6-axis load cell force calculation

[happygopher]

I feel like I am overlooking something really simple here but I could use your help to restore my sanity.

I have some data I collected using a 6-axis load cell. I would like to use this and calculate the resultant load force applied to the tool tip.

Am I correct in wanting to use the 3 force balance equations and 3 moment balance equations to find the solution? The problem I am having trouble wrapping my head around is that I would have 6 equations but I'm not sure that I have 6 unknown values.

So, given some known distance along the z-axis (longitudinal axis) that the load force is applied, how can I calculate what that load force is and its spatial orientation?

I would really appreciate any insights you can provide.

 

[BobM3]

Yes, you'll have 6 equations and 6 unknowns. Apply the 3 forces and 3 moments at a point on the cutting tool and use the geometry and the 6 measured forces to calculate the unkown forces and moments at the tool tip. You'll need to know the spacing between the force measurement points inside the transducer.

Where did you find a 6 axis unit? I've heard of 3. My guess is that the moments at the tool tip will be low compared to the forces at the tool tip.

 

[GregLocock]

We use 6 axis systems in wheels. If you are doing it by hand then draw the three views and you'll probably be able to figure out the appropriate equation. Getting the polarities right is most of the battle, that's why I recommend drawing all 3.





Cheers

Greg Locock


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

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

 

[happygopher]

The load cell actually measures the three force components and three moments at its origin.

If the tip where the load is applied is cantilevered out from the fixed load cell, wouldn't there be no moments applied at the tip?

As far as I know these things are fairly common. I've heard of them before but this is the first time I have used one.

 

[GregLocock]

That's right there should be no moment at the tip.

So for example, Force_tip_X=Force_Load_Cell_x+Y*TorqueZ

where Y is the obvious measurement and the + may be - if I've got my signs wrong.

Cheers

Greg Locock


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

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

 

[happygopher]

Sorry if I am being dense, but I'm not sure what is going on with the part of your equation "Y*TorqueZ". If this equation is a sum of forces, what does this term represent?

 

[GregLocock]

A force. that's how torques work.

Cheers

Greg Locock


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

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

 

[happygopher]

But aren't you multiplying a torque by a length.

So your units are lbf*in^2 (or N*m^2)?

 

[GregLocock]

Sorry, you are right, it should be torque divided by the distance.

I should have drawn the diagram.

Cheers

Greg Locock


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

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

 

[happygopher]

OK, I should have realized the simple typo but I have not been looking at this very clearly...

Thanks for your help, I will be looking through the data more tomorrow and see if I can make sense of what is happening.

 

[BobM3]

three moments at its origin

I'm not quite sure what that means. Where is the origin? If the origin is at the center of the Fx,Fy and Fz measurement points within the load cell, then Fx,Fy and Fz directly read the external forces applied at the tool tip (assuming no moments applied at the tool tip). In that case you don't need the Mz output from the cell (Measured Fx = Tool Tip Fx)

 

[btrueblood]

Past experience at anything similar was on wind tunnel force balances. Since models in the wind tunnel are capable of generating quite high moments, the idea that there is zero moment at the tip of a cantilever beam ("stinger") force balance doesn't ring true to me...but it might be very small and safely ignored for your case.

What I did find puzzling until I did the experiments and the math...every load cell for every axis had some degree of cross-talk to every other axis. Whether due to flexing of the structure, or simple alignment errors, or both. To put it another way, when we applied a strong sideways force (yaw) during calibration, some response would be noted on the fore/aft (drag) sensor, and some on the vertical (lift) sensor as well, ditto the moment sensors. This resulted in, e.g. for a 6-dof force balance, the correction equations forming a 6x6 matrix. To reduce the data, we multiplied the matrix by the measured force/torque "vector" to produce the true force "output" vector.

Dunno, maybe the sensor manufacturer has calibrated the device you are using, and the corrections are made internally in some black box, and the alignment is very tightly controlled by tight-fitting dowel pins and a finely polished flat surface plate...but I'd still calibrate it and verify that it is so, as any misalignment of your tool axes to the sensor axes will show up as moments or forces in funny places...

 

[BobM3]

btrueblood -

Even more incredible is that this 6 axis transducer is for a machine tool. It would have to be stiff in all 6 axes which would make the problem with cross talk even more difficult.

happygopher -

Is this for a lathe (transducer connected to tool bit) or a mill (transducer connected to work piece)? Can you say who the manufacturer is? (Just curious, multiaxis load cells are neat)