reported earlier P-51 Flap Blowup Characteristics

[Magic Zach]

TL;DR first: The flaps for the DCS P-51D get blown back up at high speed, and do not stay fixed in place like they should.  There is an unusual though inconsistent exception however, when airspeed is lower than necessary with the engine idled.

THE EXPECTATIONS:
First let's establish what is expected for flap behavior in the P-51.
I'll start with the mechanical linkages first:
The P-51 actually has a rather unique flap mechanism for the fighters of the time, and are rather automatic in nature.  Aircraft that used hydraulic power to move their flaps typically required the pilot to manually move their flap control lever in the cockpit up or down to actively make the flap strut actuate the flaps up or down.  The pilot would then return his flap control handle to the neutral position to stop the flap movement.  This is the case for the flap control in the P-47, Mosquito, and ( believe) the P-38.  If you don't have or know these aircraft, you can see how these flap handles operate here, at 11:09:

Now you may be wondering why the P-51's flap controls work so automatically, without requiring the pilot to determine the time when he needs to stop the flap movement.  It's all in the linkages:

Spoiler

When you push down on the flap handle to lower the flaps, this pulls the Follow Up Control Rod forwards, and thus hinge/point #1 forwards.  However, hinge #3 is somewhat fixed in its position.  As the Follow Up Control Arm (not the rod) is pulled forwards at hinge #1, hinge #3 acts as the fulcrum around which the Control Arm will rotate.  Note that the Follow Up Control Arm is a solid linkage that extends from point #1 to point #3, and does not bend at point #2.  Connected to the center of the solid Follow Up Control Arm is the Control Valve Rod Assembly, at hingepoint #2.  As the Follow Up Control Arm is pulled forwards at point #1, a part of this movement is transferred to the Control Valve Rod Assembly, which is connected directly to the Wing Flap Control Valve's lever.  This directs the valve to allow fluid to pass to the strut, which will push on the Torque Tube Arm, turning the Torque Tube.
But how does the flap know when to stop actuating automatically?  Well, as the torque tube rotates, it will pull hinge/point #3 aft via the Follow Up Control Linkage.  This brings hinge/point #2 back to its previous neutral position, and thus the Flap Valve's lever back to its neutral position.  This kind of automatic functionality allows the pilot to predetermine a desired setting for the flaps, and the torque tube (and flaps) will neutralize its own controls once it's found the new equilibrium between points #1 and #3, in order to bring #2 back to its neutral position.
Of interesting note, the Flap Valve's lever acts essentially exactly as the flap control handle in the P-47, Mossie, and P-38, as they are usually directly linked 1:1 to each other.
In case the official drawing above is too cluttered to follow, I also made these simplified drawings to try to show how this automatic function of the flaps works.

Spoiler

The red zig-zags show which controls/hinges are currently static, or will not move or rotate.

I want to establish here that I did make a mistake with drawing the hydraulic flow to the actuator strut, through the valve.  It does not flow from a "raise" side to the "lower" side.  I wanted to include this as just additional fluff to the drawing, but seeing how I drew it wrong here, ignore that.  The important part here is seeing which directions the mechanical linkages move.

Now we can start on the hydraulics:

Spoiler

(Source: P-51D-5 Erection and Maintenance manual)

Note that I won't cover the entire layout of the hydraulics, just the parts pertaining to flaps operation.
The engine driven pump will continuously supply pressured fluid to the Unloading and Relief Valve at 1500 PSI.  However, the Unloading and Relief Valve will maintain the hydraulic system's pressure at a lower 1050 PSI (specifically the manual says 1000-1100 PSI).  When the system pressure reaches 1050 PSI, the unloading valve will open, creating an idling circuit just between the resevoir, pump, and Unloading and Relief Valve, isolating the hydraulic system from the pump until there is a demand on the system that consumes pressure.
The wing flap selector valve does not look like this as pictured, but this is merely a visual demonstration to show roughly how the flow of fluid works.  Take note of the check valves on the pressure and return feed ends on the flap valve.
This is how the Wing Flap Selector valve operates the flaps circuit:

Spoiler

And of course, I've made another drawing that simplifies this, pertaining to just the flap operations:

