Opinions on my latest landers?

[Brainlord Mesomorph]

Hi guys,

I think I’m done designing ships for my Jool 5 Challenge mission. But I’d kind of like your opinions before I go.

First, I like to design things “science fiction style†(i.e. multipurpose, multi use, multi mission, 100% reusable) as opposed to “NASA/Apollo style†(single purpose, single use, single mission, 100% disposable). To that end, I designed this, the “Gamma 4†Single Engine Nuclear Lander.

Seen here carrying a TriSci module, on top, its surprisingly light, 1.75 tons, despite being larger than nine ton fuel tank. So it really doesn’t affect the center of gravity. And that is in its primary mission, to ferry these TriSci modules back and forth from the moons of Jool (or elsewhere) to the Orbital Science Lab. Fueled mass of 39.54 tons, Dry Mass: 13 tons, Max Delta V: 3917m/s

The segmented fuel tanks are a blessing and a curse. A blessing because they allow pinpoint control of how much we refuel the lander, and which tanks we’re consuming fuel from to keep the CoG low. A curse, because they make refueling hell. (Q: is there any way to group fuel tanks in a stock game?)

The small fuel tanks, docking ports, and other stuff of the very bottom, create a remarkably effective crumple zone. During a hard landing all of that can be destroyed, and it protects the exhaust cone of the nuclear engine , so the ship remains 100% flyable and able to complete its mission. And these Gamma 4’s are so light, and so cheap, I can afford to send the two of them.

But it turns out (thanks for the help, guys) that it doesn’t have a thrust to weight ratio capable of landing (or more importantly, returning from) Tylo.

And so, I had to go NASA/Apollo style and design of this liquid fueled, two stage lander especially to carry a TriSci module to Tylo.

I give you The Tylo Excursion Module:

(My god, it’s hideous!)

First I designed the Ascent Stage powered by 5 LV909s (four wasn’t enough and six was too many!) Its fueled mass is 22.2 tons, Dry Mass: 9.4 tons, Max Delta V: 3229 m/s. (with the triSci) Then I designed a decent stage to carry it using four Poodles.

Combined Fueled mass of 88.4 tons, Dry Mass (including Ascent Stage, fueled): 40 tons, Max Delta V (as you see it) 2921 m/s. But if you add a x200/16 drop tank, that we will shed halfway through the deorbit burn, that goes up to 3177m/s +. As for TWR: if I use both stages as you see them I can make orbit of Kerbin!

That just leaves Laythe, and for that I’d like to show you the “Hawk Command Moduleâ€Â

A tail landing, SSTO spaceplane that makes quick work of the KSC to LKO run. I think that means she should easily handle .8 gees on Laythe. Of course you also have 80% of the air, and 80% of the oxygen, so I’m not exactly sure how that’s going to work. But that’s one of a (exciting) known- unknowns of this mission.

Well, what you think?

Am I ready for Jool?

Edited November 4, 2014 by Brainlord Mesomorph

[Jouni]

The Tylo lander should work, but the landing would be easier with a shorter lander. The final moments of a Tylo landing aren't exactly the finest moments of a space program, especially if you're doing it for the first time. You basically crash the lander on the surface in a more or less controlled way, hoping that no critical components get destroyed, that the lander doesn't have too much horizontal speed to flip over, and that the thrust isn't too high for the lander to bounce and flip over.

The LV-909 isn't the best engine choice for the ascent stage, as its TWR is quite low. Six 48-7S engines should get the job done with less fuel. 1800 units of LFO should be enough for the ascent and a nice margin of error, or you can use 2000 units if you want to be on the safe side. You can also remove the decouplers used for attaching the struts between the descent stage and the ascent stage, as they're not necessary.

For the descent, you could just use four ascent stages attached radially to the ascent stage. To reduce the part count, you can use Aerospikes as nice low-profile replacements for the clusters of 48-7S engines.

Here is a quick proof-of-concept without the massless parts:

The payload is around 4.7 tonnes, the ascent stage has 3023 m/s of delta-v with 15.4 tonnes of mass, and the descent stage has 3263 m/s of delta-v with 62.8 tonnes of mass.

[Zylark]

Only thing I'd like to throw in, is to replace the stock ScienceJr and Goo thingies with Universal Storage mod equivalents. Will save you a ton of space, a bit of weight and will look much smarter.

[Brainlord Mesomorph]

Six 48-7S engines should get the job done with less fuel. 1800 units of LFO should be enough for the ascent and a nice margin of error, .

