landings suggests that Mars is a particularly nasty mix of atmosphere enough to be an issue, but not enough to be useful.  And the current procedure seems to be enormous drag chutes followed by some sort of rocket assist for the final landing.  Or giant bouncy ball.  But you can't really put astronauts in one of those bouncy ball things.  Or an ascent stage for that matter.  The fact that getting a Mars landing in RO+RSS is really hard, seems accurate rather than troubling.

Yeah.  I read a paper yesterday about Mars landing techniques and hard data, and they seem to use either aeroshells with parachutes and either rockets on the final descent or airbags (or in the case of Curiosity, a skycrane).  Either way, the way Voyage does it seems a bit simplistic and maybe inadequate.

3 hours ago, TiktaalikDreaming said:

I should mention I haven't read whatever book is being discussed, and so this MEM is not based on that, but the NAR studies.  So, the engines, the fuel supply, all that is working on the assumption of those chutes helping.  So, to use these parts, without chutes, you need to find at least a few hundred m/s in delta V of deceleration.  Based on http://www.projectrho.com/public_html/rocket/realdesigns2.php#rockwellmem, you'd need to find about 660m/s (dif between MACH 3.5 and 1.5).  That said, that descent plan does seem to assume more aero deceleration before the chutes than what is experienced, and much higher up too.  (I'm assuming an Earth surface level MACH reference, rather than local speed of sound)

The book is called Voyage by Stephen Baxter.  It's about alt-history post-Apollo NASA where Kennedy survives his assassination attempt and helps convince Nixon to commit to a Mars landing over the Space Shuttle when he's deciding what to do with NASA after Apollo.  The book itself jumps between a detailed description of the Ares mission from launch to the landing on Mars to the events that happen between Apollo 11 and the fictional 1985 when Ares is launched.  I was able to read the book online somewhere and can't find it again, but I only got a bit past the development of the MEM and the Saturn VB launch failure.

The book bases its rockets and spacecraft off of real and proposed real-world craft, and from what I gather, the MEM in the book is based on the NAR MEM except that it doesn't use parachutes or ballutes.  Given how challenging it is to land on Mars, that seems a bit ludicrous.

6 hours ago, Dafni said:

The ascent is not even described in the book, it end with York stepping out on to the surface and walking around for a bit. Previously they only describe the ascent in hardware terms: the center stack separates from the landing frame, using the same aerospike engine as it did for the descent... no mention of tanks etc, IIRC

Funny you mention that because this MEM uses two different engines.  The recreation here uses two different engines. https://forum.nasaspaceflight.com/index.php?topic=35720.0 (and I've been using this thread for references for my craft).  Chris P. Bacon's MEM also uses two different engines.  Maybe Baxter's MEM is nothing like the NAR MEM.  I don't recall if that's true, though.  I like the dual-engine design personally.  Tiktaalik's MEM handles ascents just fine as I mentioned before and I have no current plans to modify that.  It's the landings that are tricky.

Granted that fact, I guess I can take more liberty with my MEM than I'm allowing myself and not bust the recreation up.  Given how little info there is on it in the book as far as I know, maybe I can add another personal touch here and use airbrakes and/or retros for the descent.  KSP includes some airbrakes, and I can always download some from the B9 pack as well.  I feel like the real MEM would go with this.  I also thought of adding a heat shield under the MEM base that has the 70 degree angle design to improve the aerodynamics, but that may or may not work and could add considerable launch weight.  In the end, I have to keep the MEM light enough for the Saturn VB to loft it and the rest of the Ares stack into orbit (along with the Venus probe I want to make).

Part of the fun of playing KSP RSS/RO is playing aerospace engineer and figuring out these problems, even if we're going by a work of hardcore science fiction, and I wanted to do this recreation to see if the whole thing was possible in real life to begin with.

Edited September 15, 2016 by Nittany Tiger

[MaxL_1023]

Could you offset the MEM/s center of mass using the descent stage? It will be harder to land on rockets, but you will get a lot of lift in the upper atmosphere flattening out your descent profile. You should get a couple hundred m/s of effective braking due to the longer time you spend above the surface. The MEM is about the right shape for it - similar to the Apollo Command Module which did the same thing to re-enter Earth's atmosphere. Since you drop this stage on takeoff you no longer have the issue on ascent. 

