Superchargers?

[tomo pauk]

Not sure why you think that.

The intake on the Hurricane MK II was good for anywhere from 27.7hp to 14.1 hp of Air intake momentum drag from 15,000 to 35,000ft.

Nobody ever accused Hurricane to be a miracle of aerodynamics.
Hoerner's analysis of the Bf 109G pointed out to the ram air intake as source of a drag, while at the Fw 190D and Ta 152 the side-mouted intake accouted between 4.3 and 5.9 % of total drag (or about as much as weapon-related drag). For comparison, the internal intakes on the Fw 190A represented 0.6% of total drag.

You don't quite get "RAM" for free. You want to turn the forward speed of the plane into higher pressure air (than ambient) going into the carb or supercharger inlet you are going tohave to pay for it in both form drag ( intake scope external aerodynamics) and internal duct drag (internal aerodynamics) and if you are compressing the ambient air in the intake duct/scoop that compression has to be paid for somehow even if it is only 1-2 psi.

The 'no free lunch rule' applies as ever

The 109F and G may have increased the effectiveness of the intake compared to the E by moving the intake further away from the fuselage (boundary layer/turbulent air) and going to the round shape instead of square (corners don't do a lot for air flow)

Agreed.

as for the Corsair, they didn't quite fit the new engine and supercharger set up in the old fuselage.
Notice the "cheek" scoops to supplement the wing root intakes and the fact that the fuselage is bulged behind the cowl flaps over the wing roots. On an F4U-4 the cowl flaps go much further down the cowl.

Fuselage was more or less old. The engine section was changed (and obviously the powerplant itself); old air intakes were insufficient to provide enough of airflow for oil coolers, inter-coolers and engine itself all int the same time, thus the -4 gained one extra intake, and -5 two (possitioned so the air flow towards the two impellers of the 1st stage is as direct as possible).

 

[Shortround6]

Nobody ever accused Hurricane to be a miracle of aerodynamics.
Hoerner's analysis of the Bf 109G pointed out to the ram air intake as source of a drag, while at the Fw 190D and Ta 152 the side-mouted intake accouted between 4.3 and 5.9 % of total drag (or about as much as weapon-related drag). For comparison, the internal intakes on the Fw 190A represented 0.6% of total drag.

and there is the trade off.

Is 4-6% more drag worth several thousand feet of FTH on the engine?

as for the Hurricane.

Most anything would have been an improvement as far as streamlining goes.
One also wonders about the efficiency of the boundary layer splitter on the radiator with the turbulence generator air scoop only few feet in front of it.

 

[tomo pauk]

and there is the trade off.

Is 4-6% more drag worth several thousand feet of FTH on the engine?

It will depend on use? Going against the B-17s (with or without an assorted escort) will favor greater FTH. Prominent/external intakes also allow for installation of air filters (useful if the aircraft is used on dusty African or steppe airfields).
In case that much of increase in FTH is not of an interest (East front, mostly), lower drag might be preferred.

 

[Deleted member 68059]

Somebody may have said that on a two stage supercharger any mistake or problem with the first stage is multiplied by the 2nd stage.

This is from a strict theoretical perspective true. However the practicalities of actually joining the airpaths together means than in practise a 2-stage SC tends to end up having
the pressure heads added not multiplied, such are the losses incurred from bending the airflow around so many corners (180 deg turn between stages 1 & 2 in the case of something like a Merlin 61). In pure "maths-world" of little boxes on paper with the compression ratios written on them, they ought to multiply...the multiplication takes no account of any of the inefficiencies of a real system. Hence Merlin 45 might have an ACTUAL pressure ratio of something like 3.5:1 and a Merlin-61 6.5:1 (both stages considered as one bloc, in perfect-land it ought to be something like 10:1).

 

[ThomasP]

Hey Snowygrouch,

I believe you are mistaken as to how the air is traveling through the Merlin 2-stage supercharger. The only relatively sharp radius is at the 90º elbow at the turn from the inlet duct/carberetor to the eye of the 1st stage impeller. After that the air is routed to the 2nd stage by circumferential and spiral ducts to the 2nd stage, and the same from the 2nd stage to the entry into the engine intake manifold. There is kind of a large radius 90º turn at the entry to the engine intake manifold.

