Clutch_maximum_torque_down misnamed and not working

[StuLes20]

Version: *SU14±1.37.12.0

Frequency: Consistently

Severity: High

Context: *Developing/Debuging helo turbine engine to rotor connection.

Bug description:
[HELICOPER]
clutch_maximum_torque_up This is the clutch simulation maximum clutch torque when the engine [RPM] Default value is 1000. Float No
clutch_maximum_torque_down This is the clutch simulation maximum clutch torque when the engine [RPM] Default value is 1000. Float No

definitions are the same, when they should be different.

clutch_maximum_torque_down defines transfering torque from the engine to the rotor. This is working correctly. Can be verified by reducing value to ‘0’, and no torque is transfered from the engine to the rotor.

clutch_maximum_torque_up this parameter should be to adjust the torque that is transfered from the rotor to the engine. In a real heli, this is called the ‘sprag clutch’ or ‘freewheeling unit’. It is critical and is tested before takeoff to check functionality. In the critical case of an engine or gearbox siezure, the rotor is allowed to keep spinning, without drag (the adustment value on this parameter) so the when the engine stops, the rotor doesn’t stop too.

Simply: the engine should be able to drive the rotor, but the engine should not be able to drive the engine.

Repro steps: Using ‘debug engines’ dev tool:
Reducing the ‘master throttle’ from 100% to 0%
‘Engine shaft torque in’ reduces from 85lbs/ft(arbitrary idle number) to +9.2lbs/ft not the ‘0’ torque required to release the rotor from the engine/gearbox.
clutch_maximum_torque_down changing this value has no effect.

Another way to test is:
From idle, roll up the throttle. Eng and rotor needles will rise together.
From full throttle, roll off the throttle to idle, the eng needle should reduce to idle ahead of (before) the rpm needle.This is called a ‘needle split’.

Attachments:

Private attachments: Send a PM to @PrivateContent with the link to this topic and the link to download your content

[FlyingRaccoon]

Hello @StuLes20

The documentation will obviously be changed as clutch_maximum_torque_down is a limiting factor when the rotor is pulling the engine rpm down, not up. And the units are incorrect.

But I’m not sure to understand how clutch_maximum_torque_down is not working.
In the repro steps you mentioned (reducing master throttle from 100% to 0%), the engine slows down faster than the rotor, the engine rpm are being pulled up by the rotor, so it’s the clutch_maximum_torque_up that is limiting the torque.
When setting it to 0, as in the Cabri G2 flight_model.cfg, the engine debug shows “Engine shaft torque in” to 0 in this situation, as expected.

Am I missing something?

Regards,
Sylvain

[StuLes20]

Hi Sylvain
I will use the Huey in the example below.

Using in the Flight_Model.cfg
clutch_maximum_torque_up = 0.0
clutch_maximum_torque_down = 1000.0

This is 100% throttle on the ground.
GOV engaged, ‘clutch’ engaged and transmitting torque.

Here is:
Throttle roll off to idle, GOV disengaged, and needles split.
Notice that the clutch does not dissengage, and torque does not reduce to zero.It keeps transmitting a fixed low torque value.
Maybe it should reduce to 0 and dissengage, but it doesn’t.

The settings of:
clutch_maximum_torque_up = 0.0
clutch_maximum_torque_down = 1000.0
should emulate a sprag clutch, No Dev has made it work yet. There a lots of ingenious coding workarounds.

The concept is easy. Its like a bicycle.
You can pedal the bicycle, but when you stop pedalling and ‘freewheel’, the bicycle can’t pedal you.

Its a checklist item before takeoff. A no-go item if it fails the test.

How and why it woriks uniquely on the G2 is beyond me.
It would be great step foreward if this gets resolved.
The Description error in the SDK is minor.

thanks for enquiring further. It gives me hope.
Please also respond to my ‘skid contact points’ as well. That should be a quick fix.

Skid contact points

Hi For heli skid contact points: The maximum number of contact points for skid-type retractable landing gear is 6 (previously this was limited to only 4). These are: Center Front Skid Center Rear Skid Left Front Skid Left Rear Skid Right Front Skid Right Rear Skid In order to make the skid behave realistically: 2 points are used on the heal (rear) 2 points at the toe (where its turned up) 2 points before the turned up part. This leaves the entire length of the skid as a ‘non-contact’ …

Regards
Stu

[FlyingRaccoon]

Hello @StuLes20

Can you give us access to your package so we can trace what happens in your case? (or if you know a marketplace package that demonstrate the same behavior, I can investigate as well)

The only packages I have at my disposal for now are the Cabri and Bell 407 that do not demonstrate this problem.
(I forced the governor off on the Bell but it looks like the turbine has a lot of inertia and engine RPM do not decrease faster than the rotor anyway)

Regards,
Sylvain

[StuLes20]

Hi Sylvain

The Huey will be uploaded to ‘marketplace’ very soon. I think there was a holdup with info from MS.
I have passed on your request to the head developer Joss (Taog’s Hangar. Lama, Alouette and Huey).
The 407 seems to exhibit the same flaw.

