Developing Duna - Part 9

Cosmic Changes

I haven't played with KSP for about six months now, mainly because of frustration with the game Physics changing so drastically during the first versions of KSP 1.0. But recently, a friend gave me a Mcrosoft Surface 2 Pro that he no longer wanted (thanks, John!), so while testing it out I installed KSP 1.0.5 (which I hadn't even bothered with during my hiatus). Wow... you can forget a lot about how to play KSP if you don't touch it for six months! Now what was the key shortcut for that thing again?

Playing KSP wasn't particularly impressive on the Surface, but this got me to finally install version 1.0.5 on my main computer (which is a Mac), and I was having fun again. After messing around a while, I decided to dig out the old Save game and test a few ships that I had been working on in 1.0.4, as well as let the 1.0 hammer hit my universe.

The first thing I wanted to do in version 1.0 was try to fly a plane (because the aerodynamics were supposed to change radically from version 0.90). So in 0.90 I built the C724, a 24-passenger airliner that I practiced with and could land reliably. Let's see how the landing went for Kerbin Airlines veterans Pilot Billbo Kerman and Co-pilot Bobo Kerman, bringing flight KAL 100 in for an early morning landing at KSC, as the 1.0 hammer hits:

Bobo: "Whoa! What hell was that?"
Billbo: "I don't know. It felt like a strong updraft. I've got it under control...but the trim feels odd. Are all the systems OK?"
Bobo: "Umm...Yeah. Everything's green. Wait...we are using fuel at a high rate."
Billbo: "Do we still have enough to make KSC?"
Bobo: "Sure. But it's strange. No indications of any leaks. The engines are just drawing more fuel."

Billbo: "KSC tower, this is KAL 100 on final approach to runway zero nine."
KSC Tower: "Roger, KAL 100."
Bobo: "Yeah...we used a lot of fuel on that last leg. Hey, you're a little high."
Billbo: "I know. Throttle is down...we just aren't dropping fast. Again...its feels like we're in an updraft."

Bobo: "Set her down, Billbo."
Billbo: "I can't. She just wants to float."
Bobo: "Billbo..."
Billbo: "Hold on. I'll force her down. Come on... Touchdown!"
Bobo: "OK! Whew. Uh...Billbo...brakes... Here comes the end of the runway!"
Billbo: "Brakes are on full...they just aren't grabbing!"
Bobo: "Billbo! Go around!"
Billbo: "Full throttle!"
Bobo: "Too late..."

Alas, KAL 100 ran off the end of the runway and crashed into Booster Bay. There were no survivors.

Not an auspicious start. Luckily, the problem with too much lift and weak brakes got sorted out in the first update (or maybe I just got used to the differences...I don't recall the details now). But this was just the start of a Very Bad Day on Kerbin. A couple trans-oceanic flights almost didn't make it down safely. And KAL flight 033 from South City did not make it, having to ditch into the sea not far from KSC when it ran out of fuel. Hotshot pilot Valentina Kerman took out a Mallard (one of my seaplane designs) to check out the crash site, but when she tried to land the Mallard near the crash site, her plane also crashed into the sea. seems that those nice radial air intakes that made such great floatation pontoons in 0.90 no longer worked for that in 1.0. Sorry, Valentina.

Things were also not going well in space...

Gene Kerman: "We're getting anomalous power telemetry from the comsats at Duna. All of them are showing main bus undervolt readings."
Wherner von Kerman: "They haven't all started tumbling again, have they?"
Gene Kerman: "No. Attitude control is steady."
Wherner von Kerman: "See if you can contact Thompbles at Duna Base."

Aaron Kerman: "Flight, we're getting even worse power readings from the comsats at Laythe."
Gene Kerman: "Can they still operate?"
Aaron Kerman: "Umm...yes. They always had extra capacity."
Gene Kerman: "OK...there's definitely something strange going on. Contact all of the offworld mission crews and tell them to stand down all activities and start doing systems checks."

*** The lights go out ***

Glynn Kerman: "What the hell was that?"
Griffin Kerman: "Power outage. OK...the emergency generators are kicking in."
Aaron Kerman: "We just got a call from Willow Point nuclear station. They were detecting increased radiation in their main cooling loop, so they had to shut down."
Glynn Kerman: "Umm...we also just got a telex from Rockomax. They had a test stand explosion a few minutes ago during a certification firing of the 48-7S Spark engine. They're looking into the reason, but they wanted us to know right away."
Wherner von Kerman: "Looks like I picked the wrong week to quit drinking."


