Help me understand the splitter plate...

[Broncazonk]

The splitter plate on a trans / supersonic jet aircraft 'split's off the boundary air flowing along the aircraft's fuselage thereby improving the engines performance. Okay, I've got that, however, once air enters the intake doesn't another boundary air situation arise inside the intake? Seems like the inside air going down the outside edge of intake would form another boundary air problem?

Next question: Can you get too much air inside the intake? I can image a situation where so much air is inside the intake that it compesses.

Next question: What happens to the air going behind the splitter plate?

Thanks!!

Bronc

 

[saintkatanalegacy]

they use intake bleed to solve that B)

it's like a cheese grater in the inlet that sucks air

 

[Broncazonk]

Help me understand the splitter plate...

Baby steps, Saintkatanalegacy, baby steps. Would you please expand on that?

Bronc

 

[Alber Ratman]

The whole purpose of the intake is to slow down the air from transonic/supersonic speed and compress it! Air at the first compressor face cannot be at a speed of above Mach 0.5. The ram effect of the intake at speed helps in some part to increase dynamic pressure and reduce dynamic velocity and a pitot intake (like the Jaguar) will work comfortably to about Mach 1.4. Above that speed, intake design needs to be more clever, using shockwaves produced by ramps or cones to reduce speed of air in the intake. As for too much air, rarely a problem due to the intake although the Jaguar Adours were very susceptable to surging with unstable airflow into the intake as seen in spins etc.

 

[Bill Walker]

The boundary layer grows in thickness the longer the airflow has been going past a fixed object (like the fuselage in front of the inlet). A boundary layer can start to grow inside the duct downstream of the inlet, but good duct design will be increasing the dynamic pressure in the flow as the air passes along the duct (as the other post pointed out) which keeps this to a minimum. Modern designers like to keep the ducts short for this reason, this is one of the reasons you don't see F-100 or MiG-19 style inlets much any more.

Too much air in the inlet can become a problem, leading to nasty engine behaviour like surging. This is why you will see variable geometry inlets, with the change in geometry being a function of both aircraft speed and engine throttle setting. Some aircraft have also used spring loaded bypass outlets, that open up to allow some air in the duct to spill out before reaching the engine if the pressure in the duct becomes too high.

The air behind the splitter plate usually just carries on along down the fuselage. The shape of the fuselage behind the splitter will generally try to move the air smoothly without causing too much drag. This is hard to describe in words, but picture the piece of fuselage holding the splitter plate as a "V" lying on its side, pointy end facing forward. Some air goes up one side of the V, the rest goes down the other side. This length of this V is roughly the same as the height of the splitter plate.

 

[saintkatanalegacy]

this

 

[Alber Ratman]

The boundary layer grows in thickness the longer the airflow has been going past a fixed object (like the fuselage in front of the inlet). A boundary layer can start to grow inside the duct downstream of the inlet, but good duct design will be increasing the dynamic pressure in the flow as the air passes along the duct (as the other post pointed out) which keeps this to a minimum. Modern designers like to keep the ducts short for this reason, this is one of the reasons you don't see F-100 or MiG-19 style inlets much any more.

Too much air in the inlet can become a problem, leading to nasty engine behaviour like surging. This is why you will see variable geometry inlets, with the change in geometry being a function of both aircraft speed and engine throttle setting. Some aircraft have also used spring loaded bypass outlets, that open up to allow some air in the duct to spill out before reaching the engine if the pressure in the duct becomes too high.

The air behind the splitter plate usually just carries on along down the fuselage. The shape of the fuselage behind the splitter will generally try to move the air smoothly without causing too much drag. This is hard to describe in words, but picture the piece of fuselage holding the splitter plate as a "V" lying on its side, pointy end facing forward. Some air goes up one side of the V, the rest goes down the other side. This length of this V is roughly the same as the height of the splitter plate.

