Since graphs & charts are OK in General KSP Discussion, here is one on engines.

[Dispatcher]

Hi,

Edit: my Kerbal engine comparison chart may be of interest to you. It contains somewhat practical engine test results as well as data gleaned from the KSP game editor (VAB/ SPH) and Wiki. It is sorted by the max mass each engine was able to successfully launch with from sea level, what I designate as "strength". This is sometimes in line with Squad's thrust data, but sometimes not. Altitude figures of a second test at sea level are of each engine using a small tank, which sometimes is in line with iSP, but again, sometimes not. I refer to this as "efficiency". You may also compare Squad's T/WR (and other characteristics) of the engines. All engines are compared (inline, radial; liquid fueled [rocket & jet], solid boosters, monopropellant thrusters, ion; even a Kerbalnaut with EVA pack). Common questions are addressed by the info at the side and keys & comments at the bottom of the chart.

Anyone wishing to sort Squad's data may do so using the interactive page of stock parts at the KSP Wiki. Here is the chart:

For those wishing to download it:

http://kerbalspaceprogram.com/engine-compare-chart/

Details are at the Spaceport and in the chart itself. Thanks!

Here is another forum link on my engine chart, just for completeness:

http://forum.kerbalspaceprogram.com/showthread.php/47954-Kerbal-Engine-Strength-Efficiency-Comparison-Chart?p=621286#post621286

Thanks!

Edited August 30, 2013 by Dispatcher
SpacePort issues?

[Johnno]

But you're not posting a graph, you're posting an advert for a thread that itself links to the spaceport, where you'd have to download the thing...

I'm fine with charts, I'm fine with addons and utilities, but this thread is a bit over the top.

[1001sd]

I agree with Johnno, why did you do it in that way?

And secondary, what about giving us the spreadsheet you used, so we can fiddle with it ourself (sorting it by the different categories).

[sojourner]

So, what the heck is "strength"? your test seems arbitrary for this. "I loaded the engine with as much fuel as it could carry without crashing or going out of control". Define "crashing or going out of control" in regards to how much fuel you loaded. Bad science is bad.

[Dispatcher]

My apologies to you all. I had posted about the chart in the add on forum before I became aware of the graph discussion(s) in this forum, so I decided to mention it here and link to the other post. If that's bad taste, my bad. Johnno, a poster below you is in fact asking for a download, as you can see now. Sojourner, the chart itself demonstrates what is meant by "strength". However, I will go into more detail. By using the small RGU (and an air intake to keep mass and drag more equal for all engines where possible) and loading up the engine with tanks of various sizes, I tested until each engine failed to carry its starting mass or failed to fly until it ran out of fuel. The maximum altitude included inertial or momentum coasting (and for great altitudes used the apoapsis projection as the maximum). The more important figure is not the number of tanks or amount of fuel, but starting mass. One column gives max fuel, the next column give the more important mass and the next one gives the max altitude. The starting mass is I think the best measure of "strength". The altitude simply shows how far the empty tanks and engine were propelled. I do not know of a better way to demonstrate on a practical level the "strength" of an engine, since all engines can fail if loaded with too much payload. I hope you might decide that this is not "bad science". At least, that the tests might have practical applications, such as helping someone determine which engine to use for a desired purpose, which would be up to you to decide. And 1001sd, your english is excellent!

Edit: Since the tests are of differing types of engines, there were bound to be some special circumstances. For example, solid boosters were not able to use up their payloads as fuel; this was dead mass. I also had to adjust figures for boosters to, in effect, shrink them to the size of a small tank (the small booster is about as massive as two small tanks). For jets, they run out of oxidizer instead of fuel and thus have a natural altitude limit. There were several other considerations, but the chart touches on those also.

As for the spreadsheet, I use OpenOffice. If more people want that, I am open to the idea of making it available. Edit: note that you may use the interactive charts at the KSP Wiki to sort each variable of a stock part. So my providing a spreadsheet seems a bit redundant. The primary purpose of my chart is simply to generally rank engines. "Strong" tends to be toward the top of the chart. It happens that efficient engines tend to cluster toward the middle of the chart and the ones best used only in space tend to be found at the bottom of the chart.

Edited August 28, 2013 by Dispatcher

[Kerbart]

First of all, thank you for all the work you put into it. Without having even downloaded the chart it looks like you spent a lot of effort on figuring everything out.

Now, from what I gather, is that this particular chart tries to compare engines with each other "everything else being equal". That's great but the reality is that "everything" is not equal, which is why we have different engines in the first place. If you need sheer brute strength, get a mainsail. The fact that it consumes insane amounts of fuel is kind of irrelevant, as this is a typical first-stage engine. Who cares that it can empty a Rockomax Jumbo 64 tank within a minute? This is about pure, raw, uncensored thrust, and nothing else. On the other side of the spectrum is LV-N atomic rocket motor. I'm sure it's not as strong as the Mainsail, but hey, it gets me to Jool and back on half an X-200-8, so who cares? At this point I do not need thrust.

