How to know a strain gauge's accuracy?

[GProgramer]

Hello all.

I am looking over various strain gauges as I need to purchase one for my project. However, I have never worked with them before, and I cannot seem to figure out how to know its accuracy (able to detect a certain change in pressure, for example changes of the scale of 1mmHg) , seeing as none of the datasheets seem to have that kind of data.

Forgive my inexperience..

 

[jim hardy]

""...able to detect a certain change in pressure, for example changes of the scale of 1mmHg)

strain gages measure deflection

are you seeking a pressure transducer?

some of them are strain gage based.

 

[GProgramer]

I'm actually looking for a strain gauge. I was looking for a transducer at first but I can't find any that can remain unaffected in magnetic fields up to 4Tesla or so, so I'm building one of my own using a strain gauge.

 

[Studiot]

Jim was on the nail with his comments.

You need to provide considerably more detail of the mechanical arrangements before any useful comment about the electrical ones can be made.

 

[GProgramer]

I still don't have quite the clear idea yet, seeing as I don't have a good enough idea myself. This is pre project preparations so I don't have the full data yet.

I have a tube the girth of a vein, and I'm running liquid through it. I need to be able to capture the pressure using a sensor, whether it's a stain gauge or a pressure transducer with an accuracy of 0.1mmHG.
The output could be anything. Ohms, voltage, etc..

 

[Studiot]

This is more useful information.

Does this imply you can't insert anything into the flow?

What are the properties of the tube?
That is
Is it flexible enough to distort sufficiently to measure with a strain gauge?
Can strain gauges (you normally need more than one) be attached or is it some type of plastic that you can't stick to (strain gauges have to be mechanically bonded to the item being strained)?
If they can't be glued could a suitable collar be fashioned?

How fast will the pressure change?

Do you want absolute pressure or just variations?

What sort of speed and frequency of measurement is required?

For instance a resistance strain gauge will not be affected by the magnetic fields you mention. They can be manually balanced (slow + cheap) or electronically monitored by equipment ouside the mag field (fast, convenient but expensive).

What is the temperature range of the fluid?

Anything else relevant you can think of?

 

[AlephZero]

You are talking about very small changes of pressure here. The order of magnitude is
1 atmosphere = 760mm Hg = about 100 kPa
So 0.1mm Hg is about 13 Pa.
That will produce a very small strain unless you have a flexible pipe, but a flexible pipe could give calibration problems by not staying circular (i.e. the strain gauge is stiff compared with the pipe).

You might be able to use a laser vibrometer (LDV) to measure the change in shape of the pipe directly (to an accuracy of nm, or even pm with special image processing software). They can measure steady-state as well as dynamic effects.

 

[GProgramer]

First, thank you for your genuine interest in this!
I'm going to answer the question as best I can, seeing I don't have all the info.

This is more useful information.

Does this imply you can't insert anything into the flow?
Seeing as this is a preliminary study done by us (students) we are going to be using water as a liquid, but the long term goal of the project is using a liquid that has the same viscosity as blood. So in a way, it doesn't quite matter.

What are the properties of the tube?
That is
Is it flexible enough to distort sufficiently to measure with a strain gauge?

It -should- be flexible enough to be measured with a strain gauge, but so far I haven't managed to find one that is thin yet flexible, at least not flexible enough for a small gauge to measure it. That being said, it is flexible enough to easily bend by hand, I don't know if that is enough for it to be detected by the gauge.

One idea I had was to put the gauge inside the tube, to be affected by the flow, but that would probably cause short circuit problems.

Can strain gauges (you normally need more than one) be attached or is it some type of plastic that you can't stick to (strain gauges have to be mechanically bonded to the item being strained)?
If they can't be glued could a suitable collar be fashioned?

Hmm, the tube I have in mind is plastic of sorts. I have yet to try to glue a strain gauge to it.

How fast will the pressure change?

Do you want absolute pressure or just variations?

What sort of speed and frequency of measurement is required?

As fast as a human's blood pressure could change, so anywhere between 60mmHG and 120mmHG.

As for the speed and frequency, it will depend on how fast the pressure could change. So every heart beat, so.. around 1-2Hz or so?

For instance a resistance strain gauge will not be affected by the magnetic fields you mention. They can be manually balanced (slow + cheap) or electronically monitored by equipment ouside the mag field (fast, convenient but expensive).

I've been doing my research on the strain gauges and already found a strain gauge that is resistant to magnetic fields. I found that a constantan alloy, and consists of two identical grids, with one stacked directly above and insulated from, the other. That way they cancel out the noise.
Finding it has lead me to this question, as I don't know how to figure out it's measuring accuracy.

What is the temperature range of the fluid?

Either 20-30 degrees or 30-40 degrees, depending if we go with room temperature or that of the human body.

Anything else relevant you can think of?

Not really! Quite honestly you gave me a lot of questions to think about myself and which I'll be asking the prof I'm working with! So thank you for that!
Edit: Saw your reply after I posted Aleph!

You are talking about very small changes of pressure here. The order of magnitude is
1 atmosphere = 760mm Hg = about 100 kPa
So 0.1mm Hg is about 13 Pa.
That will produce a very small strain unless you have a flexible pipe, but a flexible pipe could give calibration problems by not staying circular (i.e. the strain gauge is stiff compared with the pipe).

You might be able to use a laser vibrometer (LDV) to measure the change in shape of the pipe directly (to an accuracy of nm, or even pm with special image processing software). They can measure steady-state as well as dynamic effects.

Yes unfortunately it has to be quite small.

I've never heard of LDV, so I will do some searching on that and get back to you!

 

[Studiot]

OK we seem to be talking about blood, plasma, or an infusion flow.

We seem to be measuring in a tube or vessel external to the body.