Also, in the language of physics (and mathematics is considered the language of physics) the 1st and 2nd stage in centrifugal superchargers add, they do not multiply. In the same sense as the 1st stage adds heat to the air (i.e. by doing work) and the 2nd stage then adds more heat to the air (i.e. by doing additional work), the 2nd stage adds its work of compression to the already accomplished 1st stage compression. I think where the confusion comes in is that once you have determined how much the effective increase/addition in compression is, you can then determine the resulting ratio and use that ratio as a descriptive multiplier. (If I misunderstood what you were saying I apologize.)

 

[fliger747]

It might be that the scoop in front of the radiator would help compensate for the gap in the gear doors? Wow, what miracles would a detail cleanup of the exterior of the "Hurri" have done?

 

[Shortround6]

In practical terms they multiply. Each unit operates at a pressure ratio (which varies with the rpm and a few other things)

this is from a modern turbo but the principle is the same.

the output is going to be a certain ratio of pressure compared to the input (most cars and certainly most planes don't use flow meters (measure the mass moving through the system)'

so if you take in air at 15in Hg and compress it to 37.5in Hg you have compressed it 2.5 times,

Now you add a second unit in series. It should not be identical as it won't flow the required mass. But if it takes in air at 37.5 in Hg (and I would agree with Snowygrouch that this is ideal and not real world) and it compresses it 2.5 times you will wind up with air at 93.75in Hg.
In practice you will find that no fuel that was in service use in WW II would stand up to that so there was no incentive to design such a system. That or the superchargers were running into the choke limit.

another modern turbo map running at sea level. The unit simply won't flow more than about 95-97lb/min no matter how fast you turn the impeller.

Don't confuse cause and effect. Most two stage systems in use in WW II didn't exceed an overall pressure ratio of around 6-7 to 1 but they were designed that way in order to use the lowest amount of power and heat the intake the charge the least. The different gear ratios in the V-1650-7 from the V-1650-3 were to limit the total amount of possible boost and lower the power needed to drive the supercharger and heat the intake charge less to give more power at lower altitudes. Using oversized/over capacity/higher pressure ratio superchargers than needed for the application limited power at other points in the flight envelope.

one of the few (only?) exceptions were the late model P-47s (M & N) which could pull 72in of Hg at 32,000ft including ram or to 27,500ft while climbing which is performance that probably could not be matched if you added the ratios of two superchargers instead of multiplying them, this was possible because the the turbo allowed for a much wider variation in speed of the auxiliary stage impeller than any mechanical system.

 

[ThomasP]

Hey Shortround6,

I am sorry, but your statement "In practical terms they multiply." is actually the direct opposite of how it works. In practice they cannot multiply - if they could they would have to be 100% efficient - which is an impossibility. The inability to describe compression using multiplication was discovered back in the early days of mechanical compression systems, when the engineers, physicists, and mathematicians tried to develop formulas that could predict the effect of 2 steps or more of compression, and were unable to do so using multiplication.

The actual formula for multiple stage compression is similar to the formula I used to show the pattern of heat rise in the compressed air in my post#94 earlier in this thread:

Ta + T1 - 25 + T2 - TAC = MT

If you substitute P for each T increase , and substitute i for each temperature decrease you get:

Pa + P1p - P1i + P2p - P2i = MP

simplified to

P1e + P2e = MP

where

Pa = ambient air pressure
P1p = Pressure increase due to work that could be done by 1st stage if there was no inefficiency (i.e. p = perfect)
P1i = Pressure increase not achieved due to inefficiency of 1st stage
P1e = effective Pressure after the 1st stage
P2p = Pressure increase due to work that could be done by 2nd stage if there was no inefficiency (i.e. p = perfect)
P2i = Pressure increase not achieved due to inefficiency of 2nd stage
P2e = effictive Pressure after the 2nd stage
MP = Manifold Pressure

If you substitute values for the above variables I believe you will find that there is no way to use any form of multiplication to arrive at the correct answer until after you have already used the above formula to calculate the resulting MP. Once you have done so you can then use the values for P1e, P2e, and MP to generate ratios that can be used as a descriptive multiplier.

