When looking at glass inspection machines, one of the first steps is to do a test report, so you can understand what kind of performance you can get in the field. This blog will cover some best practices for testing glass; what a good test report is like; and how to interpret the report so you can get a good sense of what will happen in the field if you were to buy that machine.
Best Practice #1: Mold Number Test Sample Size
The concept of testing for glass containers is built around the machine’s ability to build that container. If you look at the bottom of any glass container, you’ll see a “mold number”. Every glass container is going to have it. That’s important because, in the manufacturing process, there are a number of molds used that hot glass is poured into, and then it cures, and you’ve got a glass container.
Each mold is a little different. So when you do testing, it’s really important to do enough mold numbers that you’re really sure you’ve got a representative sample of containers. Otherwise, you can get over-indexed, good or bad, that won’t be achievable in the field.
Normally when you have fewer containers, you can over-tune your machine to handle just that type of container; but then your test results in the field aren’t as good. So, it’s really important, and most glass manufacturers – most food and beverage companies – work with glass manufacturers. There can be up to 60, 70 molds depending on how many bottles you’re actually getting per year, so you want to have a big data set when you do the testing.
We also recommend exhaustive testing of the most difficult areas of the glass container. Generally, that’s somewhere on the bottom. You want to make sure you get a statistically valid test result; meaning, you want to test between 100 and 600 mold numbers, to make sure you get consistent results across every mold you have in your population.
That’s an important “best practice” that you want to see in any test report from a vendor. You want to make sure the vendor recommends testing enough molds so that you have a thorough test report.
Best Practice #2: The Specific Gravity Weight of the Glass Test Piece
The type of glass that’s used for the contaminant testing itself makes a big difference. Most of the glass containers are made of something called “soda-lime glass”. Other types of glass are “pyrex”, which has a little bit lower specific gravity; and “leaded glass”, which is used to make things like fancy decanters – but you don’t find it in food and beverage industries, for obvious reasons (hint: it has lead in it).
When Peco-InspX does testing, we use two types of test pieces: Type P and Type M.
The reason we use these two types of glass is one’s got a slightly different specific gravity than the other. Soda-lime glass has a specific gravity of about 2.5. If you look for Type M soda-lime glass, you will see that its specific gravity is about 2.58 – so a little heavier. Then Type P is a little lighter at 2.5.
During the manufacturing process, the specific gravity of containers varies depending on the exact composition of materials in the glass. It’s tough to ensure that the glass stays a constant density. So, we use the two types of glass test case contaminants so you know 100% over a range of what you’d expect in production with the X-ray inspection machine. Because the test pieces have a specific gravity range (i.e., they have different weights), you’re able to test for contaminants within the range of the average contaminant weight (usually between 2.5 and 2.58).
Our Type P and Type M glass test pieces are spheres. The reason we use spheres for laboratory testing is that they roll around in the container; meaning as they roll, they’ll cover a lot of the glass container area. They are also very repeatable!
If that sounds like a lot of work – it is. To get a good glass test report is a big undertaking.
Best Practice #3: Using Your Own Glass as the Test Piece
We also encourage customers to make sure their vendor gives some test of actual samples of your own glass. When Peco-InspX gets a glass from customers, we break them and create what we call “custom test pieces”. Then we measure them. For example, if we have a glass shard that’s 13.5 mm long by roughly 2.5 mm thick, how well could we find that? We want a sense of as the pieces get smaller if detection rates decrease. Therefore, it’s best to use a real test piece – something that came from the container itself – to see what detection is like for your goods.
When you take a test piece from the container, those pieces are often asymmetric; you get different detection capability depending on how the piece is oriented.
Best Practice #4: Use a 300-series
The last thing we’ll mention: metal contaminants for X-ray systems. Generally, anything metal is equivalently detectable, except for aluminum (but that’s another blog post). You want to make sure your vendor uses a standard 300-series piece of stainless steel, which is the benchmark for testing. Even 304 or 316 are perfectly fine.
Best Practice #5: Collect Data Sheets
Finally, make sure your vendor provides the datasheets for all their samples so you can make sure you understand the density of those contaminants as well.
So, to summarize:
- Use enough containers with enough mold variation to get a representative sample size
- Make sure your test pieces are sufficient in specific gravity so that you get pure coverage
- Exhaustive testing – that means approximately 100 to 600 mold numbers tested
- Make sure you get some sense of real-life contaminants with pieces from the jar
- Your vendor should use a 300-series piece of stainless steel for X-ray systems
- Get the datasheets for the samples from the vendor
Thank you for reading! If you have any further questions, please contact us today.
If you want more helpful resources from Peco-InspX visit https://www.peco-inspx/lp/resources