The requirement is a change of 0.1 in 100.

Now one issue is that the pressure will vary within a flow from point to point acrss its cross section. The output from a blood pressure monitor is an average across the whole.

I am not a medical physicist, there are some here at PF who might usefully contribute to this.

I wonder if a diaphragm pressure cell, perhaps with piezoelectric transducers might be the way to go? These can be very sensitive, easily within your range.

I will see if I can do a sketch later tonight, I have a meeting shortly so will have to leave for now.

What is your discipline by the way?

 

[GProgramer]

OK we seem to be talking about blood, plasma, or an infusion flow.

We seem to be measuring in a tube or vessel external to the body.

The requirement is a change of 0.1 in 100.

Now one issue is that the pressure will vary within a flow from point to point acrss its cross section. The output from a blood pressure monitor is an average across the whole.

I am not a medical physicist, there are some here at PF who might usefully contribute to this.

I wonder if a diaphragm pressure cell, perhaps with piezoelectric transducers might be the way to go? These can be very sensitive, easily within your range.

What is your discipline by the way?

The problem is I don't have much experience in biomedical field either. I'm studying computer engineering, and with a focus on programming. But this was the only project available left.

So far I'm glad you have given me many options to ponder and think, even if I'm out of luck as having any more help in this. You've done more than enough to help me so far, so thank you!

I will try and search for the solutions you guys have proposed so far. And I will ask the questions you've given me to my prof.

 

[jim hardy]

well

maybe you have the sensor in hand already

do you know what is a "bourdon tube" ?

a tube wound into a spiral will try to unwind when subjected to internal pressure.

Thought experiment - take a long tube of rectangular cross section
and wind it into a spural.
Because it's circular in shape, the inside wall is now shorter than the outside wall so those two walls have unequal areas
width * 2 * pi * inside length < width *2 * pi * outside length

so the internal pressure makes unequal forces P X Area that try to straighten it out.

try winding a couple feet of your tube into a coil and blow in the end
if it moves perceptibly that's a start. you need to measure its displacement, maybe with a front surface mirror and laser pointer.

may not work at all. but it'll get your creative juices flowing.

 

[Studiot]

I bourdon tube is the diaphragm type but I doubt that it would be sensitive enough.

Here is a sketch of a pressure diaphragm transducer that would meet your needs.

It is basically a chamber with one very flexible wall, made of very thin sheet metal or plastic. This arrangement gives much greater deflection than straining the wall of a tube.

The deflection can be measure by crossed strain gauges, a piezo pickup (could be similar to an old fashioned gramophone pickup if you need to make your own in a laboratory) or a (laser) interferometer that measures displacement of the central boss in response to pressure changes.

Changes of 1 in 10⁵ can readily be detected with this system - you only require 1 in 10³.

The piezo pickup (electrostatic) and the laser (light), are inherently immune to magnetic fields.

Are you planning to build this and if so, is budget or equipment availability an issue?

 

[GProgramer]

I see! That would definitely take take care of the tube's material problem! I have a few questions, if you don't mind..

Wouldn't the rectangular surface of the cell affect the pressure somehow as opposed to the round nature of the tube? Or is it too small to have any effect?

A 10^5 sensitivity would be done even with the gauges or just the laser and piezo?

Last question, might I ask why the gauges need to be crossed? Regarding the budget, I would have to ask the prof. I can afford somewhere around $300 but the prof would probably want the whole system to be more or less reasonably priced. Though what she considers reasonably priced, I do not know.

And I definitely need to build this.

 

[Studiot]

A 10^5 sensitivity would be done even with the gauges or just the laser and piezo?

Last question, might I ask why the gauges need to be crossed?

Yes, possibly, but I would still recommend the piezo approach.

Because one measures tension (pressure increase above average) and the other compression (pressure decrease below average)

Wenk E Jr : A diaphram type gage for measuring low pressure in fluids : SESA proc 8(2): 90, 1951

At this stage you need a detailed discussion with your project tutor about what you want to find out.
College projects are meant more as a learning exercise rather than achieving the 'primary goal of a working piece of equipment' as would be the case in industry, although seeing it work is certainly nice.

So answers to questions like

Wouldn't the rectangular surface of the cell affect the pressure somehow as opposed to the round nature of the tube? Or is it too small to have any effect?

might be something to find out and discuss in your report. I can tell you that working pressure cells with diaphragms are commercially available off the shelf and that working pressure cells based on flattening a section of round tube with or without mounted strain gauges are also available this way.

If this is truly about blood then you need also to consider the effect of any flattening or constriction on the cells and particles in the blood.

go well

 

[jim hardy]

i guess you've checked these people?
http://www.allsensors.com/catalog-pdf-download

about sixty bucks at Digikey

the 4 teslas seems an unusual condition
does this need to work in a MRI machine or somethiing?

 

[Studiot]

I also asked about temperature.

This is because temperature variations of the size you are talking about will (I think but you should check) bring about pressure variations within you measurement range.

So you should think about a standard temperature to reference to and a formula to do this.

 

[GProgramer]

I'm not quite sure of the details, but I think it's somewhere along those lines Jim.
Also the temperature will be controlled, somewhere along the lines of 20-40 degrees, we still haven't decided.

Thank you again all for all your help. I realize I'm asking a bit too much. The remaining questions, as well as the ones you posed, along the sensors you proposed will be discussed with my supervising prof on Monday.

You've given me more than enough to take the first step, and for that thank you!

 

[Studiot]

Keep us posted about your progress and good luck with the project.

 

[Studiot]

You might like to look at some commercially available pressure measurring equipment.

http://www.testo.co.uk/online/abaxx-?$part=PORTAL.GBR.SimpleContentDesk&$event=show-from-menu&categoryid=1690408