Talk:Specific impulse

From Wikipedia, the free encyclopedia

This article is within the scope of WikiProject Physics, which collaborates on articles related to physics.

Physics Portal

???

This article has not yet received a rating on the assessment scale. [FAQ]

???

This article has not yet received an importance rating within physics.

Help with this template

Please rate this article, and then leave comments here to explain the ratings and/or to identify the strengths and weaknesses of the article.

I'd like to add a concrete explanation for what "seconds" are when dealing with specific impulse. Otherwise, it is quite an abstract unit. Something like this:

In this case I_sp is measured in seconds. While this unit is somewhat abstract, specific impulses measured in seconds can be seen concretely as the length of time the propellant would need to fall in response to Earth's surface gravity in order to attain the same speed at which it leaves the rocket.

However, I'm not sure this is useful or helpful (or even correct). Thoughts, anyone? P3d0 01:12 24 Jul 2003 (UTC)

Yet another way of describing "seconds": "If a rocket engine is tested by throttling it so it just barely hoovers above the Earth, and continuously fed fuel from an external source, specific impulses measured in seconds can be seen concretely as the amount of time it takes to burn a mass of fuel equal to the hoovering mass."

The article currently says:

"if an engine has a specific impulse of 10000 seconds, one kilogram of fuel can produce the same impulse as a one-earth-gravity thrust acting for 10000 seconds."

I'm pretty sure that is not correct. It would be more correct to say "If an engine has a specific impulse of 10 000 seconds, when the spacecraft has a mass of 10 000 kg, one kilogram of fuel can produce the same impulse as a one-earth-gravity thrust acting for 1 seconds." -- DavidCary 06:08, 12 May 2004 (UTC)

Use of seconds as units of specific impulse is the widely used convention. I would not try to invent a new one, the scientific literature certainly will never use it. One occasionally sees "kgf/(kg/sec)" or "kgf s/kg". DonPMitchell 09:18, 1 December 2005 (UTC)

[edit] units specifying thrust

Is there general consensus that specifying thrust in seconds is the right thing to do?

I'd say on balance that seconds are better to use. The most important property is independence of units; people in both the US and europe know what you mean.--Wolfkeeper 12:47, 3 Jun 2004 (UTC)

I'd prefer Ns/kg to the rather misleading seconds unit. At least for those types of propulsion that aren't actually about leaving the atmosphere, I don't see any point of including earth gravity in the equation.

Prumpf 22:55, 11 Nov 2003 (UTC)

Sounds fine to me, both units are equally non-intuitive to my untrained mind. As long as there's a link explaining them I'm happy. :) Bryan 20:35, 15 Feb 2004 (UTC)

See specific impulse for why to include Earth gravity. No really good reason, but not as dumb as it looks. Is this conventional though? --Andrew 04:58, Apr 17, 2004 (UTC)

Very conventional.--Wolfkeeper 12:47, 3 Jun 2004 (UTC)

Is it true that 1 Ns/kg == 1 m/s ?

Yes. 1 newton for 1 second gives an exhaust speed of 1 m/s, from Newton's third law and the definition of the newton.--Wolfkeeper 12:47, 3 Jun 2004 (UTC)

The "thrust", as measured in Ns/kg, numerically equals the average propellant exhaust speed (parallel to the direction of travel), as measured in m/s -- right ? Rocket science is already complicated enough. I'd prefer to use simple units like m/s if possible. -- DavidCary

As long as you include the units everyone should be able to work out what you mean. Americans prefer either seconds or ft/s.--Wolfkeeper 12:47, 3 Jun 2004 (UTC)

[edit] Velocity definition

i like the definition in terms of velocity: I_sp = V_e (m/s) because it intuitively relates to the physics of propulsion. engines with higher specific impulse are able to eject reaction mass at a higher velocity, e.g., ion engines have V_e~30 km/s whereas chemical rockets have V_e~3 km/s.

[edit] Summary

I'd like to ditch the Summary section. This is not an exam study aid; it's an encyclopedia. If nobody objects, I'll chop it and make sure all the info is included in the remaining parts of the article.--P3d0 19:08, Dec 5, 2004 (UTC)

I object. I find this a convenient overview. However, you could copy (perhaps in rephrased and/or expanded form) what is not yet in the rest of the article, so that the summary section actually is a summary.--Patrick 10:23, Dec 15, 2004 (UTC)

Hi Patrick. You and I often seem to disagree about what an article should look like. Personally I don't have any use for a point-form restatement of facts already found elsewhere in the article, especially when there are two dozen bullets, making the section almost as large as the rest of the article (and therefore not really a summary). If we really want a summary, it ought to be in paragraph form. --P3d0 03:24, Dec 17, 2004 (UTC)

The word summary referred to the convenient organized form of concisely listing items (like a table, we could also make it a table), but I changed it to overview, that may be better.--Patrick 09:45, Dec 17, 2004 (UTC)

Yes, I think that's a good change. It highlights that this is not supposed to be a re-statement of the remainder of the article, but is actually a listing of the many ways to interpret specific impulse concepts. I made a couple more changes that make the lists a bit more formal. --P3d0 18:02, Dec 17, 2004 (UTC)

[edit] Way too many parentheses

A hell of a lot of parentheses have crept into this article. I'm going to go rephrase some of them and just remove others... --P3d0 02:10, Dec 22, 2004 (UTC)

Holy cow, this article is getting edited pretty heavily at the moment. I think it needs major revisions, but I don't want to step on any toes, so I'll let the flurry of activity settle down a bit first. --P3d0 02:17, Dec 22, 2004 (UTC)

[edit] Propellent vs Propellant

Either or both spellings are actually completely correct... please don't keep changing it over. Thanks!

The OED lists only "propellant" as a word in its own right, although it does list "propellent" as an alternative spelling. The Merriam-Webster dictionary lists only "propellant" as a headword, with "propellent" as a variant. And so on. I'd say "propellant" is the favored spelling, if only because fewer people will feel the need to change it. --Andrew 02:58, Mar 23, 2005 (UTC)

dictionary.com lists both equally too. In the past I've seen people change stuff the other way too. You can't win.

Yeah, well, every time I read "propellent" I think "What an illiterate", but you'll notice I'm not changing pages either way. If people want to standardize on a weird spelling, well, peace is more important than pedantry. --Andrew 05:18, Mar 23, 2005 (UTC)

Don't you mean "standardise on a weird spelling"? :) The real problem is that there's multiple standards out there... (standardise is the UK spelling...) propellent is correct in the UK as an adjective...

[edit] Definition of specific impulse

I see the phrase: "The specific impulse is a measure of the thrust per unit of propellent". I have two questions:

Should 'per unit' really be stated as 'per unit mass' i.e. rather than 'per unit volume'?

For the usual definition, yes.

Yes AND no, it's actually either per unit mass OR per unit weight.WolfKeeper 22:51, 2005 May 5 (UTC)

Sometimes volume specific impulse is relevant, but it is usually referred to explicitly. Also note that in that context thrust needs to mean total impulse, not force. Evand 22:46, 5 May 2005 (UTC)

per unit volume is called 'density impulse' IRC WolfKeeper 22:51, 2005 May 5 (UTC)

Is it really true that it is 'of propellant' rather than 'of the object'. I would have thought that specific impulse includes the mass of the engine at least. Bobblewik (talk) 14:50, 5 May 2005 (UTC)

Nope, it's definitely of propellant, not including engine and other dry mass; the reasons should be clear from a careful examination of the rocket equation. Also consider the case of changing the tank size / engine size ratio; Isp doesn't change, though thrust and burn time do. If you want "of an object," the usual term is delta-v. Evand 22:46, 5 May 2005 (UTC)

[edit] Specific impulse units in articles

There seems to be some disagreement on the units that should be used for specific impulse in various articles (mostly realted to spacecraft propulsion, naturally enough). It'd be nice to come to a consensus we could point people at if they start changing them around.

The possibilities seem to be:

seconds
meters per second
composite units: kN·s/kg
composite units: lbf·s/lb

Personally I think the last is an abomination and should be stricken from articles without mercy, but I'm from a quasi-metric country. I don't like either kind of composite unit. I would prefer m/s, as I find it the easiest to work with and think about. I would also be comfortable with giving both s and m/s, because it seems people who actually do rocketry use seconds. I'd be unhappy with seconds only, but it might be possible to convince me that's the best choice.