I'd be really interested in seeing, for the various payloads, how much fuel it would cost me using the various engines. That would actually allow me to choose engine x over engine y for a compelling reason.

[Dispatcher]

Thanks for the kind words, Kerbart! The chart went through several "final" versions until I got it figured out properly. I put too many hours into testing the engines and refining. My hope is that I did the tedious work so that you don't have to.

You are precisely right. In the end, you need to tailor each mission and hence each engine for the payload at hand. What my chart does do is show you the maximum mass each engine can handle. For all non solid engines, that is also the maximum amount of fuel tank mass each can handle. In real simulation terms, the actual for your needs will always be less for payloads other than fuel (excepting those pesky solids again). The solid booster results may relate to what you wrote since most of that mass was dead. I learned a lot from the testing, so perhaps you will find a nugget here or there in the chart that you may find interesting and useful.

It seems to me that what you are suggesting would be some kind of interactive chart where you plug in one variable and get results in another cell. That's a bit beyond me but maybe someone else is up to the task. However that would be a hard project because as you know engines are typically used in combination in a tandem event.

[tavert]

These seem like fairly straightforward calculations to make by hand, you don't need to do experiments to figure these out. Max mass at liftoff? Thrust divided by Kerbin's gravitational acceleration. Amount of that mass that can be fuel tanks? Total mass minus engine mass. Fuel contained in that wet-tank mass? Multiply by 8/9, or a little less if you're using any Oscar-B or Round-8 tanks.

The "how far with a given amount of fuel" is best expressed in terms of delta-V, which is easy to calculate from the rocket equation. If you burn straight up, max height for a given amount of fuel is a function of delta-V, thrust-to-weight ratio, and specific impulse. The exact function for max height in a vertical burn is a little messy to calculate (also depends whether or not you throttle down at terminal velocity), but is less useful than delta-V and TWR.

[Dispatcher]

Very interesting information, "tav" (my spell checker keeps 'correcting' you to 'tavern'). I wanted to see what the simulation would do in lieu of earth based calculations. I actually enjoyed the testing and learned from it. The chart is an attempt to share data from those tests. You mentioned something which is important to note: the throttle setting(s). The chart only shows results with max throttle (except as noted on the chart). So, like all charts (Squad's, mine or anyone else's), these are handy for starting points but ultimately every mission simply needs to be figured out (or else throw more boosters at it).

Edit: the testing does show some quirks which are likely due to the game "engine" and physics, but still noteworthy. We are all familiar with how a mainsail needs to run throttled down a bit while directly connected to an orange tank, while two "half" tanks can be lifted at full throttle. My test results on the aerospike were that it can fly more mass using the "narrow" tanks than it can using the "wide" tanks. It also is a challenge to control even with SAS enabled, while the LVT-30 appears to fly true (without control surfaces) even though it is not a vectoring engine.

Edited August 29, 2013 by Dispatcher

[sojourner]

Yeah, this "strength" thing is still bad science. You're just making a mess of TWR and ISP. You're also confusing bad rocket design (wide tanks vs narrow tanks) with engine performance.

[tavert]

Very interesting information, "tav" (my spell checker keeps 'correcting' you to 'tavern'). I wanted to see what the simulation would do in lieu of earth based calculations. I actually enjoyed the testing and learned from it. The chart is an attempt to share data from those tests. You mentioned something which is important to note: the throttle setting(s). The chart only shows results with max throttle (except as noted on the chart). So, like all charts (Squad's, mine or anyone else's), these are handy for starting points but ultimately every mission simply needs to be figured out (or else throw more boosters at it).

Edit: the testing does show some quirks which are likely due to the game "engine" and physics, but still noteworthy. We are all familiar with how a mainsail needs to run throttled down a bit while directly connected to an orange tank, while two "half" tanks can be lifted at full throttle. My test results on the aerospike were that it can fly more mass using the "narrow" tanks than it can using the "wide" tanks. It also is a challenge to control even with SAS enabled, while the LVT-30 appears to fly true (without control surfaces) even though it is not a vectoring engine.

Great, I applaud the scientific spirit. Maybe you would've gotten a slightly better response from people with a different spreadsheet format and posting method (Google docs would've been very appropriate here). And I agree with sojourner that some more research into TWR and delta-V concepts would've helped inform your experiments.

