Question about availability and reliability in complex system

[Gintama]

Dear all,

This is my first topic in your (our ) forum. The matter is how to calculate availabily (A) in complex system.

I know that for a single components A=mtbf/(mtbf+mttr).

In some articles i've readed that:

for series of two components (just a simple example) Aseries= A1*A2 and for a redundant schemes Ared= 1-(1-A1)*(1-A1).

Is it correct all of this? This relations are mathematically correct for reliabilities but i've some problems to accept them for the availabilities, because there are already mean values.

I think that i should calculate firts the mtbf of serial or parallel and mttr of serial or paralle and than the availability as the definition.

Thanks

 

[IRstuff]

Your approach is mathematically correct. However, a few examples show that the two approaches are generally comparable until Avail < 80%

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[Gintama]

Thanks IRstuff for your reply.

I prefer to use a model mathematically correct but I'm reading, for my work , some documents in which there are calculations very approximate. For example there is mtbf of a complex system calculated as an mtbf equivalent (let me say) calculated, that is:

mtbf_eq= avail/(1-avail) * MTTR (where has been considered deterministic value)

The problem is: I'm finding often an MTBF equivalent that is greater than the minimum mtbf of a series. Can be possible ?

If I reduce the system to a RBD series of subsystems, I think that the system MTBF must be equal/lower than the minimum MTBF. It's right?

Thanks

 

[IRstuff]

"system MTBF must be equal/lower than the minimum MTBF"

That's the way it should be.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529