The physics principle illustrated above is a pattern followed by everything in the universe (at least everything so far discovered). The principle being that the universe is additive, not multiplicative.

Make sense?

 

[wuzak]

Make sense?

No.

Consider the MetroVick F2/2. The 9 stage compressor had an overall pressure ratio of 3.5:1.

If the final pressure ratio is the sum of the pr of the stages, the average pr would be about 0.4:1.

If you take the multiple of the stages, the stage pr would be about 1.15:1.

 

[wuzak]

Hey Shortround6,

I am sorry, but your statement "In practical terms they multiply." is actually the direct opposite of how it works. In practice they cannot multiply - if they could they would have to be 100% efficient - which is an impossibility.

Multiplication does not require 100% efficiency, as it does not say anything about the power required to drive the system

 

[Deleted member 68059]

Hey Snowygrouch,

I believe you are mistaken as to how the air is traveling through the Merlin 2-stage supercharger. The only relatively sharp radius is at the 90º elbow at the turn from the inlet duct/carberetor to the eye of the 1st stage impeller. After that the air is routed to the 2nd stage by circumferential and spiral ducts to the 2nd stage, and the same from the 2nd stage to the entry into the engine intake manifold. There is kind of a large radius 90º turn at the entry to the engine intake manifold.

Also, in the language of physics (and mathematics is considered the language of physics) the 1st and 2nd stage in centrifugal superchargers add, they do not multiply. In the same sense as the 1st stage adds heat to the air (i.e. by doing work) and the 2nd stage then adds more heat to the air (i.e. by doing additional work), the 2nd stage adds its work of compression to the already accomplished 1st stage compression. I think where the confusion comes in is that once you have determined how much the effective increase/addition in compression is, you can then determine the resulting ratio and use that ratio as a descriptive multiplier. (If I misunderstood what you were saying I apologize.)

I have engineering drawings (not magazine sketches) of the 61 supercharger and have spent time at Rolls-Royce poking at the real parts (below is actually my 100 series
pic as its a better photo to show this particular part of the supercharger).

The inter-stage area has turning vanes to try to make the flow have an easy time flowing between stages 1 and 2, but its still a huge loss which cannot be recouped by any means. Its a package layout problem not a design flaw. If you removed the turning vanes and the 1st stage volute exit, the air from the 1st stage would fly off into space still needing to be rotated about 180 degrees to make it back to the 2nd stage inlet eye.

Compressor work can only be ADDED when it is decribed in the form of work, i.e. for example in WW2 German practice as: "Adiabatic pressure head" (Had).

When compressor work is written as m.kg/kg (work done per mass unit) then one CAN simply add together the individual stage work to get to the actual total compressor work result.

FYI an Merlin-61 supercharger in reality makes about 12,500 m.kg/kg Had compressor work by the above equation. This is a useful representation as you can multiply this by the mass flow rate and it gives you the power required to drive the supercharger, and it also (roughly) tells you the height up to which the compressor can maintain sea-level pressure. I.e. the Merlin-61 supercharger can (by this rough measure) maintain sea level pressure to 12,500meters alititude (41,500feet), so its a useful visual guide to how effective a compressor might be in an aeroplane.


With regards to the figures discussed for axial stages by a later poster, it is not at all uncommon that each axial stage might only produce a pressure ratio of something like 1.2:1 - hence
why you do need loads of axial stages to do anything useful.

