puff engines

[farmerben]

I overlooked the puff engines for a long time.  Now I'm working on a mining vessel with xenon engines.  My mining vessel doesn't want to carry lots of fuel and oxidizer tanks, these can dock during refueling.    Monopropellant tanks are a better deal in terms of weight and usability.  It occurs to me that puff engines could provide the kick to move asteroids or takeoff from Minimus when more thrust is required.    

In early testing I just blow through all my Mono propellant.  

 

Does anybody the puff engines?  How do you use them?  Pictures please

[Freds]

No pictures because i'm not at home, but, almost all my satellites use monopropellant based engines (not the puff, but other engines provided by mods). Monoprop is light, and can give a lot of deltaV without need of electricity, which is perfect for outer planets like Jool, Eeloo, Dres (who goes to Dres anyway?)....

[Rocket In My Pocket]

Not a good idea.

Monopropellent is always worse Dv wise than any other fuel in the game. (In the stock game at least.)

Any other engine with similar weight would be better.

[Snark]

42 minutes ago, farmerben said:

Does anybody the puff engines?

I've never used them myself, ever, because math.  It all comes down to this:

7 minutes ago, Rocket In My Pocket said:

Monopropellent is always worse Dv wise than any other fuel in the game. (In the stock game at least.)

^ This.  Monopropellant has a worse Isp that liquid fuel, in general.  The Puff has a vacuum Isp of 250; the RCS blocks are 240.

That's not... abysmal, exactly, but it's significantly lower than the Isp of LFO engines.

Lower Isp means lower dV for the same amount of fuel.

So, I simply have no use for the Puff.  As far as I'm concerned, it's pretty much just a gimmick.  I use monoprop for just one thing, and that's RCS thrusters on ships that need to do docking.  That's a use case where the crappy Isp doesn't matter, because I don't need much dV, only a few centimeters per second; the convenience of multi-directional thruster blocks is the determining factor there.

[farmerben]

Mono has density 4kg per unit, where LFO is 5kg/unit.  So thats like an ISP of 312.

This is much better than the Twitch engine.

[farmerben]

 

This is an asteroid miner I'm working on.  With 6 xenon engines TWR =.1, with 4 puff TWR = .75 and the Mono is good for about 900 dV.  I have more solar panel and batteries than I actually need.

All the prototypes for this vessel had more xenon engines (as many as 64), but the performance was not much better and my computer could barely handle it.  

 

What do you have that is similar?  What improvements would you suggest?

 

 

 

 

 

[Rocket In My Pocket]

I'd suggest you get rid of all or most of the monopropellent unless you are planning on docking with something for the reasons stated above.

You may need a bit for fine tuning your meet up with the asteroid but I don't recommend using it for general propulsion needs.

Edited July 25, 2018 by Rocket In My Pocket

[Freds]

48 minutes ago, farmerben said:

Mono has density 4kg per unit, where LFO is 5kg/unit.  So thats like an ISP of 312.

This is much better than the Twitch engine.

This is why i use monopropellant in my satellites - weight saving. Small vehicles have small engines, small engines means lower thrust, which require a lighter vessel to be efficient.

I follow this rule: If it is small, very light and goes to the outer planets, monoprop. If it is small,very light and goes to the inner planets and will operate almost in the sunlight, xeon.

In my playstyle, a vessel going to another planet usually has  a probe core, big relay antenna , battery and solar panel installed in the orbital capture stage. This way, with only a small extra weight, when decoupled the 'disposable stage' turns into a relay sattelite. If it has another 3 or 4 tiny and lightweight sattelites attached, a planetary commnet to support land operations can easily created, with a only a little deltaV cost in the main vessel. 

If i will made a ship like the farmerben made, i wont be using xeon - 12T is too much for xeon or monoprop... A LVN will be perfect if high ISP is required.!
 

[farmerben]

Yeah my previous asteroid miner was and LVN, I thought this might be better, maybe not.  Xenon + Nuke has not been satisfactory for me yet.  Anything with TWR  < .2 has difficulty matching velocities with target and so forth.  

 

I like the idea of detatchable sattelites made out of the interplanetary xenon engines, kind of expensive but they can be positioned better.  It would be nice to have half a dozen extra telescopes.

[OHara]

1 hour ago, farmerben said:

Mono has density 4kg per unit, where LFO is 5kg/unit.  So thats like an ISP of 312.