The Harrier had spring loaded bypass doors on the internal walls of the intakes for such a purpose of dumping excess air, and the F15 has variable geometry intakes for the reasons you stated. Tornado GR1s had the variable ramps to induce shock wave reduction of intake air speed, but unlike the F3s the mud mover system was disabled in later service life. The original Jaguar prototypes had spitter plates, but production models had plain pitot intakes with no spitter. I'm not sure if this was because of other adverse aerodynamic problems found, or just they were not required. Not a Phantom man, but were the spitters on the Tomb variable? Typhoon has a boundry layer spitter + variable geometry ramps that open wider or close depending on speed and subsquent ram effect. Maybe a solution not to have additional intake doors that other British aircraft seemed to have required!

 

[AeroFranz]

Albert,

the springloaded intakes on the Harrier are there for the opposite reason! At low speed and high power settings, and in particular VTO, the Pegasus draws airflow from everywhere, including behind it. The inlet lips are a compromise between sharp for high speed, and blunt to allow air to "make the turn" at low speed and be swallowed without separating (the original Harrier prototypes -or was it Kestrel?- had an inflatable system that did not work too well). So the springloaded doors open up and bypass the sharper inlet lip, allowing more air in and with a gentler turn. I need to find a picture of this...

 

[Mike Pryce]

AeroFranz,

Alber is, I think, talking about the Harrier's use of boundary layer doors on the inner intake walls, by the cockpit, which close during low speed and hovering flight, to reduce losses, but are open during wing-borne flight to bleed off the boundary layer. There is no appreciable boundary layer at low/no airspeed.

One of these doors can be seen open, as a vertical line just ahead of the Pegasus's fan 'bullet', here:
http://www.airliners.net/photo/UK---Navy/British-Aerospace-Sea/1156372/L/

and closed here:
http://gb.fotolibra.com/images/previews/206224-harrier-nose-and-intake-detail.jpeg

I think what you are talking about are the eight blow-in doors on each outer intake wall, which work as you say to increase mass flow. These are open when the boundary layer doors are shut. All the doors, boundary layer and auxiliary intake, were spring loaded and opened by air pressure effects sucking them open, the boundary layer ones due to airflow over their exit apertures behind the canopy.

The same boundary layer doors were to be used on the supersonic P.1154RAF, but it was later thought, after cancellation, that they would have choked at supersonic speeds due to shockwaves, and an inlet redesign with a splitter might have been needed.

 

[Broncazonk]

this

Yep. Thanks for the drawing. I can see it now.

Bronc

 

[Broncazonk]

Why are the latest intakes rectangular? (F-15, Super Hornet, Su-27, Mig-25, Mig-29) Wouldn't an oval (F-16, F-18 Legacy) or round intake be a better arrangement?

And, what is going on with the YF-23 (photo) is that a Spitter plate as opposed to a Splitter plate?

Bronc

 

[grcoffman104]

Ok. While we're on splitter plates. Please explained the vision splitter plate if front of the pilot on an f-102.




Thanks!




GRCoffman 104

 

[Broncazonk]

Please explained the vision splitter plate if front of the pilot on an f-102.

What is a "vison splitter plate?"

 

[Sundog]

The YF-23 didn't have a splitter plate, what you see in front of the inlet is a porous material to suck off the boundary layer.

As for modern intakes, the simplest intake, such as those on an F-100, F-102, F-16, and F-18 are known as pitot inlets, as they create a shock wave perpendicular to the airflow for pressure recovery. They are the simplest intakes, but also the least efficient. They are also called one dimensional intakes because the shock wave in them is just a vertical shock.

Next up is the 2D (Two dimensional) oblique shock inlet, as first seen on the A-5 Vigilante, but also used on the MiG-25, F-14, F-15, MiG-29, Su-27 and Tornado. They set up shocks which are swept back in the vertical and they tend to have moveable ramps to match the airflow and they're more efficient than a pitot inlet. The F-4 Phantom and some other aircraft also have a 2D inlet, but the ramps on the F-4 are built into the boundary splitter plate. You can see the piano hinge for it right down the middle of the splitter plate in side view. The XB-70 also had oblique shock inlets.