 

[ThomasP]

Hey wuzak,

The problem is that if you do not already know that the MetroVick F2/2 9-stage compressor had an overall pressure ratio of 3.5:1, you would not have any way of figuring the average or any possible inverse power (i.e. 1.15^9) of the stages, let alone the actual values. If the 1st stage achieves a 1.5:1 effective compression (P1e in the formula) instead of 1.15, what is the effective compression of the 2nd stage (P2e), the 3rd stage,...? Do all the stages achieve the same multiple for the effective compression ratio? or don't they? The reason that multiplication cannot be used is because it does not take into account the wasted power (Pi in the formula), just the work (physics term meaning change in energy state - in this case the effective increase in pressure (Pe in the formula)) achieved at the last measured stage. There are only 2 ways to determine the effective increase in pressure (Pe) done by a system:

1. Build the 1st compressor stage, build the 2nd stage, test each one separately and then attach (i.e. add) them to each other and measure their combined pressure increase, repeat the process with 3rd through 9th stage. Assuming the MetroVick F2/2 3.5:1 ratio is achievable at sea level, the MP (I know, the out put of the MetroVick is not going into an intake manifold, you can substitute another acronym if you wish) value of 51.45 lb/in^2 (3.5 x 14.7 lb/in^2) would be the sum of the original 14.7 lb/in^2 (Pa) + the 1st stage work (P1e) + the 2nd stage work (P2e) + the 3rd ....., + the 9th stage work (P9e).

2. Use a formula similar to the one above, which previously accepted values for similar systems.

 

[ThomasP]

Hey Snowygrouch,

re: "why you do need loads of axial stages to do anything useful."

In axial flow turbo- and fan-jets the main reason is to spread the work (physics term) done over more material (i.e. more stages, vanes, shafts, etc. ) without overloading the parts in terms of stress and heat, while keeping the diameter of the overall engine as small as possible. In other words, to decrease the chance of breaking or melting something while keeping the potential drag to a minimum.

 

[Shortround6]

Hey Shortround6,

I am sorry, but your statement "In practical terms they multiply." is actually the direct opposite of how it works. In practice they cannot multiply - if they could they would have to be 100% efficient - which is an impossibility. The inability to describe compression using multiplication was discovered back in the early days of mechanical compression systems, when the engineers, physicists, and mathematicians tried to develop formulas that could predict the effect of 2 steps or more of compression, and were unable to do so using multiplication.

The actual formula for multiple stage compression is similar to the formula I used to show the pattern of heat rise in the compressed air in my post#94 earlier in this thread:



Make sense?

No it does not because you are measuring two different things. Amount of work done from an efficiency standpoint is not pressure rise.

A single stage compressor multiplies the pressure of the air flowing through it.
if you have 30in Hg going in and the supercharger has a pressure ratio of 2.2 at a certain rpm and mass flow you will get 66in Hg at the output. Now there is no mention of power used to drive the supercharger or the efficiency of the supercharger (what percent of the power is used to actually compress the air).

If you climb to an altitude to where you have only 17.6in Hg going in you will get 38.72 coming out (or very close)

Now what happens when you run two impellers in series?

You are asking us to believe that two 2.2 pressure ratio supercharges (they would not be identical due to airflow considerations) would only give us a total rise of 4.4 times the original inlet pressure instead of the 4.84 that multiplication would give us. It gets worse if the pressure ratios are higher. two 2.25 superchargers should give 5.086.
In practice it would be rare to have two stages operating at identical pressure ratios.

However as Calum has shown us there are losses between the two stages, despite these losses two stage systems show an ability to operate at pressure ratios higher than simple addition would account for.

Superchargers (or centrifugal compressors) got better as the war went on (doesn't mean that all engines got new superchargers) but state of the art at the close of WW II was under 4.0 for a single stage compressor in useable form (not test compressor on the bench.)

I have no idea what RR did to the impeller closest to the engine, if anything, on the two stage engines besides cropping the diameter a fraction of inch. (fewer blades, different shape/depth, etc) but they ran the impellers a lot slower on the two stage engines.

 

[wuzak]

With regards to the figures discussed for axial stages by a later poster, it is not at all uncommon that each axial stage might only produce a pressure ratio of something like 1.2:1 - hence
why you do need loads of axial stages to do anything useful.

I am aware of the small pressure ratios that axial flow compressor stages have.