The extra time in the air might also let you slow enough to deploy chutes, which while not enough to land on their own should get you below 500 m/s and let you use your engines for a final landing sequence. 

If they get the RSS version of that inflatable heatshield part it would also be very useful - it would scale to be about 20m in diameter but fit in a 4m package when deflated. Since drag scales with surface area you will end up slowing down a lot more, even on Mars. 

 

[Nittany Tiger]

39 minutes ago, MaxL_1023 said:

Could you offset the MEM/s center of mass using the descent stage? It will be harder to land on rockets, but you will get a lot of lift in the upper atmosphere flattening out your descent profile. You should get a couple hundred m/s of effective braking due to the longer time you spend above the surface. The MEM is about the right shape for it - similar to the Apollo Command Module which did the same thing to re-enter Earth's atmosphere. Since you drop this stage on takeoff you no longer have the issue on ascent. 

The extra time in the air might also let you slow enough to deploy chutes, which while not enough to land on their own should get you below 500 m/s and let you use your engines for a final landing sequence. 

If they get the RSS version of that inflatable heatshield part it would also be very useful - it would scale to be about 20m in diameter but fit in a 4m package when deflated. Since drag scales with surface area you will end up slowing down a lot more, even on Mars. 

 

I thought of offsetting the CoM of the descent stage more to get that lift and thus do a skip entry onto Mars.  I do notice when I pitch the MEM about 30 degrees, I get a drop in vertical velocity. by a couple of tenths of a meter per second.  The MEM descent engine does gimbal 17 degrees, which should correct for the offset CoM.  This is because the CoM is offset already (and it gets more offset as fuel drains out).  Just need to offset it more to get a passive angle of attack of about 10 or 20 degrees.  It would be nice to find a way to get more time in the atmosphere while not plummeting to my doom at 100 - 200 m/s.

I'm currently experimenting with airbrakes on the descent module to add drag on descent.  Haven't done any descent tests yet.  I'm still trying to figure out where to put them without adding any width to the MEM when undeployed, which would made it too wide on launch, or having the side-mounted heat shields destroy the airbrakes on jettison.  I can't attach anything to those shields as far as I know since the game treats them as fairings.  Also, I discovered what I think was a bug in the realism overhaul config that was causing the MEM to have a lower ISP than it was supposed to and to have no ISP or thrust curve.  Looking at the config file, it looked like either a possible syntax error or a purposeful commenting out of the code.  Either way, I fixed those lines of code and I have yet to experience any issues.  According to MechJeb, the MEM should have more descent delta V, so I went back to the version without extra fuel for now.

Edited September 15, 2016 by Nittany Tiger

[MaxL_1023]

How much margin do you have on the Mothership? You could attach radial drop tanks underneath the main body (next to the engine) to give you more delta-v, dropping them as they empty using fuel lines. 

[TiktaalikDreaming]

3 hours ago, Nittany Tiger said:

I can't attach anything to those shields as far as I know since the game treats them as fairings.

Actually they treat them as "don't allow surface attach" due to the zero in the forth spot of the line
attachRules = 1,0,1,0,0
Change that to a 1, and you're good to go.  eg;
attachRules = 1,0,1,1,0

I'm also wondering if there's a set of air brakes in some mod that could be used to replace that shell.

EG like this, but actually covering where the shell would otherwise be;

[Nittany Tiger]

36 minutes ago, MaxL_1023 said:

How much margin do you have on the Mothership? You could attach radial drop tanks underneath the main body (next to the engine) to give you more delta-v, dropping them as they empty using fuel lines. 

I'm not sure.  I may have the margins on the launch vehicle though.  It currently has about 9,850 m/s of delta V.  I haven't updated the launch vehicle with the new MEM though because I'm still trying to create one that doesn't crash.

I'll know how much room I'll have on the Ares ship as a whole when I start practicing putting it together and flying the mission.

6 minutes ago, TiktaalikDreaming said:

Actually they treat them as "don't allow surface attach" due to the zero in the forth spot of the line
attachRules = 1,0,1,0,0
Change that to a 1, and you're good to go.  eg;
attachRules = 1,0,1,1,0

I'm also wondering if there's a set of air brakes in some mod that could be used to replace that shell.

