C8 dimension capability analysis

This week we began working on a capability analysis for the hub height dimension also known as the C8 dimension. There is a hub in the center of each battery seal where eventually the nail is put through. This measurement must be in line with the customer’s parameters or the battery may not be able to function. However, due to the accuracy of the machines and the leniency in the parameters we believe that it is unnecessary for us to be measuring this dimension during each sample because the probability of the C8 dimension being out of the parameters is very small. Using the PPK from 17 different samples and Minitab we created a I-MR chart and a C8 Distribution Plot. PPK is the long term capability of a process. The I-MR chart shows the difference of PPK over each sample. The distribution plot gives us the probability of “out of speculation,” or the probability of the hub height being out of the customers parameters. We found that if the machine is operating on a 6 second cycle, meaning one shot every 6 seconds, that we would expect one rejectable C8 in 100 octillion years.

LR-3

This week we began the study of LR-3 sized batteries. Because the most efficient molding parameters are unknown we must experiment to find the most adequate parameters. We began by getting different shots, each with a different injection velocity. Then we measured the weight of each individual cavity and then the weight of each shot. This allows us to narrow down the range the injection velocity parameter can be in. For example we found that when the injection velocity was set at 0.2 the cavity and shot weights were not up to standard and therefore setting the injection velocity at 0.2 would create inadequate parts.

Next week we hope to continue working on the LR-3 battery seals.

Week 6: Contamination

Although this week I wanted to continue focusing on the effects of statistical analysis in production, there was a problem in the hydration tanks. Rust has contaminated about a week’s worth of hydrated parts. The problem is that rust can cause significant problems if inside a battery, especially if rust is found near the battery seal’s vent. Because of this we examined boxes of battery seals to determine whether they were contaminated with rust. Although it wasn’t the most exciting week, I was still able to learn valuable information.

I also analyzed defects in particular parts. Certain parts were slightly deformed to the point where cosmetically they were unappealing however they were still functionally sound. However there were other parts that were deformed with flash issues to the point where they were unable to function. We had to separate the deformed parts from the flash parts so the customer can determine whether they still want to buy the parts. Next week I look forward to focusing more on the statistical part of the problem

Week 5: Research

This week I researched the positives that statistical analysis can have on production. Gathering data and analyzing the data can help improve the production to where these battery seals can be made more efficiently and with less variability. In essence, the use of statistical analysis should be able to “perfect” the way battery seals are produced. Next week I plan to get more into the use of statistical analysis.

Week 4: Effects of statistical analysis on production

Because hydration is not a priority I have been working on at Microtech Southwest, I have decided to move away from the hydration process on battery seals. As most of my blogs are focusing on the benefits of statistical analysis on the production of these battery seals I have decided to shift my blog to that focus. From here on I will focus more on the statistical analysis and how it effects the productions.

This week I was using the laser mic in quality control. The customer wants sample shots to be examined to make sure there are no flaws in the production. So using a laser mic, I was able to measure the parameters that the customers want. A laser mic allows us to efficiently and accurately measure parts. The part is being measured by a laser and then the dimensions are communicated by a program automatically to an excel spreadsheet. Because all the necessary data wasn’t collected until Friday I was unable to analyze it. Next week I look forward to analyzing the data to make sure the production is adequate. 

Week 3: LR-20 DOE and Gage R&R

This week was again another step away from hydration because there are other problems to overcome before we can continue the hydration process.

Similar to the LR-14 DOE, we ran another DOE except for the LR-20s. We wanted to again find if there is was any problems with the variance between shot to shot and cavity to cavity. After measuring the weights of 5 shots of LR-20 we found that the variance was a lot smaller cavity to cavity than what we found in the LR-14s despite some outliers. We concluded that there may have been a venting problem with the molding press.

We also ran a Gage R&R, standing for Gage repeatability and reproducibility, on a new program that was made to more accurately measure the plastic parts being made. The customer wants the parts to only have a certain amount of variance. A sample of the parts must be measured and analyzed to determine if the parts fit the criteria. However there are multiple ways of measuring these parts. We have a newer program that is a lot faster and more accurate. However we must run a Gage R&R to show that this new method is just as accurate if not more. A Gage R&R is a statistical tool helps investigate the amount of variability the measurement system is causing, or in more simple terms it determines how reliant the measurement method is. Unfortunately we were unable to analyze the data from the Gage R&R due to time constraints. I’m looking forward to next week where we will be able to determine if our new program is adequate. 

Week 2: LR-14 DOE

Hello again!

This week I will take a small step away from hydration because a customer wanted an "emergency" DOE analysis of the LR-14 press. Because the pressure exerted by the press can cause bending in the mold, we wanted to run an experiment to find evidence of this in the LR-14 mold.

Here's a little bit of background on the mold before I talk about the experiment: a mold is the object that allows the injected plastic to take the correct shape, in this case the shape of an LR-14 battery seal. This mold is placed in a molding press that injects plastic into the mold at the desired pressure, velocity, and temperature.

The LR-14 mold has 32 cavities meaning every injection of plastic, or shot, creates 32 LR-14 sized battery seals. Because the mold and the press can never be perfect there will always be variance between each cavity. However, we are trying to find the best combination of factors for the molding press to create the least variability from cavity to cavity and shot to shot.

Therefore an DOE on the LR-14 must be run to find any discrepancy in the molding press. We did two DOE's on the LR-14 mold. The first one consisted of 10 shots in which we weighted all 32 cavities in each shot and looked for the correlation in weight from shot to shot and cavity to cavity. After running a grouped box plot graph, we found little variance from shot to shot however the cavity to cavity variance was unclear. We decided to run a second DOE analysis of 5 new shots to compare the variance from cavity to cavity. We found that the LR-14 mold has negligible variance from cavity to cavity and therefore the parameters on the molding press are adequate for the particular mold. However we look to improve the parameters on the mold to make it more efficient for not only variance but also functional and cosmetic problems.