It was several days later before our valiant explorers on Duna got more details:

Bob Kerman: "Bob Kerman calling Duna Base 2. Do you copy?"
Kurt Kerman: "Roger, General Kerman. What's the news?"
Bob: "OK, here's what we know so far. To put it simply: The laws of Physics appear to have changed."
Hellou Kerman: "The laws of physics have CHANGED? That can't be."
Bob: "I know, I know. But that's what all the lab boys and girls are telling us, and they are verifying it with experiments. It's not everything...there are big effects on aerodynamics, chemical reaction rates..."
Hellou: "Chemistry? Then why haven't we all dropped dead?"
Bob: "Not all of chemistry. Only noticeable for certain reactions at high temperatures...but it's a problem for the NERV engines. We aren't going to be able to run oxidizer through them anymore. Even the commercial power plants had to shut down until they can make enough coolant that is free of contaminants that are now causing corrosion problems."
Kurt: "So what does that do to the nuclear tugs that are supposed to be our ride home?"
Bob: "The engineers are already modeling advanced versions of the tugs that carry only liquid fuel. For the older tugs, we will have to dump the oxidizer and take the big delta-V hit. Plus, with only the liquid fuel, tests are showing severe overheating problems can occur. We are working on control software changes...but the old tugs may just have to be used at less than full throttle. Even worse...the changes in the physics may not have settled down yet. We should be able to send out improved tugs before you need to return home, no problem."

Kurt: "So what else is going to affect us immediately?"
Bob: "Part of the changes in aerodynamics stems from the behavior of turbulence completely changing...not that we understood turbulence well to begin with. But the changes in turbulent mixing physics has a big effect on rocket motor injector design. The 48-7S Spark engine, in particular, has taken a big Isp hit."
Kurt: "That's one of our most important engines."
Bob: "I know...and we are uploading control software changes so they won't explode on you...but they will be much less effective. Your Duna Fidos, for example, are NOT going to be able to hop back to orbit...or very far at all. Basically, they're only going to be useable as surface rovers wherever you land them. Also, they will have to rest more often while roving because all solar panels are now less effective."
Kurt: "What's the problem with the solar panels?"
Bob: "Well, you know how solar panles work, right?"
Hellou: "They generate electricity by both the Photovoltaic and Energy Harvester effects."
Bob: "Right. Except the Harvester effect has vanished. The physics is just gone."
Hellou: "Gone? But...then the panel energy output will drop off with the square of the distance from the Sun..."
Bob: "Yes. And if you think that's bad for you, think about how it's going to affect the guys on Laythe. But the upshot is that your DunaDog planes are effectively useless. But the engineers think they might be able to send you fuel-cell-powered versions."

Nelemy: "So what CAN we still do, Dude? Or are we on extended vacation? Not that I would mind."
Bob: "As the engineers get a chance to test all the equipment, we'll let you know what can and can't be used, and what the new limitations are. Your EVA suits appear to be fine, and the surface driving features of the rovers appear to work, so you'll probably be able to continue your surface exploration plans. We think your Duna Base ships can still reach orbit, so you aren't trapped there. But for a while longer...yes, you are effectively on vacation."
Hellou: "Arrgh. OK...vacation it is. I guess I might try changing my hairstyle to that new topknot ponytail look that's all the rage on Kerbin now."
Nelemy: "Aww, Hellou-Dude, I like your hair the way it is!"
Hellou: "Yeah...but having to dye it all the time is a big pain. And Emilynn says she's thinking of going with the new look, too."

Bob: "In the long run, we are going to have to unstack all of the new equipment that we were going to send to you guys and make a lot of modifications. We may have to skip a launch window to get all of it to you. Plus, the guys on Laythe are going to need redesigned equipment, too. Our engineers are notoriously fast, but things are going to take time. The first thing we need to do is make sure our basic infrastructure of crew-launch and heavy-lift vehicles work."


Crew Craft Changes

Indeed, the changes in version 1.0 made my previous SSTO Crew Carrier rocket designs unworkable, mainly because of the degrading of the RAPIER engines. And when I made a new version that worked, later fractional updates killed those (again, because of further degrading of the jet-engine performance of the RAPIERs).

BUT...with the simple application of MOAR BOOSTERS (or, more RAPIERs in this case), and a cutting of mass and fine balancing of fuel loads, I present to you: SSTO Crew Carrier 9e:

The Crew Carrier 9e can carry four passengers to Kerbin orbit with delta-V to spare for orbital operations. Also, the craft is fully reusable and can return its crew safely to Kerbin. Note that the propellant levels in some of the tanks must be changed as shown above. The Mk2 Rocket Fuel Fuselage needs to contain only liquid fuel (the forward tank is full of fuel and oxidizer), and the six FL-T400 tanks need to have oxidizer reduced to 154 units each. Without these changes, the ship boosts too slowly and is too aerodynamically unstable.