Next up, we have 3D shock structures, which are swept in both the vertical and horizontal plane, like the diamond inlets on the F-18 Superhornet and the F-22. They are more efficient than a 2D inlet and the main reason such complicated structures are used is to get high efficiency from a fixed (no moving parts) inlet. They don't want moving parts for stealth reasons. Moveable parts change the signature and make it much more difficult to maintain a low signature throughout the entire flight regime. The YF-23 has trapezoidal 3D inlet as well.

Then you have a special kind of oblique shock inlet that it is swept around a conic centerline. This would be the shock cone inlet (e.g.-SR-71 and B-58) and the half shock cone inlet (e.g.-Mirages and the F-104).

Then you have the super specialized DSI (Diverterless Supersonic Inlet) which is what the F-35 has as well as some of the newer Chinese Fighters (J-10B & J-20) This is a case of once again of optimizing the inlet for both supersonic flight and L.O., only now they are getting rid of the boundary layer splitter plate (Bad for stealth) and using the cone like hump on the fuselage to split the boundary layer and divert it away from the inlet.

As for the vision splitter in the F-102 and F-106, this was to prevent reflections from the two forward window panes from disorienting the pilot and making it difficult to see; it blocks the internal reflections between the two panes.

 

[saintkatanalegacy]

and we haven't even discussed the alpha and beta sensitivities of those inlets ;D

really brings back a lot of memories of digging books in the library

regarding the Harrier inlet, I seem to remember having inflatable "lips" tested

 

[Bill Walker]

A recent addition to the intake designers list of things to worry about is radar reflections. I suspect this is partly responsible for some of the modern rectangular intakes. A small loss of aerodynamic and structural efficiency is made up for in big gains in stealth. Aircraft design is always a series of trade offs.

I think the inflatable lips were tested on the Kestrels.

 

[Tailspin Turtle]

I think the inflatable lips were tested on the Kestrels.

To add a bit to harrier's correction to AeroFranz (I had initially misread it as well) and a picture: In order to maximize engine thrust at zero airspeed, the Pegasus engine wanted to be fed with a bellmouth-shaped inlet. This, of course, adversely affected drag/thrust at high speed flight. The inflatable inlet was the first approach. The final solution was the spring-loaded suck-in doors on the exterior of the fuselage, which opened to provide what amounted to a bellmouth shape, a very clever solution.

Other inlet thoughts:

The first attempt to navalize the F-86 Sabre—in addition to adding a tailhook, folding wings, etc—included a slightly bigger inlet in order to maximize low-speed thrust at the expense of high-speed thrust.

The F-105 had an auxiliary inlet in the main landing gear wheel well to increase mass flow to the engine at low speeds.

The F8U-3 was designed with another approach to a variable-geometry inlet. To handle the excessive airflow, there were large doors on the side of the fuselage to dump inlet air at high speeds. (If I remember correctly, there was also one small dump door on the F8U-1/2.) They were also an auxiliary air inlet at low speeds. Because of the length of radome ahead of the inlet, the F8U-3 had slots across the upper side of the intake to suck off the boundary layer air.

 

[AeroFranz]

My bad! I wasn't even aware of the presence of BL doors on the inner intake walls. It is true that they are a fairly particular installation, most subsonic aircraft would remove the BL ahead of the inlet. Well, this is turning out to be a most educational thread
Nice picture of the XF8-U, Tommy. I think it's time I paid visit to the Ginterbooks website again ;D

 

[shockonlip]

I think the inflatable lips were tested on the Kestrels.

...

The F8U-3 was designed with another approach to a variable-geometry inlet. To handle the excessive airflow, there were large doors on the side of the fuselage to dump inlet air at high speeds. ...

Because of the length of radome ahead of the inlet, the F8U-3 had slots across the upper side of the intake to suck off the boundary layer air.

Some additional info on the F8U-3:

The doors Tommy is talking about are known as the bypass doors. And their normal function was
to match freestream pressure ratio (StaticPressure/TotalPressure) to duct pressure ratio.
There was a sensor for each and the signals were compared creating an error signal, which
drove a cam that positioned the bypass doors to correct duct pressure ratio.