Of course, this article needs all of them, explanations on how to convert, and so on. --Andrew 02:56, May 6, 2005 (UTC)

My position is that

specific impulse is *always* measured in seconds (special mention to the H-1 article that accidentally used minutes for that and the amazing invention of 260 lbf-min/lb- no, the H-1 engine is not SSTO, alas!)
Ve is not Isp (ft/s m/s)
'c' is not Isp (ft/s m/s or Ns/kg lbfs/lb)
c* isn't Isp either (ft/s m/s Ns/kg lbf/lb)

I asked some bonafide rocket engineers, and they agreed. I also cracked the books; them too.

I intend to amend Specific Impulse appropriately.WolfKeeper 03:10, 2005 May 6 (UTC)

I have now removed the lbf.s/lb abomination from everwhere in the wikipedia with extreme prejudice. I may have missed some though; if so please let me know where. The only place it is at all valid is if the manufacturer quoted it, and even then conversion into sane units should be given. I'm actually not aware of any usage like that.WolfKeeper 05:55, 29 January 2006 (UTC)

You have done so sticking bull-headedly to all the false statements in your original 6 May 2005 posting, despite ample evidence of your errors below on this talk page. Gene Nygaard 00:12, 31 January 2006 (UTC)

Oh yeah, the lbf/lb thing. The problem is that lb predate even Newton, they were originally a unit of *weight* i.e. force, because they were used on pan balances in London. When Newton did his stuff they didn't create a new unit for mass, it's just lb. I have a lot of sympathy for lbf, actually even more sympathy for anyone using imperial units, since they are very broken here. The lbf is quite a clever idea, but nobody uses it for anything very much, particularly specific impulse. And to cap it all lbf/lb just looks too horrible.

WolfKeeper 03:10, 2005 May 6 (UTC)

Good grief, Wolfkeeper. Go read pound. Pounds have always been units of mass. Since 1893 in the United States we have not had any independent standards for a pound; they have been defined as an exact fraction of a kilogram. The exact number changed in 1959, when the whole world agreed on a common definition of the pound as 0.45359237 kg exactly.

P.S. That ought to tell you something about one meaning of the ambiguous word "weight." Let me just try to make this point more explicit. Note specifically that the net weight on billions of items on the grocery stores and pantry shelves and hardware stores and work benches in the United states are almost always given in both kilograms, and those pounds which are every bit as much units of mass as the kilograms are, since they are an exact fraction of a kilogram. Why in the world do you suppose the law bothers defining a pound in the first place? The law doesn't really give a damn what you use to measure specific impulse.

It is the pound force which is the recent bastardization. These units were never well defined units before the 20th century. Even today, they don't have an "official" definiton.

The only reason we ever get specific impulse in those pseudo-seconds is because early rocket scientists were sloppy, and failed to distinguish force units from mass units. The U.S. rocket scientists divided thrust in pounds (actually pounds-force) by mass flow rate in pounds per second, then just canceled out pounds with pounds leaving seconds. Of course, the German and Russian rocket scientists weren't going to be outdone by the Americans, just because they used metric units. They may even have been the first to use these "seconds". They divided thrust in kilograms (actually kilograms-force) by mass flow rate in kilograms per second, then just canceled out kilograms with kilograms leaving seconds. What they really had were lbf·s/lb or kgf·s/kg, but they used the shorthand version, calling them "seconds" (even though most of them knew all along what was going on).

See the NASA drawing, Image:H-1 rocket engine diagram.jpg -- Gene Nygaard 14:25, 6 May 2005 (UTC)

Of course, in the modern metric system, there are no kilograms force. So we can now avoid that problem.

None of the early formulas included any g_n factor; that isn't something that belongs there because of the physics. In fact, this isn't even a physics concept; it is strictly a concept of metrology, something used to define grams force or pounds force. Note that the actual, local acceleration of gravity is alsways totally irrelevant to these calculations of specific impulse. This fudge factor is only something thrown in by a later cleanup crew, to make the dimensional analysis work for those "seconds" units. Gene Nygaard 05:24, 6 May 2005 (UTC)

Both Andrew and WolfKeeper have complained about a supposed "ugliness" of lbf/lb, but that's what they are, and it is one of the simplist ways to express this. The simple fact is that there are indeed two different units of measure involved here. It is important to distinguish pounds force from normal pounds as units of mass. Some even distinguish "lbm" as well, but that is not recommended by standards agencies such as NPL and NIST, who use "lb" and "lbf" for this purpose. It could be expressed in units of acceleration, of course; say 32.16 ft/s², or 9.80665/0.3048 ft/s² (or the equivalent of the latter, 9.80665 m/s²) or whateer you choose to use to define pounds force. But then you need a number in addition to the units. Imagine how much clinkier those specific impulse numbers would look if that multiplicative factor were explicitl included. Gene Nygaard 14:25, 6 May 2005 (UTC)

[edit] Specific impulse units (discussion moved from user page to 'specific impulse' article talk page)

Measurements of specific impulse in seconds happens to looks far cleaner then measurements in lbf·s/lb or kN·s/kg, not to mention that it saves space, especially when units are placed in tables. There is also no requirement that lbf·s/lb or kN·s/kg need to supplement m/s or sec. Your original needles alteration distorted the lay out of the [1] greatly and had to be corrected, even so double line spaces in a table are also not aesthetically pleasing. People are not going to read articles that they find excessively complicated in layout. --BerserkerBen 19:58, 30 Apr 2005 (UTC)

Upon further review it has been found that Isp stated in seconds could either be stating force in metric or empirical, these means that any conversion of Isp in second could be erroneous because you can only guess what the units for force originally were. This means that many of your Isp conversions could be erroneous. I implore you to check your conversions and verify them with sources! --BerserkerBen 04:43, 2 May 2005 (UTC)

What's your point? Kilograms-force are not part of the modern metric system, but 1 kgf·s/kg ≡ 1 lbf·s/lb. (At least it is exact if you use the same acceleration to define pounds-force as is official for defining kilograms-force, the conventional way it is done now. But even if you don't, the difference is less than the precision to which thrust measurements are made, and 1.0000 kgf·s/kg = 1.0004 lbf·s/lb even if, for example, you use 32.16 ft/s² to define pounds force.)

In other words, there is absolutely no need to "guess what the units for force originally were."

The "metric" units of thrust, if you are talking about the modern metric system (i.e., SI), are newtons. The corresponding units of specific impulse are newton-seconds per kilogram. Note especially that, since there are no units of acceleration called gees in the International System of Units, measuring specific impulse in pseudoseconds is useless as tits on a boar if you stick to SI. Gene Nygaard 11:48, 2 May 2005 (UTC)

Are the seconds measured in Newton or kgf?, there you get an error of ~2%. Also that still does not mean you need to complicate an already complicated table with elongated units, there are ways of saving space and still stick to standards.--BerserkerBen 12:12, 2 May 2005 (UTC)

What do you mean? That doesn't make any sense whatsoever. Explain the calculations which you think would give you an "error of ~2%". (I made a minor adjustment in my numbers above, 1.0000 to 1.0004) Gene Nygaard 12:45, 2 May 2005 (UTC)

I've checked the books, I've checked with rocket engineers, and they all say the same things. The imperial unit of specific impulse is seconds. The metric unit of specific impulse is seconds. Specific impulses in units of m/s, ft/s or Ns/Kg are actually measurements of 'c' the effective exhaust speed; this is a useful measurement, but it isn't Isp. Please desist in removing Isp from the wikipedia. WolfKeeper 00:55, 2005 May 6 (UTC)

Those seconds are a "metric" unit only when your "metric system" includes kilograms force. SI, the modern metric system, does not include kilograms force. Gene Nygaard 13:59, 6 May 2005 (UTC)

Sorry. You are just incorrect here Gene. The metric unit of specific impulse *is* the second. I checked. Extensively. There is a figure called 'c' which is of the units you give, but that's not specific impulse, it's effective velocity. Incidentally Ns/kg is exactly equivalent to m/s due to F=Ma...