Spoiler

Lowering the flaps:

Raising the flaps:

Flaps once they reach their preset angle (blue represents stagnant fluid):

Now, with the the hydraulic and mechanical systems in mind, what should occur if the flaps are mismanaged and oversped beyond their limits?
With no relief and bracketed with check valves, the hydraulics will not yield.  The weak point is the mechanical linkages.
If you can imagine an immense pressure exerted on to the flaps, attempting to push them up, this will apply direct pressure on to the hinges of the torque tube arms.  The weakest point taking that pressure are the bolts within the hinges, which could shear.  These hinges were noted to be the weakest link in the flaps' strength.  As such it was designed so that in the event of overpressure on the mechanical system, the first hinge (bolt) to shear would be that connecting the torque tube to the hydraulic strut.  The two hinge bolts used to connect the torque tube to the two flaps were larger in diameter than the one hinge bolt connecting the strut to the torque tube.  It was critical that any failure was to occur here, as this ensured that both flaps would simultaneously fail in sync.  This helped to prevent assymetric drag and lift throughout the aircraft's flight.
If the flaps are oversped and the hinge fails (bolt shears), then both flaps will weathervane with the direction of wind.  So, if they were previously down when being overspeed, they will both snap back upwards at most speeds.  When coming in for landing, or generally very slow however, the flaps will drop on their own, as the wind force on them decreases.

Spoiler

^NOTE FOR THE ABOVE: even though the drawing makes it appear like the bolt #41 is a hexagonal type, it is in fact a clevis bolt.  This appears to have been a mistake on the artist's part.

THE DCS BUGS:
In the DCS P-51's flap behavior, there are a couple behaviors that conflict with the mechanics and hydraulics of the real aircraft.

1) Flaps blown up when they're oversped, while strut is currently actuating flaps down.
What would prevent this in reality would be the two check valves that are immediately upstream and downstream of the Wing Flap Selector Valve, to prevent flow reversal.  These would be what would stop the flaps from getting blown back up when oversped, while the strut is currently attempting to actuate the flaps down.  This results in the flaps not moving until airspeed decreases, or the flaps snapping lose when the strut's hinge bolt shears.
2) Flaps blown up when they're oversped, after strut is finished actuating flaps down.
There are now two measures in place to prevent fluid from flow reversal.  Now not only are the check valves upstream and downstream (pressure and return) of the Wing Flap Selector Valve in place, but the Wing Flap Selector Valve itself will have all its poppet valves closed.  This will isolate the wing flap circuit from the main circuit, holding the fluid in place.  The flaps will not move, unless the flaps are oversped and snap lose when the strut's hinge bolt shears.
3) The DCS exception...
However in DCS currently, if A) your engine is idled; and B) your airspeed is >100MPH, the flaps will have a chancec of finally behave correctly, and could remain fixed in place when you dive with them.
Why is it odd?  Because no matter the other external factors (aside from the pilot moving the flap handle in the cockpit), so long as the flaps achieve the angle they were determined to match, they will "lock" (the flap selector valve will close).  This, in testing, appears to be possible with full flaps as fast as 150MPH IAS.  Airspeed (and the force against the flap surfaces) is what will affect if the hydraulic system has the pressure required to lower the flaps.  What affects if the flaps will "lock" or not, is if the flaps reach their predetermined setting or not.  Not only is 100MPH IAS too slow, but also it's only an indirect reason for determining if the flaps should "lock."

Here's a video highlighting the issues:
TURN ON SUBTITLES
Timestamps of interest: 0:30 2:40 7:15

P-51 Flap Blowup test.trk

 

CORRECTIONS:
1) The flaps should NOT get blown back up uncommanded under ANY circumstances, short of a component failure.  The check valves and flap selector control valve ensure that fluid will not give or reverse flow to permit flaps to blow up against excessive force on the flap surfaces.
2) Add a component failure.  As flaps would be oversped while deployed beyond limits, the most likely point of component failure will be the torque tube arm hinge between the strut and the torque tube.  If the bolt in this hinge were to shear, this would cause both of the flaps to suddenly weathervane into the wind, and essentially flap freely.  This would create some interesting problems when it would come to landing.
Slapbladder if you have read this far, message me in Discord with the codephrase: Blue sky

Edited December 21, 2022 by Magic Zach
12/9/2022: included hinge bolt info /// changed likely point of breakage from torque tube arm to torque tube arm hinge bolt

[grafspee]

I agree with 2 and 3 point, flaps definitely should return to set position, and not remain at position where airflow blow then up. Unless it is some kind effect of mechanical damage, like rods being bent for example or speed is still too high so actuating mechanism don't have enough to force to overcome airflow.

