strain gauge out of a mile long wire 1

[ultrasoundguy2]

hello
say I have a coax wire about a mile long that is subject to strains of 1% of the total length.
This particular wire has a resistivity of about 4 ohms per 200ft.

Is there a way to make a strain measurements out of this?.
if could do a bridge into a differential amplifier if needed.

what are the pros and cons of this system
I could put an amp though the wire to boost the generated voltages too.
I know it would be subject to temperature variations and joule heating which are not a problem in this case.

yes, I know there are better ways to measure strain over long distances.
this is a niche application and has to be done this way.

thanks
Jim

 

[rb1957]

are you really getting 5' movement over 1 mile, or is this a spec thing ?

sure you could measure the change in resistance, but first you'd need a slack cable (which i'm guessing is hard to do) to give you a zero.

what are you trying to prove ? how the co-ax performs (electrically) with 1% strain ? or how the co-ax performs structurally with 1% strain ? or signal degradation ?? sounds to me like you could do this is a lab ??

Quando Omni Flunkus Moritati

 

[ultrasoundguy2]

cant think of a way to strain a wire a mile in the lab.
really looking for references or ideas so I can do some calculations on preliminary work.

even if I can record a strain, I need to know what it means as I have to know how to interpret the strain..

 

[ultrasoundguy2]

woops... need to know how to interpret the recorded voltages into strain..

 

[rb1957]

"cant think of a way to strain a wire a mile in the lab."
1) you could roll it up.
2) you could test 10' with 1% strain.

"how to interpret the recorded voltages into strain?"
1) google "strain gauge", go to the wiki article
2) buy a text book

Quando Omni Flunkus Moritati

 

[ultrasoundguy2]

don't think im going to see an effect with 10' unless I have a super accurate ohmmete3r.
rolling it wont product enough change to get the gage factor..

the terms I needed were "elastic wire strain meter", a type of "unbounded strain meter"
these are giving me the background I need such as the best wire types..

yea, Wikipedia has all the answers
thanks anyway buddy

 

[TheTick]

There's got to be better ways to measure displacement over a mile.

 

[TheTick]

Don't forget to run a second wire in parallel that is not under strain to compensate for temperature.

 

[rb1957]

"yea, Wikipedia has all the answers
thanks anyway buddy"

hence option 2),
and you're welcome.

btw, you seemed to pick on one comment in my post, so ...

1) are you really getting 50' movement over 1 mile, or is this a spec thing ?

2) sure you could measure the change in resistance, but first you'd need a slack cable (which i'm guessing is hard to do) to give you a zero.

3) what are you trying to prove ? how the co-ax performs (electrically) with 1% strain ? or how the co-ax performs structurally with 1% strain ? or signal degradation ??

4) 1% strain sounds like an enormous amount, a typical yield strain is 0.02%. i doubt the co-ax will remain intact.

Quando Omni Flunkus Moritati

 

[dhengr]

Ultrasoundguy2:
I think you should do your study and testing on a 50 or 100' length of cable, in a lab, and then factor the results by 5280ft./mile. In the lab you have a chance of controlling the tension on the cable and the ambient temperature around it. In the field, on a mile long cable, you will have variations in initial tension, and in the temps. over various unit lengths of the cable. Also, unlike a single fine wire, cables have funny manufacturing strains, and relaxation and tightening and bending of wires over each other, as a function of the way the cable is manufactured. You should study how this affects your results. Maybe you should pre-stretch the cable to take some of this manufacturing slack out of the cable.

 

[ultrasoundguy2]

yes, I would have to do about 100ft to get a measurable resistance. (2-4ohms per 200ft)
I will probably have to do this, but not in the lab, more outside in the road..
its the only way to get the gage factor

I can check temperature effects in a controlled environment and then remove the effects after data is recorded.

ive now looked in the literature.
this method can measure the strain typically in the range of 4- 20" and is accurate to 1.5 to 6 microstrain
so this would be stretching the knowing distance it works.


I need to figure out what type of wire to use though.

Im not sure what kind of stain we are going to get.

 

[rb1957]

i don't know of any material that can carry 1% strain.

eg Al (which i know) ... E = 1E7psi, 1% strain = 100ksi (> ftu)
steel E = 30E7psi, 1% strain = 300ksi (> ftu)
etc

Quando Omni Flunkus Moritati

 

[rb1957]

if you two end points are moving 50' relative to each, i'd recommend having strain relief loops (ie 50' extra length so you don't strain the cable).

Quando Omni Flunkus Moritati

 

[ultrasoundguy2]

I don't know the strain we are going to get.
Im guessing 1% might be high, more like 2-3 u strain


but im still not sure this can work.
all the references out there on unbounded strain gages only cite certain types of metals with high gage factors.
such as nickel or constantan

http://soliton.ae.gatech.edu/people/jcraig/classes/ae3145/Lab2/bridge-measure.pdf

I want to know if this can work with standard coax and steel or copper cable.
cant find the specs on gage factor for these metals.

 

[ultrasoundguy2]

I guess copper is in the table if I look carefully

 

[3DDave]

Look at the bulk resistivity for the material and the Poisson's ratio. Poisson's ratio will tell you what the expected decrease in area is for a given increase in length. Then use the bulk resistivity in an equation that relates section area and length to produce the resistance.

This will apply to the solid core of a wire, not the braided shield.

 

[ultrasoundguy2]

thanks you dave and everyone..

there is no potential on the braided shielding so it shouldn't matter if it and the jacket strain too

really not sure if this will work, its a very niche thing.
there has to be a reason it wont work, something non ideal like the wire thickness not being uniform?

 

[tbuelna]

How do know what amount of the resistance change measured can be attributed to strain solely created by displacement in the conductor length due to a change in the distance between the cable end points, and how much resistance change is missed by effects like elimination of slack in the cable? Over the distance of 1 mile, just the weight of the cable itself resting on the ground will create a substantial amount of friction drag when the cable is tensioned. The strain produced along the length of the cable will not be uniform, it will be greater towards the end the force is applied to.

 

[MikeHalloran]

If your coax has a solid core, you might be able to get a useful strain signal out of it, using the shield as your reference material. If the resistance per unit length of the shield is much different from that of the core, as i suspect, you will have an unbalanced bridge, so you may need funny business to correct for that. I'm pretty sure that you will need a bridge; no ordinary ohmmeter will have the resolution.

If your coax has a stranded core, I don't think you'll get a stable and usable signal that correlates to strain out of it.

You'll also need to take several wraps around something at each end; the usual crimped connections won't survive much pull force.




Mike Halloran
Pembroke Pines, FL, USA

 

[ultrasoundguy2]

tbulena - excellent! of course, im counting on this, I cant say why though, sorry
mike, somewhat good.. The shield wont matter much I don't think as there is no potential on it... I can make a bridge and differential amplifer if I need it from another cable of the same length without strain on it.