Langley,
i'll try to help. And it's not ignorance - there's a lot of definitions out there. Wikipedia supplied this generic definition, taken from UK Def Stan 00-56:
A Safety Case is a structured argument, supported by evidence, intended to justify that a system is acceptably safe for a specific application in a specific operating environment
00-56 is still extant and often called out in contracts. This is a perfectly usable Standard, but the MAA decided to go their own way and create a bespoke set of regulations for air safety management. Their definition is set out in RA1250 thus:
An Air System SC is an up to date, through-life body of evidence that presents a coherent safety argument that demonstrates that all credible RtL associated with an Air System have been identified, assessed and mitigated satisfactorily.
In aircraft terms, a Safety Case is written around a defined system (defined by its build standard) and its specific application is set out in the requirements against which the system (aircraft) was designed. The safety case may also include limitations against the original requirement to manage identified hazards. Evidence will include analysis, component testing, demonstrations, rig tests, ground tests, and finally flight test. It can also include read-across from other systems. It can also be supported by safety cases for individual equipments (like ejection seats) or, further down the chain, chunks of evidence sometimes called safety statements.
The chain then moves on, with Build Standard (configuration) and the Safety Case being used to build the Release to Service (RTS) recommendations, which the RTS Authority uses to build and release the RTS.
What this means is this: when you go and sign for an aircraft, you should be able to assume that the equipment you are strapping into has been thoroughly analysed, tested and certified as safe to use within the limitation in your publications. That's what the RTS should give you. Moreover, you should be able to assume that if anyone in your command and engineering chains have changed the configuration (or servicing) of the system, then that change has been fully checked, tested and analysed against the current safety case. Again, you get this via the RTS.
I hope you can see why people like Tuc and Chug (and me) have been so vocal about this one. The MoD did not hold a Safety Case for the Mk10 seat. If they didn't hold that, they should never have issued an RTS for the aircraft. As Tuc has said elsewhere, it's highly probable that Martin Baker had a perfectly good 'Safety Statement' for the seat, to get it past their own internal safety management system. However, that document may not have been kept up to date, or supplied to the MoD, as the funds to support that activity were slashed in 1993. So, somewhere between the Mk10 entering service and 2011, the MoD ceased to have a Safety Case for the seat. The public needs to know why. And ACAS needs to explain how he issued an illegal RTS.
Best Regards as ever to those working hard to keep our people safe
Engines
Edit - Tuc got in just after me with an even better response. Hope these help.