When valves are open actuating cylinder has as much force as oil pressure, at some point this may not be enough to overcome dynamic pressure acting on flaps surfaces.

Edited November 25, 2022 by grafspee

[Magic Zach]

16 minutes ago, grafspee said:

 

When valves are open actuating cylinder has as much force as oil pressure, at some point this may not be enough to overcome dynamic pressure acting on flaps surfaces.

 

It may not be enough to overcome the pressure on the flap surfaces, however, the check valves will still prevent the flaps from being blown up.  The strut will attempt to push the flaps down, however it cannot, and the check valves ensure the flaps circuit doesn't encounter flow reversal.  So the flap surface will just stay in place until either A) airspeed decreases enough that the hydraulic force overcomes the wind force on the flap surfaces; B) component failure; C) Pilot raises flaps

[Skewgear]

What's the TLDR?

[grafspee]

2 hours ago, Magic Zach said:

It may not be enough to overcome the pressure on the flap surfaces, however, the check valves will still prevent the flaps from being blown up.  The strut will attempt to push the flaps down, however it cannot, and the check valves ensure the flaps circuit doesn't encounter flow reversal.  So the flap surface will just stay in place until either A) airspeed decreases enough that the hydraulic force overcomes the wind force on the flap surfaces; B) component failure; C) Pilot raises flaps

But look at accumulator line this one don't have check valve because both way flow is required on this line so flaps can be blow up while actuating is on.

[Magic Zach]

30 minutes ago, grafspee said:

But look at accumulator line this one don't have check valve because both way flow is required on this line so flaps can be blow up while actuating is on.

The accumulator is completely unrelated to the flaps operation, unless the pump (and engine) is not operating.

Edited November 25, 2022 by Magic Zach
and engine

[Skewgear]

I'm inclined to think @Magic Zachis right.

P-51D E&M manual here: https://ww2aircraft.net/forum/threads/mustang-manuals.9051/#post-241325

Hydraulic circuit is split between PDF parts 1 and 2 and some pages are out of order, very annoyingly.

Figure 309 (pp264, PDF part 1 page 292, copy attached below) is the main hydraulic system schematic. It shows two check valves fitted to the main system reservoir. It also shows check valves fitted to the supply and return pipes immediately upstream of the flap selector valve.

Figure 329 (pp284, PDF part 2 page 13, copy attached below) is the flap system schematic. It starts from the flap selector valve and shows everything downstream of it. There are no check valves downstream of the flap selector valve, just a flow restrictor.

--

When the flap system is in a steady state (i.e. flaps are stationary at their selected position) all four poppets in the flap selector valve are closed, as shown to the bottom right of Figure 309. This means hydraulic fluid cannot flow. Therefore air pressure on the flaps cannot back drive the system through the flap operating strut shown on the right of Figure 329.

In this scenario an overspeed should lead to hydraulic overpressure in the flap system downstream of the flap selector valve. The flaps will press on the operating strut until something gives way.

If I've understood the schematics correctly, the only place for excess hydraulic pressure is bursting one of the pipes or joints shown in Figure 330 (attached). The schematics do not show any pressure relief valve downstream or inside the flap selector valve, a partial exploded drawing of which is at Figure 331 (attached).

This should lead to total loss of hydraulic pressure. Edit - but that depends on the max pressure the flap system will accept before catastrophic failure, which must be a design feature of the system and also depends on airspeed. It seems sensible to me that the flap system will withstand pressures well above those in the pilot's notes.

Expressed in a different way, the system must withstand the max flap limiting speed plus a safety margin.