Changing your story now, claiming "metric" rather than "SI". Note that the Russian space program routinely measured rocket thrust using the kilogram-force until the very late 1980s or into the 1990s. But kilogram-force is not a part SI. Note also that the Chinese space program today, and perhaps the European Space Agency as well, still measure rocket-thrust in those non-SI kilograms force, even today.

In a kilogram and kilogram-force system of units, shortening kgf·s/kg to seconds makes just as much sense as shortening lbf·s/lb to seconds does—and it even gives you the same number.

Is c* a more common way of writing that c?

c* is the exhaust velocity at the throat.WolfKeeper

Isn't that "characteristic velocity" c* sometimes considered to be the same as effective exhaust velocity, but more properly that effective exhaust velocity multiplied by some efficiency factor?

Only a moron would think that they are basically the same. You don't think that, do you Nygaard?WolfKeeper

It is v_e and not c* in the latter meaning which is effectively the same thing as specific impulse, and usually expressed in units of velocity. Gene Nygaard 11:17, 8 May 2005 (UTC)

Also 'c'.WolfKeeper

And as a matter of record, I think removing data/units is a form of vandalism (adding conversions as well is fine though). I found the H-1 rocket engine page had a specific impulse listed as multiple thousand lbf-min/lb, when the original data specific impulse from the manufacturer was 260 min- and even *that* was wrong it should have been 260 *seconds*. It had been converted multiple times... this is *very* bad. Destroying data by conversions, even from rounding errors should be a total no-no. If you want to *add* metric or imperial conversions, sure. Removing or changing data???? It's just never right to do that, unless you have a clear reference that it is wrong. WolfKeeper

You are having the same reading comprehension problems that the earlier editor of H-1 (rocket engine) had. See the NASA image at Image:H-1 rocket engine diagram.jpg. That is the source of your "minutes". Your confusion is understandable; it is not clearly written. Here is the relevant line:

"SPECIFIC IMPULSE (LB-SEC/LB) 260.5 MIN 261.0 MIN

My confusion? Are you saying that the ISP really is "15630 lbf.min/lb (153 kN·s/kg)" like the article I corrected??? An exhaust velocity of 153000 m/s????? :-)

Still no comprehension after you read the explanation below, Wolfkeeper? That "MIN" is "minimum", not "minutes". It really is that simple.

If it was that simple explain how it ended up as "15630 lbf.min/lb (153 kN·s/kg)". Quite frankly, I don't give a damn, it was an excellent example of why screwing around with units in the way you guys are doing is wrong.

Or are you just so bull-headed that you can't even go back and change your silly comments, even when you are the only one who has seen them so far, after you saw the explanation below.

Pot, kettle.WolfKeeper

I'm saying that the I_sp is actually "a minimum of" 260.5 lbf·s/lb (2,554 N·s/kg), which can equivalently be expressed, just as accurately, as an effective exhaust velocity of 2,554 m/s (8,381 ft/s). Gene Nygaard 16:37, 8 May 2005 (UTC)

In other words, it clearly states that the units used for specific impulse are lbf·s/lb, not lbf·min/lb. So how does NASA get those units, rather than "seconds". Simple. The very same way that you get your "seconds", and with the very same number as the result. There just is no pretense that the pounds in the two parts are the same unit, though that isn't made very clear when they use LB for both of them.

Even when it was written by NASA, these were still the wrong units. WolfKeeper

They were the correct units; even NASA isn't wrong all of the time. Of course, if NASA ever starts listening to the recommendations of their own Inspector General, they won't use those units any more, but instead will use N·s/kg. Gene Nygaard 16:37, 8 May 2005 (UTC)

So what does "MIN" mean here? It means minimum, not minutes. The numbers to which you changed this, being just slightly higher, are above this minimum. Gene Nygaard 11:17, 8 May 2005 (UTC)

Nevertherless, the correct units are seconds, not lb.sec/lb. The formula for defining Isp includes 'g'.

It never includes an unadorned g. As the article here explains, sometimes it does include a g₀ (more conventional modern notation g_n) factor, and sometimes it does not. It works either way; that's only a unit conversion factor, basically.

The correct units for specific impulse are seconds in both imperial and SI units. (See Rocket Propulsion Elements by George P. Sutton and Oscar Biblarz, seventh edition et al). Therefore no conversion need be employed.

So where's your quote from them making that claim, Mr. or Ms. Anonymous? Do they use all SI units? No SI units? A mixture of SI and non-SI units?

Seconds are most definitely not units of specific impulse in SI. The SI units of specific impulse are newton seconds per kilogram. http://gltrs.grc.nasa.gov/reports/1996/TP-3576.pdf

"Isp,V vacuum specific impulse, N-s/kg (lbf-s/lbm)"

Note in particular that there are no units of force called kilograms force is SI, and since there are also no units of acceleration called "gees" in SI, those numbers in "seconds" are totally useless if you stick to SI.

The way it got to be "seconds" in the first place was in sloppiness in the use of units, failing to distinguish units of force from units of mass. In the old days, none of the formulas ever included a gn factor (which has nothing to do with the local acceleration of gravity); that isn't something that belongs there because of the physics, and when you see that in the formulas it is only something thrown in a later time by some cleanup crew, trying to make the dimensional analysis work.

Note that this this is a somewhat arbitrary standard acceleration of gravity. This is a concept of metrology, not of physics. It's only purpose is to define units such as kilograms force. Gene Nygaard 04:47, 6 May 2005 (UTC)

Sometimes specific impulse is incorrectly, but not unusually listed as a velocity or as Ns/kg. This is in fact a different quantity known as 'c', the effective exhaust velocity. This can take the units m/s (or equivalently, from Newton's third law Ns/kg) or ft/s. Its use is deprecated, but may be employed if quoted by a manufacturer, or added in addition to the specific impulse.

It is not a different quantity at all, and I've never seen anybody deprecate its use, even people who don't use it very often. There certainly isn't any standards organization which has deprecated the use of effective exhaust velocity, is there?

There is only one quantity being measured here. There is always a constant conversion factor between the quantity called "specific impulse" and the quantity called "effective exhaust velocity."

When you use FFU such as the kilogram-force or pound-force and measure specific impulse in "seconds", the numbers for these two quantities are different. But you can multiply those "seconds" by a constant to get effective exhaugst velocity in meters per second, or a different constant to get feet per second. You need absolutely no additional information to convert from one to the other, because there is only one quantity being measured here.

In SI, there is also a constant conversion factor, but the constant is one. Specific impulse in newton seconds per kilogram is not merely dimensionally equal to effective exhaust velocity in meters per second, but it is numerically equal as well.

Because the fools who can't tell a pound force from a pound butchered everything up so bad, many people using SI units simply avoid using the term specific impulse at all, with any units. They just use effective exhaust veloctity instead. You don't need both, especially since they are numerically identical in SI units. Gene Nygaard 04:47, 6 May 2005 (UTC)

Note further that there are other purists out there who do not use those pseudoseconds, though they do use the same numbers. They simply use them with the proper units, kgf·s/kg or lbf·s/lb. Note that both of these are numerically identical to your "seconds". The only difference is that you don't need to throw a fudge factor into the formulas to make the dimensional analysis work. Gene Nygaard 04:47, 6 May 2005 (UTC)

Note further that you could use the absolute fps system, and then specific impulse in poundal-seconds per pound would be both dimentionally and numerically equal to effective exhaust velocity in feet per second. Or, you could use the gravitational fps system, and then specific impulse in pound-force-seconds per slug would be both dimentionally and numerically equal to effective exhaust velocity in feet per second. In other words, 1 pdl·s/lb = 1 lbf·s/slug = 1 ft/s. Gene Nygaard 04:47, 6 May 2005 (UTC)

Good god people! I brought this up here to try to avoid edit wars - please don't go changing units until a consensus is reached here. Society will not collapse if we leave articles using the "wrong" units (even if they're furlongs per fortnight).