But Isp is thrust per fuel-rate in terms of weight of fuel burned per second  -- not units of fuel per second (weight at sea level on Earth, for some historical reason).  The units used to measure fuel don't seem to mean anything specific in KSP.

So the effect you get from a kg of monopropellent is best expressed with the Isp of 250s, which is not bad.

Edited July 25, 2018 by OHara

[Snark]

1 hour ago, farmerben said:

Mono has density 4kg per unit, where LFO is 5kg/unit.  So thats like an ISP of 312.

This is much better than the Twitch engine.

Incorrect; that's not how it works.

The mass-per-unit of fuel is completely irrelevant to Isp.  The units themselves have no physical significance; it's just a convenient label to use when displaying the capacity of fuel tanks.  What really matters for a fuel tank is how many kilograms are in it, not how many units.  If I have a tank with 1 ton of fuel in it, it doesn't matter if I call that "250 units at 4 kg apiece", or "200 units at 5 kg apiece".  The only thing that matters to the rocket equation-- and, therefore, dV-- is the total mass of fuel carried, and its Isp.

It's not "like" an Isp of 312.  It's an Isp of 250.  Exactly that. No more, no less.  The engine has an Isp of 250 because you can see the engine's Isp rating in the VAB.  That's what it is.

The Puff has a lower Isp than the Twitch, the Spark, the Ant, and the Spider.  There are plenty of small LFO engines that have better Isp -- and therefore more dV per kilogram of fuel-- than the Puff.

38 minutes ago, Freds said:

This is why i use monopropellant in my satellites - weight saving. Small vehicles have small engines, small engines means lower thrust, which require a lighter vessel to be efficient.

Sure, but wanting to be lighter means not using monoprop, doesn't it?

If you want to be lighter, you want three things:

• Less fuel mass
• Less empty fuel tank mass
• Less engine mass

Small LFO engines are better than the Puff on the first two, and at least as good on the third!

• You need less fuel mass, because LFO engines have better Isp than the Puff.
• You need less empty-fuel-tank mass, because LFO tanks are better than monoprop tanks.  LFO tanks have a full/empty mass ratio of 9:1.  Compare that to anywhere from 5:1 to 7:1 for monoprop tanks.
• You need less engine mass, because there are plenty of tiny LFO engines to choose from.  Compared to the Puff, at 0.09t, there's the Spark at 0.1t, the Twitch at 0.1t, and the Spider and Ant at 0.02t.  (Note that any radially attached engine requires you to have at least a pair of them to be able to fly straight.  So taking that into account, the smallest possible engine mass with the Puff is 0.18t, compared with 0.18t for Twitch, 0.1t for Spark, 0.04t for Spider, and 0.02t for Ant.)

So... if you like using the Puff just because it appeals to you, fine.    But "saving mass" isn't a reason, because math.

 

[4x4cheesecake]

Recently, I tried to use the Puffs as RCS thruster:

But it is ugly...so I didn't continued this idea.

It is indeed not the most useful part but once I build an SSTO which got 2 Puffs at the back just to get rid of unsued Monoprop in a reasonable amount of time at the end of the mission. So I got a little bit of dV by firing the Puffs and saved some weight by loosing Monoprop. Maybe some LFO engines might be more efficent, but I had to carry some Monoprop anyway, so why not using it

[farmerben]

I tried swapping engines on my 12 ton asteroid miner.  The key factor in this design is that the kicker engines need plenty of TWR, not so much dV as that is covered by the xenon.  I'm filling this key requirement better using puff than using twitch, any weight difference affects the dV of the xenon.  With spark I can beat puff on every category.  The only reason to favor puff over spark ease of attachment and because we carry lots of mono for docking anyway.  

I just tried this engine for the first time yesterday.  I'm not in love with it.  Just trying as hard as possible to see the merit in it.

Is there any real difference in using the ISRU to create one fuel over another?

 

.... And now I find out the small ISRU can't produce xenon... I think the one in Feline Rover Utilities mod does...

 

Edited July 25, 2018 by farmerben

[Zhetaan]

@farmerben:

Stock ISRU does not, as you discovered, produce xenon.  It produces monopropellant at a loss--it turns 5 kg of ore into 5 kg of LF, O, or LFO, or 4 kg of monopropellant.  That has to do with the density difference, so if you're on a planet, it's really a matter of time spent rather than risking a shortage.  For asteroids and their finite resources, that's another matter, but I digress.