Thanks this is great, I never would have considered using clusters of tiny engines but now that I look they have a ridiculous TWR.

I’ve always had trouble trying to separate asparagus staged landers at liftoff, it suddenly occurs to me though I could separate immediately after liftoff (DUH!)

I love that its 20 tons lighter. (looks better too!)

You can also remove the decouplers used for attaching the struts between the descent stage and the ascent stage, as they're not necessary.

I’ve always been afraid struts were going to interfere with stage separation. But they do seem to sort of “disappear†How *do* they work?

Edited November 1, 2014 by Brainlord Mesomorph

[mhoram]

I’ve always been afraid struts were going to interfere with stage separation. But they do seem to sort of “disappear†How *do* they work?

The parts in a ship are organized in a treemodel:

- there is a single root part

- every other part has a parent part to which it is attached to

So it is impossible to create circular structures in the VAB.

As addition fuel lines and struts contain information about

- the part they are attached to (source) and

- the part they are linked to (target)

When a ship is split in two (via undocking or decoupling or other means) fuel lines and struts are cheched if their source and target part are on now different ships. In that case their connection is removed during the undocking or decoupling process.

There is a bug around that causes struts between stages to have an impact on decoupling forces (but I don't know the current status on this).

[Brainlord Mesomorph]

There is a bug around that causes struts between stages to have an impact on decoupling forces (but I don't know the current status on this).

I wouldn’t think that’s a bug, I would think that’s part of the physics model. It is precisely the reason I attach struts to redial decouplers. (When they go across stages)

[GoSlash27]

I understand what you're trying to do here, but these are seriously huge landers for such a little job!

Tylo is the most difficult job in the Jool system, so let me show you my lander arrangement for that just for comparison.

These 2 rovers hold enough fuel and have enough thrust to put themselves and the lander in orbit in a single stage.

Not saying you need to do it the way I do it, but it shows how little vehicle is actually required for this job. Smaller landers mean smaller mass movers and much, much smaller boosters.

Best,

-Slashy

[Starman4308]

I understand what you're trying to do here, but these are seriously huge landers for such a little job!

Tylo is the most difficult job in the Jool system, so let me show you my lander arrangement for that just for comparison.

I think this illustrates the maxim of "Think of what you need, and whether you can possibly design with less".

Your lander doesn't have the Science Jrs or goo pods, uses the 1-seat pod (the 2-seat pod is ludicrously overweight), dispenses with the large Clamp-O-Trons, and uses the lightweight legs instead of the max-size ones. That's 1.4t science payload shaved off*, 1.9t of command module shaved off, about 0.7t of lander legs shaved off, and 0.6t of Clamp-O-Tron shaved off, for a total of 4.6t less payload to the surface. There's some added back with the rovers, but if I had to guess, you just leave the rovers behind.

*My preference would be 2 of each, which would add back 0.7t. The third and fourth suffer really bad diminishing returns.

[GoSlash27]

I think this illustrates the maxim of "Think of what you need, and whether you can possibly design with less".

Your lander doesn't have the Science Jrs or goo pods, uses the 1-seat pod (the 2-seat pod is ludicrously overweight), dispenses with the large Clamp-O-Trons, and uses the lightweight legs instead of the max-size ones. That's 1.4t science payload shaved off*, 1.9t of command module shaved off, about 0.7t of lander legs shaved off, and 0.6t of Clamp-O-Tron shaved off, for a total of 4.6t less payload to the surface. There's some added back with the rovers, but if I had to guess, you just leave the rovers behind.

*My preference would be 2 of each, which would add back 0.7t. The third and fourth suffer really bad diminishing returns.

Starman,

I actually bring the rovers along; they *are* my boosters.

The idea here is that the rovers working in pairs can bring payloads to the surface, traverse terrain to a site, assemble whatever they brought in place, then refuel and take payload back to orbit with them.

It's a "surface rendezvous" scheme intended for constructing elaborate habitats on other bodies. A simple exploration lander/ lifter could be built lighter and simpler than what I've shown here. This is just presented to illustrate how small and light these systems can be if you're designing for it.

Best,

-Slashy

[technion]

These 2 rovers hold enough fuel and have enough thrust to put themselves and the lander in orbit in a single stage.

Not saying you need to do it the way I do it, but it shows how little vehicle is actually required for this job. Smaller landers mean smaller mass movers and much, much smaller boosters.