At this point both of you have made claims about what "everybody" uses. I think it's time to start listing web references. I'll start. A casual Googling reveals:

Specific impulse at NASA, though unreadable, gives an "effective velocity" Ve, and gives specific impulse as Ve/g0, where g0 is 9.8 m/s^2. One advantage is said to be that it doesn't matter whether you use imperial or metric.
Rocket Engine Specific Impulse Program from the Air Force has the comment "The program is not written to use SI units, and uses a grab-bag of American, pre-SI metric, and SI units. For example, energy is in calories rather than joules, temperature is Kelvin, and pressures are in psi." It goes on to say "The output of the program is the traditional rocket engineer's "Isp", or specific impulse. The units of Isp are seconds, and due to the way that a pound of force and a pound of mass are defined in the US. measurement system, the unit includes the reciprocal of the acceleration of gravity at the earth's surface. To convert Isp into SI units, multiply Isp by the standard acceleration of gravity, 9.807 m/s^2, to get exhaust velocity in m/s."
The Encyclopedia of Astrobiology, Astronomy, and Spaceflight by David Darling in its very brief entry on specific impulse says "Measured in seconds, specific impulse is an important gauge of the efficiency of a rocket propulsion system [...]".
How do you calculate specific impulse? on Northwestern's page about Deep Space 1 states "The specific impulse is: Isp = ueq/ge, where Isp = specific impulse, ueq = total impulse / mass of expelled propellant, ge = acceleration at Earth's surface (9.8 m/s2)".
A post by Henry Spencer to sci.space.tech in which he explains why he thinks specific impulse in seconds is used instead of effective exhaust velocity in meters per second.
MIT's Space Propulsion Laboratory has this to say: "Why seconds? Well, it is a convention. You could well multiply by earth's gravity constant (g = 9.8 m/s^2) and have the specific impulse (or Isp, as rocket scientists love to call it) units converted into m/s, or speed."

So, of these sources (which vary greatly in authoritativeness), none use a unit of the type lbf·s/lb (even the one that admits to using a mishmash of imperial, SI, and other units). All say that specific impulse is measured in seconds; a few suggest, not that specific impulse is measured in m/s, but that people wanting to use SI may want to use effective exhaust velocity in m/s. Note that there seems to be some disagreement on the conversion factor to be used! Some say 9.8, others 9.807...

That aligns exactly with my position.WolfKeeper

I did briefly wonder whether it is 'more right', or not, to add an effective exhaust velocity only in metric to an Isp specified in seconds, but there's no *requirement* that we use anything other than metric in the Wikipedia, but with use of imperial we are encouraged to have a conversion to metric. On the contrary, the fact that the metric version of Isp is the same as imperial suggests we shouldn't even list the effective exhaust velocity, but it doesn't seem to me actually wrong to have it (although it will tend to confuse people into thinking they are the same, so it seems preferable to identify it if you do add it).WolfKeeper

I am certainly not a rocket scientist, but it is the effective exhaust velocity that I find useful, for example in the Tsiolkovsky rocket equation, so I would like to have it (with, of course, adequate clarification) on most articles that talk about specific impulse. Of course, for here, we need to clarify what the correct conversion factor is! I'm really unsettled that there is no obvious standard. Maybe people don't worry too much about the third decimal place of a specific impulse? --Andrew 18:44, May 6, 2005 (UTC)

As a hobbyist rocket scientist, people only sometimes worry about the third decimal place, and almost never about the fourh. The reason is that theoretical numbers only get you so far; third and fourth decimal places will get overwhelmed in actual practice by combustion inefficiencies, start and stop transients, inexact mix ratios, tank ullage, etc etc. You really only need a third decimal place for high-efficiency engineering, and anyone doing that today will be calculating all their own numbers via a piece of software, and therefore doesn't care what conversion factor other people use. (by User:Evand)

So, in spite of my preference, evidence seems to say that specific impulses should be measured in seconds. My personal preference would then be to include an effective exhaust velocity in m/s as well. Further, I would explain the whole story of units in this article.

Of course, I don't think any of this should be done until consensus is reached. --Andrew 17:03, May 6, 2005 (UTC)

Pay attention to this article itself.

Yes, the article is actually incorrect; and it's not an uncommon mistake, and we also need to correct it, and explain that it is a common mistake.WolfKeeper

The article is not incorrect. If anything is incorrect, it is the use of "seconds" for specific impulse. Unfortunately, because of Wikipedia NPOV provisions, we are probably stuck with the incorrect but widely used versions, as well as the correct specific impulse in N·s/kg.

Just to show you that this is not something recently dreamed up by a few strange editors of Wikipedia, like me, here is something published by NASA 27 years ago:

Sloop, John L., Liquid Hydrogen as a Propulsion Fuel, 1945-1959. NASA SP-4404, 1978.

from Appendix B. PROPULSION PRIMER, PERFORMANCE PARAMETERS, AND UNITS. [2] (italics in original, bold emphasis added by Gene Nygaard)

"Specific impulse and exhaust velocity. In 1903 Tsiolkovskiy, and other Europeans after him, expressed rocket engine performance in terms of the velocity of the exhaust emerging from the nozzle in meters per second (m/s). This made sense because the rocket exhaust velocity was a term in the equation expressing the velocity of a rocket-propelled vehicle. In the United States, it became the custom to express rocket performance in terms of the measured quantities: thrust and mass flow of the propellants. The thrust divided by the total mass flow of propellant was defined as the specific impulse. Specific impulse is the inverse of specific fuel consumption used in discussing the performance of other types of propulsion systems. In English units, specific impulse is in pounds force per pounds mass per second (Ibf ^. sec/Ibm)[sic, likely an OCR error in converting to html]. On seeing pounds in both numerator and denominator, many succumbed to the temptation to cancel them and express specific impulse incorrectly in units of seconds: the two pounds represent different physical phenomena, force and mass, and are connected by the conversion factor 32.2 Ibm ^. ft /lbm-sec2. In SI, specific impulse is expressed in newtons per kilogram per second or N ^. s/kg. English values of specific impulse are converted to SI by multiplying by 9.807, which can be rounded to 10 for approximations. The numerical value of specific impulse and exhaust velocity in SI are the same; only the units are different. Since exhaust velocity is a simple concept to visualize physically and since specific impulse expressed in newtons per kilogram per second is unfamiliar to many, including the author, all performance values in this text have been converted to exhaust velocity in meters/second (m/s). Typical values of exhaust velocity for liquid propellant rockets range from 2000 to 4500 m/s. The V-2 had an exhaust velocity of about 2200 m/s, very good for 1944. High energy propellants give exhaust velocities in the range of 3000 to 4500 m/s, and the liquid hydrogen-oxygen combination is in the upper part of this range."

Gene Nygaard 22:53, 8 May 2005 (UTC)

It tells you that these are merely two different ways of looking at it. It just depends on whether you use:

$\mathrm{Thrust}=I_{\rm sp} \cdot \frac{dm} {dt} \,$

or whether you gratuitously throw in a conversion factor not called for by the physics, as is often done,

$\mathrm{Thrust}=I_{\rm sp} \cdot \frac{dm} {dt} \cdot g_{\rm 0} \,$

The former gives you specific impulse in units of N·s/kg or lbf·s/lb or kgf·s/kg, in units which are actually used for this purpose. Of course, you could also use pdl·s/lb, or lbf·s/slug, or kgf·s/hyl—but I never see anybody using any of those units.

The second one which includes a standard acceleration of gravity factor (a concept of metrology, not of physics), lets us use a "standard weight" fiction, so that we can treat a measurement of mass as if it were a measurement of force, and gives you specific impulse in "seconds".

