[IROC]

Originally Posted by TimT

Changing the air pressure to change the power output of the wrench is counter to what most standards accept.. usually an impact wrench is calibrated with a fixed length of hose at a fixed pressure on a calibrated tension measuring device.. (Skidmore/Wilhelm etc)

Vary the pressure, change the length of hose and the applied torque changes..

Keep the air pressure fixed, and since it seems that this is a very controlled environment, the length of the supply is fixed also..

Yep, this has been our problem. Initially, we had a mechanical torque tester and the intent was to put it into the cell and then calibrate/validate the torque vs pressure curves prior to an operation. We didn't put the tester in the cell because we needed it outside the cell to calibrate/validate new wrenches and once something goes in - it never comes back out. We then tried an operation where we took the torque tester into an airlock room where we could bring the wrenches out and test them. That wasn't optimal because we couldn't replicate the pressure drops in the cell and then...we broke the torque tester and the manufacturer hasn't made them in a while so they balked on fixing and re-certifying it. Very few (no one?) makes totally mechanical torque testers any more and the electronic ones won't survive the radiation environment (rad-hard video scopes last about 1 hour in there).

Use turn of nut..

You will have to determine initial torque for the fastener, and how each degree of rotation of the bolt increases the tension..

This is intriguing. One issue in our application is that it would be difficult to contol the bolt rotation accurately and even harder to measure the amount of rotation without some special tooling.

Unless you have calibrated the wrench at each pressure you have no idea what torque is being applied to the bolt..

Bingo.

Can you use ultrasound to measure elongation in that environment?

Doubtful. I've used extensometers before, but don't see any way to install the transducers into the bolts (and they have to be removed as once this component is installed, it is inserted ~30 feet into a core vessel on a 200,000 lb cart so access is impossible). Our ability to do delicate work is also very limited.

I really appreciate the discussion on this as it has been a frustrating experience trying to come up with a robust way to torque these bolts. Following torquing now, we test the seal by pressuring the interstitial region between the two knife-edges with nitrogen to about 85 psi and then isolate and run a rate-of-decay test. Our allowable pressure decay is .05 psi/hr (and the volume is very low ~ 1liter, so this is a hard test to pass). Right now, if the test passes, we call it good and go on. If the leak rate is marginal, we increase the torque slightly and re-torque the bolts. To give you an idea of how difficult this is in a remote environment, torquing these bolts takes us about 5 hours (and we've gotten really good at it).

We're currently having issues as the joint passed the test during the recent installation, but is now leaking at a rate of almost 4.7 psi/hr. This seal is a liquid mercury boundary, so we're watching it closely, needless to say. We're thinking about replacing the nitrogen with helium in the seal because the extreme neutron flux is turning the nitrogen into exotic isotopes that are resulting in very high dose rates in weird places. Lotsa fun.