Best,

-Slashy

What am I missing about how that's meant to reorbit? The lander can appears to have no fuel tanks at all?

If the idea is to reattach to those docking ports, that would mean driving those rovers sideways?

[Starman4308]

What am I missing about how that's meant to reorbit? The lander can appears to have no fuel tanks at all?

If the idea is to reattach to those docking ports, that would mean driving those rovers sideways?

I presume he drives them backwards to just alongside the pod, and then uses those landing struts to jack up the rovers until they magnetize and attach. He might lower the legs on the pod to make it easier.

[GoSlash27]

I presume he drives them backwards to just alongside the pod, and then uses those landing struts to jack up the rovers until they magnetize and attach. He might lower the legs on the pod to make it easier.

Pretty much this. The rovers and lander pod (or whatever other payload pod they're working with) are designed to attach when parked next to each other and jacked up on the landing gear. The payload doesn't necessarily have to be a landing pod as shown here, but can be (for example) habitat modules, power stations, fuel cans, etc.

The idea is that you're not going to drop payloads from orbit down to the millimeter where they're supposed to be, so you need to 1) land them on the surface, 2) traverse them across broken terrain to the build site (possibly fording/ traversing hazards), and 3) assemble them together in place once you've arrived. I designed the rovers to incorporate all these functions and more into a single assembly to ease the logistical process of building outposts off- planet.

But this is all getting away from the point I'm trying to illustrate; a lander doesn't have to be very big to work even on Tylo. If you design the payload small, then every stage that precedes it will be smaller. It's an exponential progression backwards down the chain.

Best,

-Slashy

Edited November 2, 2014 by GoSlash27

[Jouni]

But this is all getting away from the point I'm trying to illustrate; a lander doesn't have to be very big to work even on Tylo. If you design the payload small, then every stage that precedes it will be smaller. It's an exponential progression backwards down the chain.

The progression is linear, not exponential. While the required mass ratio depends exponentially on the delta-v requirements, it's independent of payload size. In practice, the progression is slightly sublinear, because the overhead from probe cores, docking ports, decouplers, and similar stuff is more significant with smaller payloads.

[GoSlash27]

The progression is linear, not exponential. While the required mass ratio depends exponentially on the delta-v requirements, it's independent of payload size. In practice, the progression is slightly sublinear, because the overhead from probe cores, docking ports, decouplers, and similar stuff is more significant with smaller payloads.

Jouni,

It is, indeed exponential (though I can understand why you'd think it wasn't).

Doubling the payload of a stage will double the mass of the stage to generate the same DV. And while I'm saying here what you just said, this is an exponential progression back down the ship.

Grossly simplifying the math to illustrate... You have a 1 ton package to deliver to orbit from the surface of a moon. That means your ascent stage is 2 tons, your descent stage is 4 tons, your transfer stage is 8 tons, your injection stage is 16 tons, your transstage is 32 tons, and your booster sitting on the pad is 64 tons.

If you double the mass of that payload, then you've doubled your launch vehicle to 128 tons. So 1 kg increase in the mass of your payload has cost you 128 kg on the pad.

Keeping things small, light, and efficient pays huge dividends back up the chain.

Best,

-Slashy

[Kryxal]

No, I think linear is the right term ... you add x kg to the payload, you add some constant times x kg to the vehicle, keeping all other terms the same. It's delta-v where it gets exponential ... you add x m/s, you multiply the mass by e to the power of some constant (probably on the order of .001) times x.

There's no question it's worth keeping the payload mass down, though.

[Starman4308]

No, I think linear is the right term ... you add x kg to the payload, you add some constant times x kg to the vehicle, keeping all other terms the same. It's delta-v where it gets exponential ... you add x m/s, you multiply the mass by e to the power of some constant (probably on the order of .001) times x.

There's no question it's worth keeping the payload mass down, though.

It's all a matter of perspective. It's linear with respect to overall rocket mass (double the payload, double the rocket), but the additional rocket mass is exponential with regards to payload (additional 1 kg payload means an exponential amount of booster, at the same ratio as the rest of the payload).

[GoSlash27]

It's gonna work out to roughly 2^(N-1) X 2Mp where N is the number of staging events and Mp is the mass of your payload. If you look at the 2Mp part, then yeah it's linear. But the 2^(N-1) part makes it exponential.

The upshot is that saving a little mass at the pointy end saves a lot of mass at the hot end, so it's in your best interest to design each stage to be as light and efficient as you can make it.

