Welding Of A Molded Part

You can't weld glass. Glass is an insulator and it doesn't melt. 30% of your part is glass. Unless you perform a secondary sealing operation, be prepared to re-grind 25% of your output as scrap.

You might try ultrasonic welding. It will require some tooling modifications to add energy directors.

If this is what you are doing now, never mind.

It can be done quite successfully. Ultrasonic welding and hot plate welding are both succesful if the following have been done.

1) Your interface surfaces must be a perfect 3D match to each other.

2) Your weld material (energy director/bead face) must be correct volume.

3) You welding process is properly controlled and scientific in nature.

4) Your parts are held together while the weld material "sets".

5) Your components are properly "dry" before the welding.

6) Your test pressure is not excessive, or applied before material "set".

7) You test seal interface is also hermetic. (We found many times the interface was the leak location and not the weld.)

8) You leak tester is appropriate to the task and the leak rate you are testing at is appropriate to the component volume.

I suggest you look look up "Tagguchi", "orthogonal arrays" and "designed experiments" for possible leads on how to arrive at the optimum parameters for your product.

It does work, we had an item (acrylic) with 95% fail that sometimes didn't appear until 24 hours after welding as fractures in material. We managed to achieve 99.95% yield following some experiments along Tagguchi style (But latin squares would have reached similar results using more trials) that gave us a stable process for the four year project life.

Don't give up.

Is welding compulsory ? Can you not only use a glue to bond ?

Goo suggestion Guru, we will try that too..........

pinto

What is your leak tested method? Perhaps your test criteria are too stringent, or your test method is flawed in some way? My background is in leak testing, so if I knew more about your application I might be able to help.

Assuming this is ultrasonic welding - I presume you have a sacrificial 'V' detail in the joint faces?

If not, you have little chance of a consistent gas tight weld.