The reason for the aerospike instability is it has a slightly lower drag coefficient than other parts in KSP. Drag is calculated in a strange not-very-physical way in the stock game, where drag on each part is proportional to that part's mass and applied at the center of mass of each part. Since the drag coefficient of the aerospike is slightly lower than other parts, less drag force per unit mass gets applied to the aerospike than all the other parts on a rocket. If the aerospike is mounted at the bottom of your rocket (as you might expect for most designs), this means the overall center of drag of your rocket is located higher up than the center of mass. Unless you're pointed exactly surface-prograde (or retrograde), this means the drag force exerts a torque on your rocket that makes it want to point backwards.

So while the "pendulum rocket fallacy" is a fallacy most of the time, due to the aerospike's drag characteristics it actually helps immensely to mount aerospikes at the front of your rocket instead of the back. Then the center of drag will be lower than the center of mass, and drag will exert a stabilizing torque instead of a destabilizing torque. And since aerospikes don't gimbal, they aren't subject to the bug where gimbaling rockets mounted ahead of the center of mass don't switch their gimbaling direction like they should.

Edited August 29, 2013 by tavert

[Dispatcher]

sojourner, I think its noteworthy that only the aerospike is sensitive to what kinds of tanks are used (see "tav"'s following post too as he addresses the reasons). As for making a mess of TWR & ISP, the chart does include KSP's data for those wanting to refer to it. As for "strength", that's my attempt at a practical simulation result which anyone is free to use or ignore. But thanks for the attention!

Edited August 29, 2013 by Dispatcher
Spelling error.

[Jesrad]

Funny I made a similarly-minded "cheat-sheet" for KSP engines a long time ago, and kept it updated.

The conclusion is, use mainsails with some optional skippers/LV-T30s for takeoff stages (with asparagus tanks on an aerospike optional), then slap a LV-N on your orbital/interplanetary stage. Everything else is either a complex-maths exercize in optimization (you may get a better payload mass at a given low deltaV target value using smaller, higher TWR engines) or an aesthetic pursuit.

Edited August 29, 2013 by Jesrad

[Dispatcher]

Very nice work, Jesrad! I think that could be quite useful and at the very least is most interesting to see. My favorite staged launch vehicles tend to use a mainsail with a mix of (turbo)jet engines and asparagus stepped LVT 30s and aerospikes on the first stage (well, I also mount jets on higher tanks but jettison them at around flame out time or just past that). In space, the atomic is great!

[Mesons]

Informative and cool. But a link to a site with a download button to download a zip file with a folder with a single .png image was not. At least the link was to a trusted site.

[Dispatcher]

Mesons, thanks! Hmmm. A folder. Using a Mac, I simply compressed the png and obtained a ... zipped file. I'm guessing you use Windows and that is perhaps why you got a folder in the process? Ooops. Not intended. Happy building!

Edit: I scan files with an up to date anti-virus app prior to uploading. It recognizes definitions of Windows as well as Mac viruses. I thought you'd like to know.

Edited August 30, 2013 by Dispatcher
Adding reassurances.

[Jesrad]

Compressing an already compressed image format provides zero or negative gain on filesize.

Keeping the file in its PNG format instead of hiding it into a .zip file, makes it possible to display natively in any browser, so the extra steps of downloading, uncompressing and opening the file in a preview application are bound to be viewed as, at the very least, unnecessary and cumbersome... and at worst they may be considered malicious: it's common practice for malware distributors to hide the nature of their stuff by compressing them into .zip files then masking the true nature of the file by changing the "supposed" extension in the file's displayed name to something innocuous, like that of a picture. The goal being, you double-click on a .zip file thinking it contains picture_of_naked_kitten.png while it really contains takes_over_your_pc_secretly.exe

Thought you'd like to know.

[Dispatcher]

Thanks! I'll see about taking care of that. I've been under the wrong impression that a zipped file is less prone to corruption than an original.

Edit: Links are updated.

Edited August 30, 2013 by Dispatcher
Completed task.

[M5000]

Dispatcher, would you please kindly upload your spreadsheet as an image of some sort, using an image hosting site like Imgur.com or Photobucket or something of the like? It's a bit odd to have it up on Spaceport, as that's not really what it's meant to be used for. If you want to give users access to the actual spreadsheet file, you can use a file hosting site like Rapidshare or Mediafire. It's nothing personal, but that's just not what Spaceport is for, not to mention that we get a lot of players complaining about the Spaceport being cluttered in the first place. Spaceport is for mods, crafts, and packs really, and shouldn't be used as an image hosting site. Thanks in advance for complying, -M5K

Edited August 30, 2013 by M5000
got dang line breaks.

[Dispatcher]

I don't see why not. I suppose its the impression I had of SpacePort being the 1 stop place for Kerbal needs; that and file formats including image formats being the allowed ones. Perhaps that's more a feature of the site software used rather than for content guidance. I'll put it up and edit my thread starts to add the image as an alternate to downloading a file from another location. Thanks for the friendly advice!

Edit: added the image at head of thread.

Edited August 30, 2013 by Dispatcher
Complied with request.