Best,

-Slashy

[Jouni]

It's gonna work out to roughly 2^(N-1) X 2Mp where N is the number of staging events and Mp is the mass of your payload. If you look at the 2Mp part, then yeah it's linear. But the 2^(N-1) part makes it exponential.

The 2^(N-1) part is a constant here, because we're scaling up the payload, not the delta-v requirements. If you need f(x) tonnes of launch mass for x tonnes of payload, linear growth means that f(x) = xf(1), while exponential growth means that f(x) = c^x f(1), for a constant c > 1.

[Dispatcher]

I did an experiment (no pics or craft files, sorry) where my first rocket payload (inc. separator) was exactly 1 ton and the portion below the separator was a single LF stage. Total dV was about 4500. Total tonnage (inc. payload) was just over 92 tons. I did a second craft where the payload was exactly 2 tons. Using a similar approach (single stage and same engine and same dV), the total tonnage (inc. payload) was just under 114 tons. Thus this design change from a 1 to 2 ton payload required a total difference of about 22 tons. You can work out the ratios for these, but for me the observational approach is a lot of fun.

Results will differ more and more with multiple stages, engines and engine types, as well as final payload. I do use KER, which allows me to more rapidly develop and test various configurations.

[Starman4308]

I did an experiment (no pics or craft files, sorry) where my first rocket payload (inc. separator) was exactly 1 ton and the portion below the separator was a single LF stage. Total dV was about 4500. Total tonnage (inc. payload) was just over 92 tons. I did a second craft where the payload was exactly 2 tons. Using a similar approach (single stage and same engine and same dV), the total tonnage (inc. payload) was just under 114 tons. Thus this design change from a 1 to 2 ton payload required a total difference of about 22 tons. You can work out the ratios for these, but for me the observational approach is a lot of fun.

Results will differ more and more with multiple stages, engines and engine types, as well as final payload. I do use KER, which allows me to more rapidly develop and test various configurations.

Your payload wasn't 1-2 tons. It was 1-2 tons plus the mass of every empty fuel tank and engine. My guess is that, while you added fuel for the 2t payload, you did not add engines. Depending on what the TWR was, you either had an inefficient 1t launcher (too much engine for the payload), an inefficient 2t launcher (not enough engine for the payload), or both (poor TWR for both).

EDIT: Really, this is important: engines and empty fuel tanks are dead weight for the delta-V calculation, so you do your best to minimize them.

[GoSlash27]

I did an experiment (no pics or craft files, sorry) where my first rocket payload (inc. separator) was exactly 1 ton and the portion below the separator was a single LF stage. Total dV was about 4500. Total tonnage (inc. payload) was just over 92 tons. I did a second craft where the payload was exactly 2 tons. Using a similar approach (single stage and same engine and same dV), the total tonnage (inc. payload) was just under 114 tons. Thus this design change from a 1 to 2 ton payload required a total difference of about 22 tons. You can work out the ratios for these, but for me the observational approach is a lot of fun.

Results will differ more and more with multiple stages, engines and engine types, as well as final payload. I do use KER, which allows me to more rapidly develop and test various configurations.

I'm thinkin' your booster wasn't optimized to your payload, so the effect of doubling the payload was heightened.

It doesn't take a 92 ton vehicle to put 1 ton in orbit, even SSTO. For example, you could do that job with a single aerospike and the total vehicle mass would be about 11.3 tons. Building the same lifter for a 2 ton payload would be about 16 tons. In this case the 1 ton increase in payload created a 4 1/2 ton increase in total mass. And that's just one stage. The more stages you use, the more the effect is exaggerated.

Best,

-Slashy

Edited November 2, 2014 by GoSlash27

[Dispatcher]

I'm thinkin' your booster wasn't optimized to your payload, so the effect of doubling the payload was heightened.

It doesn't take a 92 ton vehicle to put 1 ton in orbit, even SSTO. For example, you could do that job with a single aerospike and the total vehicle mass would be about 11.3 tons. Building the same lifter for a 2 ton payload would be about 16 tons. In this case the 1 ton increase in payload created a 4 1/2 ton increase in total mass. And that's just one stage. The more stages you use, the more the effect is exaggerated.

Best,

-Slashy

You are absolutely right about it not being optimized. If I had wanted to do that, I'd have used my Engine Chart or Graphs (see sig line). Instead I simply chose the largest single engine and the fuel tank(s) which yielded the right dV typically used for achieving stable orbit. It was simply a matter of being consistent with my engine choice. Each stage and/ or engine choice will yield varied differences.

