Do you fly test flights in career mode?

[Lukaszenko]

I test and I pay for all failures. This means that I will rarely test an entire huge rocket, but I will at least drop the lander form a height, for example, or test some aspect of the rocket I'm unsure of. This means I have to be more thoughtful and careful of my designs. When landing on another planet I have to rely on simple math and engineering choices to make sure it will work. They didn't hyperedit the Lunar Lander onto the moon before going for the full mission (although they probably can simulate it much better nowadays, but never 100%).

When launching the full-sized rocket it will usually be the final attempt, but I have to be pretty confident it will work or else it will be a total loss. It's happened many times, so I always give the Kerbals an escape. I'll launch them separately if need be.

[xnorb]

Reverting.

Time is the only currency that's of any value in video gaming.

[11of10]

I usually do test extensively, since I play Realism Overhaul career with disabled reverting and permadeath. Actually I caught myself testing my Moon rocket just as much as real Saturn V was tested, with as much automated tests as possible

[Cirocco]

Usually repeated revert-to-VAB tests, though testing rarely goes to orbit. I like to keep it at least semi-believable that I'm "running a simulation" so tests are always relatively short. When I'm testing spaceplanes and I actually want to see if I can make orbit efficiently, I usually do a full flight with landing and pay the costs.

[rynak]

I added a proposal for a mod, that deducts a configurable percentage for reverted testflights:

http://forum.kerbalspaceprogram.com/threads/107924-Revert-Costs

[cicatrix]

This reminds me of Russian "Luna" program.

I think Squad had read this when they had KSP in mind:

The 8K78 Molniya had 9 launches on the E-6 programme. Only four of them had all four stages working as planned. The fourth stage failed 5 launches, the third one was the cause of 2 failed missions and 2 missions failed because of the first and the second stage.

This frightening statistics comes out of rocket-artillery approach to testing in USSR multiplied by politics during the Space Race. Firstly, the designers were former artillerists and they were accustomed to make multiple launches to judge the design by its failures. It should be mentioned now that this approach turned out to be incompatible with spacecraft design - the price of a wasted launch and wasted money was greatly higher than detailed testing on the ground. Secondly the seeming success of interplanetary missions to Mars and Venus led to the thinking in terms 'we've done it - let's launch it, it might even fly'. Naturally the results showed that this approach doesn't work.

It should also be mentioned that failure analysis was poorly conducted. The reason for this was in the fact that the fourth stage started over the Atlantic Ocean between South America and Africa and the craft couldn't transmit the telemetry due to curvature of Earth and more than frequently the reason for another failure remained unknown. By the time E-6 programme started the special ship 'Dolinsk' had been sent in Guinea Bay to receive the telemetry and send it to the mission control.

A simple calculation shows that in these conditions the success probability is about 10% (1 success / 10 total). Though, if you think twice, this probability approaches zero due to the fourth stage failures when you had to guess what went wrong with the fourth stage.

The first E-6 launch:

Jan 4, 1963. Fourth stage failure. Dolinsk was in position and managed to receive the telemetry. The immediate reason was a failure to send a signal to start the engine. This was caused due to a failure in the current conversion unit. This part was filled with dry Nitrogen and this caused increased wearing of electric motor brushes. In spice of the fact that during testing these brushes were operational for quite a longer time no other reason was found. As a solution it was proposed to moisturize the Nitrogen and add a bit of Oxygen.

Reason: poor design

Solution: change in design

Second launch:

Feb 2, 1963. Third stage failure. The rocked deviated off course and fell in Pacific ocean. The reason - the guidance gyroscopes were installed 'upside-down' and aimed for the ocean. Also, gyroscopes precision was not sufficient since the real trajectory deviated from the planned one.

Reason: poor design, operational failure

Solution: Changes in documentation and assembly control procedures, changes in design

Third launch:

Apr 2, 1963. All four stages work as planned, the station flies towards the Moon and named 'Luna-4'. During the course correction the burn pushes the station off course and it flies past the Moon. The immediate reason was not determined it was only known that it was a failure in astronavigation system.

Reason: Supposedly poor design

Solution: design changes

Fourth launch:

Mar 21, 1964. Failure in the third stage. The oxygen valve failed to open the third stage engine did not start. The valve pin broke and all valves had to be redesigned.

Reason: poor design

Solution: design changes

Venus helps

On 27th of March 1964 a 8K78 rocket carrying an interplanetary station to Venus was launched. Yet again, the fourth stage did not start but it had a flight recorder installed and the sequence of events prior to the start of the stage became known. It turned out that designers missed the fact that control relays were not switching instantly. Due to the delay the valves received no electric power and never opened. The fourth stage couldn't get the orientation and tumbled on the transfer orbit. What amuses the most - this problem was eliminated on Earth with a simple solderer in 5 minutes.