Edited November 26, 2022 by Skewgear

[Magic Zach]

 

14 minutes ago, Skewgear said:

P-51D E&M manual here: https://ww2aircraft.net/forum/threads/mustang-manuals.9051/#post-241325

Yerp.  Here's a color coded version of that last photo btw:

[Skewgear]

I have done some more thinking about this and actually I think the way it is modelled in-game is almost right. To move from abstract discussion of real world systems to how failures work in DCS:

1. The failure mode for flaps after an overspeed should be sudden failure of the flap system to keep the flaps at their commanded position. The flaps should be pushed by airflow back to the fully retracted position when the system fails. This is already modelled in-game as the flaps being blown back up during an overspeed.

1a. What component of the flap system fails first in an overspeed scenario? The hydraulic system can accept very high pressures: the P-51 E&M manual mentions applying 2000psi from an external source during some test procedures. Therefore it seems likely that in an overspeed, a physical component in the flap system will fail (break) first rather than the hydraulics. I guess perhaps one of the hinge pins in the torque tube arm assemblies (figure 155, in Magic Zach's first post) would be a likely point of failure. Perhaps a check valve upstream of the flap selector valve (FSV) might fail if an FSV poppet is forced open by backpressure from the flaps, but I find this unlikely.

1b. Once the flaps have been oversped and blown back to the retracted position, the flap system cannot be used again. Is this modelled in DCS? If so, then in my view the flap system overspeed scenario is correctly modelled.

2. After an overspeed causes failure of the flap system, the flaps should droop at low airspeeds. This seems to already be modelled below 100mph.

[Magic Zach]

15 hours ago, Skewgear said:

I have done some more thinking about this and actually I think the way it is modelled in-game is almost right. To move from abstract discussion of real world systems to how failures work in DCS:

1. The failure mode for flaps after an overspeed should be sudden failure of the flap system to keep the flaps at their commanded position. The flaps should be pushed by airflow back to the fully retracted position when the system fails. This is already modelled in-game as the flaps being blown back up during an overspeed.

1a. What component of the flap system fails first in an overspeed scenario? The hydraulic system can accept very high pressures: the P-51 E&M manual mentions applying 2000psi from an external source during some test procedures. Therefore it seems likely that in an overspeed, a physical component in the flap system will fail (break) first rather than the hydraulics. I guess perhaps one of the hinge pins in the torque tube arm assemblies (figure 155, in Magic Zach's first post) would be a likely point of failure. Perhaps a check valve upstream of the flap selector valve (FSV) might fail if an FSV poppet is forced open by backpressure from the flaps, but I find this unlikely.

1b. Once the flaps have been oversped and blown back to the retracted position, the flap system cannot be used again. Is this modelled in DCS? If so, then in my view the flap system overspeed scenario is correctly modelled.

2. After an overspeed causes failure of the flap system, the flaps should droop at low airspeeds. This seems to already be modelled below 100mph.

1. While I'd agree that the flaps should fail in an overspeed condition, I don't agree that they should be blown smoothly back up without breakage, as they are in DCS currently.  There's no allowance for hydraulic fluid to flow in this scenario, once flaps acheive their preset angle.  To be clear, the flaps getting blown up during overspeed in DCS is not a failure of the flaps, they are merely being blown back up and return to their position once airspeed will drop.
1a. Agreed, hydraulics wouldn't fail (this contradicts with your first point though..?).  A more mechanical component seems likely to fail, and it's likely the torque tube arm attaching the strut to the torque tube, as mentioned in my final notes in the post and yourself.
1b. If you look at the video, you'll see this isn't the behavior in DCS.  Flaps are undamaged by overspeeding, and can be used repeatedly.  Even in the very odd case that they do actually lock.
2. This is not modeled.