One thing I will stress, and you have my apologies because it's been mentioned already, but I want to be sure that you understand I_sp.  Specific impulse is a function of the engine and the engine alone--bluntly, it's a measure of how efficiently an engine can use its fuel.  That's more intuitive when you understand that specific impulse in the rocket equation is derived from exhaust velocity; it should be apparent that exhaust velocity depends on the engine and not the fuel.  Granted, there are certain fuels that are more energetic than others and can make the engines' job a bit easier, but even then, that energy doesn't amount to anything but a bright flash of radiation and an expanding cloud of gas unless the engine nozzle can confine and direct it to a useful purpose--namely, accelerating the exhaust to a higher velocity.

Now, on to the more important considerations to mind regarding monopropellant as a fuel:  the density of monopropellant is not really important unless we are going to combine our ever-present mass considerations with a realistic volume consideration.  For the longest time, volume in KSP was something of a running joke; it wasn't until relatively recently that the volumes of tanks were specified in litres, or any kind of standardised unit at all.  What came of that is not realistic, exactly, but it is fairly consistent.  Now that we have that, the reason density is worth mentioning is because increased volume per unit mass of a fuel necessitates a larger tank than would be needed to store a fuel that packs its mass more compactly.  This means that monopropellant tanks have a comparatively terrible wet-to-dry mass ratio.  Where LFO tanks generally keep to a ratio of nine (excepting certain odd-size tanks), monopropellant tanks tend to between six and eight.  This means that you spend more fuel pushing the dead weight of tankage than you would with LFO, which means in turn that, given that the Puff is comparable to a sea-level lifter in its efficiency, you will always get more delta-V from an efficient LFO engine and associated tankage than you can get from the Puff.

As a side note, xenon has a similar problem (density is .1 kg per unit and the tank wet/dry ratio is about 4), but it mainly overcomes that by having an absurdly efficient engine.  If there were other engines that even approached the Dawn's I_sp class, then the Dawn would be similarly neglected as the Puff.  As it is, xenon engines are only appropriate some of the time; even with the I_sp advantage, it is not a solution that handily defeats all challengers.

All of this goes to say that the Puff is very much a special-use engine, best suited to situations where you already have a lot of monopropellant and no wish to complicate your rocket with the addition of a different type of fuel, or where you need a radial engine but don't want to deal with fuel pumping (the Puff works like the thrusters in that it pulls fuel from anywhere in the rocket).  I could see a heavy station tug that uses the Puff when it needs to wrestle a large piece into position but otherwise relies on attitude thrusters for its usual work; that would allow you to make use of better thrust while avoiding the need to add a different fuel tank ... assuming that your tug holds enough monopropellant to accomplish that, of course.  Normally for such manoeuvres I prefer to rely on Vernors and LFO rather than attitude thrusters.

You raise an interesting point about using monopropellant in concert with xenon.  You're right that you will lose xenon-derived delta-V if you use LFO, and that is, in a way, traceable to fuel density.  If LFO tanks have a wet-to-dry mass ratio of nine and monopropellant tanks have a mass ratio of less than nine, then litre for litre, the monopropellant tanks will be less massive.  The fuel makes up so much of the total mass that monopropellant's lower density compensates for the higher relative mass of the tank.  However, the engine for monopropellant is so much less efficient that it compensates back in the other direction and leaves you with less delta-V overall.

The conventional wisdom when dealing with multi-engine rockets is to use the low-I_sp engines first so that the high I_sp engines can turn their superior efficiency to moving a lower-mass rocket.  You should consider that LFO, though reducing the amount of delta-V available to the xenon stage, may increase the total delta-V of the rocket if you fly it carefully.  This is especially important if you intend to use the rocket for asteroid mining; you can make LFO while you're there, so aside from a minimum manoeuvring reserve for intercept and capture, you can fly with empty LFO tanks.  You would thus get the most out of your xenon because you'd only use LFO that you make from the asteroid, and that only as you need it.  You should also consider that LFO can be used to run fuel cells, which can keep your xenon engine running in darkness or far away from the sun, so its value is not strictly limited to engine efficiency.

Edited July 25, 2018 by Zhetaan

[farmerben]

I get the Isp error I made earlier.  Thanks for correcting me.

 

[farmerben]

So nobody loves the puff engine or uses it at all.  So the challenge becomes what is the best thing you could do with it?  What advantages could you get?