The BL slots that Tommy showed us, via one of his pictures, was next to (downstream of) the
inlet throat. The normal shock sat in the throat durring normal inlet operation.
The BL slots therefore also helped to reduce pressure differences between the duct and inlet
trying to cause the normal shock to be swallowed. On the other side of the inlet throat was the
sugar scoop inlet, which from its cut-away shape, also relieved over-pressure coming from the
duct trying to unstart the inlet by pushing the normal shock forward, out of the throat.

When Chance Vought created a duct dynamics math model, and did testing, they found
that the normal shock position was fairly stable. Unfortunately this work was late in the life of the
F8U-3 and they never were able to finish flight testing and development of their encouraging
duct dynamics work.

Another note on the F8U-3's bypass doors, In Bill Lawrence's (one of the US Navy's F8U-3 test
pilots) biography, he points out that the F8U-3's bypass doors had to be both operated in synch
or there would be a large yaw on the F8U-3's airframe.

In fact Tommy, do you have an online photo of then Lt. Bill Lawrence in pressure suit standing
next to one of the F8U-3's at Edwards? I've seen such a photo in your excellent F8U-3 book.
No problem if you can't of course. Thanks!

As a side note:
Bill Lawrence, who unfortunately passed away in 2005, retired from the USN as a Vice Admiral.
He had quite a career.

He was a finalist for the Mercury 7 astronauts, an F-4 squadron commander during the Vietnam
War, was shot down, and spent a number of heroic years as an excellent leader to other
US POWs in North Vietnam (his slogan: "Never Give In"). He was also a fleet commander later
in life before retirement.

Quite an amazing guy!

His daughter, also a US Navy officer (retired), succeeded as an astronaut by flying on a
number of Space Shuttle flights as a mission specialist. There is a new DDG named after
Vice Adm. Bill Lawrence today, U.S.S. William P Lawrence, DDG-110.

 

[Broncazonk]

On the other side of the inlet throat was the
sugar scoop inlet....

Can the sugar scoop inlet of the F-8U3 be considered a three-dimensional inlet? It's non-linear geometry fits the description and definition quite nicely. One other question: Does it make a difference whether the scoop part is under the intake inlet or can it be over (upside down to the F-8U3) or outside (scooping the air) if a pair of them were mounted on the sides of the aircraft?

After everyone got to talking about it, I bought, "Naval Fighters Number Eighty-Eight Vought F8U-3 Crusader III" by Tommy Thomason and that is a really good book. My next purchase is going to be number Sixty-Four, A-5A Vigilante.

Bronc

 

[Orionblamblam]

Broncazonk:

I notice that all your posts have the "Re:" removed from the Subject line. Do you do this? If so, why?

 

[Broncazonk]

Because I am 9/10's overpowered by Obsessive-Compulsive Disorder. The "Re:" gets removed from email subject line and everything else. Also a knife, and only a knife, goes in the peanut butter, and a spoon, and only a spoon, goes in the jelly. There is a very long list....

Obsessive–compulsive personality disorder (OCPD) is a personality disorder characterized by a pervasive pattern of preoccupation with orderliness, perfectionism, and mental and interpersonal control at the expense of flexibility, openness, and efficiency. Yep. that's pretty much me.

Bronc

 

[Tailspin Turtle]

The photo of Bill Lawrence in the pressure suit in my F8U-3 monograph was added by Steve Ginter, so I don't have it to post. Sorry.

 

[shockonlip]

On the other side of the inlet throat was the
sugar scoop inlet....

Can the sugar scoop inlet of the F-8U3 be considered a three-dimensional inlet? ...

Does it make a difference whether the scoop part is under the intake inlet or can it be
over ...
outside ...
a pair ... mounted on the sides of the aircraft?
...
Bronc

So let's look at the inlet for a moment.

What do we see?

We see a cone above a sugar scoop at the front of the airplane.

Both the cone and the sugar scoop will generate a shock.

