accelerated weathering correlation

[macPT]

Does anyone knows any method to correlate accelerated weathering tests like ASTM G53-96 (now ASTM G154-00) or ISO 4892-3 or DIN 53384, with natural weathering caused by outdoor exposure?

I intend to predict, based on test results according to those standards, real lifetime of plastic components for outdoor use.

Thanks

 

[IRstuff]

It depends on the failure rates and types for your component.

Different failures get accelerated at different rates. Most can be characterized by a rate constant or activation energy that can be used in the Arhennius equation to determine the amount of acceleration. The trick is that to determine the values of the activation energies, which either requires data from the supplier or testing done by you. TTFN

 

[macPT]

The failure (the word is too strong) is the fiberglass blooming in fiberglass reinforced polyester SMC. This blooming is mainly caused by UV that produces a surface degradation of the polyester. I need to predict lifetime before blooming.

It is not a question of mechanical strenght (UV weathered reinforced polyester still maintain allmost of its original mechanical strenght) but more a question of surface appearence and to avoid people's itch caused by possible contact with the blommed fiberglass.

Regards

 

[skczmrk]

Dear macPT:

I noticed your post to compare results between ASTM G154-00 & a european standard DIN EU ISO 4892-3. Did you ever find a way to compare. I have results in the European method and want to be able to relate these results in the ASTM method..Any suggestions...Is this Apples to Oranges?

 

[macPT]

I don't know these standards in detail. The compare problem is far too much complex than comparing standards! For example, in DIN standard you can choose diferent radiation lamps and it is not clear, for me, wich radiation to use in tests. Diferent radiations will produce diferent results, but these results have no correlation.

The only conclusion you can have is, if you have two materials tested in identical conditions, according the same standard, the one with more testing hours is the one that have more radiation strengh.

My big problem is that my client wants a warranty, in years, for the product. And there is no way to convert testing hours in normal service years.

Good Luck

 

[IRstuff]

There is, but it's absurdly expensive. The first thing to do would be to start a real-time life-test of the material. The "failure" rate for real conditions must be determined.

Then, using a minimum of two different acceleration conditions, repeat the life test, again determining the failure rate for each acceleration condition. This allows you to figure out the activation energy of the failure. Given the actual failure rate, you can now determine what the acceleration factor is for any given acceleration condition.

TTFN

 

[macPT]

IRstuff

I agree with you. The two major problems are:
- the behaviour of the samples is dependent of the atmospheric conditions. We should have several samples (near the sea, on the mountains, near heavy traffic in cities,in hot places, in cold places,...). This way we could compare the testing results with real-time results. The costs will be very high.
- real-time tests will take real-time to produce results. This could be OK in a long term development. That is not the case as we should have an aproved solution "right now". But there's no way to prove to our client that the "right now" solution (with weathering tests) will perform well during X years in certain service conditions!

Regards

 

[m777182]

Hello macPT
your problem is similar to the estimation of a half life of a wood protection paint: I would select the most important factors, probably temperature, humidity and UV(wavelength)flux.Then I would take what is called "a rotatable second order surface response experimental design"- it tells you what combination of particular factor and at what level of each perform the test. Then compare the activation energies as IRstuff pointed already and calculate the model. There is a dummy factor that is nearly impossible to measure: the air pollutants, mainly free radicals, ozone and singlet oxygen and these strong oxidants attribute most to the decay of a polymeric materials.I believe they contribute to the so called "at seaside:mountains" effect.This is now the first stage.Further you will need some tests with periodically cycling the parametra because there will be very likely a strong serial correlation effect among them.
The task is very complex, time consuming and expensive.Try to talk to somebody with the experiences from pharmaceutical industry, they have similar,though more simple problems with testing medical preparates to estimate safe lifetime.
regards m777182