We have a supplier that extrudes for us, but they are a job shop with multiple machines used to extrude the same component. Each machine can very well have a completely different set of operating parameters. With many more clients then ourselves, it would be difficult to plan a production run on a single machine since they would possibly have the machine occupied for a longer period of time with another customers job.
Depending on the complexity and volume of your components, a in-line inspection process might be more cost effective then running a multi-machine validation effort IF your current production needs can support it...but I think that greater scrutiny will be placed on data capture to show objectively that the inspection is occurring and works adequately. It really depends on what your customer needs/wants and what they are willing to pay.
As more companies such as yourself get into the mix of supplying to medical device companies, they will need to be more adept to the needs to their customers(ie, medical device speak....it's music to my ears when a supplier ALREADY understands my need for V&V activities).
The specific things you can do ahead of time, regardless of customer inputs:
-Document the installation of each machine...i suggest you dig up the user manual and create a checklist that ensures you installed it according to manufacturers instruction. If the manufacturer installed it for you, they should have given you a signed statement of installation....this could be used as your IQ...just have your own internal document or form state you are accepting this as the IQ and have the right people approve it.
-OQ is tough, since your goal is to test the operational range of the specific tool doing the work. You could go through a gamut of possibilities, but you most likely(murphy's law) won't get all of them.
I would suggest you factor in a validation cost to your piece part price and sell the client on your ability to deliver them with a qualified part ready for FDA scrutiny.
I could be wrong on this, but the day is coming when medical device companies looking for suppliers won't even engage with you unless you already understand their needs and are factoring it in.
To your specific question of extrapolating out to multiple machines, I don't recommend this, but that's not to say it can be justified. You will have to show an objective way that the machines are equivalent in processing capability.