Note also that the word "specific" in phrases such as this normally means "divided by mass". In SI, the units of impulse are newton seconds, the units which were supposed to have been used in the Mars-Climate-Orbiter-turned-Crash-Lander. If you divide that impulse by mass in kilograms, you get specific impulse in N·s/kg. Gene Nygaard 11:46, 8 May 2005 (UTC)

Here are some more references for you, showing quite clearly that this is not something that somebody just dreamed up to impose on Wikipedia, coming from the clear, blue sky:

NASA Technical Paper 3576, June 1996

I_sp,V vacuum specific impulse, N-s/kg (lbf-s/lbm)

http://www.bernd-leitenberger.de/einstuf.html

Ersetzt man den Sauerstoff durch Fluor so erhält man eine Kombination deren spezifischer Impuls um etwa 100-150 N*s/kg größer als die von Sauerstoff / Wasserstoff ist und deren Dichte ebenfalls etwas höher ist , so wird die Rakete leichter und transportiert mehr Nutzlast.

http://spacearium.aresinstitute.org/article.php?story=20040518175050117&mode=print

The design of the H-1800 had an average vacuum thrust of 1.003 kN, and average vacuum specific impulse of 2724 N*s/kg.

http://www.spaceproducts.com.cn/webclass1.asp?;modelname=sat_equip_nr&FractionNo=&titleno=products&recno=19

Specific Impulse, Vacuum: 2000-1850 Ns/kg

http://mek.kosmo.cz/zaklady/rakety/motory.htm

Výhody: vysoký specifický impuls (2500-4000N.s/kg) a tah, možnost rízení velikosti tahu, možnost restartu

http://www.fandom.sk/clanok.php?clanok_id=1738

Typickým príkladem je projekt velké vesmírné lode VISTA, kde by bylo „spalováno“ 30 D-T kapslí za sekundu a dosahováno tak Isp 170 000 N.s.kg-1.

http://www.sondasespaciales.com/modules.php?name=Sections&op=printpage&artid=10003

Impulso Específico. El impulso específico de un motor es el impulso (fuerza aplicada durante un cierto tiempo) ejercido con un kilo de propelente. La unidad para el Impulso Específico es el Newton por segundo por kilogramo (Ns/kg). De tal manera, mientras mayor sea el impulso específico de un motor, mayor será el rendimiento y menos combustible requerirá. El valor numérico del Impulso Específico corresponde a la velocidad de escape de los gases en el vacio (en metros por segundo).

http://www.astronautix.com/engines/rd858.htm

spec. vac. impulse

315 s
3089 Ns/kg

285 s
2795 Ns/kg

312 s
3060 Ns/kg

http://www.apogeerockets.com/education/newsletter46.asp

SI is specific impulse in Ns/kg (you probably measure it in sec, so you must multiply by "g", typical aluminised composite SI=2500 Ns/kg
Blackpowder SI = 735 Ns/kg

http://www.fas.org/spp/guide/china/launch/lm2c/2C_Chapter2.htm

Specific Impulse (N· s/kg)

2556.5
(On ground)

2922.37(main)
2834.11(verniers)
(In vacuum)

2804 (solid motor)

There are lots more where these came from. In particular, there are many more in English; I wanted to show, however, that this is truly a use of the International System of Units, and that it is not something just picked up from Wikipedia. Gene Nygaard 17:43, 8 May 2005 (UTC)

Is english itself logical? No, it is a total mongrel of a language, quite inconsistent; there are actually competitions to see if anyone actually knows how to spell in english and essentially nobody does. But essentially everyone uses it, at the moment it's the nearest thing to an international language. Similarly, ISP is, atleast 80% of the time and it seems to me to be growing more frequent, specified in seconds. What makes you think you are important enough, or even clever enough, to specify a different standard than 80% of the rocket engineers? Everybody speaks Isp in seconds, everybody understands it, nearly all datasheets are specified in it, nearly all textbooks do it that way; a textbook that doesn't is less useful. Encyclopedias describe the world, not try to force it into something it isn't. It doesn't have to make sense, it's the standard. Learn to deal with it. WolfKeeper

I can see two issues:

1. Physics. Where specific impulse = force*time/mass, there are only a few options for logical terms.
2. Illogical terms. As Wolfkeeper says, people will encounter illogical terms.

If I have understood it correctly, then I am sure that we can revise the text to explain both those issues. I think the reference that begins In 1903 Tsiolkovskiy... explains it plausibly. Bobblewik (talk) 14:27, 9 May 2005 (UTC)

[edit] examples table

It seems to me the examples table is both wrong and out of context. Where do the numbers for specific impulse come from? No modern solid rocket in use is as poor as 100s,

Yes SSRBs for example do much better than that.

no bipropellant ever actually used is as good as 500s.

The energy figure for a jet engine is entirely off base -- since the actual exhaust velocity is low, energy consumption is low, despite the high effective exhaust velocity. By including non-reactive air in the exhaust mass, the actual exhaust velocity goes down, the thrust goes up, energy consumed per unit impulse goes down, and effective exhaust and specific impulse go up. Evand 20:19, 6 May 2005 (UTC)

Actually, rockets carry both oxidiser and fuel in their tanks, which reduces the energy per unit mass of fuel. So jet fuel is actually more energetic per kg. Since the energy is per unit propellant, then jets come out high. The fact that jet engines are more efficient is to an extent reflected in the Isp. WolfKeeper

The energy density of aviation fuel is about 42.8-42.6 MJ/kg (http://hypertextbook.com/facts/2003/EvelynGofman.shtml). Feel free to check the maths. Impressively high huh? (and a 747 can carry much more fuel than 50 tonnes BTW) WolfKeeper

What is the basis for the mass / total energy numbers, anyway? Are the different lines intended to be comparable? Are they intended to be examples of specific rockets / airplanes? I think this table, as it currently is, is confusing and detracts from the article. Evand 15:41, 9 March 2006 (UTC)

[edit] minor changes in nomenclature

The text as written had some internal inconsistencies; I made a couple of (hopefully minor) changes to make it consistent with itself; for example, "c" was used for effective exhaust velocity in one place, while both V_e and I_sp were used elsewhere.

The subject of whether seconds or meters/second is the "proper" unit for specific impulse is a touchy one among rocket scientists and engineers; everybody more or less agrees that seconds is an awkward holdover from English units, Unfortunately, you can't define the term in both units, or the equations would be inconsistent. I tried to make the page consistent in the use of v_e for specific impulse in unit of meters/sec, and I_sp in units of seconds. (before correction, the equations on the page were inconsistent with each other, and combining any equation from one part with any equation from the other part would lead to the mathematically absurd equation "9.8=1").

I also changed the word "Thrust" in the equations to F_thrust; this is more consistent with common practice.

Thanks for weighing in here, Geoff. Good improvements. By the way, you can sign comments with the name/date block you see in common use just by typing four ~ characters in a row with no spaces at the end of your comment - the Wiki software expands it automatically. Georgewilliamherbert 20:42, 28 January 2006 (UTC)

Can I just say that at the moment the first equation on specific impulse makes no sense to me in terms of units. Only by trawling through the discussion page has any of it made any sense. Could someone with better explanation skills and understanding please explain what units each quantity can be in (e.g. is thrust in Newtons as this would be the expected SI unit?), and why it's a screwy formula. Thanks. GBM 14:58, 5 April 2006 (UTC)

I've added a couple units. Does that help? Or were you looking for something more? (There's also the difference between Isp in seconds and effective exhaust velocity Ve in m/s; does that need more explaining somewhere?) Evand 15:07, 5 April 2006 (UTC)

It's a lot better now it's got some units, now it's much more usable without worrying about them. I think the difference between exhaust velocity in m/s and Isp in s is already explained OK. Thanks! GBM 12:18, 6 April 2006 (UTC)

[edit] question about propulsion

IANARS but in the section on "Interpretations" should this "A spacecraft without propulsion follows an orbit determined by the gravitational field" actually read "A spacecraft without propulsion follows a course determined by its own momentum and the gravitational field" ? If this is a really dumb suggestion, my apologies in advance ! GeraldH 10:01, 10 April 2006 (UTC)

[edit] Weight vs. Mass

I do know that weight and mass are different. That does not justify this revert. If one kilo of propellant can produce a thrust of 9.8 newtons for 1 second, then one newton of propellant can produce a thrust of one newton for 1 second. Thus, it doesn't matter if you use weight or mass here. --P3d0 22:37, 21 April 2006 (UTC)

Well, you definitely need to include the sea level part, since weight varies with altitude. And the concept of a "newton of propellant" is very odd indeed.WolfKeeper 23:04, 21 April 2006 (UTC)

Given that specific impulse is a measure commonly used for engines used in space, where the acceleration of gravity is not constant (as opposed to the surface of the Earth, where it effectively is), the distinction between weight and mass is extremely important. "One newton of propellant" is not the same amount in different locations. siafu 00:27, 22 April 2006 (UTC)

It's not hard to say that the "weight" is as measured under a gravity of one G. Whatever we do, I'm sure we can do better than "the force equal to the unit mass of fuel at the standard acceleration of gravity". --P3d0 19:42, 22 April 2006 (UTC)

But you don't really have "one newton" of propellant. You have a mass of propellant that would exert a hypothetical, imaginary amount of force if it were at some location (e.g., somewhere near 45.5° latitude sea level on Earth) where the local acceleration of gravity is 9.80665 m/s².