Edited November 28, 2022 by Magic Zach

[Skewgear]

10 hours ago, Magic Zach said:

1. While I'd agree that the flaps should fail in an overspeed condition, I don't agree that they should be blown smoothly back up without breakage, as they are in DCS currently.  There's no allowance for hydraulic fluid to flow in this scenario, once flaps acheive their preset angle.  To be clear, the flaps getting blown up during overspeed in DCS is not a failure of the flaps, they are merely being blown back up and return to their position once airspeed will drop.
1a. Agreed, hydraulics wouldn't fail (this contradicts with your first point though..?).  A more mechanical component seems likely to fail, and it's likely the torque tube arm attaching the strut to the torque tube, as mentioned in my final notes in the post and yourself.
1b. If you look at the video, you'll see this isn't the behavior in DCS.  Flaps are undamaged by overspeeding, and can be used repeatedly.  Even in the very odd case that they do actually lock.
2. This is not modeled.

 

It seems most likely to me that the intended point of failure within the flap system during an overspeed is one of the hinge pins in the torque tube linkage assemblies - in fact given the lack of a pressure relief valve in the flap sub-system, I'm sure the pins must have been designed to shear at a given load, that load being well below overpressure and failure of any hydraulic system component in the flap circuit. The alternative is flap overspeed causing total hydraulic system pressure loss through catastrophic component failure, which is not something a sensible designer would choose.

Once that pin shears, the flap system cannot be commanded by the pilot or the flap control valve self regulating mechanism. Nothing holds the flaps in place so they, er, flap around. It doesn't matter if that's smooth or not, though it would be a big coincidence if the pins for both sides sheared at exactly the same moment.

Suggested feature change for the P-51:

Flap overspeed scenario.

Current ingame behaviour: The flaps are smoothly pushed back up by airflow during an overspeed. The flaps gradually move back to their selected position when airspeed drops below the overspeed point. The pilot can select the flaps as normal after aircraft speed drops below the overspeed point. No damage is caused to the hydraulic system.

Desired behaviour: the flaps are suddenly and sharply pushed back up by airflow. This should be assymetric at slightly different speeds per side. The flaps remain up until airspeed drops below [some plausible speed, maybe 100mph?]. At that point they gradually extend, uncommanded, unless speed increases again. The flap system is destroyed by overspeeding. The pilot can move the flap selector handle in the cockpit but the system does not respond. No damage is caused to the hydraulic system.

[grafspee]

I think that mechanical sync of the left right flap has to survive. Single flap lose at high speed would disintegrate plane.  So safe guard pin has to be on flap strut. Independent fail of right left flap would be enormous design flaw.

Then if this pin fails flaps would move freely, and at 100mph those would not go down, a little bit yes but not all way down.

And failure would be rather violent  and quite dangerous for plane and pilot.

 

But when flaps slowly giving up looks a like a leak type fail. This would explain why flaps still working at lower speed.

For example if main seal of flap strut fails, it won't be able to hold flaps down at high speed but once speed drops even with internal hydraulic leak system can manage to operate flaps as usual.

 

Edited November 28, 2022 by grafspee

[Magic Zach]

On 11/28/2022 at 11:00 PM, Skewgear said:

It seems most likely to me that the intended point of failure within the flap system during an overspeed is one of the hinge pins in the torque tube linkage assemblies - in fact given the lack of a pressure relief valve in the flap sub-system, I'm sure the pins must have been designed to shear at a given load, that load being well below overpressure and failure of any hydraulic system component in the flap circuit. The alternative is flap overspeed causing total hydraulic system pressure loss through catastrophic component failure, which is not something a sensible designer would choose.

Once that pin shears, the flap system cannot be commanded by the pilot or the flap control valve self regulating mechanism. Nothing holds the flaps in place so they, er, flap around. It doesn't matter if that's smooth or not, though it would be a big coincidence if the pins for both sides sheared at exactly the same moment.

Suggested feature change for the P-51:

Flap overspeed scenario.

Current ingame behaviour: The flaps are smoothly pushed back up by airflow during an overspeed. The flaps gradually move back to their selected position when airspeed drops below the overspeed point. The pilot can select the flaps as normal after aircraft speed drops below the overspeed point. No damage is caused to the hydraulic system.

Desired behaviour: the flaps are suddenly and sharply pushed back up by airflow. This should be assymetric at slightly different speeds per side. The flaps remain up until airspeed drops below [some plausible speed, maybe 100mph?]. At that point they gradually extend, uncommanded, unless speed increases again. The flap system is destroyed by overspeeding. The pilot can move the flap selector handle in the cockpit but the system does not respond. No damage is caused to the hydraulic system.