Edit: I agree with your numbers for the aerospike. As for landers, the same kinds of stats can be useful during their design.

Edited November 3, 2014 by Dispatcher

[Brainlord Mesomorph]

Hi guys I’m back,

While you guys are involved in a (fascinating) conversation about payload capacity and rocket design scalability, I went off and built the Tylo Lander were discussing. I just love these little thrusters; they are my new favorite engine.

That looks much better.

But, I said it before, and I’ll say it again, radial decouplers on landers are problematic. If the outer tanks are out of fuel, or those engines are damaged during landing, and you try to fire the radial decouplers on the ground, the outer stages just fall back inward, and pin the ascent stage to the ground. So instead of the 4 FLT200s I use one X200/8 on the bottom and attached the outer stages directly to it, turning it into real LEM-like 2 stage lander. It still has plenty of thrust for descent, and no matter how badly the descent stage is beaten up on landing, the ascent stage still separates cleanly, And it gives me a room for a another Clamp-o-tron Sr (yay!)

(Here’s a question: with no payload, and parachutes for descent, could do this thing be pressed into duty as a Laythe Rescue lander?)

About the large lander can and the Clamp-o-tron Sr:

The Clamp-o-tron Sr, is, quite literally, the backbone of my interplanetary fleet. The two man lander can does weigh a ton more, but it’s substantially more … well… substantial. And now that I’m designing everything basically as a cylinder, with at least one Clamp-o-tron Sr and preferably two, the large lander can is simply a design requirement.

Now about this guy:

I gave you the wrong stats. It’s not 40t with a max delta V of 3200, its 21t with max delta V of over 5300! I was trying to design a Tylo Lander before I noticed the TWR problem. (and I tend to over-design)

So with the aid of a drop tank, it’s basically an interplanetary automobile. Just get in and go. I told you I like to do things science fiction style. SSTB (Single Stage to Bop) this drastically effects the makeup of my “Joolian Exploritory Fleetâ€Â. Is going to be that - a fleet!

Here’s what I’m doing:

I’ve never gotten a guy as far as Jool before, so, nearing the end of a career game (I want to start over in the 0.25), I decided to spend every last dollar and my space budget on a truly massive expedition to Jool. By the time current contracts are complete, I should have about four million dollars – er, “funds.â€Â

I’m sending eight, or 10, or 12, ships, whatever I can afford. There are 100 days between Joolian arrival and our Kerbin Return departure window. So that’s 100 days of operations in the Joolian system.

Multiple landings on every moon (maybe not Tylo) Not just flags and footprints, but orbiters and rovers, too. Plus 2 orbital science labs. We’re hauling 50 tons of science. I’ve designed 150 ton fuel tankers. And a MK2 cargo bay- orbital payload exchange system!

Basically, this is about the biggest thing I can do in a stock game. You can’t really do colonization or anything like that without mods. After I decided to do this, I realized you guys call it the Jool 5 challenge. But I’m not just doing the Jool 5 challenge.

I’m planning on blogging the whole thing.

My next game - I’m going to wrap up with a one ship Grand Tour mission.

Edited November 3, 2014 by Brainlord Mesomorph

[Farex]

I haven't read everything yet, but there are quite some interesting designs. Keep it going

€dit after reading:

Maybe you should try out the Skipper engine for your lander. You could cut the thrust to 50% and have a very efficient vehicle. Or you go Full-Scale and use the Kerbodyne KR-2L Advanced Engine with a Kerbodyne S3-3600 Tank that has per radial decoupler 4 FL-T800 Fuel Tanks for the landing sequence. I don't know if that would work or not, but I will try this out for myself, thanks to your lander-idea.

Edited November 3, 2014 by Farex

[KerikBalm]

* that lander can is very heavy, its better to use 2 small cans

* You could radially attach the science stuff, and save of adaptor weight and extra docking port weight.

* Why does it have parachutes? its not landing on Laythe.

* Why not use a smaller, lighter probe core?

* Why not use a smaller size docking port?

* Why not use a LV-N? it will use less fuel (since this is meant to be refueled and reused)

* Do you really need all that RCS fuel?

* Do you really need all those landing legs, do they really have to be the largest size ones?

* Why are there docking ports under your lander?

* Do you really need all those solar panels?

I see a lot of weight that could be trimmed.