So in this case, if you are using the formulation with that extraeous g_n factor, something not called for by the physics, even if you call this hypothetical "standard Earth weight", it is the same number no matter where the object whose specific impulse is being calculated is located.

Of course, it also has the same mass, no matter where the object whose specific impulse is being calculated is located.

So specific impulse in the proper SI units (N·s/kg, or the equivalent m/s) are always related to specific impulse in "seconds" by the very same, constant, exact factor.

So it doesn't really matter if you actually use mass, or use some pretend amount of force that that mass would exert, at some place that isn't necessarily where it actually is.

Of course, that "sea level" part is extremely misleading; it varies more with latitude on Earth than it does with altitude on earth. Even if you limit yourself to sea level on Earth, the local g varies by 0.53%; that is, more than one part in 190. Gene Nygaard 03:03, 22 April 2006 (UTC)

But even if you throw in Mt. Chimborazo, the highest mountain on Earth in both ways relevant to this discussion, the overall variation in g at all points on the surface of the Earth is only a little bit more, about 0.72% or one part in 140. Gene Nygaard 03:06, 22 April 2006 (UTC)

BTW, the average sea-level acceleration of gravity[3] on Earth is about 9.79764 m/s², not 9.80665 m/s².

Hey guys, this is starting to look pretty good. I'm about to tweak it again. I think we're starting to converge. Thanks for your help! --P3d0 02:27, 23 April 2006 (UTC)

[edit] Poundals versus lbf

I've put the F as newtons or poundals. It's not lbf in this equation. But if you divide both sides by g0; then a force in poundals/g0 = lbf, so it works out as you would expect, well kinda.

WolfKeeper 04:42, 22 April 2006 (UTC)

Oh, good grief, WolfKeeper! I know what a poundal is, and might use it myself, but who else uses it in this context?

Everybody in the context of that particular equation.WolfKeeper 16:03, 22 April 2006 (UTC)

Google: poundals site:nasa.gov

One hit to a list of conversion factors (and if you show similar ones, a broken link to a similar one)

One pdf file of images of a report printed in 1920 dealing with "Airplane Stress Analysis", with one measurement in the 74 page report expressed as "P_y, =2,221 poundals, or 69.4 pounds of force."

That's it. Gene Nygaard 18:44, 22 April 2006 (UTC)

To be fair, by using the singular "poundal" instead, Google shows us 9 of 11 hits: One a comment posted on a forum, and the other 8 all either glossary entries or lists of conversion factors. Gene Nygaard 18:50, 22 April 2006 (UTC)

What ever happened to all your professed concerns about "what the manufacturers use"? I haven't seen anybody use poundals for much of anything in the last half-century ago—pretty much going back before the start of the space age.

Of course, that formula will work just fine with pounds-force. All you need to do is to choose an appropriate unit for either mass or acceleration or both.

Uh no, not as is. Not if your mass in pounds. That's a different form of the equation.WolfKeeper 16:03, 22 April 2006 (UTC)

For example, what is that thrust using a specific impulse of 270 seconds and a mass flow rate of 0.29 slinch per second (the base unit of mass in the inch-pound-second gravitational system of units, lbf·s²/in, are often used without any special name, but some NASA engineers call them slinches). Then use a g₀ = 386.1 in/s². What if the force, using your formula?

Answer: (270 s)((0.29 lbf·s²/in)/s)(386.1 in/s²) = 30,231.63 lbf (before rounding)

NASA engineers, for example, are not only more likely to use slugs (fps gravitational system) than poundals (fps absolute system), but they are also more likely to use slinches (ips gravitational system) than poundals. Gene Nygaard 15:36, 22 April 2006 (UTC)

Of course, if for the same example you express the mass flow rate as 112 lb/s and the specific impulse in the alternative formulation as 270 lbf·s/lb, then the calculation to get thrust in pounds force is really quite simple with the formula appropriate for those units:

$\mathrm{F_{\rm thrust}}=I_{\rm sp} \cdot \frac{dm} {dt} \,$

Yup. A slightly different equation, you've divided by g.WolfKeeper 16:03, 22 April 2006 (UTC)

No, I haven't "divided by" g. I simply didn't bother multiplying by an irrelevant conversion factor in the first place, so there was no need to divide it out again. Gene Nygaard 18:33, 22 April 2006 (UTC)

(270 lbf·s/lb)(112 lb/s) = 30,240 lbf

That's a whole lot simpler than your method that "works ..., well kinda":

$\mathrm{F_{\rm thrust}}=I_{\rm sp} \cdot \frac{dm} {dt} \cdot g_{\rm 0} \,$

(270 s)(112 lb/s)(32.174 ft/s²) = 972941.76 lb·ft/s² = 972941.76 pdl

(972941.76 pdl)(1 lbf/32.174 pdl) = 30,240 lbf

Isn't that right? Gene Nygaard 15:49, 22 April 2006 (UTC)

I'm not at all disagreeing. If you want to add that in the article as well, but that equation only works in imperial, whereas the equation in the article already worked in either (except as noted, it gives poundals.)WolfKeeper 16:03, 22 April 2006 (UTC)

I just showed you how that equation (the one with the g₀) works just fine using pounds-force for thust, "seconds" for I_sp, slinches for mass, and inches per second squared for acceleration. What don't you understand about that? Gene Nygaard 16:27, 22 April 2006 (UTC)

Furthermore, the equation without the fudge factor doesn't "just work in Imperial".

$\mathrm{F_{\rm thrust}}=I_{\rm sp} \cdot \frac{dm} {dt} \,$

Example 1:

(270 kgf·s/kg)(50 kg/s) = 13,500 kgf

Example 2:

(2650 N·s/kg)(50 kg/s) = 132,500 N

where 270 lbf·s/lb = 270 kgf·s/kg and (270 kgf·s/kg)(9.80665 N/kg) ≈ 2650 N·s/kg and 112 lb/s ≈ 50 kg/s.

Other than the minor rounding of the input numbers, that is in agreement with the 30,240 lbf (or 973,000 pdl) we got above. Gene Nygaard 16:41, 22 April 2006 (UTC)

"The kilogram-force has never been a part of the International System of Units (SI), which was introduced in 1960. The SI unit of force is the newton."WolfKeeper 16:43, 22 April 2006 (UTC)

So see example 2 in that last set.

And the SI unit of specific impulse is the newton-seconds per kilogram.

But what's your point, anyway? There is no requirement that we use SI. Kilograms-force are still in use today. You and I may wish they'd disappear, but they certainly haven't done so. But since unlike SI, the English units are no longer supported and updated, nobody is ever going to bother to tell us to stop using pounds-force, without telling us not to use pounds of any sort.

Furthermore, Sputnik 1 was launched in 1957. No SI then. Gene Nygaard 18:30, 22 April 2006 (UTC)

[edit] Removed paragraph

"It may seem odd that the acceleration or weight at the Earth's surface is in the definition, while the rocket may be far from the Earth. However, accelerations are often measured in terms of g₀; for example, astronauts should not be subjected to an acceleration more than a few times this value. Additionally, in English units the relationship between force and mass is defined to involve the acceleration due to gravity. Thus pounds (force) and pounds (mass), both used in rocketry, when divided, must be additionally multiplied by g₀ to get the acceleration in more usual units. An English unit of mass the slug. This, the common use of pounds for both force and mass, is in fact the chief reason g₀ enters so often into rocketry definitions, and is likely the reason two definitions of specific impulse are in common use."

