Need innovative solution to torque bolts...

[IROC]

OK, I have a unique application where I need to torque some bolts. We currently torque these with a pneumatic wrench, but we're not confident that the we know what the actual torque value is at the socket and that might be causing us some operational issues.

I have an 8-bolt, circular pattern that I need to torque in succesive cycles of 200, 400, 600, 800 and 900 ft-lbs. We currently attempt to control the torque output of the wrench by varying the air pressure input, but due to unquantified pressure drops in the lines between the air supply and the wrench and degradation of the air pressure/torque curve of the wrench itself due to being operated without lubrication, I am not confident of that we're actually getting.

I should mention that this torquing is accomplished in a very high radiation environment, so no humans can be present to hold the wrench, adjust torques, etc. This radiation environment also precludes the use of any electronics, but electric motors are OK.

I either need to determine a simple way to measure the torque output (a calibration tool) or need to use a wrench that gives me some confidence. I have looked briefly into electric torque wrenches, but they generally need a human to change the torque settings and I'm fresh out of humans. I also designed a calibration tool, but it requires routing transducer cables outside of the work area (difficult) to access the electronic test box (and is expensive), so we haven't pursued it at this point.

Any ideas?

Here's a picture of the current operation:

[stomachmonkey]

Sounds like you need a visual indicator.

Maybe something more old school.

Obviously not this as we know they are not highly accurate, but the concept.

[Amail]

Get hold of the folks at Rotabolt. They make tension indicating bolts. You're really not interested in torque, you're interested in the tension in the bolt that a specific torque is supposed to indicate. There is a lot of error in the relationship though, mostly due to unquantified coefficients of friction between the fasteners.

We use rotabolts for several offshore applications demanding good control of bolt tension.

[IROC]

Old school is good, but we only have the ability to apply about 50 lbs of force with our servomanipulator and master-slave maniplulators, so a good ol' beam style wrench isn't going to work.

A visual indication would be wonderful also. I thought about some sort of torque-indicating washers, but with the multiple cycles, it seems difficult...

[GH85Carrera]

I don't think the Snap On truck is going to help you.

[wdfifteen]

I've used a setup which is exactly what you need, but it was a spendy system and would require running wires outside the environment.
We bonded strain gauges to a ratchet extension, calibrated the output and used it to control a rotary hydraulic motor. This was used to put a torque on an aircraft component that we suspected of premature fatigue failure, so it was cycled at about 30hz for days on end and held within something like 2% accuracy. Everything worked as designed except we had to weld the socket to the extension and use pins to attach it to the motor so there was no "slop."

[madmmac]

Calibrated hydraulic table. The table turns not the socket.

[Burnin' oil]

Just run in really fast and do it by hand . . . and hold your breath. And squint.

[IROC]

One thing is that in my application, the accuracy of the torque isn't that critical. Our limited access and our crude ability to handle the tools means that the solution needs to be rugged.

The Rotabolt thing looks good. I need to look into that. Our current bolts are custom-made Inconel 718, but that's probably OK...

[porsche4life]

Hydraulic impact?

[IROC]

Originally Posted by porsche4life

Hydraulic impact?

We can't use hydraulics unless they are water-based and we wouldn't want to use an impact wrench. We have electric power and essentially a shop air supply.

The wrenches we use now (Norbar wrenches) require as little pressure drop as possible between the air supply and the wrench. In our application we have a solenoid valve, a ball valve, two check valves, two Staubli quick disconnects, 15 feet of stainless flex hose and assorted pipe and Swagelok fittings so our pressure drops are large.

Also running the wrenches dry degrade them over time (we can't lubricate them due to a prohibition on oil/lubricants in the cell). Once the wrench is in the cell, we can't calibrate it.

Frustrating problem.

[red-beard]

A few solutions:

Hydraulic Torque wrench: Something like a Hytorq or Aztech. You can control the torque with the hydraulic pressure.

The World's Most Trusted Industrial Torque and Tension Systems - HYTORC

You could go mechanical with a torque multiplying wrench, but it sounds like you will not have the ability to use one.

Final idea is drilled bolts with bolt heaters. There are usually use on things like Steam Turbines with huge bolts. You "stretch" the bolts by inserting an electric bolt heater into a drilled cavity. The bolt elongate a set amount, based on the temperature. You then lightly snug the nuts and turn off the heaters. The bolts contract, giving you the amount of tension needed.

[red-beard]

Originally Posted by IROC

We can't use hydraulics unless they are water-based and we wouldn't want to use an impact wrench. We have electric power and essentially a shop air supply.

The wrenches we use now (Norbar wrenches) require as little pressure drop as possible between the air supply and the wrench. In our application we have a solenoid valve, a ball valve, two check valves, two Staubli quick disconnects, 15 feet of stainless flex hose and assorted pipe and Swagelok fittings so our pressure drops are large.

Also running the wrenches dry degrade them over time (we can't lubricate them due to a prohibition on oil/lubricants in the cell). Once the wrench is in the cell, we can't calibrate it.

Frustrating problem.

Then it sounds like bolt heaters will do the job.

[RWebb]

does the torque itself need to be measured?

or can you use a "twist" or angle indication to achieve the desired tension?

[vash]

Originally Posted by RWebb

does the torque itself need to be measured?

or can you use a "twist" or angle indication to achieve the desired tension?

i like the way you think!

how about making a mockup? similar mating surfaces, friction surfaces..and putting the mockup into a skidmore to measure tensions. that way you can figure exact tension with a number of rotations of the fastener. sorry, just riffing here.

[SeanPizzle]

What's the worse that could happen if they aren't torqued to spec? They won't know it was you.

[IROC]

Originally Posted by RWebb

does the torque itself need to be measured?

or can you use a "twist" or angle indication to achieve the desired tension?

I suppose this might work, but in our torquing application we're pushing stainless steel knife edges into a soft iron gasket, so we get a decent amount of bolt rotation for the first few cycles (while we're squishing the iron, essentially). Once we get to around 800 ft-lbs, the two parts go line-to-line and we're putting the compressive loads directly into the stainless parts at that point.

What we do now works, but we're having a sealing problem that might be a function of the torque and I'm not really able to defend our torquing method and am looking at better options.

Our access is relatively limited also. Here's a picture of the bolt pattern:

[Eric Coffey]

Originally Posted by NotaBRG

Thinking out loud here, but how about something like they use on cordless drill chucks? The torque is set at the chuck, so your supply pressure could remain high?

I was thinking the same thing. If it doesn't need to be super accurate, maybe fab/retro some kind of a (variable) torque "clutch" that would limit torque to specified settings, allowing air supply/pressure to run wide open (or at least high enough for the highest torque setting needed).

[vash]

that is some fancy hardware. do they make DTI's in that material..stainless?

that way you can use DIT fasteners to the 800 foot lbs..and know that they are good when they snap. then use turn of the nut for the rest of the torque readings.

[vash]

thinking way way outside the box.

do the bolts need to be the source for the initial "clamp" can a calibrated c-clamp or something put it all together first, and then you come thru with the bolting.