Capturing collapse response of tube 3

[vtmike]

Hi,

Would a regular strain gauge in wheatstone circuit be the best option to capture the lateral collapse of a tube under external pressure? The tube will be in a pressure pit & out of view for the duration of the test. Since we cannot visually monitor it during the test, there is no way to figure out when the collapse initiates.

 

[Ron]

Yes, but use strain gages on all sides of the tube and redundant gages on the expected failure side(s). Place them at regular intervals so that you can see all of the stress distribution. Gages are cheap...use a lot of them.

 

[vtmike]

Thanks!

Another problem is that we only have one free port from the pressure chamber, & there is space to insert two wires only. So that means we can only attach one strain gage which is not sufficient to capture the stress distribution. I was wondering if there exists a ring gage that will cover the ID of the tube? That way there will be only two leads coming out of the pressure chamber.

Mike

 

[rb1957]

i think you can set up a bridge to average the stress around the perimeter ... all the gauges connected in series would be a simple way to do it (just divide the output by the number of gauges).

so the next question is are you expecting a global failure (like euler) or localised ? if the latter, then i'd expect one side of the tube to experience higher stresses than the other and the overall averaging would give you a lower result ... average where you aniticpate failure.

btw, can't you rig up a video feed ?

could you connect the s/gauges to a wireless xmitter (and get around you wires limitation)?

have you modelled this ?

 

[vtmike]

Thanks rb1957,

Like you said all bridges in series would give a lower stress if the failure is localized (maybe due to manufacturing defects), so I not too keen on using that option even though it will most likely be a global failure.

A video feed is not an option we can use.

The wireless transmitter seems to be the best option if it is available. Can you suggest any references?

I am in the process of modelling this, but am having some problems with the FEA model. So it is in process.

 

[Ron]

vtmike...how do you intend to bond the gages to the inside of the tube?...or am I missing the setup. Is this a tube confined inside something else?

Using wireless gages is a possibility, but be aware that they are not necessarily as consistent as wired gages, depending on potential interference.

 

[vtmike]

Wireless gauges is not a possibility because the transmitter cannot withstand high temperatures (the temp would be 300 F)

It is a hollow tube with sealed ends and bonding the gauge on the OD would be much more easier than bonding it on the ID

 

[Ron]

Ok, given that, I would use a couple of rosettes and a couple of linear gauges. Set all gauges in pairs, 180 degrees apart. I would use 4 rosettes, 2 above the anticipated area of collapse, and 2 below. I would then use 4 linear gauges, oriented parallel to the long axis of the tube, then 4 linear gauges perpendicular (transverse) to the long axis. The transverse gauges will tell you when the tube starts to expand internally, as will the rosettes. The data from the gauges can then be plugged into your FEA model to check stress distributions and patterns.

I realize you only have limited space to get your wire into tube area. The gauge wires are very small (20-24 gauge).

 

[btrueblood]

I guess if you want to know the buckling mode, you would use a strain gage...

But if all you want to know is the pressure at which buckline occurs, wouldn't you simply monitor pressure vs. displacement (volume of pressurant liquid pumped into the chamber)? When the pressure vs. volume curve flattens, it's a pretty good sign that either the tube buckled, or you sprang a leak somewhere.

 

[Compositepro]

How about acoustic emission? That is listen for failure with a microphone and record the audio signal.

 

[MikeHalloran]

The one tube collapse I personally know much about would have been very easy to detect. It made a >>big<< noise, and everyone in the shop heard it, even over the usual din from grinders and such.

( They didn't report it at the time, because it was the inner tube of a jacketed u-tube. The outer tube suffered no damage, and nobody looked deep inside until the customer reported unusually high pressure drop. ...Ooops... )






Mike Halloran
Pembroke Pines, FL, USA

 

[vtmike]

Ron,

Thanks for that information. You are right, I think using linear gauges looks like the sensible way to about for the test.


Monitoring the pressure drop & hearing a loud bang is normally how we do it right now, but we want to capture the failure collapse mode.

 

[hacksaw]

just curious, how can you anticipate the "expected area of collapse", am i missing something, thought with the various modes of collapse, the orientation of the collapsed state could not be known in advance.

 

[vtmike]

I did an FEA to find the modes of collapse. Usually max deflection is somewhere along the middle of the tube following the first mode of collapse, which is where I plan to place the strain gauges.

 

[rconner]

http://sfwj.blogspot.com/search/label/Water in motion