If compressors are being able to take a larger cross section of air and reducing down while also reducing the velocity of the air, such that the inlet to burner is on the order of 50m/s while the pressure ratio increases upto 20 times...I get the impression of a divergent section while most cut-away views I've seen appear to me as convergent.

What am I missing?

What am I missing?

The meaning of life is curiosity; we were put on this planet to explore opportunities.

- 727EMflyer
**Posts:**538**Joined:**

Quoting Lehpron (Thread starter):I get the impression of a divergent section while most cut-away views I've seen appear to me as convergent. |

Do you mean you think the volume between stages should increase rather than decrease? I'm curious as to what gives you that iIMpression as COMpression involves making a constant mass of gas fit into a space with a decreasing volume. The fact that the velocity of the gas is reduced is explained nicely by Bernouli.

Its very simple, what you see in the engine cut away views is that each stage of compression comprises a larger blade disc with smaller blades. The overall shape of the compressor discs are divergent while the airflow path is convergent.

In layman terms, as you pass each stage of engine compression, you're stuffing the air into a smaller opening.

In layman terms, as you pass each stage of engine compression, you're stuffing the air into a smaller opening.

Forums.AMTCentral.com

Quoting Lehpron (Thread starter):If compressors are being able to take a larger cross section of air and reducing down while also reducing the velocity of the air, such that the inlet to burner is on the order of 50m/s while the pressure ratio increases upto 20 times...I get the impression of a divergent section while most cut-away views I've seen appear to me as convergent. |

If you reduce the speed of the flow through the compressor, stages will surge. But yes, you generally want to recover as much as the dynamic head as possible before the combustor.

Your bone's got a little machine

Convergent vs divergent ducts, nozzles vs diffusers, these do nothing but trade dynamic pressure (velocity) for static pressure. Total pressure remains constant (minus any losses). The entire point of a compressor is to increase total pressure, which can only be done by using a work input. A divergent duct will not increase total pressure.

You ask why is a compressor convergent from front to back. This is simply answered that since mass flow is constant, but density increases, then the exit area must be smaller compared to the inlet area. Now as DC-10Tech pointed out, if you look at the individual compressor stages, the flow path actually does expand, just not in the axial direction.

Why?

Compressors work by using alternating rows of rotor blades and stator vanes.

Every row of stators will turn the flow away from the axial direction. Because the flow is turned to the side tangentially, the exit area of each stator is actually larger than the inlet area, which means you get a rise in static pressure and a drop in velocity. However, stators don't move so no work is put into the flow, so total pressure remains constant (once again, minus any losses).

Rotors then turn the flow back to the axial direction. But because they are moving, work is put into the flow. Two things now happen in the rotors. The rotating blades increase the dynamic pressure of the flow. Rotors also turn the flow in a way that like the stators, relative to the flow they have exit areas larger than inlet areas, which increases the static pressure. The sum of the these two pressure rises is your increase in total pressure.

Here's a diagram comparing the diffuser that you are thinking of to the cascade type of diffuser in a row of rotor blades. Both have exit areas larger than inlet areas and will increase the static pressure.

The alternating rows of rotor blades and stator vanes work by allowing you to put work in to increase total pressure (rotor), then convert this total pressure rise to an increase in static pressure (stator). Net result of the compressor is an increase in pressure and a decrease flow velocity.

[Edited 2006-06-13 16:59:39]

You ask why is a compressor convergent from front to back. This is simply answered that since mass flow is constant, but density increases, then the exit area must be smaller compared to the inlet area. Now as DC-10Tech pointed out, if you look at the individual compressor stages, the flow path actually does expand, just not in the axial direction.

Why?

Compressors work by using alternating rows of rotor blades and stator vanes.

Every row of stators will turn the flow away from the axial direction. Because the flow is turned to the side tangentially, the exit area of each stator is actually larger than the inlet area, which means you get a rise in static pressure and a drop in velocity. However, stators don't move so no work is put into the flow, so total pressure remains constant (once again, minus any losses).

Rotors then turn the flow back to the axial direction. But because they are moving, work is put into the flow. Two things now happen in the rotors. The rotating blades increase the dynamic pressure of the flow. Rotors also turn the flow in a way that like the stators, relative to the flow they have exit areas larger than inlet areas, which increases the static pressure. The sum of the these two pressure rises is your increase in total pressure.

Here's a diagram comparing the diffuser that you are thinking of to the cascade type of diffuser in a row of rotor blades. Both have exit areas larger than inlet areas and will increase the static pressure.

The alternating rows of rotor blades and stator vanes work by allowing you to put work in to increase total pressure (rotor), then convert this total pressure rise to an increase in static pressure (stator). Net result of the compressor is an increase in pressure and a decrease flow velocity.

[Edited 2006-06-13 16:59:39]

- EasternSon
**Posts:**637**Joined:**

Thanks, DarkBlue. An excellent, thorough and well-worded response.

I have to say this is a pretty good example of someone asking a relatively simple question and getting a great response from some knowledgeable people, rather than being ridiculed.

Regards,

I have to say this is a pretty good example of someone asking a relatively simple question and getting a great response from some knowledgeable people, rather than being ridiculed.

Regards,

"The only people for me are the mad ones...." Jack Kerouac

Thanks DarkBlue, very nice explanation! When Lehpron asked the question, I didn't know how it worked either.

I'd like to add something...

From a thermodynamic point of view, here are the formulae that happen in the process:

I tried to solve to see what happens. Solving for P2 = 100000, T2 = 270, w2 = 100, A2 = 2, A3 = 1, gamma = 1.4, cp = 1005, l (compression ratio) = 20, gives:

P3 = 2129078, T3 = 647, w3 = 22.5

I don't know if these values resemble to real values, but the equations are correct I believe. The ''t'' subindex means stagnation (or total, however you call it). Note that ideal gas has been supposed, and the efficiency is the unity (that is, isentropic process. Actual efficiency is about 0.9 I think). Here the areas and the compression ratio have been fixed, but you could use other ones as known variables, such as the work. If work rate is 0 (diffuser), T3t = T2t, and with the same areas you see a drop in static pressure, P3 = 62271 and an increse in velocity, w3 = 280.5, which is what you could expect for a convergent section. The input of work is what makes total pressure rise, while velocity can drop.

I'd like to add something...

From a thermodynamic point of view, here are the formulae that happen in the process:

Thermodynamic formulae for compressor process |

I tried to solve to see what happens. Solving for P2 = 100000, T2 = 270, w2 = 100, A2 = 2, A3 = 1, gamma = 1.4, cp = 1005, l (compression ratio) = 20, gives:

P3 = 2129078, T3 = 647, w3 = 22.5

I don't know if these values resemble to real values, but the equations are correct I believe. The ''t'' subindex means stagnation (or total, however you call it). Note that ideal gas has been supposed, and the efficiency is the unity (that is, isentropic process. Actual efficiency is about 0.9 I think). Here the areas and the compression ratio have been fixed, but you could use other ones as known variables, such as the work. If work rate is 0 (diffuser), T3t = T2t, and with the same areas you see a drop in static pressure, P3 = 62271 and an increse in velocity, w3 = 280.5, which is what you could expect for a convergent section. The input of work is what makes total pressure rise, while velocity can drop.

Where there's a will, there's a way

- grandtheftaero
**Posts:**247**Joined:**

Quoting GrandTheftAero (Reply 7):Looks like you've taken Dave Wisler's class too! |

Yep, after posting I realized I probably should have given him credit for the picture (and for the education in compressor design)