EG like this, but actually covering where the shell would otherwise be;

I could put your fairings on IR hinges and have them open up instead of be thrown away.  It would mean adding mass on landing though.

Good news is that the airbrakes are working, and I'm getting more drag from atmospheric entry now.  The bad news is that it's still not enough for a safe landing, but it's extremely close.  Here are pics of my last two landing attempts, and this is how close I came to landing before I ran out of descent propellant:

Airbrakes hinge at the top of the descent stage:

Airbrakes hinge at the bottom of the descent stage:

Both times, I used procedural control surfaces set only to be spoilers.  Both times, I still crashed, but some of the craft survived including the crew section, so in these cases the crew would live a little bit longer.

The first airbrake setup seem to keep my craft stable, but the second setup caused a lot of issues with wobbling on descent.  I'm not sure exactly why, but maybe the sloppy placement of the spoilers was to blame.

In either case, I'm probably going to have to add more descent fuel or retrorockets for the final descent.  I could potentially make the spoilers expendable and jettison them on final descent and giving me a bit more delta V, but I think in the end, I'm going to have to add more fuel.

[MaxL_1023]

Where are you landing? Hellas Basin or the Vallis Marineris have much denser air compared to Martian MSL. 

[Nittany Tiger]

4 hours ago, MaxL_1023 said:

Where are you landing? Hellas Basin or the Vallis Marineris have much denser air compared to Martian MSL. 

Shooting for Mangala Valles like in the book.  I think it has a low altitude.  Been landing anywhere along the equator on the light side of Mars for my tests so far.

Thing is, KSP RSS seems to put MSL on Mars at the lowest point on the planet, not at some other level.  I'll have to verify this with SCANSAT, but this may or may not affect the vertical atmospheric layout of Mars in RSS.  To verify would mean needing to record data and compare it to a standard atmospheric model (good old p(z) = p(0)exp(-z/(gH)) IIRC).  I'd love to record and plot data for my descents as well to see if there are any improvements I can make, and to show my descent parameters.

[MaxL_1023]

If you have a barometer, check the surface pressure. If MSL is set to the lowest point on the planet your pressure will be somewhat below 0.615 KPA everywhere on the surface. You can get close to 1 KPA in the lowest regions, maybe even a little more if KSP uses the real Martian atmosphere. 

[TiktaalikDreaming]

I made a shell airbrake thing.

I managed to get velocity down to 50m/s before hitting the ground, but on a really gentle angle of attack.  Used RCS instead of the de-orbit engines to drop the orbit from 130km.  And waited.  A lot.  Unfortunately, that sort of reentry has a lot of variance in where you're going to come down, and my landing site was a good 18km above reference.  I think it might have safely landed if it'd been a low point.  I still had fuel left.  I've also got chutes kinda working, and they helped a fair bit for that. 

The answer to landing on Mars does seem to be; everything you can find.  And then some.  Aerobraking, chutes, ballutes, rockets, and an anti-gravity super science would be nice too.

So, new part: Aerobraking shell piece (there's room for twelve, so two sets of 6x symmetry), https://www.dropbox.com/s/2gx57zp0qdrbnaa/AltShell.zip?dl=0

And, an updated config for RO that includes niceties like restartable engines and almost functioning parachutes; https://www.dropbox.com/s/5a6v121dcgysa7o/RealismConfigs.cfg?dl=0 (Overwrite the same file in the NAR_MEM/Config folder).

I can't update the mod just yet with these because I've started working on some 1.2 adjustments, like RCS noises and effects.  So my dev stuff is a mess at the moment.  :-)

 

[Nittany Tiger]

19 hours ago, MaxL_1023 said:

If you have a barometer, check the surface pressure. If MSL is set to the lowest point on the planet your pressure will be somewhat below 0.615 KPA everywhere on the surface. You can get close to 1 KPA in the lowest regions, maybe even a little more if KSP uses the real Martian atmosphere. 

I use MechJeb to check the pressure.  You can see the readout on my latest pics.  I'm getting over 700 Pa around Valles Marines, which is a really deep canyon, so I guess KSP RSS is doing the atmosphere properly.

11 hours ago, TiktaalikDreaming said:

I made a shell airbrake thing.