The pins are bolts, of two types.  You bring up a good point about them though.  It would appear that rather than the torque tube arms failing, it would be the hinge on said torque tube arm that would fail first.
- @grafspee RE synchronous flap failure: This appears to have been ensured by the different bolt widths used between the torque tube arms connecting the strut, and the arms connecting the flaps.  The bolts used to connect the flaps to the torque tube have a larger diameter than those connecting the strut to the torque tube.  This would ensure that if there was an overpressure on the mechanical system, the first hinge to give would be the strut to the torque tube.  This would fail both flaps, but at least both flaps would fail together, connected by the torque tube.
-The inclusion of shear nuts, and the specific difference of bolt diameters between the flap hinges and the strut hinges also would be evidence that the bolts were the intended breakage point.

 

Spoiler

The strut's attachment:


Bolt thread 1/4" dia
Note: bolt (figure #41) is drawn to be an AN5 hexagonal bolt.  However the part list pictured suggests that it's in fact a clevis bolt, like #39.  This is also evidenced by the same type of shear nut being used on both bolts.  This does not change the thread diameters, but it's an inconsistency between the drawings and the parts listed that I thought I better mention, just in case.

Both flap's attachments:


The bolts for the wing attachments are of larger diameter than those used on the strut.
A different bolt/nut assembly is used on the flaps' hinge lug as well.

Also on a neat side note, more detail on these can be found on https://military-fasteners.com/ and https://www.aircraftspruce.com/

 

Edited December 9, 2022 by Magic Zach

[grafspee]

@Magic Zach I agree. If designer put smaller bolt on arm which actuate left and right flap then bolt holding individual flap, conclusion is obvious here.

And if that is a case, DCS P-51 flaps behavior landed very far from how it should behave

Edited December 10, 2022 by grafspee

[HavaFaza]

On 12/9/2022 at 4:05 AM, Magic Zach said:

The pins are bolts, of two types.  You bring up a good point about them though.  It would appear that rather than the torque tube arms failing, it would be the hinge on said torque tube arm that would fail first.
- @grafspee RE synchronous flap failure: This appears to have been ensured by the different bolt widths used between the torque tube arms connecting the strut, and the arms connecting the flaps.  The bolts used to connect the flaps to the torque tube have a larger diameter than those connecting the strut to the torque tube.  This would ensure that if there was an overpressure on the mechanical system, the first hinge to give would be the strut to the torque tube.  This would fail both flaps, but at least both flaps would fail together, connected by the torque tube.
-The inclusion of shear nuts, and the specific difference of bolt diameters between the flap hinges and the strut hinges also would be evidence that the bolts were the intended breakage point.

 

  Hide contents

The strut's attachment:


Bolt thread 1/4" dia
Note: bolt (figure #41) is drawn to be an AN5 hexagonal bolt.  However the part list pictured suggests that it's in fact a clevis bolt, like #39.  This is also evidenced by the same type of shear nut being used on both bolts.  This does not change the thread diameters, but it's an inconsistency between the drawings and the parts listed that I thought I better mention, just in case.

Both flap's attachments:


The bolts for the wing attachments are of larger diameter than those used on the strut.
A different bolt/nut assembly is used on the flaps' hinge lug as well.

Also on a neat side note, more detail on these can be found on https://military-fasteners.com/ and https://www.aircraftspruce.com/

 

 

"Shear nut" is simply in the context of the difference between an AN310 and AN320 castle nut. The AN310 is designed for both shear and tension loads. The AN320, with its lower profile design, has about half the tensile rating and is intended for shear applications only (eg. a clevis).

[NineLine]

Sorry for the late response to this, I did a search internally and we do have a long standing report about this or similar. I have updated the report, modernized it and added some test tracks based on how this blow back works (and the fact that it shouldn't) I also conferred with Mr Grey to make sure I am understanding it all right and have submitted it. Not sure it will be a fast one as I ma not sure if it requires a major overhaul of how the system is coded or what have you. But its in and reported. Thanks!