User login


You are here

Puzzle: What caused this failure?

Here's a puzzle for our readers.  The following image is of the surface of a failed joint followed by a picture of the joint (not the same one but a similar one) before joining.  What material is it? What caused the failure?  All manner of speculation is welcome.

Failed joint

Unfailed joint


Mike Ciavarella's picture

but I am unsure about fatigue, I do not see clearly beach marks

plus what are these buttered circles?  corrosion?

tell us more about it.  What do you mean joint?  Which type of joining?  Welding?

Also what are the scales, they seem different in the two pictures? 

What is the material -- looks a metal.  Why the second 'unfailed' seems neverthless to have cracks?

Many more details for a reasonable descprition are needed to raise the case! 



I was hoping that some of the younger members of iMechanica would give it a shot before I provided more details.

The failed image has been magnified 6 times, the unfailed one 16 times (my mistake).  

The joint is not a weld and the material is not a metal.  The unfailed material does not have cracks - those are ridges generated during the cutting process.

-- Biswajit 

Not even close huh?

I think this is spot on. You can see telltale signs of the torsional failure of the bond between the alternating organic materials. My guess from the failure pattern is that the outer shell was twisted fast, causing a relatively clean failure of the binder. Perhaps this was some sort of preperation for a chemical treatment in a bath of ovine lactic fluid?


You're spot on about the torsion bit.  But I'll leave the experiment with ovine lactic fluid to you and then we can compare images on iMechanica.

-- Biswajit 


I assure you that the next time it'll be an Oreo.  That thought had never crossed my mind.

-- Biswajit 

I will do as best as this flaky Drupal and Internet connection allow me to do.

I bet it's a polymer glue and that the failure occurred in torsion (of a rotating rod), sure, but that primarily this happened because the actual cross section failed short of the required strength in torsion.

In the top picture, the circles appearing in the grayish background are due to air entrapment or perhaps, even gas release during polymer setting. The whitish patches in the center is the glue material fractured suddenly in shear. However, it may carry inclusions like crud or dust particles. The glue in the bottom also shows sudden brittle mode fracture. The bottom picture shows the glue applied with a brush, just before the two parts are axially pressed together.

I would bet that the rod material itself was non-polymeric---otherwise, one would expect dissolution of surface layers, and no gas release of this nature.

A reasonable guess would be that either wood or ceramic or metal (likelihood in that order) rods were joined using a synthetic glue, e.g., two parts of the axis of a motor or a hairdrier or a turbine or so, and that the joint failed due to above reason.


No mention of scale, material, whether it's a micro/meso/macro-graph, method of preparation of micrograph... And what else do you expect? 


I also felt initially that it could be fatigue. But, fatigue, it's not. There are no progressions characteristics of beach marks. Plus, typically, fatigue cracks grow from only one dominant point on the circumference. Here, the top picture shows a lot of shear walls. So, sure, it's a sudden and brittle failure. But not involving fatigue even in its history. (Plus, the top bubbles are not stretched in one direction, and are too big. So, they can't be dimples. They have to be gas bubbles. My best guess is air entrapment during mixing of the base and the hardner.

Up to you, Biswajit.


You analysis is amost absolutely correct.  The bottom picture doesn't have any glue on it - that's what the surface looks like just after it's cut (before joining).  The ridges are gone in the failed specimen precisely because of dissolution.  I  took the photographs while testing a light microscope that can (supposedly) magnify to 40x.  But given the vibrations in our lab only around 25x is possible.

I took the unfailed picture at a higher magnification to give you an indication of the material.  What do you think it is? 

-- Biswajit 

Mike Ciavarella's picture

michele ciavarella

Lianhua Ma's picture

I can not see the attachment, can you upload again?


There is no attachment.  Just the two images that you see in the post.

-- Biswajit 

Subscribe to Comments for "Puzzle: What caused this failure?"

Recent comments

More comments


Subscribe to Syndicate