What's following is my experience with many types of plastique and polymer which includes PTE.RGC_0767 wrote:These air pockets in molding processes, are causing these companies great grief lately lol
There are so many possibilities where an air pocket can form with TPE that the only way to "kinda" make sure that there's none is to have some kind scan that goes through the doll. (Due to the metal frame, it can't be something like an X-Ray obviously.)
And since this kind of technology is still quite on the high-cost side of the market, it mostly only started to appear for dolls in high end studios.
(In those studios, the dolls are usually handmade in a way that allow the studio to scan each "part" after being processed for air bubbles and potential stiffness. It's a long hence costy process as it might add quite a bit of additional work if some pieces doesn't pass the tests.)
The way a piece is scanned is actually quite simple in principle. You put the piece in an heating room (or box) and you got heat detection camera on all sides. You record the whole thing as you slowly raise the temperature in the room. While the temperature changes, it doesn't take long to detect some variation (spots) in multiple cameras when there's an air bubble within the TPE as its temperature will shift differently than the PTE material.
Thermal cameras aren't hard to find nowadays, but ones that can film and scan at high resolution so that even a slight <1mm bubble can be seen if zoomed in and visually detected, that's where the cost is as well as the cost of energy and space to actually have an heating room. It also have to be able to record at, at least, 60 FPS for better quality controls.
As a matter of making things easier, one process of detecting air bubbles in TPEcould be even more efficient if the TPE is, initially, cooled (not frozen, I minds you). Remember that when cooled, air's molecules extend their links, making the air less dense (so bubble grow a bit in size). Then, if you heat it up, the air bobble will contract and become dense. This can help detecting "bigger" spots during the first 1/3 of the heating process while it's being recorded.
Note that, when I write about heating up the TPE, I don't mean at high temperature. You could say that you could notice, with the high sensible thermal camera, a visual change (spots) if you were to cool the doll parts to around 8-10 degrees Celsius (41F-50F) and, then raised the temperature to approx. 35 degrees Celsius (95F) at a time based on the equivalent to the time required for the most dense part of the part to change temperature (this value varies depending on the TPE used in the doll.) Just as the air bubble gets affected by the heat going through the TPE, there would be a slight variation in the temperature between the TPE and the air trapped. After that, watching the recorded videos in sync (showing all angle at the same time in the heating process) on slow-motion (2-3x slower, meaning around 20-30FPS) would display the aforementioned air bubbles as blurry spots.
In the case of what we have seen in the picture and how we all can guess it was most likely an air bubble in the TPE, my guess is that either there were no actual thermal checkup done on the doll in the inspection. As I have previously state, it's not rare since it's quite time consuming, requires some space for the heating room and can be quite costy with the high definition thermal camera required on all sides and all. Another thing that could have make this happen as fast as it did is if there's a major difference between the temperature of the room in which you got the doll and the temperature in which the doll was manufactured and checked.
As I have previously stated, air expand when heated up. If there was a tiny minuscule air bubble (less than 1mm) and, for a moment, the room where the doll is stored was a bit too hot, that air bubble would expand just a tiny bit, raising the tension of the TPE by quite a bit (if you consider the torsion created from moving the arm away from its "default" position).
To be honest, as a game developer who does both scripting and 3D modeling & animations, arms (and legs) animation in games is actually one huge hard thing in the process because "stretching" is something quite common and seen as normal. To allow as much movement without as much stretching, it's a common practice to make the 3D model with arms set in a T position.
Something like this:
This is a lot easier to model (as it follows the X, Y and Z directions in 3D space) and allows the character, afterward, to raise the arms without stretching the underarms. The downside is that the biceps and shoulder area actually end up deformed when the arms are dropped. (That's a 3D thing.)
To counter that effect this is the "best" position when it comes to fight against stretching in 3D:
This is the best pose to have the stretching extreme force as low as possible.
Since it's recommended to leave the doll in its factory position to avoid constant stretch of the TPE, I guess that the "stick" position of the doll is coming from that so that the owner of the doll doesn't have issues with the arms taking a bit too much space. It also helps for the delivery in a compact box.
Personally, if I were to make a recommendation to the factories and shops, I would recommend making the default position I have shown above. While, year, this means that the "stick" position would mean keeping a slight amount of torsion onto the doll's TPE, the amount is not that much if the density & elasticity of the TPE is well balanced.
As to make this reply an actual answer to the "Help" request of this topic... here's my 2 cents.
For those who get such holes where glue and solvent doesn't really work due to the amount of stress being high in that zone, I could suggest the innovation of adding a bit more TPE while keeping the arm halfway toward the side. (Like 45 degrees toward the outside.) This means it's actually quite like a surgery operation where you now have to make a new armpit in the hole, but it would also means that, afterward, the armpit wouldn't have as much tension as it currently has.
The only downside to this operation would be that the arm would be slightly harder to bring into the "stick" factory position and there might be some creases if the armpit isn't correctly formed (as to fold onto itself).
If my upcoming Date does ends up with such a hole under her arm, I'll be certain to try my suggestion even through I don't really know how I would do it right now. (Gotta cross the bridge only at the bridge.)
Buy some electronic soldering tools lad, they WILL come in handy for future. I use mine more than just for dolly 
