A loadcell which only measure force

[lmuri]

Hi, All,

I am in a very strange situation. I had to measure the force along a shaft which withstand both compression force and bending moment and direction of the bending moment keeps changing.

Because I am only interested in the compression force, I used to connect a six-axis loadcell, which can give me both compression force and bending moments, on the shaft. However, our new design does not permit me to put a loadcell in and I have to provide an alternative to the loadcell. The moment is quite large can break most of the one axis loadcell which is available on the market and fit my requirement.

Good part of the problem is that the working environment is not harsh at all.

It seems that I will have to build some strain gauge based sensor myself. But I feel headache about the changing bending moment and its influence on the strain. Any suggestions?

Thank you in advance.

Best.

Ming

 

[BrianE22]

What is the ratio of (strain from bending)/(strain from axial force)?

 

[rb1957]

strain gauges, directly onto the shaft ? average strain (of a regular pattern) = axial load.

add a 2nd set of s/g to confirm (i'm assuming that the axial load isn't changing in the shaft)

 

[rb1957]

i'd probably stick all the signal conditioning etc on the shaft too ... i imagine it's whirling around at some speed.

maybe you can get s/gs that broadcast their readings (RF) ?

 

[hydtools]

Apply strain gages to measure axial strain and cancel bending strain.

http://www.omega.com/faq/pressure/pdf/positioning.pdf

Ted

 

[lmuri]

The axis compression load changes from 600N to 100N and the moment changes from 60Nm to 10Nm. It is a relative thin shaft but I am not limited to the cylinder shape for the loadcell.

Best.

Ming

 

[rb1957]

can you mount s/g on the shaft ?

 

[dhengr]

Install a single strain gage at the neutral axis.

 

[dhengr]

Install two strain gages, diametrically opposite each other. The difference in the strain readings should be double the strain caused by the bending, at any instant in time. In that way you can measure both the axial train and the bending strain if you wish. Place these paired strain gages at 10° intervals if you wish. If one reads zero, the other is reading half axial and half bending strain.

 

[rb1957]

isn't it as simple as averaging opposite pairs of s/gs ? (sounds a lot shorter that way, no?)

can we have some sort of pic of the problem ? some reason as to why the obvious (putting s/g on the shaft) doesn't work ?

 

[BrianE22]

As I mentioned earlier, you really have to compare the strain (not the loads) due to bending vs the strain due to the axial force. You'll never get a gage lined up exactly on the neutral axis. A small alignment error can cause large cross axis errors if the strain from bending is much greater than the strain from the axial force.

One of the more popular techniques for measuring axial forces in the presence of bending forces is with low profile "shear beam" cells. They're available from Transducer Techniques and InterfaceForce amongst others.

 

[rb1957]

i agree with your alignment concerns, and maybe those "shear beam" cells are the way to go. i'd've thought that setting up several pairs of s/g would average out alignment issues, or you could test load with an axial load to calibrate the s/gs.

 

[GregLocock]

Aye, you can fuss about things, or do it properly. 3,4 or 6 strain gauges will do the job, and if you tell us no more, that's it.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[lmuri]

Hi,

I attached a picture of the system. It is for a leg prosthesis, and the loadcell is labeled in the drawing. Because it is very hard to decide how big the moment will be and which direction the moment will point to when a person is walking with this device, averaging approach will not work.

Thank you for all the response.

Best.

Ming

 

[btrueblood]

3, 4 or 6 strain gages, and some work with calibrated loads (apply known, calibrated, moments and forces and measure outputs, develop a data reduction matrix (system of linear equations) that will return load given strain gage readings). There are entire textbook chapters documenting how to do this. Wind tunnels use such instrumentation (they are called "stinger" balances there) for obtaining simultaneous lift, drag, and pitch/yaw moments on models in the tunnels, with the minimum obstruction to airflow; so, you might want to pick up a good textbook on low speed wind tunnel instrumentation (hint: Pope), or find a copy in the local engineering school library.

 

[rb1957]

i wonder why you'd want to know the compression load in a shaft, and not be interested in the moment ?

 

[lmuri]

Hi,

The compression load is used for control purpose.

Best.

Ming