I have many, many issues with it:

it mentions slugs as if this explains things, but slugs aren't mentioned anywhere else in the article, so there's no obvious connection between it and the rest of the piece
the sentence 'An English unit of mass the slug' has no verb
use of the phrase 'it is likely' implies the anonymous who wrote it was guessing.
the fact that g force is important for astronauts doesn't explain why it's important for Isp.
it makes unreferenced claims
overall quality of the article seemed higher after I removed it.

WolfKeeper 15:52, 24 April 2006 (UTC)

You yourself added poundals, which are not only not mentioned elsewhere in the article,

Trivially true only. That usage is referred to as you well know.WolfKeeper 17:57, 24 April 2006 (UTC)

but which unlike slugs are not in current use among any real-life rocket scientists.

Slugs aren't mentioned anywhere else; as a matter of fact I've never seen slugs used anywhere by real-life rocket scientists; YMMV.WolfKeeper 17:57, 24 April 2006 (UTC)

They are of no more help explaining things in what you said than slugs are here, and slugs could be used just as well as poundals where you discussed them. There is no difference in the difficulty of the calculations if you use any of the three fps systems; the only difference is at which stage of the process you throw in a g_c (udimensionless) or g₀ (with units) factor.

So put in a verb.

It's only important for I_sp for those who use pounds and pounds force, or kilograms and kilgorams force.

Gene Nygaard 17:28, 24 April 2006 (UTC)

You might recall that I've already pointed out the two NASA hits for poundal or poundals, a few list of conversion factors and one article written in 1920.

Just so in the future you won't be able to say you've never seen slugs used:

Need more? Gene Nygaard 18:44, 24 April 2006 (UTC)

I've got nothing against slugs; I've never seen them used, but I don't pretend to be the ultimate knowledge about anything, that's why I use references and books and stuff. All I'm saying is that the paragraph didn't seem to reach the acceptable quality and I removed it. If you want to rewrite put it here, we can check it out, edit it as required and reinsert it as appropriate.WolfKeeper 19:04, 24 April 2006 (UTC)

Pages like this one might not get so convoluted, if you actually understood that there are three different fps systems, and how they are used. Gene Nygaard 19:56, 24 April 2006 (UTC)

[edit] standard specific impulse

The specific impulse that a rocket delivers varies with the chamber pressure and expansion ratio of the nozzle as well as the fuel that is used. To compare propellants, the standard specific impulse is determined for standard conditions. The standard chamber pressure is 1000 psia. The standard nozzle is that which results in an exit pressure of one atmosphere at sea level. Specific impulse numbers given for propellants are for standard conditions unless they say otherwise. 4.232.3.182 00:06, 16 August 2006 (UTC)

[edit] General Considerations

The middle paragraph under 'General Considerations' would be better changed as follows:

In addition it is important that thrust and specific impulse not be confused with one another. The specific impulse is a measure of the impulse per unit of propellant that is expended, while the integral of thrust times time is impulse. In fact, propulsion systems with very high specific impulses (such as ion thrusters: 3,000 seconds) are power limited to producing low thrusts, due to the relatively high weight of power generators. 4.232.3.182 00:44, 16 August 2006 (UTC)

[edit] Units for ISP using MathCad

Wikipedia correctly explains lbf, lb, and even slug. These are commonly employed in engineering calculations that use the U.S. system of units as I have done for decades including three decades in design of military rockets. Not keeping track of units can lead quickly to wrong answers. MathCad knows the difference between lb and lbf and is unforgiving of their misuse. If you use MathCad, SI units, seconds for ISP, propellant rate in kg/s and include g, both the magnitude and the units in the calculated thrust (Newton) come out right. With U.S. units, seconds for ISP, propellant rate in lb/s and include g, both magnitude and units are correct in the calculated thrust (lbf). If you delete g and have ISP in lbf-sec/lb and propellant rate in lb/s using U.S. or ISP in N-sec/kg and propellant rate in kg/s using SI, both magnitude and units are correct (lbf, Newton) in the calculated thrust. 4.232.0.111 13:31, 28 August 2006 (UTC)

[edit] Erroneous content

This article is presently erroneous and missleading. It begs for a major correction/rewrite by someone, with an engineer's understanding of units and the necessity for their correct use, that is familiar with the subject. 63.226.127.196 17:56, 24 August 2006 (UTC)

Go for it. --P3d0 20:10, 24 August 2006 (UTC)

In work. This will take a few days. 4.232.0.111 13:44, 28 August 2006 (UTC)

[edit] Rewrite

I hate to say it, but I agree with the fellow above, this article is a mess, long-winded and full of un-scientific discussion and unnecessary long "examples". The next section explains how rocket propellants are compared. General considerations is pretty much as it was except its impulse per unit of propellant rather than thrust. In the examples, the columns for fuel mass and energy expended are for specific rockets and should not be part of this article.

Impulse is the integral of force over time, in units of kgf-seconds or newton-seconds. Impulse equals momentum. What engineers mean by specific impulse is the impulse (or momentum) generated per mass of fuel consumed. Thus the commonly used units of "kgf-sec/kg", which is traditionaly abreviated to just "sec".

It's impossible to get the simple and important concept of specific impulse from this very badly written article. DonPMitchell 01:06, 29 August 2006 (UTC)

Do you have a cite for it being kgf-sec/kg?WolfKeeper 02:47, 29 August 2006 (UTC)

Note the units of specific impulse given on this page: http://www.energia.ru/english/energia/launchers/engines.html I don't mean to sound snarky, but people who don't know what the units of specific impulse are should not be writing an encyclopedia page about it. DonPMitchell 10:13, 29 August 2006 (UTC)

You do sound snarky, but you're absolutely right that people who don't know the units should tread with care. siafu 10:26, 29 August 2006 (UTC)

Of course if the unit was invented today, it would be given in units of impulse per mass, newton-sec/kg. That is not a common familiar convention though. "sec" or "isp" units have a factor of g in there. You can think of it as impulse per weight (lbs-sec/lbs), or you can think of it as impulse per mass, using kilogram-force which is oftgen used to measure rocket engine thrust.

Kilogram-force is not often used to measure anything at all; it's not an SI unit, and it's been deprecated since the 60's. Moreover, the only source you've provided is an English translation of a Russian website. The claim that the "commonly used units" for specific impulse is "kgf-sec/kg" is not all supported. siafu 13:05, 29 August 2006 (UTC)

Here is an attempt at a rewrite. I think that a discussion of units is a good place to start since they are so controversial.

I'm sorry, but I think it's a really bad place to start, of largely historical interest only. Presumably most people come to this page to find out what Specific Impulse is, rather than see some stupid argument about units.WolfKeeper 17:42, 5 September 2006 (UTC)

As appropriate, the reader gets the short answer of what specific impulse is in the first sentence of the article. The section on units is intended to be explanatory rather than argumentative. It is particularly useful to someone who knows the difference between pound force and pound mass. The numerical equivalency between m/s and N s/kg should be addressed prior to showing it in the table. Also, how you get to m/s from s should be spelled out early. Since some references report specific impulse in lbf s/lbm it seems appropriate to discuss that early. Other than that, possibly the rest of the units story could be in a section titled 'units history' or some such. Dan Pangburn 02:02, 7 September 2006 (UTC)

Also there seems to be a major error: "Originally, in the U.S. to compare propellants, a weighed quantity of propellant was expended in a rocket and the impulse that it produced was measured. Since the impulse was determined in lbf·s and the weighed quantity of propellant was in lbm, the correct units for specific impulse became lbf·s/lbm." But weight is a FORCE (lbf), as is thrust. The quantity of propellant was defined in terms of weight. So we have lfs.s/lbf and it is *correct* to cancel units in that case. I actually wonder if there may be an urban legend here.WolfKeeper 17:55, 5 September 2006 (UTC)

If it wasn't for some references using lbf s/lbm the first part of the discussion would not be needed. Note that the discussion transitions into exactly what you state above. BTW, I am very antagonistic towards urban legends. In this case, I was there. Dan Pangburn 02:02, 7 September 2006 (UTC)

And currently this is a particularly bad section since it is utterly uncited, if correctly cited IMO it might be good as an appendix type section of historical importance though. Also, wherever possible we probably should stress SI units like seconds, m/s and Ns/kg and the 'temp' article does precisely the opposite. All the big textbooks emphasis SI units right now, and I believe NASA is moving across also. Russian block work has historically use kgf units more, but I'm unclear that they still do this to any degree.WolfKeeper 17:42, 5 September 2006 (UTC)