The Crew Carrier lifts off with all six RAPIERs in air-breathing mode at 100% throttle (see below). I fly this ascent manually without MechJeb (although I assume MechJeb's Ascent Guidance parameters could be tweaked to give the needed flight profile).

The ship needs to pitch down fairly quickly. Below we see it pitched to 45 degrees at about 3000 meters, and its on its way to a pitch angle of about 20 degrees above horizontal -- the pitch is held about there for most of the ascent. When doing the pitch down, don't stray too far from the prograde marker on the NavBall, especially during the transonic region between 300-400 m/s, or the ship could flip.

Things get hot fast with six RAPIERS shoving the ship along on this flat trajectory, but nothing gets hot enough to explode at standard 1.0.5 settings...and you need the flat trajectory to give the RAPIERS longer time to burn its fuel using that sweet, sweet oxygen from the atmosphere. A somewhat-more-vertical ascent can work, but it isn't as efficient and you'll have less delta-V remaining in orbit.

The RAPIERs will begin to flag at around 24,000 meters, but just watch the speed indicator to see when it stops increasing. I usually get to around 1,580 m/s before the speed tops out, and then I press the 1 key to switch the RAPIERS over to closed-cycle rocket power mode (using an Action that I setup).

When the apoapsis hits 90 km, I cut the engines and let the rocket coast. Kerbal Engineer Redux or MechJeb displays are great for monitoring the apoapsis value. Here we are, two minutes into the flight, and all the hard work is over. This is why I like this SSTO rocket design better than flying a spaceplane; it's so much less tedious to fly.

As the red-hot meteor of a rocket coasts toward space, the apoapsis will drop to around 85 kilometers. Keep pointing prograde to minimize drag.

The circularization burn (below) took about 267 m/s of delta-V.

Four and a half minutes after liftoff, the ship and its crew of four are in an 85-kilometer circular orbit. The delta-V readout below shows that the Crew Carrier has 634 m/s of delta-V remaining for future maneuvers. I typically get a remaining delta-V value between 550 and 650 m/s, depending on how well I fly the ascent profile.

The retro burn to return the Kerbin is shown below. Alas, I miss the days when MechJeb could reliably target a pinpoint landing.

The reentry is hot, but with standard settings all the parts can survive.

For this flight, there was plenty of delta-V remaining that the Crew Carrier could make a rocket-powered landing (which MechJeb is doing for me below). The ship has no landing legs to save weight, but it sits just fine on the RAPIERs.

It's much easier to manually land if you use the parachutes (there are six of them). The RAPIERs can be used to make the landing as gentle as you want.

Note that a water landing is not advised. The ship will land safely, and if you activate the SAS to hold its attitude, it sometimes remains standing vertically in the water. However, it usually falls over, resulting in breaking some of the parts off that are ahead of the front tank (but the crew cabin always survives intact), which goes against the desired reusability, of course.

With the Crew Carrier 9e operational, my kerbals are guaranteed reliable access to Kerbin orbit.



But what of my heavy-lift capacity? In version 0.90, I used the Big Advanced Reusable Rocket to lift my big payloads, and I had taken to using the standard arcing ascent trajectory in anticipation of version 1.0's new aerodynamics that would work much better with that kind of trajectory instead of the old "straight-up-until-booster-separation" trajectory that I used with my original Reusable Rocket vehicles in order to drop the reusable boosters back onto KSC property (in theory). But I was not satisfied with the need to drop my reusable boosters into nasty salt water, so my new design is shown below.

A collaboration of the Kerbodyne Launch Consortium and the Rockomax Launch Alliance...I present to you: FUBARR!

FUBARR stands for Flyback Upgraded Booster Advanced Reusable Rocket. It can place 162 tons of payload into a 90 kilometer orbit, and all parts of the rocket are reusable ( theory...if KSP allowed flying multiple parts of a rocket...and I had enough of me to fly all the parts).

FUBARR lifts off on the power of its KS-25x4 "Mammoth" Liquid Fuel Engine in the core, plus four Rockomax "Mainsail" Liquid Engines and eight J-X4 "Whiplash" Turbo Ramjet Engines in the four side boosters. The jet engines are started first and allowed to spool up to speed before the other engines are ignited.

Below: A rear view of the mighty launch vehicle as it ascends.

At over 16,000 meters, the boosters burnout and are separated by Sepatron motors, and the sustainer presses on to orbit. (More about the boosters later.).