I managed to get velocity down to 50m/s before hitting the ground, but on a really gentle angle of attack.  Used RCS instead of the de-orbit engines to drop the orbit from 130km.  And waited.  A lot.  Unfortunately, that sort of reentry has a lot of variance in where you're going to come down, and my landing site was a good 18km above reference.  I think it might have safely landed if it'd been a low point.  I still had fuel left.  I've also got chutes kinda working, and they helped a fair bit for that. 

The answer to landing on Mars does seem to be; everything you can find.  And then some.  Aerobraking, chutes, ballutes, rockets, and an anti-gravity super science would be nice too.

So, new part: Aerobraking shell piece (there's room for twelve, so two sets of 6x symmetry), https://www.dropbox.com/s/2gx57zp0qdrbnaa/AltShell.zip?dl=0

And, an updated config for RO that includes niceties like restartable engines and almost functioning parachutes; https://www.dropbox.com/s/5a6v121dcgysa7o/RealismConfigs.cfg?dl=0 (Overwrite the same file in the NAR_MEM/Config folder).

I can't update the mod just yet with these because I've started working on some 1.2 adjustments, like RCS noises and effects.  So my dev stuff is a mess at the moment.  :-)

 

That is awesome.  I'm definitely putting that to use.  I really think the real MEM would end up doing stuff like this to slow down because of the weight and safety.  I might use chutes as well.  Anything to get the MEM to slow down with fuel to spare.  I really get the feeling Baxter's MEM was either designed differently than the NAR MEM or he took some artistic liberties with the descent.  Guess I'll footnote this in the final run, or maybe design a more book-accurate MEM afterwards though I feel that his decision to use the same engine for descent and ascent is impractical unless it had deep throttling capability.

I might still add fuel to the descent stage, but not as much as before partly to save weight and space on the MEM for a rover or two.

[MaxL_1023]

Yeah it looks like your landing site is still 4km high - if you can find a spot closer to "sea level" I bet you would have landed safely. I wonder if attaching a few retros (think small solid motors with high thrust, short burn) would let you kill off that last bit of velocity right before landing. 

[Nittany Tiger]

4 hours ago, MaxL_1023 said:

Yeah it looks like your landing site is still 4km high - if you can find a spot closer to "sea level" I bet you would have landed safely. I wonder if attaching a few retros (think small solid motors with high thrust, short burn) would let you kill off that last bit of velocity right before landing. 

Maybe.  I'll be doing more tests to see what I need.  I shouldn't need retros, but might use chutes.  I'll probably add in more fuel as well, but the aerobraking should allow me to not have to pack the descent stage with fuel and thus save some weight for a rover.

I have thought of what entry angle I should go in with.  It would be nice if the deorbit stage was liquid so I could pick an AOA from any orbit.  I've been trying to imagine how to get a descent profile that maximizes the time I spend in the atmosphere so I can get more drag, but it always seems at a certain critical velocity I start losing lift and have to use the descent engine before I pancake.

Edited September 17, 2016 by Nittany Tiger

[TiktaalikDreaming]

2 hours ago, Nittany Tiger said:

Maybe.  I'll be doing more tests to see what I need.  I shouldn't need retros, but might use chutes.  I'll probably add in more fuel as well, but the aerobraking should allow me to not have to pack the descent stage with fuel and thus save some weight for a rover.

I have thought of what entry angle I should go in with.  It would be nice if the deorbit stage was liquid so I could pick an AOA from any orbit.

I tried using a liquid deorbit stage and messed up the ullage.  Gotta remember the RCS for the descent stage can't thrust forward.

Also, the more I mess with this, the more it seems the key failing is the TWR of the descent stage/engine.  That rocket slows you down way too slowly.  I'm going to make sure I have my numbers right, in case it really should have more thrust.

And, well, there we go.  Should be 140,000 lb Thrust, which I make to be around the 621kN.  Not 256.  So, I'll be fixin that and seeing how it behaves next.

Edited September 17, 2016 by TiktaalikDreaming
checked

[Nittany Tiger]

1 minute ago, TiktaalikDreaming said:

I tried using a liquid deorbit stage and messed up the ullage.  Gotta remember the RCS for the descent stage can't thrust forward.