Reason: poor design

Solution: design changes

Fifth launch:

Apr 20, 1964. Yet again, the fourth stage engine fails to start. The telemetry records the absence of the start control signal. New tests of the I-100 guidance unit and the PT-500 current converters are conducted. They discover in the labs that in spite of generally satisfactory temperature of the I-100 operation, some of its parts are locally overheating. Converters were sent to be fixed and decided to cool it down prior to launch.

Please, note that the immediate reason of this failure is the same with what it had been during the first launch and the reason for it was unclear yet again. Because of this a conflict between the designers of PT-500 (Iosifyan) and I-100 (Pilyugin) is brewing up.

Reason: poor design

Solution: design changes

Sixth launch:

Mar 12, 1965. Fourth stage failed to start again. Both lead designers argue whose part is to blame. Pilyugin's colleagues test Iosifyan's PT-500 converter and many hours later discover the possible reason for failure - the balancing ring might interfere with the fixing bolt which overloads the drive and leads to its failure. The failure investigation commission decides to replace the P-500 converter with two older models of PT-200 which were also of Iosifyan's design.

Reason: poor design

Solution: design changes

Seventh launch:

Apr 10, 1965. The serial number of the station reached the number 8 and the launch officers joked about this (number eight - wont get to Moon). This joke turned out to be true - the third stage failed.

Reason: Manufacturing defects in the tank pressurization system.

Solution: None. A single manufacturing defect could not cause any changes in design.

Eighth launch:

May 9, 1965. All four stages work as planned and the station named 'Luna-5' flies to the Moon. However, there was too little time planned for gyroscopes heating up in control algorithm and the correction burn was performed only at the third attempt. What killed the station was the failed braking burn (because of the same reason) and it crashed onto the Moon surface.

Reason: programming error

Solution: changes in design, changes in guidance unit algorithm

Ninth launch:

Jun 8, 1965. 'Luna-6' station successfully started towards the Moon. It failed because the engine failed to turn off during the correction burn and burned all the fuel. The reason was the absence of the burn time duration in the commands that had been sent from Earth. The station flew past the Moon.

During the Moon flyby all possible tests were performed which uncovered a whole

number of new faults.

Reason: operator's mistake

Solution: not listed.

Tenth launch:

Oct 4, 1965. Successful launch as 'Luna-7'. It managed to get to the braking phase without any problem. The station switched to the real time telemetry mode and reports in the Mission Control started: 'On the Moon vertical', 'Radio altimeter reads 5000 km', '4000 km'... and a half an hour later - 'Earth is lost'. Astronavigation sensor lost the view of Earth and it meant the braking command wouldn't be executed. The station impacted on the surface. The reason was in the way the sensor was installed - Earth was on the edge of its field of view and even the slightest deviation meant the loss of orientation.

Reason: operational failure, poor design

Solution: design changes.

Eleventh launch:

Dec 3, 1965. The station is named 'Luna-8'. For further reliability a fake correction maneuver was planned when the station was supposed to turn on the astronavigational system and orient itself but not turn on the engine. However, during the brake when the impact amortization balloons were pressurized the station had lost the altitude readings. Presumably it started rotating and the altimeter 'lost' the Moon which blocked the braking burn. The engine had finally turned on but only 9 seconds before the impact and the station was too fast to survive.

The loss of orientation made the impact amortization balloons the main suspect. Five days of experiments have discovered the problem - the balloons when pressurized broke the antenna attachment which having been broken had sharp edges which in turn pierced through the balloon and the resulting gas leak lead to rotation and the loss of orientation. The fixing point of antenna broke due to manufacturing fault. There is an alternative version - the pressurization of the balloons itself caused the rotation. In this case the solution was simply a later pressurization when the braking engine already worked which would help to stabilize the flight.

Reason: manufacturing fault, design fault.

Solution: control and acceptance testing during manufacturing, changes in landing algorithm.

Twelfth launch:

Jan 31, 1966 the space station 'Luna-9' successfully started and on the 3th February of 1996 successfully landed on the Moon for the first time in the history of mankind.

Edited January 20, 2015 by cicatrix

[ArmchairGravy]

What is this "testing" you speak of?

[Tw1]

I also test my landers with Hyperedit because there's no way to simulate gravity / atmosphere.