 

It seems like if there is a use for it at all its on something like my asteroid miner, where you travel quite far but might need a kick when you get there.   Or possibly to land and take off from a small moon, where you only need the extra thrust for a few seconds and then can use xenon.  I'll work on a lander later.  Note that my asteroid miner has a docking port on the back.  Boost stages and crafts coming to refuel on LFO dock there.  I didn't see the point in carrying empty fuel tanks to solar orbit and then back to Kerbin orbit.   In my career mode I've got a number of big orange tanks with spider engines and such that I launched from Kerbin to provide refueling in LKO.  Such things can attach to the asteroid miner when necessary.  It doesn't seem necessary for getting an asteroid.  

The puff is an extremely light engine.  Puff TWR = 22.65.  This is the same as Rhino.  Worse only than vector, mainsail, and mammoth.  And those are all a completely different class. Twitch and Spark are TWR =18.12 and 18.35. 

Really the best thing to compare it to is the Sepatron.  The Sepatron is lighter and has great thrust, but worse Isp and other drawbacks.  So probably any puff mission could be done with Sepatrons or Twitch engines, but there may be a sweet spot in the middle.  

[Snark]

2 hours ago, farmerben said:

So nobody loves the puff engine or uses it at all.

Well, not necessarily nobody, but probably not most people, I'd guess.

Probably a better way of phrasing it than "it's useless" is to say "it's specialized, so is useful mainly if you have a specific situation that favors it."  Such a situation never comes up in my own gameplay, which is why I never use the engine.  About the closest thing to a likely scenario I can think of is what Zhetaan mentions,

3 hours ago, Zhetaan said:

All of this goes to say that the Puff is very much a special-use engine, best suited to situations where you already have a lot of monopropellant and no wish to complicate your rocket with the addition of a different type of fuel, or where you need a radial engine but don't want to deal with fuel pumping (the Puff works like the thrusters in that it pulls fuel from anywhere in the rocket).

 

2 hours ago, farmerben said:

The puff is an extremely light engine.  Puff TWR = 22.65.  This is the same as Rhino.  Worse only than vector, mainsail, and mammoth.  And those are all a completely different class. Twitch and Spark are TWR =18.12 and 18.35.

Yes, its TWR in vacuum is a bit higher.  In my experience, though... the TWR really doesn't matter much in this case.  There are two main scenarios, and I think it matters little in both cases:  orbital operations, and landers.

Orbital doesn't matter much because TWR doesn't really matter at all, unless it's so super low that burns take a really long time.  What matters there is usually dV much more than TWR, in which case Isp is king-- and the Puff has about the worst Isp of any engine in a vacuum, not counting SRBs.

Landers tend not to matter because except for Tylo, vacuum worlds generally have pretty low gravity, so TWR is very forgiving.  The fact that Spark and Twitch have TWR that's 20% lower than the Puff ends up not mattering at all, at least not in my own gameplay, because on lightweight vacuum worlds I generally have plenty of TWR to spare and don't miss it-- whereas I do care about dV a lot, and those engines have better dV.

Where TWR really matters is for heavy-lift engines, e.g. launching from Kerbin... but that's not something you'd use a Puff for anyway because it's truly abysmal in atmosphere.

2 hours ago, farmerben said:

Really the best thing to compare it to is the Sepatron.  The Sepatron is lighter and has great thrust, but worse Isp and other drawbacks.  So probably any puff mission could be done with Sepatrons or Twitch engines, but there may be a sweet spot in the middle.

Well, Sepratrons are kind of apples-and-oranges; they're a single-use SRB that can't be throttled or deactivated once you stage it, and can't be refueled.  It's a little SRB.  Certainly very handy for certain uses, but for the most part I think those uses are different cases from where you'd use a throttle-capable engine like the Puff.

[Freds]

@Snark: I arrived home and made some tests. 
You are right. After some tests and part changes, i realized that i get  lower mass and better efficient with monoprop only because i use MPR-5R Stratus Monopropelent from a mod (dont remember which mod added this).

I apologize for the wrong information i was giving

In the spoiler below there is a screenshot of my sattelite. Ignore the name "Untitled Space Craft", i was testing with different kind of engines and fuels. Its a simple antena + battery + probe + monoprotank, 4 solar panels and 2 scaled down MPR-5R. It was designed to be placed in low orbit and act as an relay between a rover (or something else) in the surface and another, bigger, relay sattelite connected to Kerbin, so, that 695m/s are more than enough for me.

