# Velocity issues -- experimental data from an accelerometer and a gyroscope

1. Dec 4, 2014

### BicBiro

I'm trying to get an understanding on what's going on with some data of mine. I've attached some plots below.

This data is from and accelerometer and gyroscope where I've calculated the global accelerations and velocities (top and bottom plot respectively in each figure). The data is from the arm of someone lifting an item which is increasing in weight, and thus their speed becomes slower as the weight gets heavier. My understanding on what I should see is simple; as weight increases the acceleration and velocity decreases. However, I don't think this is what I'm seeing.

This is a light weight. I've put lines where the start and end of the lift begins (I've only highlighted one of four lifts). Notice the amount of time elapsed in both plots and the maximum velocity (black horizontal line) in the bottom plot is around 0.9m/s.

Now look at a much heavier weight. Having watched the person perform the task I know they performed it slower, and the elapsed time proves this as it's taken twice as long to perform the task, but the maximum velocity is higher than that of a lighter weight at ~1.2m/s.

Have I performed a calculation wrong and what I'm seeing is incorrect, or is it correct and I'm just plain not understanding? I expect the velocity plot to be as wide as it is, as I know the task took longer to perform due to the additional exertion required from the lifter, but I also expected a lower velocity to reflect this. Any help would be appreciated.

2. Dec 4, 2014

### Danger

I really can't understand any of that stuff, so I'm just going to ask you what might be a silly question flat-out. Since the leverage and thus effective weight will increase as the arm swings up, might that have an effect that isn't linearly proportional to the actual weight? Also, won't the transition from horizontal to vertical motion during the arc have an accelerative effect upon the speed of the upward motion? If those are actually covered in your charts, my apologies for redundancy.

3. Dec 4, 2014

### BicBiro

Thanks for your reply. Sorry I was a little vague. Think of the lifting activity as a benchpress of sorts, so the weight is only being lifted in one dimension. Unfortunately the device recording the session wasn't mounted in a perfect perpendicular way so the three dimensional planes aren't perfectly straight with gravity, but the sessions were performed in the same way for each addition of weight so there shouldn't be a discrepancy between weights.

4. Dec 4, 2014

### Danger

Thank you for the clarification. Unfortunately, it means that I'm pretty much hooped as far as helping out goes. Someone else should be along shortly.

5. Dec 4, 2014

### BicBiro

No problem, thanks anyway.

6. Dec 4, 2014

### A.T.

What is the justification for these assumptions? Acceleration? Maybe, if they try the maximize it in both cases. Velocity? Not so sure. It is easy to lift a light object slowly, while a very heavy object is usually thrown upwards.

Your graphs seem bogus. How can there be zero velocity periods, despite varying acceleration?

7. Dec 4, 2014

### Danger

I just thought of something else. Even in a bench-press situation, the leverage changes significantly as the angle of the elbows straightens.

8. Dec 5, 2014

### BicBiro

The zero points in the velocity are to remove velocity drift from the calculations using MEMS components. It's probably best to ignore those points as they're where I calculated stationary (or some percentage either side of stationary).

As I mentioned, it was visibly slower, and calculating the elapsed time of the curve which is the lift of the object is gets slower as the weight increases. Should the velocity not coincide with this?

9. Dec 5, 2014

### A.T.

It still doesn't make sense.

If the weight is stationary in these phases, how come there is such significant acceleration? In the first case, the last 0.5s before the red line has an avg of ~4 m/s^2, which would result in 2m/s velocity change, much higher than anything you have on the velocity graph.

Did you remove 1g upwards from the accelerometer data, to get acceleration relative to the ground, instead relative to free fall?

Why is the start of the lift defined to be at peak acceleration, not begin of acceleration?