Hey all. On my gauge journey, I have learned things along the way working in the automotive industry, but lately it seems to have all piled on at once, and I am still trying to figure out what is correct. To give a little backstory, I worked as a quality technician in a plant making automotive parts, and having prior gauge management experience from another job, I became well versed in the bench gauges we had on the production floor. Eventually, I applied for the "gauge specialist" role because I was already performing maintenance and repair on the gauges to an extent, and I was building more knowledge working with the gauge manufacturers to fix the gauges. At that time, I was being introduced into what a Gauge R&R was, and we just knew that it was 3x3x10. I have recently learned what Type 1 studies are, and how they not only check repeatability of the gauge, but also how close to the reference it is.
We have a document at work that tells us all about the different types of studies, but I am wanting to know more, such as what study is required for what types of gauges, and whether or not you can use linearity verifications instead of MSA studies for certain gauges? For instance, we have many length and runout gauges that measure runouts using Solartron LVDT probes. You put the probe onto the shaft with a little preload, start the measurement which logically sets the probe to 0 and reads the min/max. We have other gauges that fixture parts by air actuated bushings that sit against the part while also pressing against an LVDT to read depth. For that gauge, we use a setting master to set the probes where they should be based on the values of the master.
My questions are:
We have a document at work that tells us all about the different types of studies, but I am wanting to know more, such as what study is required for what types of gauges, and whether or not you can use linearity verifications instead of MSA studies for certain gauges? For instance, we have many length and runout gauges that measure runouts using Solartron LVDT probes. You put the probe onto the shaft with a little preload, start the measurement which logically sets the probe to 0 and reads the min/max. We have other gauges that fixture parts by air actuated bushings that sit against the part while also pressing against an LVDT to read depth. For that gauge, we use a setting master to set the probes where they should be based on the values of the master.
My questions are:
- For those two gauge types, what studies are required?
- Do you do a Type 1 AND Type 2, or Type 1 for both, Type 2 for length and runout machine, and Type 3 for air actuated fixturing since it essentially removes operator influence?
- When doing Type 1 studies, is it required to do them at different lengths (linearity study)? If that were the case, could you just do a verification and potential calibration using some kind of reference standard so you don’t have to run another part 25 times?
- Since we are required to do ANOVA studies with our company, what do some of you do for parts where you have to change machine settings to modify parts to cover the range? I could see if you were just turning out something on a lathe, but these are drive shafts that are friction welded using huge friction welding machines, therefore adjusting them to apply more or less pressure while also having to shim to get different runouts is a bit of a painful and lengthy process. Also having to do this for multiple machines on multiple lines can get costly and take up a lot of resources. Is there an alternative?
- Do the MSA studies HAVE to be done on components, or do you think you could get away with creating a fixture that simulates parts with different readings? For example we measure bearing deflection using a handheld indicator with a sleeve on it. The issue is, the machine stakes two bearings at a time, so it’s not like you can modify each of the 4 bearings to have a different reading. Rather than doing it on a bearing, I was wondering if I could just make a plate that has 10 different holes to simulate bearings and use that to run the study, and then we just verify that plate every year as a reference standard?