Below: The sustainer pops off its front aerodynamic cone and continues to orbit. The payload for this test is a large orbital refueling payload (not meant to be reusable... just there as a reference weight for now).

Upon reacing its 90 km orbit, the sustainer delivers the 162-ton payload full of tasty propellants to feed my hungry orbital projects.

To recover the sustainer from orbit, it needs to have some fuel left. After boosting the 162-ton payload into orbit, the sustainer still had over 350 units of fuel. This is not enough to allow a rocket-powered landing (which requires over 800 units of fuel remaining), but the sustainer is also equipped with parachutes that can help land it safely with less fuel. Below is the retro burn to return the sustainer.

The sustainer has 16 drag brakes to help slow and stabilize it during reentry. The sustainer must NOT have too much excess fuel onboard (more than about a 1000 units), because that would make the sustainer too heavy and it wouldn't slow down rapidly enough to prevent overheating. (So if the sustainer ends up in orbit with too much fuel onboard, the excess would need to be "vented" using a mod or manual adjustment of the sfs file...or you could do orbital maneuvers to burn off the excess.

The six parachutes are insufficient to slow the sustainer down to a safe landing speed, so a little rocket braking was needed to land it safely. The sustainer can land safely in water (it is very bottom heavy, so it will remain vertical in the water), but, again, I would like to avoid dropping it into salt water.

OK... But what about the side boosters? With KSP as it currently is, it's not possible to follow the boosters and control them all manually for landing, but they are designed to flyback and land at KSC under the control of the guidance computers (i.e., me). To see if it was feasible, I did a series of launches (all using MechJeb to guide the rocket on the same ascent trajectory) and attempted to land each booster back at KSC.

At booster separation, I used the bracket keys to select one of the boosters.

The first step is turn on SAS to keep the booster stable, and then quickly right-click on the Remote Guidance Unit of the booster to select "Control From Here." By default after separation, the control point seems to be the staging separator, and trying to fly the booster with that as the control point just leads to nasty tumbling crashes. Once the forward Remote Guidance Unit is selected, it's safe to put the booster in auto-point-to-prograde hold. If you let the booster start tumbling, it makes performing the remaining steps difficult.

The booster has two FL-T400 Fuel Tanks (containing no oxidizer) to which the jet engines are mounted. Fuel flow from these tanks was disabled in the VAB, so none of this fuel would be used during the launch. As the booster coasts upward, right-click on these tanks to activate the fuel flow (I make sure the throttle is off before doing this).

The booster will coast upward to over 28 km before reaching peak altitude. The booster has an Advanced Inline Stabilizer unit to allow it to be controlled in the thin atmosphere. At this point I roll the booster on its back and set the throttle to about 1/3 (the booster is currently too high for the jet engines to fire).

When the booster falls back into thicker air, I begin to pull "up" to complete a loop back toward KSC. The jet engines will eventually fire, and their thrust vectoring makes it easier to control the booster.

I continue pulling "up" to complete the loop to level flight (don't try to do it too fast, or you can cause the booster to tumble -- and don't worry...the control gets easier as you get below 10 km).

After a leisurely flight back to KSC (don't try any exciting maneuvering with the booster -- it's too easy to tumble it at high angles of attack), the automatic guidance control (i.e., me) lines it up on the runway for final approach.

And...touchdown! I allowed this booster to roll out to the far end of the runway and parked it on the left side.

The second booster (after a second identical launch, actually) I landed and parked along the right side of the runway.

The third booster (yet another launch in reality, of course) gave me a little trouble since it started tumbling up high. As you can see below, it took longer to get under control and was still going mostly straight down at 8,500 meters...but I was able to pull it out of the dive and fly it back to KSC for a fine landing.

The fourth booster gave me no problems, and I landed it and parked it by the spaceplane hanger.

The touchdown time for the four boosters varied from 7 to 14 minutes after launch, and I wan't particularly trying to vary their landing times with different throttle settings or slow S-curves, which could be used by the automatic guidance systems to space out their landing times as needed. If necessary, the boosters can land just fine in the grass area north of the runway.

OK! Heavy lifter confirmed! And, in the real world where all the rocket parts could be tracked and controlled separately, we would have 100% reusability of the launch vehicle. In practice, of course, I will only follow the sustainer during a normal launch, happy in the 'knowledge' that my boosters can make it back to KSC on their own and not get dunked into the sea.

I also tested the FUBARR with a 6.5-meter diameter aerodynamic payload fairing in place (since I'm building all of my new Duna equipment to fit inside that size fairing, just for giggles), and it worked fine.


