Talk:Specific speed

Figures are actually unitless
As I find it, the article says the figure is unitless BUT is the speed of an ideal machine. You can't have this both ways. It's only possible to think of a unitless number as a speed because IN THIS CASE speed happens to be linearly in the numerator. Thus a pump which operates 5 times as well as an ideal reference pump would have a specific "speed" of 5 (no units), but this also happens to imply that the reference pump would have to turn at 5 rpm (or turbine at 5 radians/sec) or whatever, to give this performance. That has to be made clear. The number is a performance number because there is an unwritten speed of "1 unit" in the denominator which isn't written, and has been used to divide out the actual device speed in the numerator. It is THAT number which must be increased to turn the figure into the equivalent speed of the reference device.

As you see from the dab page on specific I've been working on these terms that involve the word "specific." What the word means is you divide by some quantity. Often that's mass or volume, but sometimes it's something more complicated, like fuel-consumption. Whenever it happens, the correct terminology would be to write the divided quanity up front, before the word "specific". So when you divide by mass, it's really a "mass-specific" quantity. Sometimes you divide by an entire reference quanity to get a unitless ratio, like "specific gravity" but there it really should be called "water-density-specific density." Often the divided-by quantity is omitted or partly omitted, so you see the problem with all these terms. Even some scientists are confused about what they mean (I've encountered people who thought "specific" always always meant division by mass or volume).

Here, "specific speeds" are really ideal-device-specific performances. They can be thought of as speeds of the ideal reference device needed to match the device being considered, but in that case the "specific" doesn't really apply to "speed." So it's a misnomer. Pump-specific performance and turbine-specific performance are closer, and pump-specific speed is only an equivalent number with different units (units of speed not a unitless ratio number).

It's not obvious to me from the units analysis why all these pump and turbine performances depend in odd power-law ways on impellar velocity and pressure-head. Dimentionally, in absolute terms, not a single figure-of-merit of the three presented here comes out dimentionless, which means that somewhere a dimentioned factor has been introduced but hasn't been written down. An obvious mark of that is that "g" appears in the turbine equation but not in the pump equations. Are we to infer then that the pump equations would provide valid numbers on the moon, but the turbine one would have to be adjusted? I don't believe it. A better article would explain some of this stuff. Are some of these power-dependences empirically derived, or are they all dependent on ab initio physics? S B Harris 15:30, 21 April 2010 (UTC)

Specific speed
The term "specific speed" has been used for well over a hundred years for both pumps and turbines. Although it gives a number which indicates a machine's performance and suitability for a given application, it is not a dimensionless number, and care must be taken to state the units used. The clue is there when the author after saying that the value is dimensionless then says "In Imperial Units..."

If a number is dimensionless then the units in the expression cancel out, as in Reynolds number (V.D/ν = [m/s][m]/[m²/s] = [-]. This does not happen with specific speed, nor with several other numbers used in the turbine industry. For pumps flow might be in US gpm, while in UK it might be in Imperial gpm. For turbines 30 years ago power was often given in metric horsepower rather than kW. I have used the term "quasi-dimensionless", as although these numbers are often treated as if they are dimensionless, they are not.

The person who wrote under Turbine specific speed: "It can also be defined as the "speed" (blade angular velocity) of an ideal, geometrically similar turbine, which would yield one unit of power per one unit of head per one radian/second of blade angular velocity." is technically correct, but in 30 years in the hydro industry I have never come across these units, which I feel just confuse the issue. This is probably an attempt to use SI units rather than rpm, but it goes against standard practice where engineers have "a feel" for rpm, and do not have a feel for radians/s. I can understand the desire to standardise to one set of units, but unfortunately convention has a large inertia, and has to be respected. Boving (talk) 12:21, 1 October 2010 (UTC)


 * Fair enough. All that should be explained in the article. Do you want to take a crack at it? All these speeds come out in rad/sec or rpm, BUT you're saying that this is for ideal pumps that are rated in imperial or metric units. So the units of specific speed should be given in each section on pumps, with caveat that this means rpm or rad/sec for a metric pump or turbine, or an imperial pump, or something. Right? Incidentally, I tried to perform a dimensional analysis from the equations, and got junk. So "quasi-non-dimensional must be right. I'd like to see the dimensional analysis done in the article someplace, perhaps as the first section after the intro.  S  B Harris 18:47, 1 October 2010 (UTC)

Equation for specific speed
Any reference I can find does not include the gravitation constant in the calculation of specific speed. I would suggest the g be removed. Davemody (talk) 19:39, 17 November 2012 (UTC)

Specific speed
Imho a few things are missing in this article. I have a physics background but I get really confused by this article, e.g. the number is not even dimensionless. I would edit it, but maybe there are some reasons to put it like this.

It is also in the 'start-class' quality; did the one who assigned this class, leave any comment?

suggestions: there are more uses than finding the suction specific speed easiest is probably by eliminating the diameter from the flow and head rise coefficient. this will also add the gravitational constant (g) to the equation to make it really dimensionless. as stated already, there are multiple variants of the number. A short overview and background would be nice. Maising (talk) 09:30, 22 March 2017 (UTC)
 * clearly state its practical use
 * clear definition and derivation
 * non-dimensionless variants
 * more references