FADEC = complex

Newps writes:

There is no change in complexity when things go wrong with FADEC.

Yes, there is, because digital systems have different failure modes
from those of analog systems, and digital failure modes are often
catastrophic failure modes.

--
Transpose mxsmanic and gmail to reach me by e-mail.

Mxsmanic wrote:
Newps writes:


There is no change in complexity when things go wrong with FADEC.


Yes, there is, because digital systems have different failure modes
from those of analog systems, and digital failure modes are often
catastrophic failure modes.


Spoken like a sim pilot. If you had the slightest idea what the hell
you were talking about you would know that when the FADEC fails, usually
the computer goes belly up but it could be an electrical power loss, the
engine continues to run but does so at a very rich setting. If the
FADEC controls the prop then that goes to high RPM. The pilot doesn't
have to do anything except land and landing immediately isn't necessary
either. These failure modes are a requirement for certification.

Newps wrote:
Mxsmanic wrote:
Newps writes:
There is no change in complexity when things go wrong with FADEC.

Yes, there is, because digital systems have different failure modes
from those of analog systems, and digital failure modes are often
catastrophic failure modes.


Spoken like a sim pilot. If you had the slightest idea what the hell
you were talking about you would know that when the FADEC fails, usually
the computer goes belly up but it could be an electrical power loss, the
engine continues to run but does so at a very rich setting. If the
FADEC controls the prop then that goes to high RPM. [..]

Seems like you just proved his statement that digital failure modes are
different. If my analog (mechanical in this case) prop control fails,
it doesn't affect my mixture at the same time, as you're saying FADEC
does.

At the same time: Mxsmanic, apparently they've designed FADEC to fail
without being catastrophic.

My own personal worry is coming automobiles with totally electronic
steering and brakes. I'm sorry, even thoughI design reliable embedded
systems and I still would hate owning a car like that :-)

Kev

Kev writes:

At the same time: Mxsmanic, apparently they've designed FADEC to fail
without being catastrophic.

The designed-for failure modes are never a problem. The problem with
digital systems is with unforeseen failures, which usually have no
correlation with real-world constraints and are often catastrophic
failures in consequence.

You always need a way to disconnect the computer, and it has to be a
mechanical disconnection, not just an option on the screen menu.

My own personal worry is coming automobiles with totally electronic
steering and brakes. I'm sorry, even thoughI design reliable embedded
systems and I still would hate owning a car like that :-)

If you design embedded systems, you know why such a car would be
risky.

--
Transpose mxsmanic and gmail to reach me by e-mail.

Kev wrote:




Seems like you just proved his statement that digital failure modes are
different.



I never said they weren't different, only that the complexity of the
situation doesn't change. The pilot will be notified, probably by a red
light, that the FADEC failed. That's it, there's nothing for him to do.
Indeed in most cases there's nothing he could do even if he wanted.


If my analog (mechanical in this case) prop control fails,
it doesn't affect my mixture at the same time, as you're saying FADEC
does.

I didn't say that either. I meant to say that if the part of the FADEC
fails that controls the prop then the exact same thing happens there as
will happen to you, high RPM.

Newps writes:

Spoken like a sim pilot.

No, spoken like a computer specialist.

If you had the slightest idea what the hell
you were talking about you would know that when the FADEC fails, usually
the computer goes belly up but it could be an electrical power loss, the
engine continues to run but does so at a very rich setting.

FADECs can fail in all sorts of ways, depending on the software bugs
they contain.

--
Transpose mxsmanic and gmail to reach me by e-mail.

Mxsmanic wrote:



FADECs can fail in all sorts of ways, depending on the software bugs
they contain.

No, they cannot.

Newps writes:

No, they cannot.

One day, you may be surprised.

--
Transpose mxsmanic and gmail to reach me by e-mail.

Mxsmanic wrote:

Newps writes:


No, they cannot.


One day, you may be surprised.



No I will not. Once again you're clueless.

Newps writes:

No I will not.

Famous last words.

--
Transpose mxsmanic and gmail to reach me by e-mail.