The cone will generate a cone shaped shock. This shock is definitely
a 3-D shock. You should get a good basic aerodynamics book. I
would recommend one of John D. Anderson's books. I really like
his "Modern Compressible Flow", but he has many books that are
good. The losses in the flow through a 3-D cone shock are smaller
than the losses through a 2-D wedge shock at the same angle.
For example, flow through a cone shock generated by a 30 deg. cone
are less than losses due to flow through a 2-D wedge shock caused
by a 30 deg wedge.

The sugar scoop will also generate a 3-D shock.

And then at the throat of the inlet will be the normal shock that takes the
flow subsonic.

Now inside the inlet there is a duct from the throat back to the turbojet's
compressor face.

Because the cockpit is on the top part of the airframe, at the front, that
duct must run through the bottom part of the airframe.

If at the front of the airplane, you put the sugar scoop on the top, and the
cone at the bottom, or the sugar scoop on the sides, the duct we are
discussing would have a tougher time clearing the cockpit volume, plus the
duct would be longer, and more complex adding drag and complexity.

The radar for the production versions of the F8U-3 fit inside the cone, and I
am sure there were benefits from that because the cone is on the top, right
in front of the cockpit and its associated electronics and instrument panel.

At higher mach, there is potentially cooling that has to be done, perhaps for
the electronics. So having this stuff close together probably has
benefits as well because you can cool one electronics section that
is all together, instead of an electronics bay in the cockpit and a radar
electronics bay on a bottom mounted cone.

And the purpose of the inlet is to pass a certain air mass flow rate to the
turbojet. So calculations are done that determine how wide and high the inlet
has to be to pass the required air mass flow rate at different flight speeds
and throttle settings.

I hope this helps.

So it isn't necessarily that something can be done, but does doing that fit in
with everything else.

 

[shockonlip]

The photo of Bill Lawrence in the pressure suit in my F8U-3 monograph was added by Steve Ginter, so I don't have it to post. Sorry.

Thanks for responding TT!

 

[Orionblamblam]

So the "Re:" gets removed from email subject line

Can't speak for others, but it's annoying. The "Re:" let's you know at a glance whether or not a post is part of an existing thread or an all-new one. By removing it, you're basically gaming the system.

 

[Broncazonk]

Can't speak for others, but it's annoying. The "Re:" let's you know at a glance whether or not a post is part of an existing thread or an all-new one. By removing it, you're basically gaming the system.

I will do my best to stop doing it.

Bronc

 

[Orionblamblam]

Re: Re: Re: Help me Re: understand Re:the Re: splitter Re: plate... Re: Re: Re:

Re:

I will do my best to stop doing it.

So far not so good...

Re:

 

[Tailspin Turtle]

On the other side of the inlet throat was the
sugar scoop inlet....

Does it make a difference whether the scoop part is... a pair of them were mounted on the sides of the aircraft?

Bronc

The F-105 had forward swept inlets on the sides of the aircraft. However, unlike the F8U-3, it had a sliding plug to vary the inlet area. The plug was fully forward at Mach 2. See http://www.clubhyper.com/reference/f105jr_1.htm. It also had and auxiliary engine air inlet in the main landing gear wheel well to provide more air at low speed for maximum thrust on takeoff and go-arounds.

 

[AeroFranz]

The auxiliary F-105 inlet is plainly visible on the bird at NASM Udvar-Hazy (although few docents seem to know its purpose). It's rather small, and circular, with a mesh to prevent FOD (that's my guess).
The swept forward inlet was the brainchild of Antonio Ferri, of NASA (AFAIK). Ferri and Katveli had previously collaborated on ramjets and the XF-103 sported the same swept-forward inlet.

 

[Mark Nankivil]

Excellent thread - thanks to all. Add to this the inlet on the F-35 and and JF-17 - a diverterless supersonic intake. I believe this was first tested on the Grumman F-11F-1F Super Tiger. For the F-35, it was tested on a modified F-16.

An interesting paper on subsonic UCAV inlet design:

www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA428819


Enjoy the Day! Mark

 

[DSE]

The cone will generate a cone shaped shock. This shock is definitely
a 3-D shock. You should get a good basic aerodynamics book. I
would recommend one of John D. Anderson's books. I really like
his "Modern Compressible Flow", but he has many books that are
good. The losses in the flow through a 3-D cone shock are smaller
than the losses through a 2-D wedge shock at the same angle.
For example, flow through a cone shock generated by a 30 deg. cone
are less than losses due to flow through a 2-D wedge shock caused
by a 30 deg wedge.