English is not my primary language, so i ask: If you see something wrong (grammatical error, syntax error, etc... ), please correct-me.

Spoiler

 

[Reactordrone]

I've built an entirely monoprop fueled, two stage Munar lander for fun in sandbox but I've never found a reason to use the puff in a career game.

[farmerben]

 

 

Here is a simple proof that there is something to the concept.  Each image has 5 stock parts and a mechjeb.

Image one: TWR = 1.31, dV = 3257

Image two TWR = 1.33, dV = 3262

 

[Snark]

1 hour ago, farmerben said:

Here is a simple proof that there is something to the concept.  Each image has 5 stock parts and a mechjeb.

Image one: TWR = 1.31, dV = 3257

Image two TWR = 1.33, dV = 3262

Well, sure, but that's because you're muddying the waters by including the LV-N in the same stage as the lower-Isp engine, not because the Puff is somehow better than the Spark.  The LV-N's Isp is so much higher than the others that it sees them basically as dead weight. 

By the same token, if you try a third build-- which is simply a lander can, two-ton LF fuel tank, and LV-N, without having the extra fuel tank and engine attached on the side-- then that has a dV of 3283 m/s.  It actually has more dV by getting rid of the extra fuel tank (and engine) entirely.

In other words:  your Puff craft has 5 m/s more dV than the Spark simply because both of them are slowing down the LV-N.  And the Puff one is slowing down the craft by less, because it's got a smaller fuel tank with less fuel in it and is therefore lighter.  Your monoprop tank is 0.1t, full.  The Oscar LFO tank is more than double that mass and therefore slows down the LV-N more.

Not really a valid comparison.  You could have just attached a small 0.05t boat anchor to the LV-N ship instead of the monoprop + Puff, and it would have better dV.  Doesn't mean that the boat anchor makes a better engine.    Just that there's less luggage getting in the way of the nuke.

[ExtremeSquared]

The one scenario in which I find them useful are small jumping rovers on low-grav bodies, especially if there's a fuel depot in orbit or isru nearby. If they are built with rcs blocks for safety anyway, then avoiding LF/Ox can be worth the extra simplicity at the cost of a little dV. One fuel tank instead of two, which is actually worthwhile on small rovers.

[farmerben]

1 hour ago, Snark said:

 

By the same token, if you try a third build-- which is simply a lander can, two-ton LF fuel tank, and LV-N, without having the extra fuel tank and engine attached on the side-- then that has a dV of 3283 m/s.  It actually has more dV by getting rid of the extra fuel tank (and engine) entirely.

 

And a TWR of 1.03

[OHara]

6 hours ago, farmerben said:

Here is a simple proof that there is something to the concept.  Each image has 5 stock parts and a mechjeb.

Image one [using spark]: TWR = 1.31, dV = 3257

Image two [using puff]: TWR = 1.33, dV = 3262

So, one spark and the smallest applicable tank gives 20kN for 31s, about 100m/s for a 6-tonne craft, leaving  0.125 tonnes dead weight
while one puff and its smallest applicable tank gives 20kN for 39s, about 130m/s for a 6-tonne craft, leaving 0.14 tonnes dead weight

I can see why the extra burn-time of the puff might overcome its heavier empty tank, but isn't that just an artifact of what tanks are available ?   If you have 4 tanks on the spark and 3 on the puff -- so each design has the booster burn for 2 minutes -- the spark would come out ahead. 

Also, I don't see how you computed a (slightly) higher TWR with the puff.  Puff and spark have the same thrust, and the build with the puff is (slightly) heavier. 

In real life, monopropellant is good when you need to stop and start a thruster repeatedly.  KSP doesn't simulate the difficulties of restarting combustion engines, so monoprop seems to lose its main advantage.  I think the only 'Puff' engine is only useful if you want a monopropellant-only craft.

[Edit: never mind all those numbers from version 1.3.1, you're using 1.4.x where the monopropellant tanks are much smaller-capacity, and the puff should have even more of a disadvantage.
Are you using the monopropellant in the lander can for your puff test, and leaving it unused in the case of the spark ? ]

Edited July 26, 2018 by OHara
remembered that monopropellant tanks changed in 1.4.1

[Blaarkies]

14 hours ago, farmerben said:

It seems like if there is a use for it at all its on something like my asteroid miner, where you travel quite far but might need a kick when you get there. 

You won't need a "kick" or a higher thrust to rendezvous with an asteroid...just the same thrust as you used to leave kerbin would suffice.