T6 aluminum fatigue 2

[Durableguy]

Hello all. First of all, I apologize for asking what I expect may be a simple question to you. I am not an engineer, but would like to ask those who are and versed in what the issue is

I am a powder coater. I have a set of motorcycle wheels cast in T6 aluminum. Since they are cast, SOP says they should be outgassed above the powder cure temp, in this case 400* F, prior to powder application.

The wheels have been outgassed and saw a PMT of 430* F for about an hour. My customer is very uneasy about the wheels seeing these temps in this grade of aluminum. Monday I hope to have them Rockwell tested.

If their integrity has been compromised, is it possible to anneal them and return them to where they should be?

Apologies for coming here like this, but you go where the water is!

Thank you.

 

[Durableguy]

Should have stats to post tomorrow. ustomer sent me 1/5th of a simlilar wheel to run tests on. Hit a piece teh size of a quarter tonight 7-8 times. Eliminating the high (47.5) and the low (29), they stayed in teh 36-42 range.

Simulated teh original bake tonight and will get those parts retested tomorrow.

I was very suprised at the softness of that aluminum. Fully epxected high 40's to low 50's.

 

[Durableguy]

OK, here is whats happened. Customer sent em a sample piece of a similar era wheel, which I cut a small piece off of (about the size of a quater) and had tested. Hit it arounf 6-8 times and the results came back 29-47. Most stayed around 32-37. I was very suprised at the variance in such a small sample.

Took that part back and tried to simulate the original bake. I wasn't able to start until late at night and didn't get the mass exactly right and wanted to err on the light side but got thigns fairly close. In retrospect I wished I'd not been as compliant in getting it done. I believe the ramp up was steeper and cool down was slower.

Retested the original sample piece and got a low of 19 and a high of 29. Another unbaked sample produced similar results but also showed a hit of 19. At teh same time a baked part showed a reading of 64. Both the professor and I believe those readings may/probaly have been skewed as the parts were not parallel and may have been the result of an imperfect sample surface (cast). We were both suprised at the results.

As much as the mass in the oven wasn't spot on and heat/cool ramps were not exact either, the temps remained spot on at 60 and 90 minutes. 430.

This has sparked a bit of interest in me and I hope to get another whole wheel I can chop up and run more detailed testing on.

Again, thanks for all the replies and guidance. Its been a great help to me!

 

[Metalguy]

See if you can get a bigger, thicker piece, get the surface smooth, and do a 500kg Brinell test--as previously recommended.

"You see, wire telegraph is like a very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? Radio operates the same way: You send signals here, they receive them there. The only difference is there is no cat." A. Einstein

 

[Durableguy]

Thats giong to be part of the further testing I'd like to do. The nature of these parts means the thickness is maybe 3/16" and full of curves.

What does the Brinell do the Rockwell test does not?

 

[Metalguy]

As previously written, the diff. is in the indenter size and load. It appears you don't have enough thickness for a Brinell 500kg test, so see if you can have it tested using the Rockwell "E" scale (100 g, 1/8" ball).

That should give a better reading than the "B" scale--bigger area being tested.

"You see, wire telegraph is like a very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? Radio operates the same way: You send signals here, they receive them there. The only difference is there is no cat." A. Einstein

 

[Durableguy]

I'm assuming then the 'E' scale could be converted to an equivelant 'B' scale or just use that to show a percentage in change?

 

[Metalguy]

Here's one conversion chart. And yes, you are looking for the reduction in hardness/strength.

http://www.carbidedepot.com/formulas-hardness.htm

"You see, wire telegraph is like a very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? Radio operates the same way: You send signals here, they receive them there. The only difference is there is no cat." A. Einstein

 

[dtdynamics]

any updat on this? there is a few places over here who powder coat motorcycle wheels and am curious if there is a problem with the curing method are they even aware!?

 

[swall]

Dtdynamics--in my experience,as a metallurgist, it is a lack of awareness. This subject of powder coating/thermal curing heat treated aluminum tends to get glossed over in the powder coat pubs I am familiar with. You get the oft repeated "its not a problem" mantra, but never any supporting data.

 

[Durableguy]

Sorry for being away for so long. The professor who helped me wants to do further testing. Customer who has a fit over this decided to have them powder coated.

Seeing what I have from all of this, I completely agree wtih swall. There is a lack of awareness on this. At teh same time, no one has any supporting data that load bearing items (wheels) have failed because of being powder coated. A few powder coaters (of whom I am now one ) will agree these is something to this and should be aware of and not abused.

Looking at previous posts I realize I haven't posted up any data. Testing a sample piece apporx. the size of a quater yielded results varying from 29-42. Again, this was the size of a quarter. Cycled that piece along with the rest of the sample piece through the oven which had a PMT of 430 after 30-45 minutes. total bake time was 90 minutes.

I had cut chunks out of the center (hub) part with a sawzall. Looking back, this may have created some heat and altered thigns a bit. Still, 2-3 different pieces and the results varied from 12 on an unbaked section from teh hub to 62 on a baked section.

Kedp in mind these parts were not perfectly flat, let alon parralel surfaces. Even the tested surfaces had casting flaws we had tried to file smooth.

The general results did show that parts that had tested around 36-37 dropped to low-mid 20's. That was really the most accurate we could get considering the shapes of what we had.

I believe we will do further testing after the holidays.

 

[swall]

If one were to do this right, one would find a willing aluminum foundry to work with. Have them supply some A356-T6 permanent mold cast test bars and obtain the UTS, TYS and %E via a tensile test.Then powder coat a small group of bars and re-test. Publish the results in one of the powder coating pubs or the AFS publication, giving credit to all that participated.

 

[Metalguy]

Swall,

Excellent answers! That's why you got a red star.

"You see, wire telegraph is like a very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? Radio operates the same way: You send signals here, they receive them there. The only difference is there is no cat." A. Einstein

 

[Durableguy]

I would be more than happy to do this for free if the metal is supplied. Trying to get this alloy just to test seems a bit complicated. I tried, talking to a supply shop and heat treatment facility.

Tell me exactly what you want tested and how and I'll do it!