The last mechanical step before LabVIEW, please help, pls give some suggestion

[GSR]

Dear all,

I need some suggestions of this mechanical design (I understand it is not related to LabVIEW , but this is the last step before LabVIEW, please help ). I think there are some problems.

The idea is to use load cell to measure the tension and compression force across two metal structures. The load cell gives a linear reading for compression and tension, the structure in between should not affect the reading

The support frame is designed to prevent twisted force act on the load cell

One of the load cell hole is a tap hole, the other one is a REGULAR hole

Please give some ideas, do you think the structure in between does not affect the reading??

[vugie]

Hard too be sure without strict technical drawing, but it seems that two black parts you want to measure force between are stiffly connected with thin silver disc.

Think of load cell as of a spring. Measuring the force is nothing more than measuring deformation of the spring in some way. If spring is not allowed to be deformed you will not measure anything. So when you have a mechanical chain and want to measure force between its two parts, you have to cut the chain in this point and put load cell between connecting INDEPENDENTLY both sides of a cell to respective parts of a chain.

[JasonXCX]

It is difficult to answer to your question. It lacks so many details about the whole system.

where this problem come from? Are you a student?

the quality of the picture is very low . out of focus.

take good pictures of the differents parts , make drawing with a pencil and paper

Have you made some drawing? .

First of all you must know:

intensity of forces

rigidity of load cell.

it looks like an hyperstatic problem. Not too easy to solve.

The long answer is :

- full drawing (pencil and paper )

- schematic view

- stress and strain calculation or approximation.

The short one is use some

http://www.skf.com/portal/skf/home/products?maincatalogue=1〈=fr&newlink=3_3_11

it looks like the load cell have thread use it for the skf parts.

NOta ;

if you have only compression you could use ball from ball bearing .

jason

[GSR]

Think of load cell as of a spring. Measuring the force is nothing more than measuring deformation of the spring in some way. If spring is not allowed to be deformed you will not measure anything. So when you have a mechanical chain and want to measure force between its two parts, you have to cut the chain in this point and put load cell between connecting INDEPENDENTLY both sides of a cell to respective parts of a chain.

I agree with what you said . Exactly what I am thinking, the load cell is not connected independently to those two metal pieces because of the inter-connection of those screw and silver disk

The short one is use some

http://www.skf.com/p...&newlink=3_3_11

The biggest problem is that I am not the designer and I don't have the CAD drawing.

I should not be lazy and I should use make a drawing using pencil

BTW, I only understand English. Do you have English version of the above link? Thanks

[vugie]

I agree with what you said . Exactly what I am thinking, the load cell is not connected independently to those two metal pieces because of the inter-connection of those screw and silver disk

But there are situations where such parallel (to load cell) connection makes sense. For example you expect very high forces, but the sensor has very small range. With inter-connection sensor feels only a part of total force and the factor (F_sensor/F_total) is equal to k_s/(k_s+k_i), where k_s and k_i are stiffnesses of sensor and inter-connection respectively. Typically any inter-connection will has much much higher stiffness than any sensor and consequently force felt by sensor will be below its sensitivity. But when k_s and k_i become comparable it may be usable.

Edited August 12, 2009 by vugie

[JasonXCX]

you have only to choose the language in the right:

http://www.skf.com/portal/skf/home/products?maincatalogue=1〈=zh&newlink=3_3_11

or

http://www.skf.com/portal/skf/home/products?maincatalogue=1〈=en&newlink=3_3_11

or in USA:

http://www.aurorabearing.com/

You can also find this kind of spherical bearing in car suspension front wheel . You can find them in a junk yard for exemple for cheap.

But may be it is too big for your application?

I agree with what you said . Exactly what I am thinking, the load cell is not connected independently to those two metal pieces because of the inter-connection of those screw and silver disk

The biggest problem is that I am not the designer and I don't have the CAD drawing.

I should not be lazy and I should use make a drawing using pencil

BTW, I only understand English. Do you have English version of the above link? Thanks

[Of course you must cut the chain. And have only traction / flexion and no kind of bending.

The spherical bearing prevent the propagation of bending. If the load cell is bended the spring is subjected to coupound sollication it is bad for the precision. ( and it can damage the cell )

in other world You must have an idea of the sollicitations. What are the component of the screwer ( torseur ) on the cell?

Some numerical value could be interesting.

Hard too be sure without strict technical drawing, but it seems that two black parts you want to measure force between are stiffly connected with thin silver disc.

Think of load cell as of a spring. Measuring the force is nothing more than measuring deformation of the spring in some way. If spring is not allowed to be deformed you will not measure anything. So when you have a mechanical chain and want to measure force between its two parts, you have to cut the chain in this point and put load cell between connecting INDEPENDENTLY both sides of a cell to respective parts of a chain.

The spherical bearing prevent the propagation of bending. If the load cell is bended the spring is subjected to coupound sollication it is bad for the precision.( and it can damage the cell )

in other world You must have an idea of the sollicitations. What are the component of the screwer ( torseur ) on the cell?

Some numerical value could be interesting.

[GSR]

But there are situations where such parallel (to load cell) connection makes sense. For example you expect very high forces, but the sensor has very small range. With inter-connection sensor feels only a part of total force and the factor (F_sensor/F_total) is equal to k_s/(k_s+k_i), where k_s and k_i are stiffnesses of sensor and inter-connection respectively. Typically any inter-connection will has much much higher stiffness than any sensor and consequently force felt by sensor will be below its sensitivity. But when k_s and k_i become comparable it may be usable.

My situation is having 500lb load cell to measure the force that in between 0 - 250lb. I understand what you say. Your idea is like the potential divider in Electrical term. I really have no clue why my boss and his friends consider that is right . The biggest problem is the compression and tension forces have different offsets and gain value because of the "Middle sturture" bypass part of those forces. Moreover, how tight the screws will also change the gain of the reading and the "Middle sturture" also creates an idle region. My understand is that the output is Y = G1*compress if compress > theshold1 and Y = G2*tension if tension < thershold2. Those thersholds can be changed by turning those screws.

[JasonXCX]

The spherical bearing prevent the propagation of bending. If the load cell is bended the spring is subjected to coupound sollication it is bad for the precision.( and it can damage the cell )

in other world You must have an idea of the sollicitations. What are the component of the screwer ( torseur ) on the cell?

Some numerical value could be interesting.

[GSR]

The spherical bearing prevent the propagation of bending. If the load cell is bended the spring is subjected to coupound sollication it is bad for the precision.( and it can damage the cell )

in other world You must have an idea of the sollicitations. What are the component of the screwer ( torseur ) on the cell?

Some numerical value could be interesting.

Thanks for your suggestion

However, on the other hand, do you think this prevent bending feature affect the reading??

[JasonXCX]

It is all about numerical value. Most of the load cell are designed to measure only the force in one direction.

If the load cell is subjected to force to other direction and to bending the measurement is false and the load cell can be destroyed.

Look for the documentation of the load cell.

Have you pictures of all the system.

What is the context?

Thanks for your suggestion

However, on the other hand, do you think this prevent bending feature affect the reading??

[lecroy]

My situation is having 500lb load cell to measure the force that in between 0 - 250lb.

If that is true, where is the +/-? Are you trying to preload (offset the force) the cell so is goes through the 0 point? If so, why?

What sort of resolution, accuracy and repeatablity are you required to meet?

I am currently working with an 'S'-type load cell. The twist in my case is prevented by the system design and the amount of deflection has no effect on the data. Maybe there is something that can be done in your case that would allow you to remove all that added hardware around the loadcell. Do you have any pictures or can you post some details about what your trying to do at a higher level?