Here is 100% throttle, clutch engaged and engine torque trnasmitted to the rotor.

Throttle rolled down to idle. Clutch still engaged (should have disengaged).
Positive torque still being transmitted to the rotor.
Needles still joined. Needles should have split, with the eng needle going to idle, and thre rotor slowing not as quick.

To me, the B407 is not working correctly either.
Thank you.
Symptoms are the same in both Huey and 407. Lama and Alouette are the same too.
I think that if you looked at the H500 and H125 of Cowansim, it would be identical when you use ‘engine debug’ tool like above.

Regards Stuart

[StuLes20]

Hi Sylvain
Currently you are trying to understand the issue in the post: This, I think shows great maturity in an engineering sense. I will try to helpl you undrstand.
We have been discussing:

“Clutch_maximum_torque_down misnamed and not working”

But, there is also this post, which is the same issue from a different perspective:
The following is a post on the "Sprag clutch for Helicopters "

Currently, helicopter developers are hindered seriously by the lack of a ‘Sprag clutch’.
Current ‘workarounds’ are unsatisfactory.

What is a "sprag clutch’.
A sprag clutch is a one-way It resembles a roller bearing but, instead of cylindrical rollers, non-revolving asymmetric figure-eight shaped sprags or other elements allowing single direction rotation, are used. When the unit rotates in one direction the rollers slip or free-wheel, but when a torque is applied in the opposite direction, the sprags tilt slightly, producing a wedging action and binding because of friction.
Helicopters
A sprag clutch is used in many designs to transfer power from the engine to the main rotor. In the event of an engine failure, the sprag clutch allows the main rotor to continue rotating faster than the engine so that the helicopter can enter autorotation.
Wikipedia.

In a nutshell.
The engine can power the rotor.
The rotor cannot power the engine.
In the event of an engine or gearbox seizure, the rotor doesn’t seize up as well.
After starting the engine, it is a ‘checklist item’ that is NO-GO if it doesn’t function. Functionality is checked by looking for a needle split, when the throttle is rolled off from 100% towards idle.

Solution for Asobo.

parameter
sprag_clutch =
description
When this parameter is set to 1 (TRUE) engine torque drives the rotor. The rotor cannot drive the engine.
Default value is 0 (FALSE).

parameter
sprag_clutch_direction =
description
A value of 0 (FALSE) here will maintain the default rotational direction of the helicopter blades, which is clockwise (when viewed from above). Setting this to 1 (TRUE) will reverse that rotation, so anti-clockwise.
Default value is 0…

Please take the time to click the ‘VOTE’ button and show interest in heli sim developement towards realism.
Stu

Please note that this is the same problem explained 2 ways.

The first way is the current parameters should emulate a ‘sprag clutch’. using maximum torque up/down.

The second post is about a sprag clutch
Adding a working sprag clutch is one solution.
Fixing the current clutch torque up/down sdk is another.

Rgards
Stuart

[FlyingRaccoon]

Hello @StuLes20

I was able to investigate further and was able to locate the problem.
clutch_maximum_torque_up/down parameters are not at fault. When setting clutch_maximum_torque_up is set to 0, rotor will be freewheeling and turbine will not pull down rotor rpm.

The problem lies in the fact helicopter turbine shaft torque never goes in negative values.
This is visible when comparing with the piston engine Cabri G2. Piston engine has negative shaft torque for a short period of time when cutting throttle, leading to engine RPM decreasing significantly faster than rotor rpm.

This does not occur with turbine engines. Rotor and engine rpm end up decreasing synchronously.
This does not slow down rotor speed. If anything, it prevents it from slowing down as fast as it should when engine rpm is slightly higher than rotor rpm.
This has been reported to our core sim engineers and added to our backlog.

Thank you for your patience and detailed feedback.
Regards,
Sylvain

[b1e1n7e]

Hey Sylvain,

is there anything new regarding this topic? I cannot get my autorotation behavior right. The rotor rpm is rising as expected but takes the n2 engine rpms up with it.
Also due to the fact that the rotor seems to be connected with the engines and transmission the rotor rpm is not rising as fast as it should since it cant spin freely.

[FlyingRaccoon]

Hello @b1e1n7e

No progress on this so far.
I don’t have much visibility for now on when the next MSFS 2020 update will be and what issues it will address.

Regards,
Sylvain

[b1e1n7e]

The issue is present in MSFS 24 as well, which is my main concern. Should I create a bug report so the issue can be logged for 2024?

[FlyingRaccoon]

Yes, you should.