My personal challenge: Offer advise if you like

[togfox]

I\'ve been to the Mun a dozen times. I\'ve got countless satelites around both celestial bodies. I\'ve created janitors and I\'ve docked several times. I\'ve orbited the sun. I have a mun base. I\'ve done just about everything once - except:

Get a mun rover to the mun, deploy it safely and return the Kerbalnaughts to Kerbin safely, starting from scratch and NEVER losing any life during the entire R&D, testing and mission execution stages.

In short, I want to design and test, over several test flights, a rocket that will deploy a rover and return to Kerbin. I need to do this safely so that no Kerbalnaught is sacrificed during this space program. That means slow and steady, building the rocket up over several test flights and building in as many safety features as possible to preserve my pilots.

Restrictions

Stock parts only (except for the CART)

Add-ons

The auto-ascent plugin (because I\'m lazy)

Mech-Jeb (because I\'m lazy and the rover will need it)

The dodADA for assisted landing (because throttle control during landing requires more than a mere mortal can provide)

The emergency ejection decoupler (forget its name) as this is the only insurance policy my Kerbalnaughts will have!!

I\'ll post craft designs prior to launch so feel free to give advice on my more silly design decisions. This will take multiple launches and if you see me about to make a catastrophic mistake then please warn me in advance!

Project plan

a) Design a \'chasis\' that can carry a rover and deliver it to the surface safely. No flight. Launch pad tests only.

Design a delivery system that can lift the rover. Low altitude test flight (30,000 m alt)

c) Design a delivery system that can reach orbit and re-enter (100,000 m alt)

d) Design a delivery system that can orbit the Mun and return

e) Design a delivery system that can land on the Mun, deploy the rover and return to Kerbin for re-entry.

Step b is a test flight but there might be multiple test flights executed before the objective is achieved.

That\'s my mission - rover to Mun and back, starting from scratch, with ZERO casualties. I\'ll post screenies and feel free to comment. Please do NOT post your craft file. I want to do this from scratch using the parts I outlined above. Offer tips. Help. Suggestions. But please no spoilers. Thx.

Program progress

Test 1:

Design of Rover/return craft to deploy rover and return to Kerbin - launch pad test - fail. Rover is obstructed by return craft.

[Kosmo-not]

My advice is to design 'backwards.'

Basically, design the return stage first and end with the lifter stage. This is how I design all my Mun mission craft. Everything except for escaping the atmosphere is easy to calculate as far as delta-V budget. So, I can design a craft that will get me from Kerbin orbit to the Mun and back without having to test it. I then only have to spend time figuring out a good way to get it to orbit.

I hope that piece of advice helps.

[togfox]

Thanks Kosmo. I agree - it is good to do it that way but the \'lander/returner\' is often the LAST thing that gets exercised yet the first thing to be designed. A conundrum. I\'m not sure how to approach this with zero casalties. Perhaps a lander that can reach a few thousand meters and land on Kerbin as a start? If so then my project plan above will need to be amended.

[Kosmo-not]

The zero casualties thing usually depends on one\'s piloting skills.

What is your method of landing on the Mun? Direct descent or low periapsis?

[togfox]

I have tried both in the past and I\'ve found a low PE provides the highest success rate and less fuel.

As to piloting skill - noted - but it is also caused by missing strut or an absense of SAS (ie poor design).

[Inquisitor]

Quickly, [i\'ll have more commentary tomorrow...] You should already know that the two most lethal and dangerous parts are the assent stage where you have the greatest stresses, and load; And the decent stage to the mun. with the mech-jeb this should introduce some safety and reduce pilot error. Surviving both in a bad situation mean being able to safety extract your self form from the current failing stage. The most difficult part in building is merging a heavy lift rocket with a viable lander. To solve many of these problems means having enough control over your rocket, and this often comes down to RCS, you probably want a little extra here because this will allow you to have more control over your craft >> and this is how you survive the assent.

speaking of control i would not use solid boosters until unless you need just a little more thrust at the beginning. the reason being that you cant shut them down, and detaching lit solids often smash into your ship. It is this lack of control that would prevent you from safety extracting to a higher stage.

Question: have you considered the possibility of a rescue mission of a stranded capsule? Also when nasa was developing their rockets, many were unmanned, have you considered this. Finally have you considered using a radial parachute mod to give you a little more safety?

[Kosmo-not]

In my opinion, a good ascent stage should not require any RCS at all. Admittedly, I\'ve had to use RCS during ascents when I was too heavy on the controls. You don\'t have to pitch over quickly to have an effective ascent.