But also remember that the conical shock also produces less compression than a 2-d oblique shock for the same amount of flow turning. Addionally the flow behind an oblique shock is uniform, whereas the flow behind a conical shock is only constant along rays which emanate from the vertex. Hence as you move away from the body towards the shock flow properties change with the most compression occurring next to the body. See info regarding the solution of the Taylor-Maccoll eqn, such as www.ryerson.ca/~jvl/2008-09/ae8121/conenotes.pdf or your Anderson reference.

 

[AeroFranz]

...so the problem being pressure distortion at the compressor face?

 

[shockonlip]

The cone will generate a cone shaped shock. This shock is definitely
a 3-D shock. You should get a good basic aerodynamics book. I
would recommend one of John D. Anderson's books. I really like
his "Modern Compressible Flow", but he has many books that are
good. The losses in the flow through a 3-D cone shock are smaller
than the losses through a 2-D wedge shock at the same angle.
For example, flow through a cone shock generated by a 30 deg. cone
are less than losses due to flow through a 2-D wedge shock caused
by a 30 deg wedge.

But also remember that the conical shock also produces less compression than a 2-d oblique shock for the same amount of flow turning. Addionally the flow behind an oblique shock is uniform, whereas the flow behind a conical shock is only constant along rays which emanate from the vertex. Hence as you move away from the body towards the shock flow properties change with the most compression occurring next to the body. See info regarding the solution of the Taylor-Maccoll eqn, such as www.ryerson.ca/~jvl/2008-09/ae8121/conenotes.pdf or your Anderson reference.

Yes, I am familiar with this.

I personally think this is much more detail than was asked for, but since you've started this, let's make sure they understand where you're going.

One can think of these rays of constant flow properties emanating from the cone vertex, from the cone shock down to the cone surface, as compression Mach waves,
or successively smaller cone-shaped Mach waves, which as a flow streamline continues to pass through these successively smaller compression mach wave cones, that
interaction continuously turns the streamline to eventually be parallel with the cone surface at infinity. The cone shock is also straight implying other aspects, but I think
that's getting too detailed.

 

[shockonlip]

...so the problem being pressure distortion at the compressor face?

No.

You're confusing the mach wave cones themselves, or the rays of constant property waves,
with a flow streamlne penetrating these waves.

See the picture I enclosed which portrays the cone shock and the mach waves and
a flow streamline.

A flow streamline penetrating the shock from the outside, will penetrate a number of these
infinitesimal strong mach wave cones curving inward as it interacts with these successively
smaller mach wave cones, until eventually it is parallel to the cone surface.

 

[shockonlip]

Another aspect I neglected to mention which I also think may be a factor, but I'm not
actually sure about, and which I haven't seen mentioned in a propulsion book, is the
mixing and turbulence and impact of other inlet components on the flow between the
cone and the compressor face. In other words, if there was some pressure or temperature
distortion even after streamline penetration of the mach wave cones, I wonder how long
it lasts.

 

[Broncazonk]

Has anyone experimented with a NACA duct / scoop in front of (in the fuselage next to) the jet intake / inlet? The idea would be for a low-drag NACA duct to strip off the boundary layer without the need of a splitter plate.

Bronc

 

[Bill Walker]

From my experience using NACA ducts in low speed applications, I think you would suck off a lot more than just the boundary layer. Also, the length to width (or height) ratio of a NACA duct is set by NACA 5102 and so on. Making a duct wide (or high) enough to cover your inlet may reuslt in a veeerrry long duct.

 

[Broncazonk]

Making a duct wide (or high) enough to cover your inlet may reuslt in a veeerrry long duct.

Understand. Could smaller ones be lined up vertically to span the distance?

How efficient / effective are NACA ducts, meaning: how much drag do they create anyway?

Bronc