timsrv wrote:
Jeeze, so demanding ;) Okay I went and cut out a cross section & polished so you can see both sides of the crack. I couldn't get it to show in the pix, so I bent the element pocket slightly away from the boiler to open the crack. Even then this type of shot with my camera is tricky as it does not have a manual focus. Out of about 50 pix I got 2 that aren't bad. Yes, it is at the toe of the weld. As you can see it starts right where the weld was terminated and curves around kind of a "U" shape. The crack isn't as big on the inside of the boiler tube. I'm guessing it started on the exterior weld and probably took a couple years and many heating cycles to progress into the inside surface of the boiler. Of the failures I've seen, they have almost all been coaches owned by full timers that ran the refer constantly on 120v electric. The average time frame for a failure to occur (on the ones I've seen) is between 2-3 years of using in this manner. I hope these pix help explain things better. Tim
Tim
Thank you very much. This is greatly appreciated. I'll do some digging at work and post back.
What you are showing exhibits signs of thermal expansion stresses that are concentrated as the end of the weld due to the different metal thickness. Thermal expansion stresses are one of those hidden lucking stresses that can really sneak up on a manufacture. I have seen some pretty bazarr machine breaks by just heating and cooling metal of different thickness over time. It is sort of shocking the first time you see it. How can that happen?? I'm just heating and cooling it? But it can and does. Most of the time it takes years for the stresses to fester, build and then "poof" it broke.
I'm assuming Dometic has already done this, but the failed part you have was a perfect metallurgical sample. It is common in industry to take failed parts like these, send them to a metallurgical lab, send the working history of the part and let them do their thing. They have many techniques at their disposal including destructive testing to look down at the metal grain structure and they can tell you exactly what, how and why the broke the part.
Once the machinery builder knows why the part broke, they can then better redesign the system to not do it again. Dometic can do the same to solve this problem for good if they have not already done so.
To give you a similar look at this testing see this post. This failure listed was low cyclic fatigue caused by an impact loading condition. I have some thermal expansion ones we just finished that I will dig up. This post was on a TV receiver but the metallurgical techniques are similar to what would be used on this cooling unit. For those who are into this kind of stuff see here:
Metallurgical AnalysisThanks again for helping educate us. We all now know better what to be on the look out for.
John