[Inquisitor]

In my opinion, a good ascent stage should not require any RCS at all. Admittedly, I\'ve had to use RCS during ascents when I was too heavy on the controls. You don\'t have to pitch over quickly to have an effective ascent.

Cannot agree more, [in-fact my finished rockets don\'t need rcs, they just use a mix of gimbaling and non gimbaling engines to get stability] Having said that though, togfox is going for max safety, so he should go for more control than what is normally required.

[Kosmo-not]

Having a high safety factor means more vehicle mass, and thus more fuel and thrust to get it to orbit. Of course, we need some factor of safety to ensure survival.

Want to know the factor of safety on a commercial airliner jet? You don\'t want to know, lol. A high safety factor would mean it wouldn\'t be able to fly effectively.

[Inquisitor]

Having a high safety factor means more vehicle mass, and thus more fuel and thrust to get it to orbit. Of course, we need some factor of safety to ensure survival.

Want to know the factor of safety on a commercial airliner jet? You don\'t want to know, lol. A high safety factor would mean it wouldn\'t be able to fly effectively.

I dont understand how mass is related to safety? Could you explain, is this like the difference between a freight train and a car [only one gets totaled ???]?

[togfox]

Solids make me nervous for the reasons Inquistor points out. This is where the escape decoupler comes in and I don\'t intend on using solids till stages C or D in my plan (see OP).

I typically don\'t use RCS for Mun missions as I don\'t need them for launch and I don\'t need them to land on the Mun. The autopilots for both exercises will be ample. I usually don\'t have SAS either but I do use symmetry and SOME gimbalded engines. I\'ll use wings/fins if necessary but I\'m not sure if that is a sound (or efficient) approach. I can not afford an unstable launch 1000m from the pad with no time to counteract or eject. If RCS can help in this area then please advise!

Integrating the lander with my lifting stage will be a challenge I suspect and I fear this is where the greatest stress will be.

Rescue mission: I had not considered that but I guess any space program needs a back up plan! The rescue mission would be, I\'m guessing would not need to carry a rover so that should be a lot simplier. That would be a standard land-on-the-mun and return mission. I\'ve done this a million times but I\'ll put it on my plan.

Kosmo - Kerbals are light on safety equipment anyway so I\'m not too worried. ;P

Edit: I just considered my early test flights can\'t use the auto-ascent autopilot as I\'d need full control during these flights. Maybe RCS will be a good idea!!

[Kosmo-not]

Mass is related to safety as in more fuel for RCS (or duel for RCS at all), and extra fuel for human error. That sort of thing. In real world, it would also relate to strength of materials, and using heavier materials to withstand heavier loading.

@ togfox

I thought you were worried about safety, lol.

BTW, it\'s always good to know how to fly a good manual ascent. It provides a great backup if the autopilot fails.

[togfox]

Here is my concept for the rover delivery module. I haven\'t tried this yet but it is essentially the rover in the middle straddled by the Mun descent engines. Two full LF tanks with the \'half\' engines should be enough for a gentle descent from a PE of 5000m or so. the rover will then disengage and drop onto the ground. I\'m hoping there will be enough fuel for the return to Kerbin. I\'m not sure if there is a safe way to test this is sufficient fuel.

Edited August 10, 2012 by togfox

[Kosmo-not]

I don\'t know how much mass you\'re dealing with there. Is the descent stage also used as the return stage?

[togfox]

That\'s the plan. It need not be though. I have experience in leaving lander modules on the surface and sending just a capsule, LFT and LFE home. I can do that also but that means the weight of an extra engine and coupler. Though, admittedly, it is way cooler to leave the lander on the surface. It can be my rover carport during solar storms. 8)

[Inquisitor]

I was thinking: perhaps the way to test safely is to re-stage your rocket... So you want to test stage 1 with solids attached as well, try this: instead of stage 2 being the decoupling of the solids and stage 3 being the lighting & decoupling of stage 1. Instead keep everything [i.e keep the same rocket] but make stage 2 activate a single solid booster that decouples from the rest of the rocket with just another decoupler + capsule + parachute. This is an intended emergency extraction, except you can get out at any time of you want. So the end of each test flight is this extraction from the rocket [be things going well or not]. Importantly you would be flying a multi stage rocket with only two functional stages. Therefore if you are testing solids, the next stage is not to decouple them, [you want to keep them attached] the very next stage is the single solid & capsule ejection. So wonce stage 1 is deemed safe you can then include the normal stage 2+3, [assuming you are using solids, if not then just stage 2] and then move you capsule ejection stage to right after that. Systematically you can test every stage safety.