My point is that in order for the method espoused by Thomas of adding the pressure ratios together to get the overall ratio, since it would require ratios less than 1:1

 

[wuzak]

Hey wuzak,

The problem is that if you do not already know that the MetroVick F2/2 9-stage compressor had an overall pressure ratio of 3.5:1, you would not have any way of figuring the average or any possible inverse power (i.e. 1.15^9) of the stages, let alone the actual values. If the 1st stage achieves a 1.5:1 effective compression (P1e in the formula) instead of 1.15, what is the effective compression of the 2nd stage (P2e), the 3rd stage,...? Do all the stages achieve the same multiple for the effective compression ratio? or don't they? The reason that multiplication cannot be used is because it does not take into account the wasted power (Pi in the formula), just the work (physics term meaning change in energy state - in this case the effective increase in pressure (Pe in the formula)) achieved at the last measured stage. There are only 2 ways to determine the effective increase in pressure (Pe) done by a system:

1. Build the 1st compressor stage, build the 2nd stage, test each one separately and then attach (i.e. add) them to each other and measure their combined pressure increase, repeat the process with 3rd through 9th stage. Assuming the MetroVick F2/2 3.5:1 ratio is achievable at sea level, the MP (I know, the out put of the MetroVick is not going into an intake manifold, you can substitute another acronym if you wish) value of 51.45 lb/in^2 (3.5 x 14.7 lb/in^2) would be the sum of the original 14.7 lb/in^2 (Pa) + the 1st stage work (P1e) + the 2nd stage work (P2e) + the 3rd ....., + the 9th stage work (P9e).

2. Use a formula similar to the one above, which previously accepted values for similar systems.

There is the method MetroVick actually used. Calculate the output conditions based on the input conditions, the profile and length of the blades for the rotors and stators, the rpm, etc, long hand iteratively.

All this before cutting metal and being able to check their calculations.

Probably all stages were considered part of one system, since each stage affects the last.

 

[Deleted member 68059]

2. Use a formula similar to the one above, which previously accepted values for similar systems.

This is a WW2 German set of compressor maps for a 2 stage supercharger (centirifugal) showing the performance of each stage
and the combined performance. Theses are in Had, adiabatic pressure head (work) in which the total compressor work can be
represented as the work of each stage ADDED. which you can see as its basically 6000+6000=12,000 m.kg/kg

This does not work for pressure ratio, and needs to be caculated as per the equation I previously posted

(look for "314" which is the impeller tip speed in m/s)

 

[pbehn]

This is a WW2 German set of compressor maps for a 2 stage supercharger (centirifugal) showing the performance of each stage
and the combined performance. Theses are in Had, adiabatic pressure head (work) in which the total compressor work can be
represented as the work of each stage ADDED. which you can see as its basically 6000+6000=12,000 m.kg/kg

This does not work for pressure ratio, and needs to be caculated as per the equation I previously posted

(look for "314" which is the impeller tip speed in m/s)
View attachment 535124

I appreciate your posts Callum, I just wish I could understand them. Great stuff.

 

[ThomasP]

Hey Snowygrouch,

Thank you for supporting my post#105, 108, and 112 on the additive nature of supercharging.

re: my statement "I believe you are mistaken as to how the air is traveling through the Merlin 2-stage supercharger."

My description is based on what I read a number of years ago as to how the air was channeled from the output of the 1st stage to the 2nd stage inlet. Maybe I misunderstood? I can not tell from the actual engine cut-away picture you posted.

 

[Deleted member 68059]

Hey Snowygrouch,

Thank you for supporting my post#105, 108, and 112 on the additive nature of supercharging.

re: my statement "I believe you are mistaken as to how the air is traveling through the Merlin 2-stage supercharger."

My description is based on what I read a number of years ago as to how the air was channeled from the output of the 1st stage to the 2nd stage inlet. Maybe I misunderstood? I can not tell from the actual engine cut-away picture you posted.

Be a bit careful, I`m stating that multi-stage compressor WORK is additive when presented in that manner...

I think the fact that the air has to travel around a "u" bend is fairly clear from the picture. The fact that the flow exits the impeller with a strong tangential component, does not alter the fact that its flow direction must be altered by 180 degrees to make it back to the 2nd stage inlet eye.