You are probably far better than me at determining where citations are appropriate. They can be added or possibly the wording revised to eliminate the need. (I added a couple already where you requested). I agree that SI should be used first (except in the units section) with U.S. units put as equivalent. I will fix. Maybe the Russian use of kgf should be mentioned in the units section (only). Dan Pangburn 02:02, 7 September 2006 (UTC)

The next section describes the standard conditions for comparing rocket propellants. General considerations is pretty much unchanged except it's impulse that is measured not thrust. In the examples, the columns for fuel mass and energy expended are for specific propulsion systems and should not be part of an article on specific impulse. I adjusted all numbers to be consistent with exact standard gravity. Rounding then becomes the choice of the user. I could not determine where some of the previous numbers came from. I increased the specific impulse of a solid rocket to 250 sec since that is closer to what is actually being done these days with current propellants in big rockets. Throughout I changed to make, IMO, more clear and/or compact. Duplications were eliminated. The heading 'Specific impulse in seconds' was renamed to reflect closer to what the section contains and the discussion following was changed. I believe that the content of the section headed 'Rocketry - specific impulse in seconds' is adequately covered under the new 'Units'. I replaced it with a brief discussion of how specific impulse is determined. A section on the effect of ambient pressure on specific impulse was added. The rest of the article was deleted as being already covered or not being needed for the article.

Dan Pangburn 00:50, 31 August 2006 (UTC)

Since the rewrite was rather huge (too big for this talk page), I moved it to Specific impulse/temp. siafu 01:50, 31 August 2006 (UTC)

I don't like the prominence of MathCad in the ref article. It's a commercial product and doesn't need to be there at all in my opinion. Seems non notable in this context.WolfKeeper 17:46, 5 September 2006 (UTC)

It is included to help the MathCad user. There is an article in Wikipedia on MathCad. It is a terrific tool that is commonly used in engineering. It took me a while to discover that MathCad uses lb for pound mass and lbf for pound force. How about moving this section to the end of the article to make it less prominent? The wording can probably be made less contentious and shorter since there is no issue in SI. If it's not mentioned in the TOC it would probably never be used anyway so if it's not in the TOC, then might as well delete that section and let the MathCad Users figure it out on their own.

Since I didn't discover these comments right away, maybe it would be better to address future discussion on the Specific impulse/temp discussion page. Dan Pangburn 02:02, 7 September 2006 (UTC)

[edit] Article inconsistency

The article claims that N.s/kg is a unit of speed. I claim that this is false, the equations don't have correct units if you do that.WolfKeeper 19:21, 13 September 2006 (UTC)

I think the article was right. One newton is one kg•m/s². Multiply by one s/kg and you get one m/s. That's a unit of speed. --P3d0 19:27, 13 September 2006 (UTC)

Of course, but that's an extra algebraic step. You should just be able to plug the figures and units into the equation and it should work out, rather than having to essentially substitute F=ma in to make it work.WolfKeeper 19:34, 13 September 2006 (UTC)

I think that m/s is 'effective exhaust velocity', whereas N.s/kg is specific impulse. And that seems to be the convention in most cases; although it is not so uncommon to see m/s listed as specific impulse.WolfKeeper 19:34, 13 September 2006 (UTC)

Also, effective exhaust velocity and specific impulse (mass) are not the same thing in imperial units.WolfKeeper 19:36, 13 September 2006 (UTC)

To your first point: I don't agree that there's a difference, but even if there were, these minutae don't enhance the article, and only serve to make it more verbose. --P3d0 20:25, 13 September 2006 (UTC)

Don't bother to be accurate??? No that doesn't work.WolfKeeper 20:47, 13 September 2006 (UTC)

If you can give an example of an incorrect conclusion someone might draw from the claim that N•s/kg is a speed, then I'd agree that we should make a distinction. To your second point: what equation won't work out? --P3d0 20:25, 13 September 2006 (UTC)

Any equation; the units are incorrect. In SI units, by 'coincidence' it numerically works out, but the units are screwed. Yeah, you can fix it by substituting, but they're logically different numbers, with different unitsWolfKeeper 20:47, 13 September 2006 (UTC)

Wait... what? N*s/kg is definitely equivalent to m/s. 1 Newton is defined as 1 kg*m/(s^2). Cancel terms, and you get m/s. This is neither coincidence nor incorrect. Isp measured in mass terms as Ns/kg is exactly equivalent to effective exhaust speed, in both units and value (though usage may differ). Effective exhaust velocity (vector not scalar) also takes into account the direction the exhaust is pointed, though normally that isn't relevant. Now, given what is being measured, I would say that it can make sense (or not; personal preference) to think and work in terms of Ns/kg, not m/s. One must also realize that effective exhaust speed does not necessarily have any direct physical counterpart, given that the exhaust is non-uniform in both speed and direction, and that any pressure thrust corrections are usually included in Isp or Ve. Evand 00:41, 14 September 2006 (UTC)

(And also, I'd like to call it the effective exhaust speed, but that's beside the point.) To your third point: they are the same thing in imperial units if you use imperial units of mass. They have to be. This is not mere number juggling; specific impulse in the mass sense really is the effective exhaust velocity. --P3d0 20:25, 13 September 2006 (UTC)

No, they actually aren't. An example of specific impulse (mass) in imperial is: 450 lbf.s/lb. It's rarely used, but it is (rarely) used. An example of effective exhaust velocity (speed) is 14,500 ft/s. They might be equivalent, but they are not the same.WolfKeeper 20:47, 13 September 2006 (UTC)

They are the same after you multiply by the unit conversions, in exactly the same sense as 1 inch and 2.54 cm are the same. If you count those as different, then I'm sure you agree that it's a difference in semantics and usage, not in what is being measured. Evand 00:41, 14 September 2006 (UTC)

Exactly. This unit "lbf•s/lb" is a speed unit; one such unit equals 32.17405 ft/s. 450 of these units equal about 14500 ft/s. So in Wolfkeeper's example, both are speed units, and they are equal. You can read about this here. --P3d0 02:30, 14 September 2006 (UTC)

Yes... obviously. Obviously to us. But is this necessarily obvious to the reader? Shouldn't this all be made very, very explicit? I could imagine somebody not realising this and coming up with all kinds of weird results. Also whilst I utterly disagree with the idea that lbf.s/lb is more correct than other units I think we've narrowly gone too far in the other direction. As in lbf.s/lb deserves about 2 sentences somewhere, because it is used (albeit rarely.)WolfKeeper 02:55, 14 September 2006 (UTC)

OK, we seem to be in agreement then. I think a brief mention of lbf*s/lbm as a unit is correct. I don't think ft/s are worth mentioning, since they're rarely used and should be obviusly equivalent to the metric m/s version. Perhaps the conversion from lbf*s/lbm into both m/s and ft/s should be given? I'll leave it to you guys to actually get the text the way you want it. Evand 12:53, 14 September 2006 (UTC)

Let's get specific. What is an example of a misunderstanding someone could come away with if we were to state that N•s/kg is a unit of speed? --P3d0 21:40, 14 September 2006 (UTC)

If you're doing stuff in imperial units (and Apollo used ft/s extensively) you can get very wrong answers. But nobody is claiming that the world will end. But mostly I think it's that explaining it clearly and logically and symmetrically builds understanding. Maybe a table like this might be worth adding:WolfKeeper 23:05, 14 September 2006 (UTC)

	SI	Imperial
Specific Impulse (by weight)	=1 Second	=1 Second
Specific Impulse (by mass)	=9.8 N.s/kg	=1 lbf.s/lb
Effective exhaust velocity	=9.8 m/s	=32.16 ft/s

(Restarting at a sane indentation level...) Wolfkeeper, I like your recent edit, as well as your table (above). We seem to have reached a good compromise and consensus here. --P3d0 02:20, 18 September 2006 (UTC)

Retrieved from "http://en.wikipedia.org../../../s/p/e/Talk%7ESpecific_impulse_f4fe.html"

Category: Unassessed physics articles