This works because you can test a rocket that has bad/unsafe stage, however you will only get to it when the very next stage gets allows you to GTFO, safely extract one\'s crew from unforeseen catastrophic technical difficulties.... =P

In simpler terms, you have an abort stage that you will use at the end of the stage test, or if things take a turn for the worst, you abort early.

This does require quite a few launches, any you may consider trying a good launch more than once to get a better feel of the rocket itself. I don\'t think I would try to test the rocket to its limits, because that might result is a partial failure that sends a single part hurling into your capsule. S:| Ship breaking in half or a single engine cluster detaching causing a complete destruction of the rocket, something you don\'t want risk

Cheers

P.S. Edit:

-I would test the abort stage first, just to get an idea of how it handles, you are going to want to see how quickly it can turn [so that you don\'t turn right back into your doomed rocket if that be the situation.

-Also I would test the lander in orbit of Kearth first before heading to the mun, so that you know how much delta V it has, and how long it burns.... This way you don\'t find out that you don\'t have enough fuel 1000m above the surface, and traviling 150m/s down :\'(

[togfox]

(not ignoring inquisitor - just posting this before I get distracted!)

So, following my lander concept, here is my first design (see image). A duel engine Mun descent rig craddling the rover with the stack holding it all together. The stack is currently:

parachute

capsule

escape decoupler

Mech Jeb

Ascent autopilot

stock decoupler

Mech Jeb (for rover)

Cart (which is the rover)

The engines are the stock tiny variety that can support the short stock legs.

PROBLEM

I field tested this on the pad. No flight. I set the staging right and dropped the rover. The height was sufficient but unfortunately the wheels are wider than the body and I couldn\'t clear the LFT or the LFE.

I\'m going to have to provide more clearance for the rover. That means I need struts. OR a half tank stuck straight to the sides of the stack and then another half-tank beside that for the engines.

Like this:

oo||oo

^ ^

(the o\'s are half tanks stuck directly without couplers).

First test: Fail!

Edited August 10, 2012 by togfox

[chimpychimp]

Can I ask what pack you are using for the lander? Also what are 'mechjebs', does this refer to the advanced SAS module? Ive been out of the loop for a couple of patches. If that is the ASAS, what is the ascent autopilot?

[togfox]

See first post and then do a search of these forums.

[Kosmo-not]

If you opt for an RCS return stage, you could eject the side tanks and drive around as a mobile launch platform.

[togfox]

That method of strategic placement of the ejection decoupler is a good one. I\'ll have to build that into each test flight.

Attached is my new design. I\'m not convinced it is the best though!! My first design didn\'t work because the side tanks were touching the mechjeb on the rover so it would never actually drop to the ground. I need some distance and clearance from the ground. I haven\'t crunched the numbers but the weight and drag should be almost exactly the same.

Edited August 10, 2012 by togfox

[togfox]

First test flight - aimed for 10,000m and ended up at 20,000m. All went well. Forgot to power-down the LFT before ejecting. Some vid of me trying to lose it!

Next mission will be with more tanks and engines. Still not ready for solids yet. I\'ll have to invent names for my delivery and return craft.

http://youtu.be/8EysRyTZznI

[mincespy]

If you succeed with this, land on the sun with stock parts.

[DonLorenzo]

Here is my concept for the rover delivery module. I haven\'t tried this yet but it is essentially the rover in the middle straddled by the Mun descent engines. Two full LF tanks with the \'half\' engines should be enough for a gentle descent from a PE of 5000m or so. the rover will then disengage and drop onto the ground. I\'m hoping there will be enough fuel for the return to Kerbin. I\'m not sure if there is a safe way to test this is sufficient fuel.

To safely test the fuel budget: Replace the rover in the test flight with an appropriate amount of RCS tanks + thrusters under your capsule. See if you can deliver that back without touching the RCS fuel. If yes: success. If no: you\'ve got the RCS fuel to get you home anyway.

[togfox]

That\'s a good idea Don, though I have to admit I\'ve never tried to escape the Mun with RCS so I\'m not sure how much is enough and I\'ve heard that multiple RCS can actually affect stability in a negative way? Still, some RCS is better than none and worth considering.

Below is my second test flight. I\'ve attached a whole lot more LFT and LFE and reached a height over 100,000m. It is really stable but I used pretty much all the fuel to reach this height. I don\'t have enough for the Mun injection, de-orbit and return trip. I also noticed my \'real\' boosters used up fuel before my gimbals. This is not efficient so I\'ll think about that one as well. The vid shows the staging sequence and I\'m happy with it.

Am I ready for solids yet?

My space program needs to be lean and calculating and so too are my vids. I\'ll keep each one no more than 60 seconds so enjoy.

http://youtu.be/U5pbhjOwNlA