Can anyone advise a method to research 'Measurement System' for testing a tire?

V

Vlad12

Hello everybody,

Our plant produces tires. When testing the tire its maximal rolling speed at the bench is referred as a key characteristic.
The matter of test method: operator mounts tire on a rim, pumps it with compressed air, and installs mounted tire in the bench. Tire is automatically pressed to the drum and the drum starts to rotate. Drum speed increases stepwise in certain intervals until the tire destroys. During testing the interval until tire failure is automatically recorded. Testing frequency is one tire in two weeks.
Usually measurement system analysis for destructive testing can be done with ANOVA Nested method.
For example, testing of 10 parts with 2 trials, but there are some problems:
1. it is impossible to make two so much similar tires, so that you can make an assumption of their homogeneity (assumption of samples identity within a part);
2. it will be an expensive research, as at least 20 tires will be destroyed (to provide sufficient sampling of tires for research it might need more than 20 tests).
Can anyone advise a method to research this measurement system? How can we confirm appropriateness of this measurement system for Supplier and during audit?
 

Stijloor

Leader
Super Moderator
Re: Can anyone advise a method to research measurement system for testing a tire?

:bigwave:
Hello everybody,

Our plant produces tires. When testing the tire its maximal rolling speed at the bench is referred as a key characteristic.
The matter of test method: operator mounts tire on a rim, pumps it with compressed air, and installs mounted tire in the bench. Tire is automatically pressed to the drum and the drum starts to rotate. Drum speed increases stepwise in certain intervals until the tire destroys. During testing the interval until tire failure is automatically recorded. Testing frequency is one tire in two weeks.
Usually measurement system analysis for destructive testing can be done with ANOVA Nested method.
For example, testing of 10 parts with 2 trials, but there are some problems:
1. it is impossible to make two so much similar tires, so that you can make an assumption of their homogeneity (assumption of samples identity within a part);
2. it will be an expensive research, as at least 20 tires will be destroyed (to provide sufficient sampling of tires for research it might need more than 20 tests).
Can anyone advise a method to research this measurement system? How can we confirm appropriateness of this measurement system for Supplier and during audit?

Hello Vlad,

I am not an expert on this topic. I am sure someone will respond.

In the mean time, I did a search here at the Cove Forums on: "MSA for Destructive Tests."

Could any of this help you?

Stijloor.
 

Caster

An Early Cover
Trusted Information Resource
Re: Can anyone advise a method to research measurement system for testing a tire?

We make wheels and use a similar test to validate the wheel, not the tire.

Its sure wakes you up when a tire explodes right next to you doesn't it.

We have never done MSA on our device, and our customers have never asked for it.

Do you have the AIAG MSA manual. It offers some hints on MSA for destructive tests (can not retest the part)

Much better is a book by Donald Wheeler - This seems to be a newer edition than I have, and mine had many ideas on how to do MSA on destructive tests, I assume this does too

http://www.patonprofessional.com/pm-65-1-emp-iii-using-imperfect-data.aspx

Also google is always your friend

http://www.google.ca/search?source=ig&hl=en&rlz=&q=msa+destructive+test&meta=

If you are allowed, I'd sure like to see the results if you do manage to complete a study

Good luck
 
Top Bottom