Also, the more I mess with this, the more it seems the key failing is the TWR of the descent stage/engine.  That rocket slows you down way too slowly.  I'm going to make sure I have my numbers right, in case it really should have more thrust.

Yeah, the low descent TWR has always been the biggest issue for me.  It means that I have to activate the descent stage at around 35 km - 40 km altitude at peak shock heating in order to even have a chance to prevent pancaking.  I go by time to impact on MechJeb to see when I should fire, and with the current engine, I end up having to fire at no less than 2 min 30 sec until impact.

I tried to look at the thrust myself from the Atomic Rockets page, but when I convert their kilograms of thrust into newtons, I get crazy numbers in the millions of newtons, and this was after converting it in two different ways (assuming the thrust was kg force and assuming it was N/(gravity)).  I'm not sure where Atomic Rockets gets their number, and I can't find any documentation on the NAR MEM that would allow me to figure out what units they were using and convert from there.

[TiktaalikDreaming]

2 minutes ago, Nittany Tiger said:

Yeah, the low descent TWR has always been the biggest issue for me.  It means that I have to activate the descent stage at around 35 km - 40 km altitude at peak shock heating in order to even have a chance to prevent pancaking.  I go by time to impact on MechJeb to see when I should fire, and with the current engine, I end up having to fire at no less than 2 min 30 sec until impact.

I tried to look at the thrust myself from the Atomic Rockets page, but when I convert their kilograms of thrust into newtons, I get crazy numbers in the millions of newtons, and this was after converting it in two different ways (assuming the thrust was kg force and assuming it was N/(gravity)).  I'm not sure where Atomic Rockets gets their number, and I can't find any documentation on the NAR MEM that would allow me to figure out what units they were using and convert from there.

Checking the Atomic Rockets page, that looks a lot like the newtons I worked out incorrectly tagged as kg.  Which would be a hell of a ride.

[Nittany Tiger]

1 hour ago, TiktaalikDreaming said:

Checking the Atomic Rockets page, that looks a lot like the newtons I worked out incorrectly tagged as kg.  Which would be a hell of a ride.

Yeah.  The thrust I calculated assuming their units were correct was 2000kN shy of an F-1 engine I think.  Imagine bolting an F-1 to the MEM and trying to land with that.

[MaxL_1023]

1 hour ago, Nittany Tiger said:

Yeah.  The thrust I calculated assuming their units were correct was 2000kN shy of an F-1 engine I think.  Imagine bolting an F-1 to the MEM and trying to land with that.

If you were doing "direct ascent" (One ship takes off, flies to mars, lands, then returns to earth with no docking) you would probably need one. 

I am running my own Mars mission using "stock" parts (procedural tanks, pods and RP-0 engines). I needed to put over 1000 tons into LEO for my mission profile (Mars orbiter with a separate lander docked with it). I did it in a single launch, but I needed something like the equivalent of 40 F1's to get it off the ground (in the form of 4 radial Pyrios boosters and a sustainer using 24 RD-171Ms (more powerful then the F1). My second stage had 18 SSMEs. Not really practical in real life, but in the best of Kerbal tradition. Jeb was flying it, after all. 

 

Edited September 18, 2016 by MaxL_1023

[TiktaalikDreaming]

"Landed"!

Updated RO Config https://www.dropbox.com/s/5a6v121dcgysa7o/RealismConfigs.cfg?dl=0

That was chutes (drogue ripped off in less than 1sec, Ballute lasted well) and upgraded to spec descent engine.  I suspect you can switch the chutes for the airbrakes with no great loss.  And as you can deploy the airbrakes much higher up, they may actually be better.

I was struggling a bit for time at the end, and forgot to switch SAS from retro to upwards.  But there was spare fuel and enough time to slow down.  I'm going to change the material for the drogue, and try that.  And retry using Airbrakes and no chutes.  But not right now, 'cos I need a break. 

[MaxL_1023]

At least you can right the ship by using a mining drill, some cables and maybe a hole under the nose.

 

Also, somehow the first 5 times I reloaded my Mars mission to continue it I somehow ended up on a solar escape trajectory somewhat inside of Jupiter's orbit. I think that Mars Kraken has migrated into KSP. 

Edited September 18, 2016 by MaxL_1023