Cockpit Calamity

While messing around on Duna, I quickly found a problem with version 1.0 and my old ships: The new Mark 1 Cockpit "upgrade" makes many of my old vehicles unusable. See the Duna Fido rover below:

The problem is the relocated entry hatch of the new cockpit. The kerbalnaut can no longer exit the hatch (because it is blocked in its new location). In this case, I was able to get Kurt out of the ship by having him "transfer" to the rear cabin module, and exit from there. But I have lots of rovers that use the Mk1 Cockpit (because of its high crash tolerance for safety), and some of them (the ones on Laythe, for example) don't have extra cabins to allow kerbalnaut entrance and egress. Bugger. Not only that, but all of my BirdDog planes and Laythe Spaceplanes and DunaDog planes have wings, canards, or intakes that have problems with the new hatch location. Double bugger.

What to do, what to do. I've heard people in the past say "If you don't like some new part, just add the old part back in so you can keep it, too." OK... I haven't liked the idea of retaining old parts before, but I can't see any other way around the problem. So I tried putting the old Mk1 Cockpit (from 1.0.4) into version 1.0.5 ...but I'm apparently not doing something right. (see below).

I changed all the names of Mark1 Cockpits in my old save file to Mark1a Cockpits, and tried to change all the references from 1 to 1a in the Part Config file that I brought over from 1.0.4 and renamed 1a (and brought over the old internal Space as well)... but I ended up with the all-white cockpit seen above. Also...Kerbals can go into the white cockpit, but their pictures don't show up down in the corner of the screen. So I guess I have more work to do. Or, if you are wise in the ways of saving old parts, maybe you can clue me in on what I'm doing wrong. Oh...and I see I have a flat tire that needs to be fixed, too (that probably happened when Kurt and Hellou drove down from the top of the Duna Pyramid).


Future Fido

The problem with the new Mk1 Cockpit inspired me to re-work the Duna Fido 3 rovers that I was going to send out with the next armada to Duna. These will be simpler rovers because they wouldn't have the "hop back to orbit" capability of the previous Duna Fido...they would be cheaper surface-driving-only rovers that could be dropped onto Duna at various locations as the explorers needed them. But the design I had previously made for the new rovers had the same hatch access problem with the Mk1 Cockpit (which really isn't suited for rover use in its new form).

And, since the new Mk1 Inline Cockpit now has an internal view (finally) and top canopy access, I redesigned the Duna Fido 3 to use that cockpit. It still has very good crash safety tolerance, and it has an even better IVA with a great view of the surrounding terrain. As long as I was at it, I removed the girder truss-work that formed much of the structure of my older Fidos (I think that truss work contributes to lag). I aso removed the solar panels and went solely with RTG power (all the best red planet rovers use RTG power!). Below we see the new Duna Fido (without its detachable landing hardware) being tested at the KSC. It still has room for a crew of two (who can take turns in the comfy rear cabin), and a Granny Klampett chair on top for better visibility of the landscape. Plus, lots of Science instruments.

If you are wondering about the unusual wheel arrangement, that is because a stack of these rovers is designed to fit inside a 6.5-meter fairing on top of a FUBARR. Yes, there is clipping and offset involved in the structure of the rover...but in real life engineers would cut the appropriate slots in those plates as needed. It is quite stable in driving tests, and I like the new cockpit. Hmm... maybe I should change all the old Mark 1 Cockpits in my save game to Mark 1 Inline Cockpits... No. That might work fine for rovers, but not for planes.

Ooooo...wait, wait, wait! A new inline crew cabin...fits two kerbals. But if I only use one kerbal in there at a time, I can use that as the mini-hab/lab for the new Fido. The advantage? I never did like the very low crash tolerance of the lander was even safer to have the second kerbal riding in the chair on top of the rover instead of being in the cabin. But the Mk1 Inline Crew Cabin has a great impact tolerance. Safety first, kiddies! Pardon me while I re-make a rover in the middle of writing this report....

...annnnd, back! Below is the new Duna Fido 3c for your enjoyment and edification:

Same basic structure as the other one, but safer for my precious kerbalnauts. When building a rover with this cabin, test which end has the main hatch (they look the same to me, but only one works when you click on the EVA button on the kerbal's image in the corner, otherwise you get "hatch is blocked". And stick a ladder under the hatch. Hmmm...I could leave off the front nose cone and use the hatch on the front of the Mk1 Inline Cockpit (with a ladder) to let the kerbal get in the cockpit from the surface...but on Duna, the kerbal can just fly up on top to the canopy entrance, or transfer from the rear hab...and I've gotten used to the look of the nose cone.

That wraps it up for now!