Verification or Validation for Glue Dispenser Units ?

[GillianG]

Hi

I have to pull a plan together to validate / verify glue dispenser units for the bonding of various silicone joints to tubing. I will have 10 dispensers going into production for approximately 12 different joints / components. I'm trying to work out how best to approach this.

The dispenser is a constant flow of glue that the operator inserts the tubing into the glue "waterfall". I therefore cannot perform a capability analysis on the volume of glue dispensed. Also, all bonded components will be subject to 100% inspection post the bonding process.

Do I perform a TMV / Gage R&R on each component - 10 separate TMVs - to verify the bonding process is successful?
Once I've passed the TMV do I perform it on each dispenser or if they are all the same model can we "assume" (I know...) that the other dispensers will provide the same result?
Or do I perform a TMV on each dispenser using a different component for each one?

Thank you in advance for any help / guidance with this.

 

[Ronen E]

Re: Verification Studies for Glue Dispenser Units - Query.

Hello Gillian and welcome to the Cove

What does the 100% inspection consist of? Is it just a visual or do you also test for functionality (strength, sealing etc.)?

In principle, if a process is verified 100% there's no formal need to validate it. Doing both represents a redundancy and possibly wasted resources.

Cheers,
Ronen.

 

[GillianG]

Hi RonenE

Thank you for your response. I agree 100% that this is a waste of time and energy. Historically we have not validated any of the processes we use for this bonding procedure as we 100% inspect - visual and functional. However, with the roll-out of the new dispensers, our customer is pushing for some form of validation. I therefore am trying to keep it as simple as possible. If we have 10 dispensers, can I perform a TMV on one and prove equivalency?

 

[Ronen E]

Hi,

If it all comes from a customer requirement, and they're not too specific about it, you have more freedom as to the scope and method. I believe we're discussing a medical device, right? If so, I think that GR&R / Test Method Validation is less applicable, and instead you better relate to production process validation. It sounds as though you're trying to validate the production equipment, not the subsequent verification testing.

For guidance, you could look at the FDA's medical devices process validation guidance, and since you're not obliged by the regulation to do it, you could modify it to suit your needs. Perhaps an IQ and an OQ would suffice. You could create a protocol draft and see whether the customer is satisfied.

Addressing variability - if you can show that all 10 dispensers are made to the same specification you could probably get away with an IQ for all, an OQ for just one or two, and documenting equivalence. The more complicated part is that you use the dispensers for a range of different products. If that range can be divided into categories, you might be able to test just one representative product for each.

HTH
Ronen.

 

[somashekar]

I see more to this than what I read. You have a scope here to address the complete process. What is the glue viscosity ? how do you measure and maintain it ? How much length of the tube does he wet under the glue 'waterfall'. How long he must hold the tube in the 'gluefall' for the complete wetting of the tube surface. Are there any controls for this., like a work holder or fixture ? You are talking about operator qualification as well as process validation (IQ / OQ / PQ).