If I want to perform a Design of Experiment on a blister machine, what would be the measure responses or output (dependent) variables?
I am telling this because I want some help with a problem I have in a package process on a blister machine.
In the packaging process I have the parameter (factors, inputs) like forming temperature, sealing temperature, speed, sealing pressure and forming pressure. The measure response (output) that I have are leak test, yield, etc. My problem is my product is returned back by the customer because the leak test didn’t pass. When I perform the leak test during my process, the test passes.
My question is: what measure response (output), a numerical variable besides leak test do I have to consider?
I consider that I need an output that can be measure in order to perform a design of experiment.
Attentive to your prompt response. Thanks in advance.
If I want to perform a Design of Experiment on a blister machine, what would be the measure responses or output (dependent) variables?
I am telling this because I want some help with a problem I have in a package process on a blister machine.
In the packaging process I have the parameter (factors, inputs) like forming temperature, sealing temperature, speed, sealing pressure and forming pressure. The measure response (output) that I have are leak test, yield, etc. My problem is my product is returned back by the customer because the leak test didn’t pass. When I perform the leak test during my process, the test passes.
My question is: what measure response (output), a numerical variable besides leak test do I have to consider?
I consider that I need an output that can be measure in order to perform a design of experiment.
Attentive to your prompt response. Thanks in advance.
Regards,[/quote]
If you have qualified the machine and then the process are the parameters when you run within both qualified ranges?
You say you test your product on site and it is fine, but when it gets to the customer it doesn’t pass their test? Is there a variation in the way you test and the way your customer test?
What about shipping conditions? This could also play a role in the root cause.
There are just a few points to consider while you perform your root cause analysis.
I did not understand your first question. What I want to do is to predict a leak on the blisters, so what would be the measure response (output) to perform a design of experiment (DOE)? What numerical variable besides leak test do I have to consider? Is there any numerical variable that can hepl me to predict a leak on a blister?
We are working in the way we test and the way the customer test to see if we can standardize the methods.
We began considering shipping conditions after you told me that, so thanks a lot.
I did not understand your first question. What I want to do is to predict a leak on the blisters, so what would be the measure response (output) to perform a design of experiment (DOE)? What numerical variable besides leak test do I have to consider? Is there any numerical variable that can hepl me to predict a leak on a blister?
We are working in the way we test and the way the customer test to see if we can standardize the methods.
We began considering shipping conditions after you told me that, so thanks a lot.
Attentive to your prompt respond.
Regards,[/quote]
Henry,
I am having a hard time understanding the goal here.
If you qualified the machine(s) you have the information you need. The critical parameters are what you indicated above the pre heating, forming temp/pressure, sealing temp/pressure. All impact the sealing quality. You would have your ranges from the OQ and then you would have your product specific through your process performance qualification.
Excuse me if I did not tell correctly. The goal here is to predict a leak on a blister using statistical method.
What continuous variable do I have to consider to predict a leak on a blister ? Is there any continuous variable that help me to predict a leak on a blister?
Excuse me if I did not tell correctly. The goal here is to predict a leak on a blister using statistical method.
What continuous variable do I have to consider to predict a leak on a blister ? Is there any continuous variable that help me to predict a leak on a blister?
Attentive to your comment.
Regards,[/quote]
I think you are looking at this in the wrong manner. The leak of a blister should have been performed during the OQ if you went to testing to failure which isn’t typical anymore.
However, what you should be doing is taking your validation information and then evaluate your results (via in process checks) to see if there is any change in your ppk
[quote=seal]I think you are looking at this in the wrong manner. The leak of a blister should have been performed during the OQ if you went to testing to failure which isn’t typical anymore.
However, what you should be doing is taking your validation information and then evaluate your results (via in process checks) to see if there is any change in your ppk[/quote]
Dear Seal,
Which variable do you consider to measure or calculate ppk?
Do you have any OQ document that I can follow to validate the machine?
Which variable do you consider to measure or calculate ppk?
Do you have any OQ document that I can follow to validate the machine?
Attentive to your comments.[/quote]
It goes beyond just having an OQ document especially in the pharmaceutical world. It seems like you do not have a validation program where you are. The first steps would be creating that (SOP and VMP). Then from there you can make URS, DQ and go into the IQ, OQ and PQ needed to qualify machines and processes.