I also test with hyperedit, but only to check how something will handle gravity and/or air density. I use kerbal engineer stats for everything else. I use the planet editor to change kerbin, rather than teleporting to the other planets,

[gazzareth]

I test and revert if I need to (or if it was a pure test and not meant to go anywhere). That said if I strand a Kerbal somewhere then I tend to try and launch a rescue mission, which is much more fun than reverting, although I haven't attempted stranding one on Eve yet.......

G

[Rosco P. Coltrane]

Test and revert if it's a simple test, usually just launch and get to orbit. If it's something more complex, then I go for sandbox because sometimes (didn't bother to research why) you just can't revert. That's how I got an 8 satellite carrier in orbit around Mün... hyperedited the thing there, tested decoupling one sat and suddenly I couldn't revert! Big lose of money because I have no use for Münar satellites anymore in the stage I am in my career.

[magico13]

As mentioned by several others, I use Kerbal Construction Time (link in sig ) to test things (even when reverts are turned off in the save). I will admit that I have been cheating lately and have turned off costs for simulations and unlocked the Body Tracker so that I can test my Duna lander around Duna before I send my OKS station there. 23,000 funds for an unlimited time simulation in orbit of Duna was a bit too cost prohibitive at the moment.

The cool thing is that I can also turn off build times (with basically one click) so that I just have the normal game + optional simulations. Works great on Sandbox so I can test things in orbit without dealing with build times or installing hyperedit, but works just as well in a hard mode career game.

[Wanderfound]

In .24 , when funds were so gratuitously oversupplied, I ran testflights in Career as a way of burning off my bank balance. Now I'm back to testing in Sandbox.

[Superfluous J]

...sometimes (didn't bother to research why) you just can't revert.

If you leave your ship at any point, you lose the chance to revert. I don't think there is any other restriction.

[Fleb]

I normally play on moderate so I just use revert. When I tried hard difficulty though, I hired a "test pilot" after Jeb was almost killed, after a pre-flight staging mishap.

[FlyingPete]

Especially for things like planes (which I'm not very good at building) I'll use revert if it doesn't work correctly (call it R&D using a CAD system). The flight where it works is considered a manned test flight.

[Engineer of Stuff]

I just revert. I pretend that the Kerbals are running simulation on a computer.

[Torquemadus]

I test fly my fly-back stages to ensure that they are capable of runway landings. Since they are meant to be launched vertically as a shuttle, I could do a vertical launch for the test flight. However, I find it more interesting to have the stage perform a runway take off so I can assess how it flies as a plane. Since the stage recovers to the runway, it feels a bit stingy not to pay for the fuel.

Once I'm satisfied that the stage flies properly as a plane. I save it as a subassembly for vertical stacking. If I waited until the first orbital flight before testing the fly-back stage, I might have a nasty surprise waiting for me when I try to return to the KSC! In such a case, I'd then have to redesign the stage, then re-stack it in the VAB, then perform another orbital flight, then have another go at returning to the KSC to see if the changes worked. Even with the revert button enabled, that's still a lot of messing about!

[Hodo]

I often build a craft and fly it doing test flights, I have a routine of 5 flights for testing before I consider a craft ready for service at that point I change its craft name from X-whatever number to SP-whatever number or F-whatever number I am up to.

Unless it is a purely atmospheric craft non-combat craft then it is listed as S-whatever number or P-whatever number.

F-Fighter

F/A-Fighter/Attack

F/B-Fighter/bomber

SP- Space Plane SSTO

xxxV-VTOL capable craft

S- Science/Scout aircraft

P- Passenger aircraft

R- Rover

U- unmanned vehicle

H- Helicopter

[kinnison]

I had to do a save/revert with an Eve ascent rocket in my career mode. Tho honestly I could have done it in a sandbox version

I think I had 3 versions that failed in the design phase, and had three versions tested until I was successful. Now I have to figure out a way to get my Kerbal to the top, so I can actually send the science bad to Kerbin

[bdito]

I test and revert in the same save. In real life you'd have months upon months and run simulation after simulation to ensure your craft is perfect. Obviously we don't have that here so the revert button does that.

[Kesa]

For my full stock game, I could just overpower my craft and go, but I feel like getting the better out of my craft, so lots of testing.

Using Kerbal Konstruction Time, I can and have to test, but the cost for it is so low (at least for launching to LKO/LEO) considered to what it should be (engineering cost should be far more than construction of one prototype, and is ~ 1/100th), I voted "avoid the cost". I'm very new to RSS, so I really can't have a good launch at my first tries, so I do many tests with KCT. Sounding rockets are dirty cheap though, and I've crashed some of it, but their unmanned, so I don't bother.

[SorryDave]

I never test rockets in career. I usually just wing it. Usually with no problem. but almost every game my first rocket ignites the engines and deploys chutes in the first stage,.