What is your design process?

[falofonos]

1. Decide to mine kethane on Mun or around Jool.

1a. Build and launch kethane scanning probe to target.

2. Build miner / harvester landers and maybe even test them.

3. Build launch system.

4. Launch. Explode, refine launcher, repeat until success.

5. Realise I left something off, end flight and correct. back to 4.

6. make mid course correction, realise I left something else off. back to 4.

6a. leave computer running for hours to scan for deposits using probe, assuming it works and doesn't require redesign.

7. Crash on target body, back to 4.

8. Land on target body, realise I left something off (kethane processor usually) back to 2.

9. Success!! Produce full tanks of fuel, mine full tanks of kethane. Wonder what I need them for. Version update of game or mods, start again.

[alacrity]

1. Come up with a bone head final product that will be a giant pain to build...

Examples...

Single launch vehicle for deploying 4 probes in a race around Karbol Space.

Launch vehicle for take Station parts up that aurrounds the parts instead of just stacking them on top.

"Spindle Station" designed merely to showoff the in line docking ports in such a way to fuel pods radiate from a central stalk.

Build as close a replica as I can to the "Rifter" in EVE Online.

http://i.imgur.com/USCxbT0.jpg

2. Get a basic design setup, add a heavy lift stage and see what happens...

Don't need an example for this... really just four options.

a) Should have used more struts option. Ship rips itself apart on launch, usually starting with the solid rocket ripping of the sides and colliding randomly with other part.

A little thin in the middle. Connecting the upper stages to the main first lift stage, I left it too thinly attached. Usually ends quickly and snaps in half at initial launch, or rips in half when the main engines cut off and wobbles apart.

c) Came up short on the math. Underestimated my weight load and don't make orbit. usually I try this design flaw twice. I run it up again to see if I can try a different flight profile to claw my way up to orbit...and then fail... again. Ends in a swim home.

3. Fix the obvious problems.

a) Add struts. This process probably last for a few launch cycles. Strut balancing is a tricky biz and I don't like ending up with a ship that is more strut then ship.

Spread the contact point between stage 1 and 2 out with struts, different engine configuration, and weight redistribution.

c)Thin out any useless mass and add some more punch the the second stage.

4. Get to orbit and celebrate. This is the part of the where things go really wrong. i get to orbit and then I have to decide what the application of the crazy idea is.

a) Too many parts. I watch my insane creation go through a process of becoming a flip book.

using the design take longer then building it. The multi probe deployment idea to 4 hours to actually deploy all the probes.

c)New version of Kerbal comes out making my design impossible, useless, or fataly flawed... assuming a fatal systems crash does not kill off my whole inventory and I have to start over.

6.Get bored with the current design download all the mods I can and make everything from scratch with weirder pieces.

Alacrity Fitz

[ApocSurvivor713]

1: Objective. Find what I want to do.

2: Know how. How do I get there? What angles, how much dV, etc.

3: Looks. Come up with a design that satisfies my eyes.

4: Works. Tweak the design's functions. Add panels, fuel, landing gear, etc. Balance CoM and CoL if needed.

5: Launcher. If it's not an SSTO, come up with something to actually get it into orbit.

6. Revisions, revisions, revisions... Tweak the design further so that it looks amazing and functions flawlessly, or close to it.

7. Launch! The mission begins. I'll probably keep revising the design, but for now it's good!

[AmpsterMan]

1: Objective

2: Delta V Required

3: Design Lander Stage

4: Design Command/Service Stage/Transfer Stage

5: Pick Launcher

6: ???

7: Explode

[johnnyhandsome]

Designing rovers:

Design in VAB.

Test on VAB grounds.

Fix problems

Land rover on Mun

Test Rover on Mun

Fix problems

Rinse and repeat.

[ruiluth]

First, I design the payload according to its purpose, and usually way overdo it (add landing lights, floodlights, extra parachutes, extra other stuff, etc.).

Second, I carefully calculate the delta-V needed to deliver it to the target world using charts and past attempts.

Third, I design a launch vehicle with the required delta-V.

Fourth, I go over everything to make sure it will work, then test it and troubleshoot the problems.

Fifth, I use MechJeb to fly it to the target to make the flight as fuel-efficient as possible.

Sixth, I realize that I left out a crucial part and start over.

[tk5800thesecond]

Backwards, I know X fuel tank with y engine can get me into orbit, but to get those up there I need z tank and z1 engin, but that isnt enough to lift the capsule so I put boosters on

[redteddy23]

I have two design processes.

The Ideal Approach

1. Think of mission (Deploy a small base on the mun with a nice rover and a small robot mun hopping scout to find interesting stuff, then return crew home.)

2. Build components and test them. (Design rover and test at Kerbin weight then work out if it will be controllable at Mun weight and so on.)

3. Build rocket for all the stages carefully calculating for delta/v requirements and T/W ratio for the gravity of each body visited.

4. Check for essential components. (Power, landing gear that actually reaches the ground, just enough RCS for each docking and maneuver of non-kerballed craft etc).

5. A cycle of refining craft to minimize weight and part count.

What actually happens

1. Chuck something together that looks cool.

2. Reach orbit.

3. Check how much fuel I have left.

4. Think of somewhere to go.

5. Go there!

5b. Possibly come back...

[Kenobi McCormick]

1: I decide where it's going.

2: I decide what's going there.

3: I spend two hours designing the rover(That's inevitably what it is, haha)

4: I toss the rover on a ready-made, proven rocket.

5: Whoooooosh!