Strain Gauge Mounting with Rod End Bearing

From my own experience strain gauge mounting is a tricky job.

In your case I am (and I imagine others are as well) confused as to what your application is and how you intend to gauge whatever it is you want to gauge. What you appear to be after is tension and compression, but not torsion. I just don't know the mechanicals details.

http://www.vishay.com/strain-gages/

If you go to the right of the page you will see links to their knowledge base. They provide a lot of data and technique for strain gauge use.

If that doesn't answer your question, provide us with a drawing and force vectors where /what you want to measure and how much force. At least with a little more information we might be able to help effectively.

You will need to do a lot of stress calculations and material analysis if you want to do this right.

This is still in concept stage, so I don't have exact dimensions. The forces are around 2 Kg max. As you see the Cyan color piece is guided on LM guideway & is pushing & pulling the blue color piece (also mounted on the same guideway thru Strain gauge & rod end bearings.

Rod end bearing is provided to take care of any X or Y misalignment between the cyan & blue piece.

The first answer was based on a misunderstanding: you do not bond a strain gage but use a force transducer which makes use of strain gages. It is a difference and of importance.

The error comes from the industrial wrong usage of "strain gage" for a "transducer". To make it clear the "strain gage" is the element bonded to a support with the scope to determine the local deformation of mentioned body. The "transducer" is the complex of a body + as many as necessary strain gages bonded on it to determine the magnitude of a mechanical load via the deformations of the supporting body.

The question is which is the uncertainty you can afford?

The transducer has - in the design you show and I know very well - a rather low one. The transducer body design in it self makes the result quite insensitive to transverse loads (as well torques as forces) and the cross sensitivity of the used strain gages is also low. If your measurements can afford a let say uncertainty of 2x to 2.5x the one of the transducer alone you do not have reasons to be worried.

If you want to have a very low added uncertainty then you have to make some efforts.

What you have to do is :

- decrease as much as possible the friction in the bearings as well via bearing design (ball bearings in stead of sliding bearings) as using the lowest possible dimension since friction torque is proportional to the bearing diameter.

- increase as much as allowed the distance between the bearings since the transverse forces are inversely proportional to this distance.

The solutions suggested with elastic hinges are very good and practiced in many lab high level applications but requires a not neglectable amount of work for their design especially at the low force level you want to work.

You can determine the level of added uncertainty considering following equations:

- friction torque in a bearing T= µ*F*D with:

- µ friction coefficient in the bearing / F axial force / Bearing sliding or rolling diameter

Radial force Fr= 2*T/ L with L= distance between bearings

Sensitivity of strain gages in transverse direction for same strain 1...3% which leads to same order of magnitude for a bridge output (in worse case) of output in active direction.

If questions I shall be pleased to answer.

A "rosette" strain gage will reveal lateral forces, if that's the kind of thing you're worried about. Conversely, if you want to eliminate lateral forces (off-axis), then you'll need to design your machine and grips with very low fricition bearings (able to handle the loads) so that you won't have any.

The problem is measurement in compression. In tension with bearings at both ends the rod will align and if one of the bearings can slide in the plane of the end plates you can get almost a pure traction in the rod. In compression it is different because a slight eccentricity will lead to a moment which tries to increase eccentricity. For such situations there are devices which convert a compression in a traction on the transducer and thus reduce the risc of transverse effects. It is also possible to use your rod with several strain gages measuring as well in axial as in transverse direction and build up a compensation for induced errors. The best as the other comment writes is to make a sketch with dimensions and explicit goal it is the only way to obtain a professional answer.

Hi,

these rod end bearings have a spherical part that allows angular movements in two orthogonal directions.

The diameter of this sphere is much too big to allow precise measurements as the friction at first very slight loading is high and will not allow any further adjustment to elastically deforming parts.

You need a bearing that has very low friction irrespective of load.

a. The friction of the sphere can be avoided by making a spherical hydrostatic or aerostatic bearing: not really a cheap solution but very good.

b. Allow two ball bearings per axis and two axis with intermediate gimbal per clamping. This two times for upper and lower clamp.

c. Replace the ball bearings by metallic strips of high quality flexure material: any misalignment will be removed by bending of flexures, only slight residual bending moment by flexure bending stiffness.

d. replace the rod end bearing by a long and thin wire of sufficient strength (if tension only).

Look at the fixing elements used at material testing: they have the same problems.

Your approach to relieve some of the misalignments by the cuts into the material to be tested is a good approach but needs some further improvements: the slots shall be designed not to introduce stress concentrations at the edges. And allow angular movement in two directions!

RHABE

As already said, the "z" form piece seems the typical load cell used in many tensile test machines. If this is the case, I agree all comments about centering loads. If you are to install the strain gages, the best solution is to install full bridge (two "T" rosettes") on opposite sides at same section of load measuring member (Poisson mounting). This will compensate any torsional or bending effect and give you just the axial load.

Best regards